{
  "schema_version": 1,
  "generated_utc": "2026-09-07T07:27:16Z",
  "experiment": {
    "name": "Deterministic full-tree vanilla CFR on two-player Kuhn Poker",
    "implementation": "JavaScript ES module; no sampling and no random seed",
    "iterations": 100000,
    "chance_deals": [
      "JQ",
      "JK",
      "QJ",
      "QK",
      "KJ",
      "KQ"
    ],
    "chance_probability_each": 0.16666666666666666,
    "ranks_low_to_high": [
      "J",
      "Q",
      "K"
    ],
    "ante_per_player": 1,
    "bet_size": 1,
    "utility_unit": "net antes per deal",
    "game_value_p0": -0.05555555555555555,
    "averaging": "reach-weighted behavioral average, accumulated from iteration 1",
    "update_schedule": "simultaneous regret update after all six ordered deals",
    "current_strategy_timing": "regret-matched policy after each completed update",
    "nash_conv_definition": "sum of both information-set-consistent unilateral best-response values in this zero-sum game",
    "exploitability_definition": "NashConv divided by two, matching OpenSpiel convention for two-player constant-sum games",
    "best_response_method": "exhaustive enumeration of 64 pure behavioral plans per responder; one action per information set"
  },
  "checkpoints": [
    {
      "iteration": 1,
      "average": {
        "on_policy_value_p0": 0.125,
        "best_response_value_p0": 0.5,
        "best_response_value_p1": 0.416666666666667,
        "unilateral_improvement_p0": 0.375,
        "unilateral_improvement_p1": 0.541666666666667,
        "nash_conv": 0.916666666666667,
        "exploitability": 0.458333333333333
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 2,
      "average": {
        "on_policy_value_p0": -0.03125,
        "best_response_value_p0": 0.25,
        "best_response_value_p1": 0.375,
        "unilateral_improvement_p0": 0.28125,
        "unilateral_improvement_p1": 0.34375,
        "nash_conv": 0.625,
        "exploitability": 0.3125
      },
      "current": {
        "on_policy_value_p0": -0.041666666666667,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.375,
        "unilateral_improvement_p1": 0.125,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 5,
      "average": {
        "on_policy_value_p0": -0.025984848484849,
        "best_response_value_p0": 0.175757575757576,
        "best_response_value_p1": 0.201388888888889,
        "unilateral_improvement_p0": 0.201742424242424,
        "unilateral_improvement_p1": 0.17540404040404,
        "nash_conv": 0.377146464646465,
        "exploitability": 0.188573232323232
      },
      "current": {
        "on_policy_value_p0": -0.095430107526882,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.262096774193548,
        "unilateral_improvement_p1": 0.071236559139785,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 10,
      "average": {
        "on_policy_value_p0": -0.035192761187016,
        "best_response_value_p0": 0.057000333736143,
        "best_response_value_p1": 0.135416666666667,
        "unilateral_improvement_p0": 0.092193094923159,
        "unilateral_improvement_p1": 0.100223905479651,
        "nash_conv": 0.19241700040281,
        "exploitability": 0.096208500201405
      },
      "current": {
        "on_policy_value_p0": -0.069444444444444,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.215277777777778,
        "unilateral_improvement_p0": 0.236111111111111,
        "unilateral_improvement_p1": 0.145833333333333,
        "nash_conv": 0.381944444444444,
        "exploitability": 0.190972222222222
      }
    },
    {
      "iteration": 20,
      "average": {
        "on_policy_value_p0": -0.048620104396379,
        "best_response_value_p0": -0.004208910273227,
        "best_response_value_p1": 0.136771668270058,
        "unilateral_improvement_p0": 0.044411194123153,
        "unilateral_improvement_p1": 0.088151563873679,
        "nash_conv": 0.132562757996831,
        "exploitability": 0.066281378998416
      },
      "current": {
        "on_policy_value_p0": 0.096738232130044,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.23659510120329,
        "unilateral_improvement_p1": 0.26340489879671,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 50,
      "average": {
        "on_policy_value_p0": -0.050570022375569,
        "best_response_value_p0": 0.003396547783115,
        "best_response_value_p1": 0.069039945711588,
        "unilateral_improvement_p0": 0.053966570158685,
        "unilateral_improvement_p1": 0.018469923336019,
        "nash_conv": 0.072436493494703,
        "exploitability": 0.036218246747352
      },
      "current": {
        "on_policy_value_p0": -0.211577704536647,
        "best_response_value_p0": 0.211577704536647,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.423155409073294,
        "unilateral_improvement_p1": 0.121755628796686,
        "nash_conv": 0.54491103786998,
        "exploitability": 0.27245551893499
      }
    },
    {
      "iteration": 100,
      "average": {
        "on_policy_value_p0": -0.055987211609916,
        "best_response_value_p0": -0.028620299574557,
        "best_response_value_p1": 0.079969771268453,
        "unilateral_improvement_p0": 0.02736691203536,
        "unilateral_improvement_p1": 0.023982559658537,
        "nash_conv": 0.051349471693896,
        "exploitability": 0.025674735846948
      },
      "current": {
        "on_policy_value_p0": 0.123274462475016,
        "best_response_value_p0": 0.123274462475016,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.456607795808349,
        "nash_conv": 0.456607795808349,
        "exploitability": 0.228303897904175
      }
    },
    {
      "iteration": 200,
      "average": {
        "on_policy_value_p0": -0.054423786261989,
        "best_response_value_p0": -0.029262058909649,
        "best_response_value_p1": 0.063873159165081,
        "unilateral_improvement_p0": 0.02516172735234,
        "unilateral_improvement_p1": 0.009449372903092,
        "nash_conv": 0.034611100255432,
        "exploitability": 0.017305550127716
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 500,
      "average": {
        "on_policy_value_p0": -0.055260296874071,
        "best_response_value_p0": -0.050200112298986,
        "best_response_value_p1": 0.071562371284827,
        "unilateral_improvement_p0": 0.005060184575085,
        "unilateral_improvement_p1": 0.016302074410757,
        "nash_conv": 0.021362258985841,
        "exploitability": 0.010681129492921
      },
      "current": {
        "on_policy_value_p0": 0.093207851365374,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.240125481967959,
        "unilateral_improvement_p1": 0.259874518032041,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 1000,
      "average": {
        "on_policy_value_p0": -0.055557219504907,
        "best_response_value_p0": -0.047681226763959,
        "best_response_value_p1": 0.062219439581088,
        "unilateral_improvement_p0": 0.007875992740947,
        "unilateral_improvement_p1": 0.006662220076181,
        "nash_conv": 0.014538212817129,
        "exploitability": 0.007269106408564
      },
      "current": {
        "on_policy_value_p0": 0.096766188403199,
        "best_response_value_p0": 0.096766188403199,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.430099521736532,
        "nash_conv": 0.430099521736532,
        "exploitability": 0.215049760868266
      }
    },
    {
      "iteration": 2000,
      "average": {
        "on_policy_value_p0": -0.055462773599963,
        "best_response_value_p0": -0.048308105600117,
        "best_response_value_p1": 0.058883768453381,
        "unilateral_improvement_p0": 0.007154667999846,
        "unilateral_improvement_p1": 0.003420994853418,
        "nash_conv": 0.010575662853264,
        "exploitability": 0.005287831426632
      },
      "current": {
        "on_policy_value_p0": -0.02610810430578,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.447783791388441,
        "unilateral_improvement_p0": 0.192774770972446,
        "unilateral_improvement_p1": 0.421675687082661,
        "nash_conv": 0.614450458055107,
        "exploitability": 0.307225229027554
      }
    },
    {
      "iteration": 5000,
      "average": {
        "on_policy_value_p0": -0.055506400503875,
        "best_response_value_p0": -0.053871953449214,
        "best_response_value_p1": 0.060476849562444,
        "unilateral_improvement_p0": 0.001634447054661,
        "unilateral_improvement_p1": 0.004970449058569,
        "nash_conv": 0.00660489611323,
        "exploitability": 0.003302448056615
      },
      "current": {
        "on_policy_value_p0": -0.109647036774031,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.276313703440698,
        "unilateral_improvement_p1": 0.057019629892635,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 10000,
      "average": {
        "on_policy_value_p0": -0.055546395829272,
        "best_response_value_p0": -0.054345107016834,
        "best_response_value_p1": 0.058980679646773,
        "unilateral_improvement_p0": 0.001201288812438,
        "unilateral_improvement_p1": 0.003434283817501,
        "nash_conv": 0.00463557262994,
        "exploitability": 0.00231778631497
      },
      "current": {
        "on_policy_value_p0": 0.034333022839931,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.132333643826735,
        "unilateral_improvement_p0": 0.132333643826735,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.299000310493402,
        "exploitability": 0.149500155246701
      }
    },
    {
      "iteration": 20000,
      "average": {
        "on_policy_value_p0": -0.055536440314122,
        "best_response_value_p0": -0.053997250601598,
        "best_response_value_p1": 0.057262502702602,
        "unilateral_improvement_p0": 0.001539189712524,
        "unilateral_improvement_p1": 0.00172606238848,
        "nash_conv": 0.003265252101004,
        "exploitability": 0.001632626050502
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 50000,
      "average": {
        "on_policy_value_p0": -0.05555595695098,
        "best_response_value_p0": -0.054575375618386,
        "best_response_value_p1": 0.056573033227029,
        "unilateral_improvement_p0": 0.000980581332594,
        "unilateral_improvement_p1": 0.001017076276048,
        "nash_conv": 0.001997657608643,
        "exploitability": 0.000998828804321
      },
      "current": {
        "on_policy_value_p0": 0.086892731187916,
        "best_response_value_p0": 0.107245859523251,
        "best_response_value_p1": 0.292627076662662,
        "unilateral_improvement_p0": 0.020353128335335,
        "unilateral_improvement_p1": 0.379519807850578,
        "nash_conv": 0.399872936185914,
        "exploitability": 0.199936468092957
      }
    },
    {
      "iteration": 100000,
      "average": {
        "on_policy_value_p0": -0.05555470109818,
        "best_response_value_p0": -0.05476979504503,
        "best_response_value_p1": 0.056122234405666,
        "unilateral_improvement_p0": 0.00078490605315,
        "unilateral_improvement_p1": 0.000567533307486,
        "nash_conv": 0.001352439360636,
        "exploitability": 0.000676219680318
      },
      "current": {
        "on_policy_value_p0": 0.126127974981742,
        "best_response_value_p0": 0.217263089911864,
        "best_response_value_p1": 0.317729770139755,
        "unilateral_improvement_p0": 0.091135114930122,
        "unilateral_improvement_p1": 0.443857745121497,
        "nash_conv": 0.53499286005162,
        "exploitability": 0.26749643002581
      }
    }
  ],
  "trace": [
    {
      "iteration": 1,
      "average": {
        "on_policy_value_p0": 0.125,
        "best_response_value_p0": 0.5,
        "best_response_value_p1": 0.416666666666667,
        "unilateral_improvement_p0": 0.375,
        "unilateral_improvement_p1": 0.541666666666667,
        "nash_conv": 0.916666666666667,
        "exploitability": 0.458333333333333
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 2,
      "average": {
        "on_policy_value_p0": -0.03125,
        "best_response_value_p0": 0.25,
        "best_response_value_p1": 0.375,
        "unilateral_improvement_p0": 0.28125,
        "unilateral_improvement_p1": 0.34375,
        "nash_conv": 0.625,
        "exploitability": 0.3125
      },
      "current": {
        "on_policy_value_p0": -0.041666666666667,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.375,
        "unilateral_improvement_p1": 0.125,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 5,
      "average": {
        "on_policy_value_p0": -0.025984848484849,
        "best_response_value_p0": 0.175757575757576,
        "best_response_value_p1": 0.201388888888889,
        "unilateral_improvement_p0": 0.201742424242424,
        "unilateral_improvement_p1": 0.17540404040404,
        "nash_conv": 0.377146464646465,
        "exploitability": 0.188573232323232
      },
      "current": {
        "on_policy_value_p0": -0.095430107526882,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.262096774193548,
        "unilateral_improvement_p1": 0.071236559139785,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 10,
      "average": {
        "on_policy_value_p0": -0.035192761187016,
        "best_response_value_p0": 0.057000333736143,
        "best_response_value_p1": 0.135416666666667,
        "unilateral_improvement_p0": 0.092193094923159,
        "unilateral_improvement_p1": 0.100223905479651,
        "nash_conv": 0.19241700040281,
        "exploitability": 0.096208500201405
      },
      "current": {
        "on_policy_value_p0": -0.069444444444444,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.215277777777778,
        "unilateral_improvement_p0": 0.236111111111111,
        "unilateral_improvement_p1": 0.145833333333333,
        "nash_conv": 0.381944444444444,
        "exploitability": 0.190972222222222
      }
    },
    {
      "iteration": 20,
      "average": {
        "on_policy_value_p0": -0.048620104396379,
        "best_response_value_p0": -0.004208910273227,
        "best_response_value_p1": 0.136771668270058,
        "unilateral_improvement_p0": 0.044411194123153,
        "unilateral_improvement_p1": 0.088151563873679,
        "nash_conv": 0.132562757996831,
        "exploitability": 0.066281378998416
      },
      "current": {
        "on_policy_value_p0": 0.096738232130044,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.23659510120329,
        "unilateral_improvement_p1": 0.26340489879671,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 50,
      "average": {
        "on_policy_value_p0": -0.050570022375569,
        "best_response_value_p0": 0.003396547783115,
        "best_response_value_p1": 0.069039945711588,
        "unilateral_improvement_p0": 0.053966570158685,
        "unilateral_improvement_p1": 0.018469923336019,
        "nash_conv": 0.072436493494703,
        "exploitability": 0.036218246747352
      },
      "current": {
        "on_policy_value_p0": -0.211577704536647,
        "best_response_value_p0": 0.211577704536647,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.423155409073294,
        "unilateral_improvement_p1": 0.121755628796686,
        "nash_conv": 0.54491103786998,
        "exploitability": 0.27245551893499
      }
    },
    {
      "iteration": 100,
      "average": {
        "on_policy_value_p0": -0.055987211609916,
        "best_response_value_p0": -0.028620299574557,
        "best_response_value_p1": 0.079969771268453,
        "unilateral_improvement_p0": 0.02736691203536,
        "unilateral_improvement_p1": 0.023982559658537,
        "nash_conv": 0.051349471693896,
        "exploitability": 0.025674735846948
      },
      "current": {
        "on_policy_value_p0": 0.123274462475016,
        "best_response_value_p0": 0.123274462475016,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.456607795808349,
        "nash_conv": 0.456607795808349,
        "exploitability": 0.228303897904175
      }
    },
    {
      "iteration": 200,
      "average": {
        "on_policy_value_p0": -0.054423786261989,
        "best_response_value_p0": -0.029262058909649,
        "best_response_value_p1": 0.063873159165081,
        "unilateral_improvement_p0": 0.02516172735234,
        "unilateral_improvement_p1": 0.009449372903092,
        "nash_conv": 0.034611100255432,
        "exploitability": 0.017305550127716
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 300,
      "average": {
        "on_policy_value_p0": -0.055637657306869,
        "best_response_value_p0": -0.037513805198584,
        "best_response_value_p1": 0.064593381485363,
        "unilateral_improvement_p0": 0.018123852108284,
        "unilateral_improvement_p1": 0.008955724178494,
        "nash_conv": 0.027079576286778,
        "exploitability": 0.013539788143389
      },
      "current": {
        "on_policy_value_p0": 0.090466603744068,
        "best_response_value_p0": 0.090466603744068,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.423799937077401,
        "nash_conv": 0.423799937077401,
        "exploitability": 0.211899968538701
      }
    },
    {
      "iteration": 400,
      "average": {
        "on_policy_value_p0": -0.054864644596016,
        "best_response_value_p0": -0.048356308921647,
        "best_response_value_p1": 0.073618616036926,
        "unilateral_improvement_p0": 0.006508335674368,
        "unilateral_improvement_p1": 0.01875397144091,
        "nash_conv": 0.025262307115278,
        "exploitability": 0.012631153557639
      },
      "current": {
        "on_policy_value_p0": -0.147767573492852,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.15237634075863,
        "unilateral_improvement_p0": 0.314434240159518,
        "unilateral_improvement_p1": 0.004608767265779,
        "nash_conv": 0.319043007425297,
        "exploitability": 0.159521503712649
      }
    },
    {
      "iteration": 500,
      "average": {
        "on_policy_value_p0": -0.055260296874071,
        "best_response_value_p0": -0.050200112298986,
        "best_response_value_p1": 0.071562371284827,
        "unilateral_improvement_p0": 0.005060184575085,
        "unilateral_improvement_p1": 0.016302074410757,
        "nash_conv": 0.021362258985841,
        "exploitability": 0.010681129492921
      },
      "current": {
        "on_policy_value_p0": 0.093207851365374,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.240125481967959,
        "unilateral_improvement_p1": 0.259874518032041,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 600,
      "average": {
        "on_policy_value_p0": -0.055548396992466,
        "best_response_value_p0": -0.043632253906033,
        "best_response_value_p1": 0.062165611604151,
        "unilateral_improvement_p0": 0.011916143086433,
        "unilateral_improvement_p1": 0.006617214611686,
        "nash_conv": 0.018533357698118,
        "exploitability": 0.009266678849059
      },
      "current": {
        "on_policy_value_p0": 0.083657022788456,
        "best_response_value_p0": 0.083657022788456,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.41699035612179,
        "nash_conv": 0.41699035612179,
        "exploitability": 0.208495178060895
      }
    },
    {
      "iteration": 700,
      "average": {
        "on_policy_value_p0": -0.055310827448669,
        "best_response_value_p0": -0.044524907864767,
        "best_response_value_p1": 0.061764841746995,
        "unilateral_improvement_p0": 0.010785919583902,
        "unilateral_improvement_p1": 0.006454014298326,
        "nash_conv": 0.017239933882227,
        "exploitability": 0.008619966941114
      },
      "current": {
        "on_policy_value_p0": 0.006657326252393,
        "best_response_value_p0": 0.120644263200189,
        "best_response_value_p1": 0.329225038638876,
        "unilateral_improvement_p0": 0.113986936947796,
        "unilateral_improvement_p1": 0.335882364891269,
        "nash_conv": 0.449869301839065,
        "exploitability": 0.224934650919533
      }
    },
    {
      "iteration": 800,
      "average": {
        "on_policy_value_p0": -0.055514512877685,
        "best_response_value_p0": -0.044023634107864,
        "best_response_value_p1": 0.060583949689441,
        "unilateral_improvement_p0": 0.011490878769821,
        "unilateral_improvement_p1": 0.005069436811756,
        "nash_conv": 0.016560315581577,
        "exploitability": 0.008280157790789
      },
      "current": {
        "on_policy_value_p0": 0.054841863293638,
        "best_response_value_p0": 0.056982940079391,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.002141076785753,
        "unilateral_improvement_p1": 0.388175196626971,
        "nash_conv": 0.390316273412724,
        "exploitability": 0.195158136706362
      }
    },
    {
      "iteration": 900,
      "average": {
        "on_policy_value_p0": -0.055452080958117,
        "best_response_value_p0": -0.046132209023174,
        "best_response_value_p1": 0.060654432160482,
        "unilateral_improvement_p0": 0.009319871934943,
        "unilateral_improvement_p1": 0.005202351202364,
        "nash_conv": 0.014522223137307,
        "exploitability": 0.007261111568654
      },
      "current": {
        "on_policy_value_p0": 0.112269437913411,
        "best_response_value_p0": 0.192872692148779,
        "best_response_value_p1": 0.338793491529264,
        "unilateral_improvement_p0": 0.080603254235368,
        "unilateral_improvement_p1": 0.451062929442675,
        "nash_conv": 0.531666183678043,
        "exploitability": 0.265833091839021
      }
    },
    {
      "iteration": 1000,
      "average": {
        "on_policy_value_p0": -0.055557219504907,
        "best_response_value_p0": -0.047681226763959,
        "best_response_value_p1": 0.062219439581088,
        "unilateral_improvement_p0": 0.007875992740947,
        "unilateral_improvement_p1": 0.006662220076181,
        "nash_conv": 0.014538212817129,
        "exploitability": 0.007269106408564
      },
      "current": {
        "on_policy_value_p0": 0.096766188403199,
        "best_response_value_p0": 0.096766188403199,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.430099521736532,
        "nash_conv": 0.430099521736532,
        "exploitability": 0.215049760868266
      }
    },
    {
      "iteration": 1100,
      "average": {
        "on_policy_value_p0": -0.055387937301166,
        "best_response_value_p0": -0.052625013783706,
        "best_response_value_p1": 0.06647994928399,
        "unilateral_improvement_p0": 0.002762923517461,
        "unilateral_improvement_p1": 0.011092011982823,
        "nash_conv": 0.013854935500284,
        "exploitability": 0.006927467750142
      },
      "current": {
        "on_policy_value_p0": 0.075258801826043,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.25807453150729,
        "unilateral_improvement_p1": 0.24192546849271,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 1200,
      "average": {
        "on_policy_value_p0": -0.055417966335593,
        "best_response_value_p0": -0.052666778934956,
        "best_response_value_p1": 0.066309112761781,
        "unilateral_improvement_p0": 0.002751187400636,
        "unilateral_improvement_p1": 0.010891146426189,
        "nash_conv": 0.013642333826825,
        "exploitability": 0.006821166913412
      },
      "current": {
        "on_policy_value_p0": -0.061369092129422,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.228035758796088,
        "unilateral_improvement_p1": 0.105297574537245,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 1300,
      "average": {
        "on_policy_value_p0": -0.055311699978819,
        "best_response_value_p0": -0.0457894867034,
        "best_response_value_p1": 0.058229540595933,
        "unilateral_improvement_p0": 0.009522213275419,
        "unilateral_improvement_p1": 0.002917840617114,
        "nash_conv": 0.012440053892533,
        "exploitability": 0.006220026946266
      },
      "current": {
        "on_policy_value_p0": -0.186840487689708,
        "best_response_value_p0": 0.186840487689708,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.373680975379417,
        "unilateral_improvement_p1": 0.146492845643625,
        "nash_conv": 0.520173821023042,
        "exploitability": 0.260086910511521
      }
    },
    {
      "iteration": 1400,
      "average": {
        "on_policy_value_p0": -0.055426376597356,
        "best_response_value_p0": -0.047130751521647,
        "best_response_value_p1": 0.059675095528273,
        "unilateral_improvement_p0": 0.008295625075709,
        "unilateral_improvement_p1": 0.004248718930917,
        "nash_conv": 0.012544344006626,
        "exploitability": 0.006272172003313
      },
      "current": {
        "on_policy_value_p0": -0.053375005081742,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.393249989836517,
        "unilateral_improvement_p0": 0.220041671748408,
        "unilateral_improvement_p1": 0.339874984754775,
        "nash_conv": 0.559916656503183,
        "exploitability": 0.279958328251592
      }
    },
    {
      "iteration": 1500,
      "average": {
        "on_policy_value_p0": -0.055570928838532,
        "best_response_value_p0": -0.050417203871714,
        "best_response_value_p1": 0.062533777818143,
        "unilateral_improvement_p0": 0.005153724966818,
        "unilateral_improvement_p1": 0.006962848979611,
        "nash_conv": 0.01211657394643,
        "exploitability": 0.006058286973215
      },
      "current": {
        "on_policy_value_p0": 0.093009986022837,
        "best_response_value_p0": 0.133524513327416,
        "best_response_value_p1": 0.252304278724176,
        "unilateral_improvement_p0": 0.040514527304578,
        "unilateral_improvement_p1": 0.345314264747014,
        "nash_conv": 0.385828792051592,
        "exploitability": 0.192914396025796
      }
    },
    {
      "iteration": 1600,
      "average": {
        "on_policy_value_p0": -0.055323567950608,
        "best_response_value_p0": -0.050295180117014,
        "best_response_value_p1": 0.062029609106455,
        "unilateral_improvement_p0": 0.005028387833594,
        "unilateral_improvement_p1": 0.006706041155847,
        "nash_conv": 0.011734428989441,
        "exploitability": 0.005867214494721
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 1700,
      "average": {
        "on_policy_value_p0": -0.055415315903866,
        "best_response_value_p0": -0.046627346949802,
        "best_response_value_p1": 0.058116066912108,
        "unilateral_improvement_p0": 0.008787968954065,
        "unilateral_improvement_p1": 0.002700751008241,
        "nash_conv": 0.011488719962306,
        "exploitability": 0.005744359981153
      },
      "current": {
        "on_policy_value_p0": -0.081522847290977,
        "best_response_value_p0": 0.112318101806016,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.193840949096992,
        "unilateral_improvement_p1": 0.418477152709023,
        "nash_conv": 0.612318101806016,
        "exploitability": 0.306159050903008
      }
    },
    {
      "iteration": 1800,
      "average": {
        "on_policy_value_p0": -0.055560904639095,
        "best_response_value_p0": -0.050145344185289,
        "best_response_value_p1": 0.061006223576564,
        "unilateral_improvement_p0": 0.005415560453806,
        "unilateral_improvement_p1": 0.00544531893747,
        "nash_conv": 0.010860879391276,
        "exploitability": 0.005430439695638
      },
      "current": {
        "on_policy_value_p0": 0.082618127479611,
        "best_response_value_p0": 0.109082492315165,
        "best_response_value_p1": 0.280404603662225,
        "unilateral_improvement_p0": 0.026464364835554,
        "unilateral_improvement_p1": 0.363022731141836,
        "nash_conv": 0.38948709597739,
        "exploitability": 0.194743547988695
      }
    },
    {
      "iteration": 1900,
      "average": {
        "on_policy_value_p0": -0.055352111767867,
        "best_response_value_p0": -0.050455924662894,
        "best_response_value_p1": 0.06121380971544,
        "unilateral_improvement_p0": 0.004896187104973,
        "unilateral_improvement_p1": 0.005861697947574,
        "nash_conv": 0.010757885052547,
        "exploitability": 0.005378942526273
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 2000,
      "average": {
        "on_policy_value_p0": -0.055462773599963,
        "best_response_value_p0": -0.048308105600117,
        "best_response_value_p1": 0.058883768453381,
        "unilateral_improvement_p0": 0.007154667999846,
        "unilateral_improvement_p1": 0.003420994853418,
        "nash_conv": 0.010575662853264,
        "exploitability": 0.005287831426632
      },
      "current": {
        "on_policy_value_p0": -0.02610810430578,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.447783791388441,
        "unilateral_improvement_p0": 0.192774770972446,
        "unilateral_improvement_p1": 0.421675687082661,
        "nash_conv": 0.614450458055107,
        "exploitability": 0.307225229027554
      }
    },
    {
      "iteration": 2100,
      "average": {
        "on_policy_value_p0": -0.055552553913902,
        "best_response_value_p0": -0.052566044390562,
        "best_response_value_p1": 0.062473336892374,
        "unilateral_improvement_p0": 0.00298650952334,
        "unilateral_improvement_p1": 0.006920782978473,
        "nash_conv": 0.009907292501812,
        "exploitability": 0.004953646250906
      },
      "current": {
        "on_policy_value_p0": 0.089777459371342,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.179554918742684,
        "unilateral_improvement_p0": 0.076889207295325,
        "unilateral_improvement_p1": 0.269332378114026,
        "nash_conv": 0.34622158540935,
        "exploitability": 0.173110792704675
      }
    },
    {
      "iteration": 2200,
      "average": {
        "on_policy_value_p0": -0.055420727264184,
        "best_response_value_p0": -0.047961204774949,
        "best_response_value_p1": 0.05728755544878,
        "unilateral_improvement_p0": 0.007459522489236,
        "unilateral_improvement_p1": 0.001866828184596,
        "nash_conv": 0.009326350673831,
        "exploitability": 0.004663175336916
      },
      "current": {
        "on_policy_value_p0": -0.164599123805093,
        "best_response_value_p0": 0.164599123805093,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.329198247610185,
        "unilateral_improvement_p1": 0.168734209528241,
        "nash_conv": 0.497932457138426,
        "exploitability": 0.248966228569213
      }
    },
    {
      "iteration": 2300,
      "average": {
        "on_policy_value_p0": -0.055517053649072,
        "best_response_value_p0": -0.050694048623618,
        "best_response_value_p1": 0.06047651354229,
        "unilateral_improvement_p0": 0.004823005025454,
        "unilateral_improvement_p1": 0.004959459893218,
        "nash_conv": 0.009782464918672,
        "exploitability": 0.004891232459336
      },
      "current": {
        "on_policy_value_p0": 0.20697303386972,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.247279401072774,
        "unilateral_improvement_p0": 0.126360299463613,
        "unilateral_improvement_p1": 0.454252434942494,
        "nash_conv": 0.580612734406107,
        "exploitability": 0.290306367203054
      }
    },
    {
      "iteration": 2400,
      "average": {
        "on_policy_value_p0": -0.055451990171955,
        "best_response_value_p0": -0.053171495887589,
        "best_response_value_p1": 0.062826059941626,
        "unilateral_improvement_p0": 0.002280494284367,
        "unilateral_improvement_p1": 0.007374069769671,
        "nash_conv": 0.009654564054038,
        "exploitability": 0.004827282027019
      },
      "current": {
        "on_policy_value_p0": -0.107426509981356,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.274093176648023,
        "unilateral_improvement_p1": 0.059240156685311,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 2500,
      "average": {
        "on_policy_value_p0": -0.055536517071997,
        "best_response_value_p0": -0.049339600778387,
        "best_response_value_p1": 0.058601121213161,
        "unilateral_improvement_p0": 0.00619691629361,
        "unilateral_improvement_p1": 0.003064604141163,
        "nash_conv": 0.009261520434774,
        "exploitability": 0.004630760217387
      },
      "current": {
        "on_policy_value_p0": 0.04699966886031,
        "best_response_value_p0": 0.072667328946047,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.025667660085738,
        "unilateral_improvement_p1": 0.380333002193643,
        "nash_conv": 0.406000662279381,
        "exploitability": 0.20300033113969
      }
    },
    {
      "iteration": 2600,
      "average": {
        "on_policy_value_p0": -0.055422354731063,
        "best_response_value_p0": -0.052752570358318,
        "best_response_value_p1": 0.061894276317128,
        "unilateral_improvement_p0": 0.002669784372746,
        "unilateral_improvement_p1": 0.006471921586065,
        "nash_conv": 0.00914170595881,
        "exploitability": 0.004570852979405
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 2700,
      "average": {
        "on_policy_value_p0": -0.055493355730567,
        "best_response_value_p0": -0.048899003522322,
        "best_response_value_p1": 0.057992566520284,
        "unilateral_improvement_p0": 0.006594352208245,
        "unilateral_improvement_p1": 0.002499210789717,
        "nash_conv": 0.009093562997962,
        "exploitability": 0.004546781498981
      },
      "current": {
        "on_policy_value_p0": -0.004375142312276,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.491249715375447,
        "unilateral_improvement_p0": 0.171041808978943,
        "unilateral_improvement_p1": 0.486874573063171,
        "nash_conv": 0.657916382042114,
        "exploitability": 0.328958191021057
      }
    },
    {
      "iteration": 2800,
      "average": {
        "on_policy_value_p0": -0.055526287234874,
        "best_response_value_p0": -0.053198990767608,
        "best_response_value_p1": 0.062060583816788,
        "unilateral_improvement_p0": 0.002327296467265,
        "unilateral_improvement_p1": 0.006534296581915,
        "nash_conv": 0.00886159304918,
        "exploitability": 0.00443079652459
      },
      "current": {
        "on_policy_value_p0": 0.041530728615329,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.125135938051337,
        "unilateral_improvement_p0": 0.125135938051337,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.291802604718004,
        "exploitability": 0.145901302359002
      }
    },
    {
      "iteration": 2900,
      "average": {
        "on_policy_value_p0": -0.055504540623188,
        "best_response_value_p0": -0.049072875759434,
        "best_response_value_p1": 0.057771658857856,
        "unilateral_improvement_p0": 0.006431664863754,
        "unilateral_improvement_p1": 0.002267118234667,
        "nash_conv": 0.008698783098421,
        "exploitability": 0.004349391549211
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 3000,
      "average": {
        "on_policy_value_p0": -0.055468389310939,
        "best_response_value_p0": -0.053937969227073,
        "best_response_value_p1": 0.062327021798845,
        "unilateral_improvement_p0": 0.001530420083866,
        "unilateral_improvement_p1": 0.006858632487906,
        "nash_conv": 0.008389052571772,
        "exploitability": 0.004194526285886
      },
      "current": {
        "on_policy_value_p0": -0.084912769341722,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.115705545431398,
        "unilateral_improvement_p0": 0.418246102675056,
        "unilateral_improvement_p1": 0.030792776089676,
        "nash_conv": 0.449038878764731,
        "exploitability": 0.224519439382366
      }
    },
    {
      "iteration": 3100,
      "average": {
        "on_policy_value_p0": -0.05548579815209,
        "best_response_value_p0": -0.049647434236636,
        "best_response_value_p1": 0.058106737889112,
        "unilateral_improvement_p0": 0.005838363915453,
        "unilateral_improvement_p1": 0.002620939737022,
        "nash_conv": 0.008459303652476,
        "exploitability": 0.004229651826238
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 3200,
      "average": {
        "on_policy_value_p0": -0.055534752093492,
        "best_response_value_p0": -0.053556025742513,
        "best_response_value_p1": 0.061727823948842,
        "unilateral_improvement_p0": 0.001978726350979,
        "unilateral_improvement_p1": 0.00619307185535,
        "nash_conv": 0.008171798206329,
        "exploitability": 0.004085899103165
      },
      "current": {
        "on_policy_value_p0": 0.057693898266314,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.115387796532627,
        "unilateral_improvement_p0": 0.108972768400353,
        "unilateral_improvement_p1": 0.173081694798941,
        "nash_conv": 0.282054463199294,
        "exploitability": 0.141027231599647
      }
    },
    {
      "iteration": 3300,
      "average": {
        "on_policy_value_p0": -0.055519869192705,
        "best_response_value_p0": -0.049537715458905,
        "best_response_value_p1": 0.057684916926611,
        "unilateral_improvement_p0": 0.0059821537338,
        "unilateral_improvement_p1": 0.002165047733906,
        "nash_conv": 0.008147201467706,
        "exploitability": 0.004073600733853
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 3400,
      "average": {
        "on_policy_value_p0": -0.055465181762414,
        "best_response_value_p0": -0.053881896396089,
        "best_response_value_p1": 0.061665908859221,
        "unilateral_improvement_p0": 0.001583285366326,
        "unilateral_improvement_p1": 0.006200727096807,
        "nash_conv": 0.007784012463133,
        "exploitability": 0.003892006231566
      },
      "current": {
        "on_policy_value_p0": -0.208999267553007,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.208999267553007,
        "unilateral_improvement_p0": 0.54233260088634,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.54233260088634,
        "exploitability": 0.27116630044317
      }
    },
    {
      "iteration": 3500,
      "average": {
        "on_policy_value_p0": -0.055511883290288,
        "best_response_value_p0": -0.049874761342294,
        "best_response_value_p1": 0.05776953650564,
        "unilateral_improvement_p0": 0.005637121947994,
        "unilateral_improvement_p1": 0.002257653215352,
        "nash_conv": 0.007894775163346,
        "exploitability": 0.003947387581673
      },
      "current": {
        "on_policy_value_p0": 0.012137639273847,
        "best_response_value_p0": 0.144805276566755,
        "best_response_value_p1": 0.436829718148046,
        "unilateral_improvement_p0": 0.132667637292909,
        "unilateral_improvement_p1": 0.448967357421893,
        "nash_conv": 0.581634994714801,
        "exploitability": 0.290817497357401
      }
    },
    {
      "iteration": 3600,
      "average": {
        "on_policy_value_p0": -0.055508542156421,
        "best_response_value_p0": -0.053751233265649,
        "best_response_value_p1": 0.061508851765761,
        "unilateral_improvement_p0": 0.001757308890772,
        "unilateral_improvement_p1": 0.00600030960934,
        "nash_conv": 0.007757618500112,
        "exploitability": 0.003878809250056
      },
      "current": {
        "on_policy_value_p0": -0.055510163394707,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.222176830061374,
        "unilateral_improvement_p1": 0.111156503271959,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 3700,
      "average": {
        "on_policy_value_p0": -0.055544926689346,
        "best_response_value_p0": -0.050609659023022,
        "best_response_value_p1": 0.058129108919166,
        "unilateral_improvement_p0": 0.004935267666324,
        "unilateral_improvement_p1": 0.002584182229821,
        "nash_conv": 0.007519449896144,
        "exploitability": 0.003759724948072
      },
      "current": {
        "on_policy_value_p0": 0.054926776373752,
        "best_response_value_p0": 0.056813113919164,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.001886337545412,
        "unilateral_improvement_p1": 0.388260109707085,
        "nash_conv": 0.390146447252497,
        "exploitability": 0.195073223626248
      }
    },
    {
      "iteration": 3800,
      "average": {
        "on_policy_value_p0": -0.055455519697412,
        "best_response_value_p0": -0.05131382577285,
        "best_response_value_p1": 0.058879839828803,
        "unilateral_improvement_p0": 0.004141693924562,
        "unilateral_improvement_p1": 0.00342432013139,
        "nash_conv": 0.007566014055953,
        "exploitability": 0.003783007027976
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 3900,
      "average": {
        "on_policy_value_p0": -0.055534484137098,
        "best_response_value_p0": -0.052068258306634,
        "best_response_value_p1": 0.05950093153985,
        "unilateral_improvement_p0": 0.003466225830463,
        "unilateral_improvement_p1": 0.003966447402753,
        "nash_conv": 0.007432673233216,
        "exploitability": 0.003716336616608
      },
      "current": {
        "on_policy_value_p0": 0.212611538653019,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.258556410639371,
        "unilateral_improvement_p0": 0.120721794680315,
        "unilateral_improvement_p1": 0.471167949292389,
        "nash_conv": 0.591889743972704,
        "exploitability": 0.295944871986352
      }
    },
    {
      "iteration": 4000,
      "average": {
        "on_policy_value_p0": -0.055474512565644,
        "best_response_value_p0": -0.051874742914757,
        "best_response_value_p1": 0.059322842917278,
        "unilateral_improvement_p0": 0.003599769650887,
        "unilateral_improvement_p1": 0.003848330351634,
        "nash_conv": 0.007448100002521,
        "exploitability": 0.00372405000126
      },
      "current": {
        "on_policy_value_p0": -0.449153034925089,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.449153034925089,
        "unilateral_improvement_p0": 0.615819701591755,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.615819701591755,
        "exploitability": 0.307909850795878
      }
    },
    {
      "iteration": 4100,
      "average": {
        "on_policy_value_p0": -0.055558843326326,
        "best_response_value_p0": -0.052815452388334,
        "best_response_value_p1": 0.059965850336707,
        "unilateral_improvement_p0": 0.002743390937992,
        "unilateral_improvement_p1": 0.004407007010381,
        "nash_conv": 0.007150397948373,
        "exploitability": 0.003575198974187
      },
      "current": {
        "on_policy_value_p0": 0.090741954026289,
        "best_response_value_p0": 0.151175407116606,
        "best_response_value_p1": 0.212466427152698,
        "unilateral_improvement_p0": 0.060433453090318,
        "unilateral_improvement_p1": 0.303208381178987,
        "nash_conv": 0.363641834269304,
        "exploitability": 0.181820917134652
      }
    },
    {
      "iteration": 4200,
      "average": {
        "on_policy_value_p0": -0.055512254219917,
        "best_response_value_p0": -0.05002048857822,
        "best_response_value_p1": 0.057211976549377,
        "unilateral_improvement_p0": 0.005491765641697,
        "unilateral_improvement_p1": 0.00169972232946,
        "nash_conv": 0.007191487971157,
        "exploitability": 0.003595743985579
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 4300,
      "average": {
        "on_policy_value_p0": -0.055485961932521,
        "best_response_value_p0": -0.054304498001005,
        "best_response_value_p1": 0.061046102132408,
        "unilateral_improvement_p0": 0.001181463931516,
        "unilateral_improvement_p1": 0.005560140199887,
        "nash_conv": 0.006741604131403,
        "exploitability": 0.003370802065702
      },
      "current": {
        "on_policy_value_p0": -0.172072284364471,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.172072284364471,
        "unilateral_improvement_p0": 0.505405617697805,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.505405617697805,
        "exploitability": 0.252702808848902
      }
    },
    {
      "iteration": 4400,
      "average": {
        "on_policy_value_p0": -0.055523426463211,
        "best_response_value_p0": -0.051020514460607,
        "best_response_value_p1": 0.057754109408086,
        "unilateral_improvement_p0": 0.004502912002604,
        "unilateral_improvement_p1": 0.002230682944875,
        "nash_conv": 0.006733594947479,
        "exploitability": 0.00336679747374
      },
      "current": {
        "on_policy_value_p0": 0.034788391398921,
        "best_response_value_p0": 0.117096741552258,
        "best_response_value_p1": 0.371297082340208,
        "unilateral_improvement_p0": 0.082308350153337,
        "unilateral_improvement_p1": 0.406085473739129,
        "nash_conv": 0.488393823892467,
        "exploitability": 0.244196911946233
      }
    },
    {
      "iteration": 4500,
      "average": {
        "on_policy_value_p0": -0.055497241617788,
        "best_response_value_p0": -0.053824933085066,
        "best_response_value_p1": 0.060792129192899,
        "unilateral_improvement_p0": 0.001672308532722,
        "unilateral_improvement_p1": 0.005294887575111,
        "nash_conv": 0.006967196107833,
        "exploitability": 0.003483598053917
      },
      "current": {
        "on_policy_value_p0": -0.187608894278185,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.187608894278185,
        "unilateral_improvement_p0": 0.354275560944852,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.354275560944852,
        "exploitability": 0.177137780472426
      }
    },
    {
      "iteration": 4600,
      "average": {
        "on_policy_value_p0": -0.055561021172362,
        "best_response_value_p0": -0.052127921329819,
        "best_response_value_p1": 0.058812486483161,
        "unilateral_improvement_p0": 0.003433099842543,
        "unilateral_improvement_p1": 0.0032514653108,
        "nash_conv": 0.006684565153343,
        "exploitability": 0.003342282576671
      },
      "current": {
        "on_policy_value_p0": 0.104799653764207,
        "best_response_value_p0": 0.104799653764207,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.438132987097541,
        "nash_conv": 0.438132987097541,
        "exploitability": 0.21906649354877
      }
    },
    {
      "iteration": 4700,
      "average": {
        "on_policy_value_p0": -0.055505572727182,
        "best_response_value_p0": -0.050262089636921,
        "best_response_value_p1": 0.05699624065075,
        "unilateral_improvement_p0": 0.005243483090261,
        "unilateral_improvement_p1": 0.001490667923569,
        "nash_conv": 0.006734151013829,
        "exploitability": 0.003367075506915
      },
      "current": {
        "on_policy_value_p0": -0.068951669218592,
        "best_response_value_p0": 0.132190832057371,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.201142501275962,
        "unilateral_improvement_p1": 0.264381664114742,
        "nash_conv": 0.465524165390704,
        "exploitability": 0.232762082695352
      }
    },
    {
      "iteration": 4800,
      "average": {
        "on_policy_value_p0": -0.05550369098406,
        "best_response_value_p0": -0.054478897495461,
        "best_response_value_p1": 0.061010759765233,
        "unilateral_improvement_p0": 0.001024793488599,
        "unilateral_improvement_p1": 0.005507068781173,
        "nash_conv": 0.006531862269772,
        "exploitability": 0.003265931134886
      },
      "current": {
        "on_policy_value_p0": -0.09825124142887,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.105498387008047,
        "unilateral_improvement_p0": 0.431584574762203,
        "unilateral_improvement_p1": 0.007247145579177,
        "nash_conv": 0.43883172034138,
        "exploitability": 0.21941586017069
      }
    },
    {
      "iteration": 4900,
      "average": {
        "on_policy_value_p0": -0.055528234254448,
        "best_response_value_p0": -0.050848729884163,
        "best_response_value_p1": 0.057404189215131,
        "unilateral_improvement_p0": 0.004679504370285,
        "unilateral_improvement_p1": 0.001875954960683,
        "nash_conv": 0.006555459330968,
        "exploitability": 0.003277729665484
      },
      "current": {
        "on_policy_value_p0": 0.046295718629647,
        "best_response_value_p0": 0.124543370432551,
        "best_response_value_p1": 0.42409825232283,
        "unilateral_improvement_p0": 0.078247651802904,
        "unilateral_improvement_p1": 0.470393970952477,
        "nash_conv": 0.548641622755381,
        "exploitability": 0.274320811377691
      }
    },
    {
      "iteration": 5000,
      "average": {
        "on_policy_value_p0": -0.055506400503875,
        "best_response_value_p0": -0.053871953449214,
        "best_response_value_p1": 0.060476849562444,
        "unilateral_improvement_p0": 0.001634447054661,
        "unilateral_improvement_p1": 0.004970449058569,
        "nash_conv": 0.00660489611323,
        "exploitability": 0.003302448056615
      },
      "current": {
        "on_policy_value_p0": -0.109647036774031,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.276313703440698,
        "unilateral_improvement_p1": 0.057019629892635,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 5100,
      "average": {
        "on_policy_value_p0": -0.055558289304577,
        "best_response_value_p0": -0.052408990536233,
        "best_response_value_p1": 0.058760329537259,
        "unilateral_improvement_p0": 0.003149298768344,
        "unilateral_improvement_p1": 0.003202040232682,
        "nash_conv": 0.006351339001026,
        "exploitability": 0.003175669500513
      },
      "current": {
        "on_policy_value_p0": 0.083802190020788,
        "best_response_value_p0": 0.108055938034307,
        "best_response_value_p1": 0.284825837306294,
        "unilateral_improvement_p0": 0.024253748013519,
        "unilateral_improvement_p1": 0.368628027327082,
        "nash_conv": 0.392881775340602,
        "exploitability": 0.196440887670301
      }
    },
    {
      "iteration": 5200,
      "average": {
        "on_policy_value_p0": -0.055514783965713,
        "best_response_value_p0": -0.050588926223588,
        "best_response_value_p1": 0.057035491023767,
        "unilateral_improvement_p0": 0.004925857742125,
        "unilateral_improvement_p1": 0.001520707058054,
        "nash_conv": 0.006446564800179,
        "exploitability": 0.00322328240009
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 5300,
      "average": {
        "on_policy_value_p0": -0.055495970589481,
        "best_response_value_p0": -0.054226659179071,
        "best_response_value_p1": 0.060416966631192,
        "unilateral_improvement_p0": 0.00126931141041,
        "unilateral_improvement_p1": 0.004920996041711,
        "nash_conv": 0.006190307452121,
        "exploitability": 0.00309515372606
      },
      "current": {
        "on_policy_value_p0": -0.202417472165139,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.202417472165139,
        "unilateral_improvement_p0": 0.535750805498472,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.535750805498472,
        "exploitability": 0.267875402749236
      }
    },
    {
      "iteration": 5400,
      "average": {
        "on_policy_value_p0": -0.055535607087689,
        "best_response_value_p0": -0.051942420535036,
        "best_response_value_p1": 0.057565349814842,
        "unilateral_improvement_p0": 0.003593186552653,
        "unilateral_improvement_p1": 0.002029742727153,
        "nash_conv": 0.005622929279806,
        "exploitability": 0.002811464639903
      },
      "current": {
        "on_policy_value_p0": 0.089205231057806,
        "best_response_value_p0": 0.172666747708852,
        "best_response_value_p1": 0.333076966697907,
        "unilateral_improvement_p0": 0.083461516651046,
        "unilateral_improvement_p1": 0.422282197755713,
        "nash_conv": 0.505743714406759,
        "exploitability": 0.25287185720338
      }
    },
    {
      "iteration": 5500,
      "average": {
        "on_policy_value_p0": -0.05549721647828,
        "best_response_value_p0": -0.052889545752658,
        "best_response_value_p1": 0.059148535255785,
        "unilateral_improvement_p0": 0.002607670725622,
        "unilateral_improvement_p1": 0.003651318777505,
        "nash_conv": 0.006258989503127,
        "exploitability": 0.003129494751563
      },
      "current": {
        "on_policy_value_p0": -0.40473996821105,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.40473996821105,
        "unilateral_improvement_p0": 0.571406634877716,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.571406634877716,
        "exploitability": 0.285703317438858
      }
    },
    {
      "iteration": 5600,
      "average": {
        "on_policy_value_p0": -0.055555000158079,
        "best_response_value_p0": -0.053632523853199,
        "best_response_value_p1": 0.059629032424268,
        "unilateral_improvement_p0": 0.00192247630488,
        "unilateral_improvement_p1": 0.00407403226619,
        "nash_conv": 0.00599650857107,
        "exploitability": 0.002998254285535
      },
      "current": {
        "on_policy_value_p0": 0.091646990640121,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.183293981280242,
        "unilateral_improvement_p0": 0.075019676026545,
        "unilateral_improvement_p1": 0.274940971920364,
        "nash_conv": 0.349960647946909,
        "exploitability": 0.174980323973455
      }
    },
    {
      "iteration": 5700,
      "average": {
        "on_policy_value_p0": -0.055539025862031,
        "best_response_value_p0": -0.051083974287282,
        "best_response_value_p1": 0.05724814034681,
        "unilateral_improvement_p0": 0.004455051574749,
        "unilateral_improvement_p1": 0.001709114484779,
        "nash_conv": 0.006164166059528,
        "exploitability": 0.003082083029764
      },
      "current": {
        "on_policy_value_p0": 0.01480642886629,
        "best_response_value_p0": 0.137053808934088,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.122247380067798,
        "unilateral_improvement_p1": 0.348139762199623,
        "nash_conv": 0.470387142267421,
        "exploitability": 0.23519357113371
      }
    },
    {
      "iteration": 5800,
      "average": {
        "on_policy_value_p0": -0.055490102092342,
        "best_response_value_p0": -0.052361910702349,
        "best_response_value_p1": 0.058463837890831,
        "unilateral_improvement_p0": 0.003128191389993,
        "unilateral_improvement_p1": 0.002973735798489,
        "nash_conv": 0.006101927188482,
        "exploitability": 0.003050963594241
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 5900,
      "average": {
        "on_policy_value_p0": -0.055537421578859,
        "best_response_value_p0": -0.053415982615025,
        "best_response_value_p1": 0.059495568107266,
        "unilateral_improvement_p0": 0.002121438963834,
        "unilateral_improvement_p1": 0.003958146528407,
        "nash_conv": 0.006079585492241,
        "exploitability": 0.00303979274612
      },
      "current": {
        "on_policy_value_p0": 0.14166757774815,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.191665755585184,
        "unilateral_improvement_p1": 0.308334244414816,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 6000,
      "average": {
        "on_policy_value_p0": -0.055516984328547,
        "best_response_value_p0": -0.051082588378641,
        "best_response_value_p1": 0.057112803185783,
        "unilateral_improvement_p0": 0.004434395949906,
        "unilateral_improvement_p1": 0.001595818857236,
        "nash_conv": 0.006030214807142,
        "exploitability": 0.003015107403571
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 6100,
      "average": {
        "on_policy_value_p0": -0.055526513390785,
        "best_response_value_p0": -0.054270698728161,
        "best_response_value_p1": 0.060161078651406,
        "unilateral_improvement_p0": 0.001255814662624,
        "unilateral_improvement_p1": 0.004634565260621,
        "nash_conv": 0.005890379923245,
        "exploitability": 0.002945189961623
      },
      "current": {
        "on_policy_value_p0": -0.062853778303635,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.229520444970301,
        "unilateral_improvement_p1": 0.103812888363032,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 6200,
      "average": {
        "on_policy_value_p0": -0.055557736018011,
        "best_response_value_p0": -0.052424037151169,
        "best_response_value_p1": 0.058054114393923,
        "unilateral_improvement_p0": 0.003133698866842,
        "unilateral_improvement_p1": 0.002496378375911,
        "nash_conv": 0.005630077242754,
        "exploitability": 0.002815038621377
      },
      "current": {
        "on_policy_value_p0": 0.093285997991411,
        "best_response_value_p0": 0.093285997991411,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.426619331324745,
        "nash_conv": 0.426619331324745,
        "exploitability": 0.213309665662372
      }
    },
    {
      "iteration": 6300,
      "average": {
        "on_policy_value_p0": -0.055522241058027,
        "best_response_value_p0": -0.051050667222656,
        "best_response_value_p1": 0.056898946164455,
        "unilateral_improvement_p0": 0.00447157383537,
        "unilateral_improvement_p1": 0.001376705106428,
        "nash_conv": 0.005848278941799,
        "exploitability": 0.002924139470899
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 6400,
      "average": {
        "on_policy_value_p0": -0.055502811266741,
        "best_response_value_p0": -0.053964975185231,
        "best_response_value_p1": 0.059742841574087,
        "unilateral_improvement_p0": 0.00153783608151,
        "unilateral_improvement_p1": 0.004240030307346,
        "nash_conv": 0.005777866388857,
        "exploitability": 0.002888933194428
      },
      "current": {
        "on_policy_value_p0": -0.284444658961008,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.284444658961008,
        "unilateral_improvement_p0": 0.617777992294342,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.617777992294342,
        "exploitability": 0.308888996147171
      }
    },
    {
      "iteration": 6500,
      "average": {
        "on_policy_value_p0": -0.055544486456877,
        "best_response_value_p0": -0.052624404118789,
        "best_response_value_p1": 0.058224472760007,
        "unilateral_improvement_p0": 0.002920082338088,
        "unilateral_improvement_p1": 0.00267998630313,
        "nash_conv": 0.005600068641219,
        "exploitability": 0.002800034320609
      },
      "current": {
        "on_policy_value_p0": 0.180695849697772,
        "best_response_value_p0": 0.281510814360894,
        "best_response_value_p1": 0.298370070673757,
        "unilateral_improvement_p0": 0.100814964663122,
        "unilateral_improvement_p1": 0.479065920371528,
        "nash_conv": 0.57988088503465,
        "exploitability": 0.289940442517325
      }
    },
    {
      "iteration": 6600,
      "average": {
        "on_policy_value_p0": -0.055509757358078,
        "best_response_value_p0": -0.051686392876928,
        "best_response_value_p1": 0.05729226345262,
        "unilateral_improvement_p0": 0.00382336448115,
        "unilateral_improvement_p1": 0.001782506094542,
        "nash_conv": 0.005605870575692,
        "exploitability": 0.002802935287846
      },
      "current": {
        "on_policy_value_p0": -0.326949801208962,
        "best_response_value_p0": 0.108983267069654,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.435933068278616,
        "unilateral_improvement_p1": 0.173050198791038,
        "nash_conv": 0.608983267069654,
        "exploitability": 0.304491633534827
      }
    },
    {
      "iteration": 6700,
      "average": {
        "on_policy_value_p0": -0.055539991793738,
        "best_response_value_p0": -0.054133297317923,
        "best_response_value_p1": 0.059835897018705,
        "unilateral_improvement_p0": 0.001406694475815,
        "unilateral_improvement_p1": 0.004295905224967,
        "nash_conv": 0.005702599700782,
        "exploitability": 0.002851299850391
      },
      "current": {
        "on_policy_value_p0": 0.024572841615125,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.142093825051541,
        "unilateral_improvement_p0": 0.142093825051541,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.308760491718208,
        "exploitability": 0.154380245859104
      }
    },
    {
      "iteration": 6800,
      "average": {
        "on_policy_value_p0": -0.055554942492524,
        "best_response_value_p0": -0.052190247627649,
        "best_response_value_p1": 0.057520227095064,
        "unilateral_improvement_p0": 0.003364694864875,
        "unilateral_improvement_p1": 0.001965284602541,
        "nash_conv": 0.005329979467416,
        "exploitability": 0.002664989733708
      },
      "current": {
        "on_policy_value_p0": 0.076884705223767,
        "best_response_value_p0": 0.076884705223767,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.410218038557101,
        "nash_conv": 0.410218038557101,
        "exploitability": 0.20510901927855
      }
    },
    {
      "iteration": 6900,
      "average": {
        "on_policy_value_p0": -0.055518036582404,
        "best_response_value_p0": -0.051206661281504,
        "best_response_value_p1": 0.056686600615425,
        "unilateral_improvement_p0": 0.0043113753009,
        "unilateral_improvement_p1": 0.001168564033022,
        "nash_conv": 0.005479939333922,
        "exploitability": 0.002739969666961
      },
      "current": {
        "on_policy_value_p0": -0.099809168859892,
        "best_response_value_p0": 0.116762082236721,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.216571251096613,
        "unilateral_improvement_p1": 0.233524164473442,
        "nash_conv": 0.450095415570054,
        "exploitability": 0.225047707785027
      }
    },
    {
      "iteration": 7000,
      "average": {
        "on_policy_value_p0": -0.055510463141939,
        "best_response_value_p0": -0.054406090963996,
        "best_response_value_p1": 0.059775264612937,
        "unilateral_improvement_p0": 0.001104372177943,
        "unilateral_improvement_p1": 0.004264801470998,
        "nash_conv": 0.005369173648941,
        "exploitability": 0.002684586824471
      },
      "current": {
        "on_policy_value_p0": -0.199824066646156,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.199824066646156,
        "unilateral_improvement_p0": 0.53315739997949,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.53315739997949,
        "exploitability": 0.266578699989745
      }
    },
    {
      "iteration": 7100,
      "average": {
        "on_policy_value_p0": -0.055542226777545,
        "best_response_value_p0": -0.05260225916641,
        "best_response_value_p1": 0.057924751234254,
        "unilateral_improvement_p0": 0.002939967611135,
        "unilateral_improvement_p1": 0.002382524456709,
        "nash_conv": 0.005322492067843,
        "exploitability": 0.002661246033922
      },
      "current": {
        "on_policy_value_p0": 0.133548781541644,
        "best_response_value_p0": 0.24024238668704,
        "best_response_value_p1": 0.286612789709209,
        "unilateral_improvement_p0": 0.106693605145396,
        "unilateral_improvement_p1": 0.420161571250853,
        "nash_conv": 0.526855176396249,
        "exploitability": 0.263427588198124
      }
    },
    {
      "iteration": 7200,
      "average": {
        "on_policy_value_p0": -0.055514792423917,
        "best_response_value_p0": -0.051758963180451,
        "best_response_value_p1": 0.056885960383852,
        "unilateral_improvement_p0": 0.003755829243466,
        "unilateral_improvement_p1": 0.001371167959935,
        "nash_conv": 0.005126997203401,
        "exploitability": 0.002563498601701
      },
      "current": {
        "on_policy_value_p0": -0.249529657513652,
        "best_response_value_p0": 0.083176552504551,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.332706210018202,
        "unilateral_improvement_p1": 0.250470342486348,
        "nash_conv": 0.583176552504551,
        "exploitability": 0.291588276252275
      }
    },
    {
      "iteration": 7300,
      "average": {
        "on_policy_value_p0": -0.055538359579893,
        "best_response_value_p0": -0.054352309912115,
        "best_response_value_p1": 0.059808053862325,
        "unilateral_improvement_p0": 0.001186049667778,
        "unilateral_improvement_p1": 0.004269694282431,
        "nash_conv": 0.005455743950209,
        "exploitability": 0.002727871975105
      },
      "current": {
        "on_policy_value_p0": -0.010409477301198,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.150852464114473,
        "unilateral_improvement_p0": 0.177076143967864,
        "unilateral_improvement_p1": 0.140442986813276,
        "nash_conv": 0.31751913078114,
        "exploitability": 0.15875956539057
      }
    },
    {
      "iteration": 7400,
      "average": {
        "on_policy_value_p0": -0.055558130091835,
        "best_response_value_p0": -0.052551029356312,
        "best_response_value_p1": 0.057658641818964,
        "unilateral_improvement_p0": 0.003007100735524,
        "unilateral_improvement_p1": 0.002100511727129,
        "nash_conv": 0.005107612462653,
        "exploitability": 0.002553806231326
      },
      "current": {
        "on_policy_value_p0": 0.089738450400128,
        "best_response_value_p0": 0.089738450400128,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.423071783733461,
        "nash_conv": 0.423071783733461,
        "exploitability": 0.21153589186673
      }
    },
    {
      "iteration": 7500,
      "average": {
        "on_policy_value_p0": -0.055528787688961,
        "best_response_value_p0": -0.051395389733299,
        "best_response_value_p1": 0.056750982577141,
        "unilateral_improvement_p0": 0.004133397955662,
        "unilateral_improvement_p1": 0.00122219488818,
        "nash_conv": 0.005355592843842,
        "exploitability": 0.002677796421921
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 7600,
      "average": {
        "on_policy_value_p0": -0.055508334840636,
        "best_response_value_p0": -0.053703449565503,
        "best_response_value_p1": 0.059015798697371,
        "unilateral_improvement_p0": 0.001804885275134,
        "unilateral_improvement_p1": 0.003507463856734,
        "nash_conv": 0.005312349131868,
        "exploitability": 0.002656174565934
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 7700,
      "average": {
        "on_policy_value_p0": -0.055544623100444,
        "best_response_value_p0": -0.053404459262667,
        "best_response_value_p1": 0.058702783343613,
        "unilateral_improvement_p0": 0.002140163837777,
        "unilateral_improvement_p1": 0.003158160243169,
        "nash_conv": 0.005298324080945,
        "exploitability": 0.002649162040473
      },
      "current": {
        "on_policy_value_p0": 0.195034016829705,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.223401366992743,
        "unilateral_improvement_p0": 0.138299316503629,
        "unilateral_improvement_p1": 0.418435383822447,
        "nash_conv": 0.556734700326076,
        "exploitability": 0.278367350163038
      }
    },
    {
      "iteration": 7800,
      "average": {
        "on_policy_value_p0": -0.055526767719642,
        "best_response_value_p0": -0.051572326625426,
        "best_response_value_p1": 0.056849710330092,
        "unilateral_improvement_p0": 0.003954441094216,
        "unilateral_improvement_p1": 0.00132294261045,
        "nash_conv": 0.005277383704666,
        "exploitability": 0.002638691852333
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 7900,
      "average": {
        "on_policy_value_p0": -0.055527714773096,
        "best_response_value_p0": -0.054385928739548,
        "best_response_value_p1": 0.059524553671189,
        "unilateral_improvement_p0": 0.001141786033547,
        "unilateral_improvement_p1": 0.003996838898093,
        "nash_conv": 0.00513862493164,
        "exploitability": 0.00256931246582
      },
      "current": {
        "on_policy_value_p0": -0.090662443342674,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.257329110009341,
        "unilateral_improvement_p1": 0.076004223323993,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 8000,
      "average": {
        "on_policy_value_p0": -0.055556680132837,
        "best_response_value_p0": -0.053400961916007,
        "best_response_value_p1": 0.058469054191091,
        "unilateral_improvement_p0": 0.00215571821683,
        "unilateral_improvement_p1": 0.002912374058254,
        "nash_conv": 0.005068092275084,
        "exploitability": 0.002534046137542
      },
      "current": {
        "on_policy_value_p0": 0.073922978363162,
        "best_response_value_p0": 0.126234191789775,
        "best_response_value_p1": 0.228710906480107,
        "unilateral_improvement_p0": 0.052311213426613,
        "unilateral_improvement_p1": 0.302633884843268,
        "nash_conv": 0.354945098269882,
        "exploitability": 0.177472549134941
      }
    },
    {
      "iteration": 8100,
      "average": {
        "on_policy_value_p0": -0.055542376287943,
        "best_response_value_p0": -0.051873501476339,
        "best_response_value_p1": 0.057037679669423,
        "unilateral_improvement_p0": 0.003668874811604,
        "unilateral_improvement_p1": 0.00149530338148,
        "nash_conv": 0.005164178193084,
        "exploitability": 0.002582089096542
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 8200,
      "average": {
        "on_policy_value_p0": -0.055511024640468,
        "best_response_value_p0": -0.052254829705756,
        "best_response_value_p1": 0.057358728999051,
        "unilateral_improvement_p0": 0.003256194934712,
        "unilateral_improvement_p1": 0.001847704358583,
        "nash_conv": 0.005103899293295,
        "exploitability": 0.002551949646648
      },
      "current": {
        "on_policy_value_p0": -0.283173665275366,
        "best_response_value_p0": 0.283173665275366,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.566347330550732,
        "unilateral_improvement_p1": 0.050159668057967,
        "nash_conv": 0.6165069986087,
        "exploitability": 0.30825349930435
      }
    },
    {
      "iteration": 8300,
      "average": {
        "on_policy_value_p0": -0.055531352858617,
        "best_response_value_p0": -0.054606671919104,
        "best_response_value_p1": 0.05961162069071,
        "unilateral_improvement_p0": 0.000924680939513,
        "unilateral_improvement_p1": 0.004080267832093,
        "nash_conv": 0.005004948771606,
        "exploitability": 0.002502474385803
      },
      "current": {
        "on_policy_value_p0": 0.061948772794291,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.271384560539042,
        "unilateral_improvement_p1": 0.228615439460958,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 8400,
      "average": {
        "on_policy_value_p0": -0.055542414306111,
        "best_response_value_p0": -0.052389651182592,
        "best_response_value_p1": 0.057061466625232,
        "unilateral_improvement_p0": 0.003152763123519,
        "unilateral_improvement_p1": 0.001519052319121,
        "nash_conv": 0.00467181544264,
        "exploitability": 0.00233590772132
      },
      "current": {
        "on_policy_value_p0": 0.077402677441956,
        "best_response_value_p0": 0.144044762919528,
        "best_response_value_p1": 0.366715829044857,
        "unilateral_improvement_p0": 0.066642085477572,
        "unilateral_improvement_p1": 0.444118506486813,
        "nash_conv": 0.510760591964384,
        "exploitability": 0.255380295982192
      }
    },
    {
      "iteration": 8500,
      "average": {
        "on_policy_value_p0": -0.055517722562796,
        "best_response_value_p0": -0.052742340583678,
        "best_response_value_p1": 0.057817528063451,
        "unilateral_improvement_p0": 0.002775381979119,
        "unilateral_improvement_p1": 0.002299805500655,
        "nash_conv": 0.005075187479773,
        "exploitability": 0.002537593739887
      },
      "current": {
        "on_policy_value_p0": -0.5,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 8600,
      "average": {
        "on_policy_value_p0": -0.055544392352781,
        "best_response_value_p0": -0.054337763939193,
        "best_response_value_p1": 0.059344451471187,
        "unilateral_improvement_p0": 0.001206628413589,
        "unilateral_improvement_p1": 0.003800059118406,
        "nash_conv": 0.005006687531995,
        "exploitability": 0.002503343765997
      },
      "current": {
        "on_policy_value_p0": 0.033805803727521,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.132860862939146,
        "unilateral_improvement_p0": 0.132860862939146,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.299527529605813,
        "exploitability": 0.149763764802906
      }
    },
    {
      "iteration": 8700,
      "average": {
        "on_policy_value_p0": -0.055557159818919,
        "best_response_value_p0": -0.052754015885313,
        "best_response_value_p1": 0.05744419621696,
        "unilateral_improvement_p0": 0.002803143933606,
        "unilateral_improvement_p1": 0.001887036398041,
        "nash_conv": 0.004690180331647,
        "exploitability": 0.002345090165824
      },
      "current": {
        "on_policy_value_p0": 0.088304324837332,
        "best_response_value_p0": 0.088304324837332,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.421637658170665,
        "nash_conv": 0.421637658170665,
        "exploitability": 0.210818829085332
      }
    },
    {
      "iteration": 8800,
      "average": {
        "on_policy_value_p0": -0.05553419476896,
        "best_response_value_p0": -0.051749188684535,
        "best_response_value_p1": 0.056691486743791,
        "unilateral_improvement_p0": 0.003785006084425,
        "unilateral_improvement_p1": 0.001157291974832,
        "nash_conv": 0.004942298059257,
        "exploitability": 0.002471149029628
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 8900,
      "average": {
        "on_policy_value_p0": -0.055512920333856,
        "best_response_value_p0": -0.053386928773911,
        "best_response_value_p1": 0.058297018553106,
        "unilateral_improvement_p0": 0.002125991559945,
        "unilateral_improvement_p1": 0.00278409821925,
        "nash_conv": 0.004910089779196,
        "exploitability": 0.002455044889598
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 9000,
      "average": {
        "on_policy_value_p0": -0.055540977713196,
        "best_response_value_p0": -0.054097585568053,
        "best_response_value_p1": 0.058952981481495,
        "unilateral_improvement_p0": 0.001443392145143,
        "unilateral_improvement_p1": 0.003412003768299,
        "nash_conv": 0.004855395913442,
        "exploitability": 0.002427697956721
      },
      "current": {
        "on_policy_value_p0": 0.113201554688885,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.220131778644448,
        "unilateral_improvement_p1": 0.279868221355552,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 9100,
      "average": {
        "on_policy_value_p0": -0.055537901300093,
        "best_response_value_p0": -0.052195930551928,
        "best_response_value_p1": 0.0570400965814,
        "unilateral_improvement_p0": 0.003341970748165,
        "unilateral_improvement_p1": 0.001502195281308,
        "nash_conv": 0.004844166029473,
        "exploitability": 0.002422083014736
      },
      "current": {
        "on_policy_value_p0": -0.021962850494489,
        "best_response_value_p0": 0.158097857351298,
        "best_response_value_p1": 0.421799061749547,
        "unilateral_improvement_p0": 0.180060707845787,
        "unilateral_improvement_p1": 0.399836211255058,
        "nash_conv": 0.579896919100845,
        "exploitability": 0.289948459550423
      }
    },
    {
      "iteration": 9200,
      "average": {
        "on_policy_value_p0": -0.055521042733183,
        "best_response_value_p0": -0.053708222426003,
        "best_response_value_p1": 0.058461707573863,
        "unilateral_improvement_p0": 0.00181282030718,
        "unilateral_improvement_p1": 0.00294066484068,
        "nash_conv": 0.00475348514786,
        "exploitability": 0.00237674257393
      },
      "current": {
        "on_policy_value_p0": -0.376675125891051,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.376675125891051,
        "unilateral_improvement_p0": 0.543341792557718,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.543341792557718,
        "exploitability": 0.271670896278859
      }
    },
    {
      "iteration": 9300,
      "average": {
        "on_policy_value_p0": -0.055548768064898,
        "best_response_value_p0": -0.054313694169323,
        "best_response_value_p1": 0.059050422749107,
        "unilateral_improvement_p0": 0.001235073895575,
        "unilateral_improvement_p1": 0.003501654684209,
        "nash_conv": 0.004736728579784,
        "exploitability": 0.002368364289892
      },
      "current": {
        "on_policy_value_p0": 0.056985274288083,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.113970548576167,
        "unilateral_improvement_p0": 0.109681392378583,
        "unilateral_improvement_p1": 0.17095582286425,
        "nash_conv": 0.280637215242834,
        "exploitability": 0.140318607621417
      }
    },
    {
      "iteration": 9400,
      "average": {
        "on_policy_value_p0": -0.055555789700341,
        "best_response_value_p0": -0.052732884680958,
        "best_response_value_p1": 0.057251004444736,
        "unilateral_improvement_p0": 0.002822905019383,
        "unilateral_improvement_p1": 0.001695214744394,
        "nash_conv": 0.004518119763777,
        "exploitability": 0.002259059881889
      },
      "current": {
        "on_policy_value_p0": 0.077189398885431,
        "best_response_value_p0": 0.077189398885431,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.410522732218764,
        "nash_conv": 0.410522732218764,
        "exploitability": 0.205261366109382
      }
    },
    {
      "iteration": 9500,
      "average": {
        "on_policy_value_p0": -0.055534709645769,
        "best_response_value_p0": -0.051894990975308,
        "best_response_value_p1": 0.056648362292685,
        "unilateral_improvement_p0": 0.003639718670461,
        "unilateral_improvement_p1": 0.001113652646916,
        "nash_conv": 0.004753371317377,
        "exploitability": 0.002376685658689
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 9600,
      "average": {
        "on_policy_value_p0": -0.055516603481788,
        "best_response_value_p0": -0.053500740572315,
        "best_response_value_p1": 0.058226142974074,
        "unilateral_improvement_p0": 0.002015862909473,
        "unilateral_improvement_p1": 0.002709539492286,
        "nash_conv": 0.004725402401759,
        "exploitability": 0.00236270120088
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 9700,
      "average": {
        "on_policy_value_p0": -0.055542713442875,
        "best_response_value_p0": -0.054266850601502,
        "best_response_value_p1": 0.058933898908309,
        "unilateral_improvement_p0": 0.001275862841373,
        "unilateral_improvement_p1": 0.003391185465434,
        "nash_conv": 0.004667048306808,
        "exploitability": 0.002333524153404
      },
      "current": {
        "on_policy_value_p0": 0.110625815933741,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.222707517399593,
        "unilateral_improvement_p1": 0.277292482600407,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 9800,
      "average": {
        "on_policy_value_p0": -0.055541447599646,
        "best_response_value_p0": -0.052239808552227,
        "best_response_value_p1": 0.056873413980637,
        "unilateral_improvement_p0": 0.003301639047419,
        "unilateral_improvement_p1": 0.001331966380991,
        "nash_conv": 0.004633605428409,
        "exploitability": 0.002316802714205
      },
      "current": {
        "on_policy_value_p0": 0.035926313599776,
        "best_response_value_p0": 0.131107406060692,
        "best_response_value_p1": 0.429615584775653,
        "unilateral_improvement_p0": 0.095181092460916,
        "unilateral_improvement_p1": 0.465541898375429,
        "nash_conv": 0.560722990836345,
        "exploitability": 0.280361495418173
      }
    },
    {
      "iteration": 9900,
      "average": {
        "on_policy_value_p0": -0.055523392599697,
        "best_response_value_p0": -0.053566862510249,
        "best_response_value_p1": 0.058176923482382,
        "unilateral_improvement_p0": 0.001956530089448,
        "unilateral_improvement_p1": 0.002653530882684,
        "nash_conv": 0.004610060972132,
        "exploitability": 0.002305030486066
      },
      "current": {
        "on_policy_value_p0": -0.389758292466357,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.389758292466357,
        "unilateral_improvement_p0": 0.556424959133024,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.556424959133024,
        "exploitability": 0.278212479566512
      }
    },
    {
      "iteration": 10000,
      "average": {
        "on_policy_value_p0": -0.055546395829272,
        "best_response_value_p0": -0.054345107016834,
        "best_response_value_p1": 0.058980679646773,
        "unilateral_improvement_p0": 0.001201288812438,
        "unilateral_improvement_p1": 0.003434283817501,
        "nash_conv": 0.00463557262994,
        "exploitability": 0.00231778631497
      },
      "current": {
        "on_policy_value_p0": 0.034333022839931,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.132333643826735,
        "unilateral_improvement_p0": 0.132333643826735,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.299000310493402,
        "exploitability": 0.149500155246701
      }
    },
    {
      "iteration": 10100,
      "average": {
        "on_policy_value_p0": -0.055556367248843,
        "best_response_value_p0": -0.053033687579827,
        "best_response_value_p1": 0.057390095269596,
        "unilateral_improvement_p0": 0.002522679669016,
        "unilateral_improvement_p1": 0.001833728020753,
        "nash_conv": 0.004356407689769,
        "exploitability": 0.002178203844885
      },
      "current": {
        "on_policy_value_p0": 0.089193538201601,
        "best_response_value_p0": 0.089193538201601,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.422526871534934,
        "nash_conv": 0.422526871534934,
        "exploitability": 0.211263435767467
      }
    },
    {
      "iteration": 10200,
      "average": {
        "on_policy_value_p0": -0.055539732717784,
        "best_response_value_p0": -0.05212765895681,
        "best_response_value_p1": 0.056717767051431,
        "unilateral_improvement_p0": 0.003412073760974,
        "unilateral_improvement_p1": 0.001178034333647,
        "nash_conv": 0.004590108094621,
        "exploitability": 0.00229505404731
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 10300,
      "average": {
        "on_policy_value_p0": -0.055518592524919,
        "best_response_value_p0": -0.052974627550693,
        "best_response_value_p1": 0.057540310945763,
        "unilateral_improvement_p0": 0.002543964974226,
        "unilateral_improvement_p1": 0.002021718420844,
        "nash_conv": 0.00456568339507,
        "exploitability": 0.002282841697535
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 10400,
      "average": {
        "on_policy_value_p0": -0.055536455383698,
        "best_response_value_p0": -0.054672456329575,
        "best_response_value_p1": 0.059130151493488,
        "unilateral_improvement_p0": 0.000863999054123,
        "unilateral_improvement_p1": 0.00359369610979,
        "nash_conv": 0.004457695163913,
        "exploitability": 0.002228847581957
      },
      "current": {
        "on_policy_value_p0": 0.064417629001353,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.26891570433198,
        "unilateral_improvement_p1": 0.23108429566802,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 10500,
      "average": {
        "on_policy_value_p0": -0.055545744844035,
        "best_response_value_p0": -0.052978905206207,
        "best_response_value_p1": 0.056997408339295,
        "unilateral_improvement_p0": 0.002566839637827,
        "unilateral_improvement_p1": 0.001451663495261,
        "nash_conv": 0.004018503133088,
        "exploitability": 0.002009251566544
      },
      "current": {
        "on_policy_value_p0": 0.093577155952037,
        "best_response_value_p0": 0.175290221569588,
        "best_response_value_p1": 0.336573868764898,
        "unilateral_improvement_p0": 0.081713065617551,
        "unilateral_improvement_p1": 0.430151024716935,
        "nash_conv": 0.511864090334486,
        "exploitability": 0.255932045167243
      }
    },
    {
      "iteration": 10600,
      "average": {
        "on_policy_value_p0": -0.055528421417742,
        "best_response_value_p0": -0.052388480535222,
        "best_response_value_p1": 0.05661645161602,
        "unilateral_improvement_p0": 0.00313994088252,
        "unilateral_improvement_p1": 0.001088030198278,
        "nash_conv": 0.004227971080798,
        "exploitability": 0.002113985540399
      },
      "current": {
        "on_policy_value_p0": -0.243397612161395,
        "best_response_value_p0": 0.081132537387132,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.324530149548526,
        "unilateral_improvement_p1": 0.256602387838605,
        "nash_conv": 0.581132537387132,
        "exploitability": 0.290566268693566
      }
    },
    {
      "iteration": 10700,
      "average": {
        "on_policy_value_p0": -0.055538137181019,
        "best_response_value_p0": -0.054670749407881,
        "best_response_value_p1": 0.059059536094086,
        "unilateral_improvement_p0": 0.000867387773138,
        "unilateral_improvement_p1": 0.003521398913067,
        "nash_conv": 0.004388786686205,
        "exploitability": 0.002194393343103
      },
      "current": {
        "on_policy_value_p0": -0.072729559398212,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.239396226064879,
        "unilateral_improvement_p1": 0.093937107268455,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 10800,
      "average": {
        "on_policy_value_p0": -0.055557483954379,
        "best_response_value_p0": -0.053756078458951,
        "best_response_value_p1": 0.058139423122087,
        "unilateral_improvement_p0": 0.001801405495428,
        "unilateral_improvement_p1": 0.002581939167707,
        "nash_conv": 0.004383344663136,
        "exploitability": 0.002191672331568
      },
      "current": {
        "on_policy_value_p0": 0.085216985046879,
        "best_response_value_p0": 0.135168511017243,
        "best_response_value_p1": 0.233430281392606,
        "unilateral_improvement_p0": 0.049951525970364,
        "unilateral_improvement_p1": 0.318647266439485,
        "nash_conv": 0.368598792409849,
        "exploitability": 0.184299396204924
      }
    },
    {
      "iteration": 10900,
      "average": {
        "on_policy_value_p0": -0.055549922097305,
        "best_response_value_p0": -0.052473254067773,
        "best_response_value_p1": 0.05687567456576,
        "unilateral_improvement_p0": 0.003076668029532,
        "unilateral_improvement_p1": 0.001325752468455,
        "nash_conv": 0.004402420497987,
        "exploitability": 0.002201210248994
      },
      "current": {
        "on_policy_value_p0": 0.037771193491434,
        "best_response_value_p0": 0.091124279683799,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.053353086192365,
        "unilateral_improvement_p1": 0.371104526824767,
        "nash_conv": 0.424457613017132,
        "exploitability": 0.212228806508566
      }
    },
    {
      "iteration": 11000,
      "average": {
        "on_policy_value_p0": -0.055528879881865,
        "best_response_value_p0": -0.052212111921884,
        "best_response_value_p1": 0.056358622978798,
        "unilateral_improvement_p0": 0.003316767959981,
        "unilateral_improvement_p1": 0.000829743096933,
        "nash_conv": 0.004146511056914,
        "exploitability": 0.002073255528457
      },
      "current": {
        "on_policy_value_p0": -0.156195353240144,
        "best_response_value_p0": 0.156195353240144,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.312390706480287,
        "unilateral_improvement_p1": 0.17713798009319,
        "nash_conv": 0.489528686573477,
        "exploitability": 0.244764343286738
      }
    },
    {
      "iteration": 11100,
      "average": {
        "on_policy_value_p0": -0.055525160007722,
        "best_response_value_p0": -0.054372755861891,
        "best_response_value_p1": 0.058738858673915,
        "unilateral_improvement_p0": 0.001152404145831,
        "unilateral_improvement_p1": 0.003213698666192,
        "nash_conv": 0.004366102812024,
        "exploitability": 0.002183051406012
      },
      "current": {
        "on_policy_value_p0": -0.267495661730551,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.267495661730551,
        "unilateral_improvement_p0": 0.600828995063884,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.600828995063884,
        "exploitability": 0.300414497531942
      }
    },
    {
      "iteration": 11200,
      "average": {
        "on_policy_value_p0": -0.055546871917186,
        "best_response_value_p0": -0.053904957101915,
        "best_response_value_p1": 0.058301981396732,
        "unilateral_improvement_p0": 0.001641914815271,
        "unilateral_improvement_p1": 0.002755109479546,
        "nash_conv": 0.004397024294816,
        "exploitability": 0.002198512147408
      },
      "current": {
        "on_policy_value_p0": 0.169842370435189,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.174827139618515,
        "unilateral_improvement_p0": 0.163490962898144,
        "unilateral_improvement_p1": 0.344669510053704,
        "nash_conv": 0.508160472951848,
        "exploitability": 0.254080236475924
      }
    },
    {
      "iteration": 11300,
      "average": {
        "on_policy_value_p0": -0.055540613710556,
        "best_response_value_p0": -0.052445415510748,
        "best_response_value_p1": 0.056815912461599,
        "unilateral_improvement_p0": 0.003095198199808,
        "unilateral_improvement_p1": 0.001275298751043,
        "nash_conv": 0.004370496950851,
        "exploitability": 0.002185248475426
      },
      "current": {
        "on_policy_value_p0": -0.019668070946689,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.460663858106623,
        "unilateral_improvement_p0": 0.186334737613355,
        "unilateral_improvement_p1": 0.440995787159935,
        "nash_conv": 0.62733052477329,
        "exploitability": 0.313665262386645
      }
    },
    {
      "iteration": 11400,
      "average": {
        "on_policy_value_p0": -0.055527644283968,
        "best_response_value_p0": -0.053833990077145,
        "best_response_value_p1": 0.05810161985298,
        "unilateral_improvement_p0": 0.001693654206823,
        "unilateral_improvement_p1": 0.002573975569013,
        "nash_conv": 0.004267629775836,
        "exploitability": 0.002133814887918
      },
      "current": {
        "on_policy_value_p0": -0.37910182824908,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.37910182824908,
        "unilateral_improvement_p0": 0.545768494915746,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.545768494915746,
        "exploitability": 0.272884247457873
      }
    },
    {
      "iteration": 11500,
      "average": {
        "on_policy_value_p0": -0.055547065641841,
        "best_response_value_p0": -0.054473417804972,
        "best_response_value_p1": 0.058793862685523,
        "unilateral_improvement_p0": 0.001073647836869,
        "unilateral_improvement_p1": 0.003246797043682,
        "nash_conv": 0.004320444880551,
        "exploitability": 0.002160222440276
      },
      "current": {
        "on_policy_value_p0": 0.033043979701266,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.133622686965401,
        "unilateral_improvement_p0": 0.133622686965401,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.300289353632067,
        "exploitability": 0.150144676816034
      }
    },
    {
      "iteration": 11600,
      "average": {
        "on_policy_value_p0": -0.055556975953274,
        "best_response_value_p0": -0.053338125406506,
        "best_response_value_p1": 0.057459675826749,
        "unilateral_improvement_p0": 0.002218850546768,
        "unilateral_improvement_p1": 0.001902699873475,
        "nash_conv": 0.004121550420243,
        "exploitability": 0.002060775210121
      },
      "current": {
        "on_policy_value_p0": 0.097010620298949,
        "best_response_value_p0": 0.097010620298949,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.430343953632283,
        "nash_conv": 0.430343953632283,
        "exploitability": 0.215171976816141
      }
    },
    {
      "iteration": 11700,
      "average": {
        "on_policy_value_p0": -0.055545104150571,
        "best_response_value_p0": -0.052445886121117,
        "best_response_value_p1": 0.056731643645538,
        "unilateral_improvement_p0": 0.003099218029454,
        "unilateral_improvement_p1": 0.001186539494968,
        "nash_conv": 0.004285757524421,
        "exploitability": 0.002142878762211
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 11800,
      "average": {
        "on_policy_value_p0": -0.055526142474856,
        "best_response_value_p0": -0.052639746831437,
        "best_response_value_p1": 0.056824908359813,
        "unilateral_improvement_p0": 0.002886395643419,
        "unilateral_improvement_p1": 0.001298765884957,
        "nash_conv": 0.004185161528376,
        "exploitability": 0.002092580764188
      },
      "current": {
        "on_policy_value_p0": -0.236889217048331,
        "best_response_value_p0": 0.236889217048331,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.473778434096662,
        "unilateral_improvement_p1": 0.096444116285002,
        "nash_conv": 0.570222550381664,
        "exploitability": 0.285111275190832
      }
    },
    {
      "iteration": 11900,
      "average": {
        "on_policy_value_p0": -0.055531715560267,
        "best_response_value_p0": -0.054917316878421,
        "best_response_value_p1": 0.058930579767617,
        "unilateral_improvement_p0": 0.000614398681847,
        "unilateral_improvement_p1": 0.00339886420735,
        "nash_conv": 0.004013262889196,
        "exploitability": 0.002006631444598
      },
      "current": {
        "on_policy_value_p0": -0.167957436132937,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.167957436132937,
        "unilateral_improvement_p0": 0.50129076946627,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.50129076946627,
        "exploitability": 0.250645384733135
      }
    },
    {
      "iteration": 12000,
      "average": {
        "on_policy_value_p0": -0.055549270522168,
        "best_response_value_p0": -0.0536546959303,
        "best_response_value_p1": 0.057841637076868,
        "unilateral_improvement_p0": 0.001894574591868,
        "unilateral_improvement_p1": 0.0022923665547,
        "nash_conv": 0.004186941146568,
        "exploitability": 0.002093470573284
      },
      "current": {
        "on_policy_value_p0": 0.214947702969056,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.263228739271445,
        "unilateral_improvement_p0": 0.118385630364277,
        "unilateral_improvement_p1": 0.478176442240501,
        "nash_conv": 0.596562072604778,
        "exploitability": 0.298281036302389
      }
    },
    {
      "iteration": 12100,
      "average": {
        "on_policy_value_p0": -0.055539520990061,
        "best_response_value_p0": -0.052429562082715,
        "best_response_value_p1": 0.056656077739332,
        "unilateral_improvement_p0": 0.003109958907346,
        "unilateral_improvement_p1": 0.00111655674927,
        "nash_conv": 0.004226515656617,
        "exploitability": 0.002113257828308
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 12200,
      "average": {
        "on_policy_value_p0": -0.05553123444072,
        "best_response_value_p0": -0.054302021657457,
        "best_response_value_p1": 0.058353230157796,
        "unilateral_improvement_p0": 0.001229212783263,
        "unilateral_improvement_p1": 0.002821995717076,
        "nash_conv": 0.004051208500339,
        "exploitability": 0.002025604250169
      },
      "current": {
        "on_policy_value_p0": -0.318170617036237,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.318170617036237,
        "unilateral_improvement_p0": 0.484837283702904,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.484837283702904,
        "exploitability": 0.242418641851452
      }
    },
    {
      "iteration": 12300,
      "average": {
        "on_policy_value_p0": -0.055549633180812,
        "best_response_value_p0": -0.054455220131955,
        "best_response_value_p1": 0.058596948449556,
        "unilateral_improvement_p0": 0.001094413048856,
        "unilateral_improvement_p1": 0.003047315268744,
        "nash_conv": 0.004141728317601,
        "exploitability": 0.0020708641588
      },
      "current": {
        "on_policy_value_p0": 0.047698482460432,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.118968184206234,
        "unilateral_improvement_p0": 0.118968184206234,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.285634850872901,
        "exploitability": 0.142817425436451
      }
    },
    {
      "iteration": 12400,
      "average": {
        "on_policy_value_p0": -0.055556717608718,
        "best_response_value_p0": -0.053313953373495,
        "best_response_value_p1": 0.057268029561514,
        "unilateral_improvement_p0": 0.002242764235223,
        "unilateral_improvement_p1": 0.001711311952796,
        "nash_conv": 0.003954076188019,
        "exploitability": 0.001977038094009
      },
      "current": {
        "on_policy_value_p0": 0.092877770566252,
        "best_response_value_p0": 0.092877770566252,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.426211103899586,
        "nash_conv": 0.426211103899586,
        "exploitability": 0.213105551949793
      }
    },
    {
      "iteration": 12500,
      "average": {
        "on_policy_value_p0": -0.055544959106704,
        "best_response_value_p0": -0.0525041414558,
        "best_response_value_p1": 0.056648686642677,
        "unilateral_improvement_p0": 0.003040817650904,
        "unilateral_improvement_p1": 0.001103727535973,
        "nash_conv": 0.004144545186878,
        "exploitability": 0.002072272593439
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 12600,
      "average": {
        "on_policy_value_p0": -0.055527446729594,
        "best_response_value_p0": -0.052752446329997,
        "best_response_value_p1": 0.056819648621109,
        "unilateral_improvement_p0": 0.002775000399597,
        "unilateral_improvement_p1": 0.001292201891515,
        "nash_conv": 0.004067202291112,
        "exploitability": 0.002033601145556
      },
      "current": {
        "on_policy_value_p0": -0.246816000443076,
        "best_response_value_p0": 0.246816000443076,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.493632000886153,
        "unilateral_improvement_p1": 0.086517332890257,
        "nash_conv": 0.58014933377641,
        "exploitability": 0.290074666888205
      }
    },
    {
      "iteration": 12700,
      "average": {
        "on_policy_value_p0": -0.055532497862984,
        "best_response_value_p0": -0.054898241090177,
        "best_response_value_p1": 0.058780080674279,
        "unilateral_improvement_p0": 0.000634256772807,
        "unilateral_improvement_p1": 0.003247582811296,
        "nash_conv": 0.003881839584103,
        "exploitability": 0.001940919792051
      },
      "current": {
        "on_policy_value_p0": -0.166188245069335,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166188245069335,
        "unilateral_improvement_p0": 0.499521578402668,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.499521578402668,
        "exploitability": 0.249760789201334
      }
    },
    {
      "iteration": 12800,
      "average": {
        "on_policy_value_p0": -0.055548351211074,
        "best_response_value_p0": -0.053766101188847,
        "best_response_value_p1": 0.05785959476368,
        "unilateral_improvement_p0": 0.001782250022227,
        "unilateral_improvement_p1": 0.002311243552606,
        "nash_conv": 0.004093493574833,
        "exploitability": 0.002046746787417
      },
      "current": {
        "on_policy_value_p0": 0.194798655104895,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.222930643543122,
        "unilateral_improvement_p0": 0.138534678228439,
        "unilateral_improvement_p1": 0.417729298648017,
        "nash_conv": 0.556263976876456,
        "exploitability": 0.278131988438228
      }
    },
    {
      "iteration": 12900,
      "average": {
        "on_policy_value_p0": -0.05554149861735,
        "best_response_value_p0": -0.052680441595081,
        "best_response_value_p1": 0.056771449394105,
        "unilateral_improvement_p0": 0.002861057022268,
        "unilateral_improvement_p1": 0.001229950776755,
        "nash_conv": 0.004091007799024,
        "exploitability": 0.002045503899512
      },
      "current": {
        "on_policy_value_p0": -0.011058678639539,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.477882642720921,
        "unilateral_improvement_p0": 0.177725345306206,
        "unilateral_improvement_p1": 0.466823964081382,
        "nash_conv": 0.644549309387588,
        "exploitability": 0.322274654693794
      }
    },
    {
      "iteration": 13000,
      "average": {
        "on_policy_value_p0": -0.055530959693559,
        "best_response_value_p0": -0.053860532108101,
        "best_response_value_p1": 0.057888615430247,
        "unilateral_improvement_p0": 0.001670427585458,
        "unilateral_improvement_p1": 0.002357655736688,
        "nash_conv": 0.004028083322146,
        "exploitability": 0.002014041661073
      },
      "current": {
        "on_policy_value_p0": -0.39934535698058,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.39934535698058,
        "unilateral_improvement_p0": 0.566012023647246,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.566012023647246,
        "exploitability": 0.283006011823623
      }
    },
    {
      "iteration": 13100,
      "average": {
        "on_policy_value_p0": -0.055547184671107,
        "best_response_value_p0": -0.054619965854537,
        "best_response_value_p1": 0.058673801923128,
        "unilateral_improvement_p0": 0.000927218816571,
        "unilateral_improvement_p1": 0.00312661725202,
        "nash_conv": 0.004053836068591,
        "exploitability": 0.002026918034296
      },
      "current": {
        "on_policy_value_p0": 0.02417550211953,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.14022404766582,
        "unilateral_improvement_p0": 0.142491164547137,
        "unilateral_improvement_p1": 0.16439954978535,
        "nash_conv": 0.306890714332487,
        "exploitability": 0.153445357166243
      }
    },
    {
      "iteration": 13200,
      "average": {
        "on_policy_value_p0": -0.055557635846715,
        "best_response_value_p0": -0.053637975787149,
        "best_response_value_p1": 0.05757464626298,
        "unilateral_improvement_p0": 0.001919660059566,
        "unilateral_improvement_p1": 0.002017010416265,
        "nash_conv": 0.00393667047583,
        "exploitability": 0.001968335237915
      },
      "current": {
        "on_policy_value_p0": 0.095590074471387,
        "best_response_value_p0": 0.110905741710648,
        "best_response_value_p1": 0.302701998854813,
        "unilateral_improvement_p0": 0.01531566723926,
        "unilateral_improvement_p1": 0.3982920733262,
        "nash_conv": 0.41360774056546,
        "exploitability": 0.20680387028273
      }
    },
    {
      "iteration": 13300,
      "average": {
        "on_policy_value_p0": -0.05554941871628,
        "best_response_value_p0": -0.052705631532519,
        "best_response_value_p1": 0.056714435793516,
        "unilateral_improvement_p0": 0.002843787183762,
        "unilateral_improvement_p1": 0.001165017077236,
        "nash_conv": 0.004008804260998,
        "exploitability": 0.002004402130499
      },
      "current": {
        "on_policy_value_p0": 0.021652958812118,
        "best_response_value_p0": 0.12336074904243,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.101707790230312,
        "unilateral_improvement_p1": 0.354986292145452,
        "nash_conv": 0.456694082375763,
        "exploitability": 0.228347041187882
      }
    },
    {
      "iteration": 13400,
      "average": {
        "on_policy_value_p0": -0.055533686887978,
        "best_response_value_p0": -0.052538455968604,
        "best_response_value_p1": 0.056291193735355,
        "unilateral_improvement_p0": 0.002995230919374,
        "unilateral_improvement_p1": 0.000757506847377,
        "nash_conv": 0.003752737766752,
        "exploitability": 0.001876368883376
      },
      "current": {
        "on_policy_value_p0": -0.156265909609259,
        "best_response_value_p0": 0.156265909609259,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.312531819218517,
        "unilateral_improvement_p1": 0.177067423724075,
        "nash_conv": 0.489599242942592,
        "exploitability": 0.244799621471296
      }
    },
    {
      "iteration": 13500,
      "average": {
        "on_policy_value_p0": -0.055529434848883,
        "best_response_value_p0": -0.054281638714679,
        "best_response_value_p1": 0.05825781764969,
        "unilateral_improvement_p0": 0.001247796134204,
        "unilateral_improvement_p1": 0.002728382800807,
        "nash_conv": 0.003976178935011,
        "exploitability": 0.001988089467506
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 13600,
      "average": {
        "on_policy_value_p0": -0.055546677880967,
        "best_response_value_p0": -0.054407010034841,
        "best_response_value_p1": 0.058351068116472,
        "unilateral_improvement_p0": 0.001139667846127,
        "unilateral_improvement_p1": 0.002804390235505,
        "nash_conv": 0.003944058081631,
        "exploitability": 0.001972029040816
      },
      "current": {
        "on_policy_value_p0": 0.115921695544159,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.217411637789174,
        "unilateral_improvement_p1": 0.282588362210826,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 13700,
      "average": {
        "on_policy_value_p0": -0.055547043552106,
        "best_response_value_p0": -0.052951597276223,
        "best_response_value_p1": 0.056696718660936,
        "unilateral_improvement_p0": 0.002595446275883,
        "unilateral_improvement_p1": 0.001149675108831,
        "nash_conv": 0.003745121384714,
        "exploitability": 0.001872560692357
      },
      "current": {
        "on_policy_value_p0": 0.067799506653651,
        "best_response_value_p0": 0.125010395995381,
        "best_response_value_p1": 0.385578221316541,
        "unilateral_improvement_p0": 0.05721088934173,
        "unilateral_improvement_p1": 0.453377727970192,
        "nash_conv": 0.510588617311922,
        "exploitability": 0.255294308655961
      }
    },
    {
      "iteration": 13800,
      "average": {
        "on_policy_value_p0": -0.055533657526158,
        "best_response_value_p0": -0.052893964474054,
        "best_response_value_p1": 0.056730968304965,
        "unilateral_improvement_p0": 0.002639693052104,
        "unilateral_improvement_p1": 0.001197310778807,
        "nash_conv": 0.003837003830911,
        "exploitability": 0.001918501915456
      },
      "current": {
        "on_policy_value_p0": -0.318919512628124,
        "best_response_value_p0": 0.106306504209374,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.425226016837498,
        "unilateral_improvement_p1": 0.181080487371876,
        "nash_conv": 0.606306504209374,
        "exploitability": 0.303153252104687
      }
    },
    {
      "iteration": 13900,
      "average": {
        "on_policy_value_p0": -0.055540839132092,
        "best_response_value_p0": -0.054746300920998,
        "best_response_value_p1": 0.058558561981877,
        "unilateral_improvement_p0": 0.000794538211094,
        "unilateral_improvement_p1": 0.003017722849784,
        "nash_conv": 0.003812261060878,
        "exploitability": 0.001906130530439
      },
      "current": {
        "on_policy_value_p0": -0.081358989557666,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.248025656224333,
        "unilateral_improvement_p1": 0.085307677109,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 14000,
      "average": {
        "on_policy_value_p0": -0.055554696727764,
        "best_response_value_p0": -0.054301298137249,
        "best_response_value_p1": 0.058042613338115,
        "unilateral_improvement_p0": 0.001253398590515,
        "unilateral_improvement_p1": 0.002487916610351,
        "nash_conv": 0.003741315200866,
        "exploitability": 0.001870657600433
      },
      "current": {
        "on_policy_value_p0": 0.08297024106252,
        "best_response_value_p0": 0.161782349878176,
        "best_response_value_p1": 0.175709115702021,
        "unilateral_improvement_p0": 0.078812108815656,
        "unilateral_improvement_p1": 0.258679356764541,
        "nash_conv": 0.337491465580197,
        "exploitability": 0.168745732790099
      }
    },
    {
      "iteration": 14100,
      "average": {
        "on_policy_value_p0": -0.055554278117842,
        "best_response_value_p0": -0.053153994028796,
        "best_response_value_p1": 0.05690348965813,
        "unilateral_improvement_p0": 0.002400284089046,
        "unilateral_improvement_p1": 0.001349211540288,
        "nash_conv": 0.003749495629334,
        "exploitability": 0.001874747814667
      },
      "current": {
        "on_policy_value_p0": 0.066415426571271,
        "best_response_value_p0": 0.066415426571271,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.399748759904604,
        "nash_conv": 0.399748759904604,
        "exploitability": 0.199874379952302
      }
    },
    {
      "iteration": 14200,
      "average": {
        "on_policy_value_p0": -0.055542796563601,
        "best_response_value_p0": -0.052620191983135,
        "best_response_value_p1": 0.056502760857721,
        "unilateral_improvement_p0": 0.002922604580466,
        "unilateral_improvement_p1": 0.00095996429412,
        "nash_conv": 0.003882568874586,
        "exploitability": 0.001941284437293
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 14300,
      "average": {
        "on_policy_value_p0": -0.055529382214287,
        "best_response_value_p0": -0.053233323060002,
        "best_response_value_p1": 0.05710331474387,
        "unilateral_improvement_p0": 0.002296059154285,
        "unilateral_improvement_p1": 0.001573932529583,
        "nash_conv": 0.003869991683868,
        "exploitability": 0.001934995841934
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 14400,
      "average": {
        "on_policy_value_p0": -0.055537547754559,
        "best_response_value_p0": -0.05492849898491,
        "best_response_value_p1": 0.058677971135094,
        "unilateral_improvement_p0": 0.000609048769649,
        "unilateral_improvement_p1": 0.003140423380535,
        "nash_conv": 0.003749472150184,
        "exploitability": 0.001874736075092
      },
      "current": {
        "on_policy_value_p0": -0.109276044173488,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.109276044173488,
        "unilateral_improvement_p0": 0.442609377506821,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.442609377506821,
        "exploitability": 0.221304688753411
      }
    },
    {
      "iteration": 14500,
      "average": {
        "on_policy_value_p0": -0.055550013191834,
        "best_response_value_p0": -0.053812215347424,
        "best_response_value_p1": 0.057624188871812,
        "unilateral_improvement_p0": 0.00173779784441,
        "unilateral_improvement_p1": 0.002074175679979,
        "nash_conv": 0.003811973524388,
        "exploitability": 0.001905986762194
      },
      "current": {
        "on_policy_value_p0": 0.212001423012671,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.257336179358676,
        "unilateral_improvement_p0": 0.121331910320662,
        "unilateral_improvement_p1": 0.469337602371347,
        "nash_conv": 0.590669512692009,
        "exploitability": 0.295334756346005
      }
    },
    {
      "iteration": 14600,
      "average": {
        "on_policy_value_p0": -0.055543393950716,
        "best_response_value_p0": -0.052790982180352,
        "best_response_value_p1": 0.056635478455153,
        "unilateral_improvement_p0": 0.002752411770364,
        "unilateral_improvement_p1": 0.001092084504437,
        "nash_conv": 0.003844496274801,
        "exploitability": 0.001922248137401
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 14700,
      "average": {
        "on_policy_value_p0": -0.055533826093084,
        "best_response_value_p0": -0.053865179329767,
        "best_response_value_p1": 0.057663858877675,
        "unilateral_improvement_p0": 0.001668646763318,
        "unilateral_improvement_p1": 0.002130032784591,
        "nash_conv": 0.003798679547909,
        "exploitability": 0.001899339773954
      },
      "current": {
        "on_policy_value_p0": -0.409968138011344,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.409968138011344,
        "unilateral_improvement_p0": 0.576634804678011,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.576634804678011,
        "exploitability": 0.288317402339005
      }
    },
    {
      "iteration": 14800,
      "average": {
        "on_policy_value_p0": -0.055546619449278,
        "best_response_value_p0": -0.054691759494639,
        "best_response_value_p1": 0.058502437234181,
        "unilateral_improvement_p0": 0.00085485995464,
        "unilateral_improvement_p1": 0.002955817784902,
        "nash_conv": 0.003810677739542,
        "exploitability": 0.001905338869771
      },
      "current": {
        "on_policy_value_p0": -0.030411905539945,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.197078572206612,
        "unilateral_improvement_p1": 0.136254761126721,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 14900,
      "average": {
        "on_policy_value_p0": -0.055556949519698,
        "best_response_value_p0": -0.053948785452598,
        "best_response_value_p1": 0.057667429896229,
        "unilateral_improvement_p0": 0.0016081640671,
        "unilateral_improvement_p1": 0.002110480376531,
        "nash_conv": 0.003718644443631,
        "exploitability": 0.001859322221815
      },
      "current": {
        "on_policy_value_p0": 0.081955357561851,
        "best_response_value_p0": 0.126932596539199,
        "best_response_value_p1": 0.243378855378637,
        "unilateral_improvement_p0": 0.044977238977348,
        "unilateral_improvement_p1": 0.325334212940488,
        "nash_conv": 0.370311451917836,
        "exploitability": 0.185155725958918
      }
    },
    {
      "iteration": 15000,
      "average": {
        "on_policy_value_p0": -0.055552438773513,
        "best_response_value_p0": -0.053009306695156,
        "best_response_value_p1": 0.056727269696169,
        "unilateral_improvement_p0": 0.002543132078357,
        "unilateral_improvement_p1": 0.001174830922656,
        "nash_conv": 0.003717963001013,
        "exploitability": 0.001858981500506
      },
      "current": {
        "on_policy_value_p0": 0.047805046485906,
        "best_response_value_p0": 0.071056573694854,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.023251527208948,
        "unilateral_improvement_p1": 0.38113837981924,
        "nash_conv": 0.404389907028188,
        "exploitability": 0.202194953514094
      }
    },
    {
      "iteration": 15100,
      "average": {
        "on_policy_value_p0": -0.055540311535221,
        "best_response_value_p0": -0.052605985162864,
        "best_response_value_p1": 0.056351592413413,
        "unilateral_improvement_p0": 0.002934326372357,
        "unilateral_improvement_p1": 0.000811280878192,
        "nash_conv": 0.003745607250549,
        "exploitability": 0.001872803625275
      },
      "current": {
        "on_policy_value_p0": -0.055123133277629,
        "best_response_value_p0": 0.139105100027852,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.194228233305481,
        "unilateral_improvement_p1": 0.278210200055705,
        "nash_conv": 0.472438433361186,
        "exploitability": 0.236219216680593
      }
    },
    {
      "iteration": 15200,
      "average": {
        "on_policy_value_p0": -0.055530430305517,
        "best_response_value_p0": -0.053705799715495,
        "best_response_value_p1": 0.057455875252716,
        "unilateral_improvement_p0": 0.001824630590022,
        "unilateral_improvement_p1": 0.001925444947199,
        "nash_conv": 0.003750075537221,
        "exploitability": 0.00187503776861
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 15300,
      "average": {
        "on_policy_value_p0": -0.055541634089869,
        "best_response_value_p0": -0.054850693279191,
        "best_response_value_p1": 0.058546193985548,
        "unilateral_improvement_p0": 0.000690940810677,
        "unilateral_improvement_p1": 0.00300455989568,
        "nash_conv": 0.003695500706357,
        "exploitability": 0.001847750353179
      },
      "current": {
        "on_policy_value_p0": 0.045200459986738,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.162483570817348,
        "unilateral_improvement_p0": 0.288132873346595,
        "unilateral_improvement_p1": 0.207684030804087,
        "nash_conv": 0.495816904150682,
        "exploitability": 0.247908452075341
      }
    },
    {
      "iteration": 15400,
      "average": {
        "on_policy_value_p0": -0.055550050317617,
        "best_response_value_p0": -0.053638685559389,
        "best_response_value_p1": 0.057278421208839,
        "unilateral_improvement_p0": 0.001911364758229,
        "unilateral_improvement_p1": 0.001728370891222,
        "nash_conv": 0.00363973564945,
        "exploitability": 0.001819867824725
      },
      "current": {
        "on_policy_value_p0": 0.161840757956272,
        "best_response_value_p0": 0.271151853546901,
        "best_response_value_p1": 0.281377808818741,
        "unilateral_improvement_p0": 0.10931109559063,
        "unilateral_improvement_p1": 0.443218566775012,
        "nash_conv": 0.552529662365642,
        "exploitability": 0.276264831182821
      }
    },
    {
      "iteration": 15500,
      "average": {
        "on_policy_value_p0": -0.055542654345772,
        "best_response_value_p0": -0.052870857484781,
        "best_response_value_p1": 0.056599311651373,
        "unilateral_improvement_p0": 0.002671796860992,
        "unilateral_improvement_p1": 0.001056657305601,
        "nash_conv": 0.003728454166593,
        "exploitability": 0.001864227083296
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 15600,
      "average": {
        "on_policy_value_p0": -0.055535241097779,
        "best_response_value_p0": -0.054117390618989,
        "best_response_value_p1": 0.057757185487316,
        "unilateral_improvement_p0": 0.00141785047879,
        "unilateral_improvement_p1": 0.002221944389537,
        "nash_conv": 0.003639794868326,
        "exploitability": 0.001819897434163
      },
      "current": {
        "on_policy_value_p0": -0.379843934298634,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.379843934298634,
        "unilateral_improvement_p0": 0.546510600965301,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.546510600965301,
        "exploitability": 0.273255300482651
      }
    },
    {
      "iteration": 15700,
      "average": {
        "on_policy_value_p0": -0.055547575072096,
        "best_response_value_p0": -0.054707665408578,
        "best_response_value_p1": 0.058411261388337,
        "unilateral_improvement_p0": 0.000839909663517,
        "unilateral_improvement_p1": 0.002863686316241,
        "nash_conv": 0.003703595979759,
        "exploitability": 0.001851797989879
      },
      "current": {
        "on_policy_value_p0": -0.009418428101458,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.155791713429124,
        "unilateral_improvement_p0": 0.176085094768124,
        "unilateral_improvement_p1": 0.146373285327666,
        "nash_conv": 0.322458380095791,
        "exploitability": 0.161229190047895
      }
    },
    {
      "iteration": 15800,
      "average": {
        "on_policy_value_p0": -0.05555680175119,
        "best_response_value_p0": -0.053985773213965,
        "best_response_value_p1": 0.057591449765853,
        "unilateral_improvement_p0": 0.001571028537225,
        "unilateral_improvement_p1": 0.002034648014663,
        "nash_conv": 0.003605676551888,
        "exploitability": 0.001802838275944
      },
      "current": {
        "on_policy_value_p0": 0.080818442415928,
        "best_response_value_p0": 0.123483656447542,
        "best_response_value_p1": 0.248002905270106,
        "unilateral_improvement_p0": 0.042665214031614,
        "unilateral_improvement_p1": 0.328821347686034,
        "nash_conv": 0.371486561717647,
        "exploitability": 0.185743280858824
      }
    },
    {
      "iteration": 15900,
      "average": {
        "on_policy_value_p0": -0.055552786692384,
        "best_response_value_p0": -0.05310921342395,
        "best_response_value_p1": 0.056720939147344,
        "unilateral_improvement_p0": 0.002443573268433,
        "unilateral_improvement_p1": 0.00116815245496,
        "nash_conv": 0.003611725723394,
        "exploitability": 0.001805862861697
      },
      "current": {
        "on_policy_value_p0": 0.047337260960357,
        "best_response_value_p0": 0.071992144745953,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.024654883785596,
        "unilateral_improvement_p1": 0.38067059429369,
        "nash_conv": 0.405325478079286,
        "exploitability": 0.202662739039643
      }
    },
    {
      "iteration": 16000,
      "average": {
        "on_policy_value_p0": -0.055542098653138,
        "best_response_value_p0": -0.052713818878483,
        "best_response_value_p1": 0.056366433277674,
        "unilateral_improvement_p0": 0.002828279774654,
        "unilateral_improvement_p1": 0.000824334624536,
        "nash_conv": 0.003652614399191,
        "exploitability": 0.001826307199595
      },
      "current": {
        "on_policy_value_p0": -0.017499001284854,
        "best_response_value_p0": 0.15791716602424,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.175416167309094,
        "unilateral_improvement_p1": 0.315834332048479,
        "nash_conv": 0.491250499357573,
        "exploitability": 0.245625249678787
      }
    },
    {
      "iteration": 16100,
      "average": {
        "on_policy_value_p0": -0.055532107928094,
        "best_response_value_p0": -0.053607230686422,
        "best_response_value_p1": 0.057250897437884,
        "unilateral_improvement_p0": 0.001924877241672,
        "unilateral_improvement_p1": 0.001718789509789,
        "nash_conv": 0.003643666751461,
        "exploitability": 0.001821833375731
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 16200,
      "average": {
        "on_policy_value_p0": -0.055540981015269,
        "best_response_value_p0": -0.054955403660228,
        "best_response_value_p1": 0.058523598545224,
        "unilateral_improvement_p0": 0.000585577355041,
        "unilateral_improvement_p1": 0.002982617529956,
        "nash_conv": 0.003568194884996,
        "exploitability": 0.001784097442498
      },
      "current": {
        "on_policy_value_p0": -0.060042645634701,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.119126372359033,
        "unilateral_improvement_p0": 0.393375978968034,
        "unilateral_improvement_p1": 0.059083726724333,
        "nash_conv": 0.452459705692367,
        "exploitability": 0.226229852846183
      }
    },
    {
      "iteration": 16300,
      "average": {
        "on_policy_value_p0": -0.055551063919276,
        "best_response_value_p0": -0.053901142880413,
        "best_response_value_p1": 0.057479214956145,
        "unilateral_improvement_p0": 0.001649921038864,
        "unilateral_improvement_p1": 0.001928151036868,
        "nash_conv": 0.003578072075732,
        "exploitability": 0.001789036037866
      },
      "current": {
        "on_policy_value_p0": 0.21764748696485,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.268628307263033,
        "unilateral_improvement_p0": 0.115685846368483,
        "unilateral_improvement_p1": 0.486275794227883,
        "nash_conv": 0.601961640596367,
        "exploitability": 0.300980820298183
      }
    },
    {
      "iteration": 16400,
      "average": {
        "on_policy_value_p0": -0.055545220232231,
        "best_response_value_p0": -0.052912287382564,
        "best_response_value_p1": 0.056537431969867,
        "unilateral_improvement_p0": 0.002632932849667,
        "unilateral_improvement_p1": 0.000992211737636,
        "nash_conv": 0.003625144587303,
        "exploitability": 0.001812572293651
      },
      "current": {
        "on_policy_value_p0": -0.000051579152515,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.499896841694971,
        "unilateral_improvement_p0": 0.166718245819181,
        "unilateral_improvement_p1": 0.499845262542456,
        "nash_conv": 0.666563508361638,
        "exploitability": 0.333281754180819
      }
    },
    {
      "iteration": 16500,
      "average": {
        "on_policy_value_p0": -0.055536074020142,
        "best_response_value_p0": -0.05375641088895,
        "best_response_value_p1": 0.057365164435839,
        "unilateral_improvement_p0": 0.001779663131192,
        "unilateral_improvement_p1": 0.001829090415698,
        "nash_conv": 0.00360875354689,
        "exploitability": 0.001804376773445
      },
      "current": {
        "on_policy_value_p0": -0.440828788936198,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.440828788936198,
        "unilateral_improvement_p0": 0.607495455602864,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.607495455602864,
        "exploitability": 0.303747727801432
      }
    },
    {
      "iteration": 16600,
      "average": {
        "on_policy_value_p0": -0.055545598093948,
        "best_response_value_p0": -0.054755420008886,
        "best_response_value_p1": 0.058314510122744,
        "unilateral_improvement_p0": 0.000790178085062,
        "unilateral_improvement_p1": 0.002768912028796,
        "nash_conv": 0.003559090113858,
        "exploitability": 0.001779545056929
      },
      "current": {
        "on_policy_value_p0": -0.055527383856011,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.222194050522677,
        "unilateral_improvement_p1": 0.111139282810656,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 16700,
      "average": {
        "on_policy_value_p0": -0.055555475713576,
        "best_response_value_p0": -0.054289481345283,
        "best_response_value_p1": 0.057746072406826,
        "unilateral_improvement_p0": 0.001265994368294,
        "unilateral_improvement_p1": 0.00219059669325,
        "nash_conv": 0.003456591061543,
        "exploitability": 0.001728295530772
      },
      "current": {
        "on_policy_value_p0": 0.077806845576161,
        "best_response_value_p0": 0.148209573689285,
        "best_response_value_p1": 0.192527877107084,
        "unilateral_improvement_p0": 0.070402728113125,
        "unilateral_improvement_p1": 0.270334722683245,
        "nash_conv": 0.340737450796369,
        "exploitability": 0.170368725398185
      }
    },
    {
      "iteration": 16800,
      "average": {
        "on_policy_value_p0": -0.055554498091971,
        "best_response_value_p0": -0.053350460321978,
        "best_response_value_p1": 0.056779280481123,
        "unilateral_improvement_p0": 0.002204037769992,
        "unilateral_improvement_p1": 0.001224782389152,
        "nash_conv": 0.003428820159145,
        "exploitability": 0.001714410079572
      },
      "current": {
        "on_policy_value_p0": 0.066832596037821,
        "best_response_value_p0": 0.066832596037821,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.400165929371155,
        "nash_conv": 0.400165929371155,
        "exploitability": 0.200082964685577
      }
    },
    {
      "iteration": 16900,
      "average": {
        "on_policy_value_p0": -0.055545701981819,
        "best_response_value_p0": -0.052886210906084,
        "best_response_value_p1": 0.056443308407271,
        "unilateral_improvement_p0": 0.002659491075735,
        "unilateral_improvement_p1": 0.000897606425452,
        "nash_conv": 0.003557097501187,
        "exploitability": 0.001778548750593
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 17000,
      "average": {
        "on_policy_value_p0": -0.055534530449593,
        "best_response_value_p0": -0.053182288525406,
        "best_response_value_p1": 0.056689667161395,
        "unilateral_improvement_p0": 0.002352241924187,
        "unilateral_improvement_p1": 0.001155136711802,
        "nash_conv": 0.003507378635989,
        "exploitability": 0.001753689317995
      },
      "current": {
        "on_policy_value_p0": -0.255070625035375,
        "best_response_value_p0": 0.255070625035375,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.51014125007075,
        "unilateral_improvement_p1": 0.078262708297958,
        "nash_conv": 0.588403958368708,
        "exploitability": 0.294201979184354
      }
    },
    {
      "iteration": 17100,
      "average": {
        "on_policy_value_p0": -0.055536821733515,
        "best_response_value_p0": -0.054779850974526,
        "best_response_value_p1": 0.058234662129303,
        "unilateral_improvement_p0": 0.000756970758989,
        "unilateral_improvement_p1": 0.002697840395788,
        "nash_conv": 0.003454811154777,
        "exploitability": 0.001727405577388
      },
      "current": {
        "on_policy_value_p0": -0.217620775279051,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.217620775279051,
        "unilateral_improvement_p0": 0.550954108612384,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.550954108612384,
        "exploitability": 0.275477054306192
      }
    },
    {
      "iteration": 17200,
      "average": {
        "on_policy_value_p0": -0.05554887206293,
        "best_response_value_p0": -0.054405107656891,
        "best_response_value_p1": 0.057927849718784,
        "unilateral_improvement_p0": 0.001143764406039,
        "unilateral_improvement_p1": 0.002378977655854,
        "nash_conv": 0.003522742061893,
        "exploitability": 0.001761371030947
      },
      "current": {
        "on_policy_value_p0": 0.130842229173543,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.20249110415979,
        "unilateral_improvement_p1": 0.297508895840209,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 17300,
      "average": {
        "on_policy_value_p0": -0.055548646917902,
        "best_response_value_p0": -0.053328011358737,
        "best_response_value_p1": 0.056622537403001,
        "unilateral_improvement_p0": 0.002220635559165,
        "unilateral_improvement_p1": 0.001073890485099,
        "nash_conv": 0.003294526044264,
        "exploitability": 0.001647263022132
      },
      "current": {
        "on_policy_value_p0": 0.063498065702524,
        "best_response_value_p0": 0.127815677786577,
        "best_response_value_p1": 0.371364775831895,
        "unilateral_improvement_p0": 0.064317612084053,
        "unilateral_improvement_p1": 0.434862841534419,
        "nash_conv": 0.499180453618471,
        "exploitability": 0.249590226809236
      }
    },
    {
      "iteration": 17400,
      "average": {
        "on_policy_value_p0": -0.055540016386524,
        "best_response_value_p0": -0.053001759243794,
        "best_response_value_p1": 0.056399117661699,
        "unilateral_improvement_p0": 0.00253825714273,
        "unilateral_improvement_p1": 0.000859101275174,
        "nash_conv": 0.003397358417905,
        "exploitability": 0.001698679208952
      },
      "current": {
        "on_policy_value_p0": -0.187648386971794,
        "best_response_value_p0": 0.062549462323931,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.250197849295725,
        "unilateral_improvement_p1": 0.312351613028206,
        "nash_conv": 0.562549462323931,
        "exploitability": 0.281274731161966
      }
    },
    {
      "iteration": 17500,
      "average": {
        "on_policy_value_p0": -0.055540597782019,
        "best_response_value_p0": -0.054688366810656,
        "best_response_value_p1": 0.058163123135401,
        "unilateral_improvement_p0": 0.000852230971363,
        "unilateral_improvement_p1": 0.002622525353382,
        "nash_conv": 0.003474756324745,
        "exploitability": 0.001737378162372
      },
      "current": {
        "on_policy_value_p0": -0.201498237606252,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.201498237606252,
        "unilateral_improvement_p0": 0.368164904272918,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.368164904272918,
        "exploitability": 0.184082452136459
      }
    },
    {
      "iteration": 17600,
      "average": {
        "on_policy_value_p0": -0.055551966517644,
        "best_response_value_p0": -0.054666648773547,
        "best_response_value_p1": 0.058098723818009,
        "unilateral_improvement_p0": 0.000885317744097,
        "unilateral_improvement_p1": 0.002546757300365,
        "nash_conv": 0.003432075044462,
        "exploitability": 0.001716037522231
      },
      "current": {
        "on_policy_value_p0": 0.055431139093704,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.111235527572962,
        "unilateral_improvement_p0": 0.111235527572962,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.277902194239629,
        "exploitability": 0.138951097119814
      }
    },
    {
      "iteration": 17700,
      "average": {
        "on_policy_value_p0": -0.055557043001472,
        "best_response_value_p0": -0.053821391216095,
        "best_response_value_p1": 0.057189247950971,
        "unilateral_improvement_p0": 0.001735651785378,
        "unilateral_improvement_p1": 0.001632204949498,
        "nash_conv": 0.003367856734876,
        "exploitability": 0.001683928367438
      },
      "current": {
        "on_policy_value_p0": 0.104642279585512,
        "best_response_value_p0": 0.104642279585512,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.437975612918845,
        "nash_conv": 0.437975612918845,
        "exploitability": 0.218987806459423
      }
    },
    {
      "iteration": 17800,
      "average": {
        "on_policy_value_p0": -0.055551530964683,
        "best_response_value_p0": -0.053123416645276,
        "best_response_value_p1": 0.056572628509157,
        "unilateral_improvement_p0": 0.002428114319407,
        "unilateral_improvement_p1": 0.001021097544474,
        "nash_conv": 0.003449211863881,
        "exploitability": 0.001724605931941
      },
      "current": {
        "on_policy_value_p0": 0.03043752980043,
        "best_response_value_p0": 0.105791607065806,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.075354077265375,
        "unilateral_improvement_p1": 0.363770863133764,
        "nash_conv": 0.439124940399139,
        "exploitability": 0.21956247019957
      }
    },
    {
      "iteration": 17900,
      "average": {
        "on_policy_value_p0": -0.055541704218431,
        "best_response_value_p0": -0.052858200202372,
        "best_response_value_p1": 0.056256580305196,
        "unilateral_improvement_p0": 0.002683504016059,
        "unilateral_improvement_p1": 0.000714876086765,
        "nash_conv": 0.003398380102825,
        "exploitability": 0.001699190051412
      },
      "current": {
        "on_policy_value_p0": -0.092134072647532,
        "best_response_value_p0": 0.120599630342901,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.212733702990433,
        "unilateral_improvement_p1": 0.241199260685801,
        "nash_conv": 0.453932963676234,
        "exploitability": 0.226966481838117
      }
    },
    {
      "iteration": 18000,
      "average": {
        "on_policy_value_p0": -0.055534450924783,
        "best_response_value_p0": -0.053866725067225,
        "best_response_value_p1": 0.057310217371871,
        "unilateral_improvement_p0": 0.001667725857558,
        "unilateral_improvement_p1": 0.001775766447088,
        "nash_conv": 0.003443492304646,
        "exploitability": 0.001721746152323
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 18100,
      "average": {
        "on_policy_value_p0": -0.055542901085678,
        "best_response_value_p0": -0.054949217258741,
        "best_response_value_p1": 0.058342817962911,
        "unilateral_improvement_p0": 0.000593683826938,
        "unilateral_improvement_p1": 0.002799916877233,
        "nash_conv": 0.003393600704171,
        "exploitability": 0.001696800352085
      },
      "current": {
        "on_policy_value_p0": -0.027444625319261,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.132490580298508,
        "unilateral_improvement_p0": 0.360777958652595,
        "unilateral_improvement_p1": 0.105045954979247,
        "nash_conv": 0.465823913631841,
        "exploitability": 0.232911956815921
      }
    },
    {
      "iteration": 18200,
      "average": {
        "on_policy_value_p0": -0.055551095294284,
        "best_response_value_p0": -0.05399248837686,
        "best_response_value_p1": 0.057395075681459,
        "unilateral_improvement_p0": 0.001558606917424,
        "unilateral_improvement_p1": 0.001843980387176,
        "nash_conv": 0.0034025873046,
        "exploitability": 0.0017012936523
      },
      "current": {
        "on_policy_value_p0": 0.209597256622512,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.252527846578357,
        "unilateral_improvement_p0": 0.123736076710822,
        "unilateral_improvement_p1": 0.462125103200868,
        "nash_conv": 0.58586117991169,
        "exploitability": 0.292930589955845
      }
    },
    {
      "iteration": 18300,
      "average": {
        "on_policy_value_p0": -0.055546718119135,
        "best_response_value_p0": -0.053126748677168,
        "best_response_value_p1": 0.056546882446757,
        "unilateral_improvement_p0": 0.002419969441967,
        "unilateral_improvement_p1": 0.001000164327623,
        "nash_conv": 0.003420133769589,
        "exploitability": 0.001710066884795
      },
      "current": {
        "on_policy_value_p0": -0.024490758555312,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.451018482889376,
        "unilateral_improvement_p0": 0.191157425221979,
        "unilateral_improvement_p1": 0.426527724334063,
        "nash_conv": 0.617685149556042,
        "exploitability": 0.308842574778021
      }
    },
    {
      "iteration": 18400,
      "average": {
        "on_policy_value_p0": -0.055538304364728,
        "best_response_value_p0": -0.053479609253597,
        "best_response_value_p1": 0.056886142636219,
        "unilateral_improvement_p0": 0.002058695111131,
        "unilateral_improvement_p1": 0.001347838271491,
        "nash_conv": 0.003406533382622,
        "exploitability": 0.001703266691311
      },
      "current": {
        "on_policy_value_p0": -0.410502858151057,
        "best_response_value_p0": 0.136834286050352,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.547337144201409,
        "unilateral_improvement_p1": 0.089497141848943,
        "nash_conv": 0.636834286050352,
        "exploitability": 0.318417143025176
      }
    },
    {
      "iteration": 18500,
      "average": {
        "on_policy_value_p0": -0.055544337243695,
        "best_response_value_p0": -0.054863903609115,
        "best_response_value_p1": 0.058167881272832,
        "unilateral_improvement_p0": 0.000680433634581,
        "unilateral_improvement_p1": 0.002623544029136,
        "nash_conv": 0.003303977663717,
        "exploitability": 0.001651988831859
      },
      "current": {
        "on_policy_value_p0": -0.083633203705036,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.250299870371703,
        "unilateral_improvement_p1": 0.08303346296163,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 18600,
      "average": {
        "on_policy_value_p0": -0.055553956851692,
        "best_response_value_p0": -0.054584545302737,
        "best_response_value_p1": 0.057841427206681,
        "unilateral_improvement_p0": 0.000969411548956,
        "unilateral_improvement_p1": 0.002287470354989,
        "nash_conv": 0.003256881903945,
        "exploitability": 0.001628440951972
      },
      "current": {
        "on_policy_value_p0": 0.075540933490946,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.151081866981893,
        "unilateral_improvement_p0": 0.09112573317572,
        "unilateral_improvement_p1": 0.226622800472839,
        "nash_conv": 0.31774853364856,
        "exploitability": 0.15887426682428
      }
    },
    {
      "iteration": 18700,
      "average": {
        "on_policy_value_p0": -0.055556099762485,
        "best_response_value_p0": -0.053696782286559,
        "best_response_value_p1": 0.056870480615574,
        "unilateral_improvement_p0": 0.001859317475926,
        "unilateral_improvement_p1": 0.001314380853088,
        "nash_conv": 0.003173698329015,
        "exploitability": 0.001586849164507
      },
      "current": {
        "on_policy_value_p0": 0.087156518819416,
        "best_response_value_p0": 0.087156518819416,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.42048985215275,
        "nash_conv": 0.42048985215275,
        "exploitability": 0.210244926076375
      }
    },
    {
      "iteration": 18800,
      "average": {
        "on_policy_value_p0": -0.055550009061526,
        "best_response_value_p0": -0.053140625191234,
        "best_response_value_p1": 0.056514776005984,
        "unilateral_improvement_p0": 0.002409383870292,
        "unilateral_improvement_p1": 0.000964766944458,
        "nash_conv": 0.00337415081475,
        "exploitability": 0.001687075407375
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 18900,
      "average": {
        "on_policy_value_p0": -0.055540759155258,
        "best_response_value_p0": -0.052994725990635,
        "best_response_value_p1": 0.056215756027117,
        "unilateral_improvement_p0": 0.002546033164623,
        "unilateral_improvement_p1": 0.000674996871859,
        "nash_conv": 0.003221030036483,
        "exploitability": 0.001610515018241
      },
      "current": {
        "on_policy_value_p0": -0.13372308985243,
        "best_response_value_p0": 0.13372308985243,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.26744617970486,
        "unilateral_improvement_p1": 0.199610243480903,
        "nash_conv": 0.467056423185764,
        "exploitability": 0.233528211592882
      }
    },
    {
      "iteration": 19000,
      "average": {
        "on_policy_value_p0": -0.055535775110904,
        "best_response_value_p0": -0.054099794975011,
        "best_response_value_p1": 0.057449330480621,
        "unilateral_improvement_p0": 0.001435980135893,
        "unilateral_improvement_p1": 0.001913555369717,
        "nash_conv": 0.003349535505611,
        "exploitability": 0.001674767752805
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 19100,
      "average": {
        "on_policy_value_p0": -0.055544855751985,
        "best_response_value_p0": -0.054960336043707,
        "best_response_value_p1": 0.058254907765983,
        "unilateral_improvement_p0": 0.000584519708278,
        "unilateral_improvement_p1": 0.002710052013998,
        "nash_conv": 0.003294571722276,
        "exploitability": 0.001647285861138
      },
      "current": {
        "on_policy_value_p0": 0.045274561848252,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.160341943439599,
        "unilateral_improvement_p0": 0.288058771485082,
        "unilateral_improvement_p1": 0.205616505287851,
        "nash_conv": 0.493675276772932,
        "exploitability": 0.246837638386466
      }
    },
    {
      "iteration": 19200,
      "average": {
        "on_policy_value_p0": -0.055551740206055,
        "best_response_value_p0": -0.053965097179101,
        "best_response_value_p1": 0.05724303566909,
        "unilateral_improvement_p0": 0.001586643026953,
        "unilateral_improvement_p1": 0.001691295463035,
        "nash_conv": 0.003277938489989,
        "exploitability": 0.001638969244994
      },
      "current": {
        "on_policy_value_p0": 0.219541576010716,
        "best_response_value_p0": 0.329925278139456,
        "best_response_value_p1": 0.279232595742521,
        "unilateral_improvement_p0": 0.11038370212874,
        "unilateral_improvement_p1": 0.498774171753237,
        "nash_conv": 0.609157873881977,
        "exploitability": 0.304578936940988
      }
    },
    {
      "iteration": 19300,
      "average": {
        "on_policy_value_p0": -0.055546978388171,
        "best_response_value_p0": -0.053142929414689,
        "best_response_value_p1": 0.056480619628185,
        "unilateral_improvement_p0": 0.002404048973481,
        "unilateral_improvement_p1": 0.000933641240014,
        "nash_conv": 0.003337690213495,
        "exploitability": 0.001668845106748
      },
      "current": {
        "on_policy_value_p0": -0.006771586863438,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.486456826273123,
        "unilateral_improvement_p0": 0.173438253530105,
        "unilateral_improvement_p1": 0.479685239409685,
        "nash_conv": 0.65312349293979,
        "exploitability": 0.326561746469895
      }
    },
    {
      "iteration": 19400,
      "average": {
        "on_policy_value_p0": -0.055539142739939,
        "best_response_value_p0": -0.053628003351824,
        "best_response_value_p1": 0.05696373911321,
        "unilateral_improvement_p0": 0.001911139388115,
        "unilateral_improvement_p1": 0.001424596373271,
        "nash_conv": 0.003335735761385,
        "exploitability": 0.001667867880693
      },
      "current": {
        "on_policy_value_p0": -0.476956090854324,
        "best_response_value_p0": 0.158985363618108,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.635941454472433,
        "unilateral_improvement_p1": 0.023043909145676,
        "nash_conv": 0.658985363618108,
        "exploitability": 0.329492681809054
      }
    },
    {
      "iteration": 19500,
      "average": {
        "on_policy_value_p0": -0.055545232035161,
        "best_response_value_p0": -0.054887159862563,
        "best_response_value_p1": 0.058098236680157,
        "unilateral_improvement_p0": 0.000658072172598,
        "unilateral_improvement_p1": 0.002553004644996,
        "nash_conv": 0.003211076817594,
        "exploitability": 0.001605538408797
      },
      "current": {
        "on_policy_value_p0": -0.080991036545656,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.247657703212322,
        "unilateral_improvement_p1": 0.085675630121011,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 19600,
      "average": {
        "on_policy_value_p0": -0.055554072645861,
        "best_response_value_p0": -0.054607123332653,
        "best_response_value_p1": 0.057783732728209,
        "unilateral_improvement_p0": 0.000946949313207,
        "unilateral_improvement_p1": 0.002229660082348,
        "nash_conv": 0.003176609395556,
        "exploitability": 0.001588304697778
      },
      "current": {
        "on_policy_value_p0": 0.074513770481083,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.149027540962167,
        "unilateral_improvement_p0": 0.092152896185583,
        "unilateral_improvement_p1": 0.22354131144325,
        "nash_conv": 0.315694207628834,
        "exploitability": 0.157847103814417
      }
    },
    {
      "iteration": 19700,
      "average": {
        "on_policy_value_p0": -0.055556216553777,
        "best_response_value_p0": -0.053761290919874,
        "best_response_value_p1": 0.056873883132976,
        "unilateral_improvement_p0": 0.001794925633902,
        "unilateral_improvement_p1": 0.001317666579199,
        "nash_conv": 0.003112592213102,
        "exploitability": 0.001556296106551
      },
      "current": {
        "on_policy_value_p0": 0.090504085345249,
        "best_response_value_p0": 0.090504085345249,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.423837418678582,
        "nash_conv": 0.423837418678582,
        "exploitability": 0.211918709339291
      }
    },
    {
      "iteration": 19800,
      "average": {
        "on_policy_value_p0": -0.055550634842611,
        "best_response_value_p0": -0.053204134959148,
        "best_response_value_p1": 0.056489604568211,
        "unilateral_improvement_p0": 0.002346499883462,
        "unilateral_improvement_p1": 0.0009389697256,
        "nash_conv": 0.003285469609062,
        "exploitability": 0.001642734804531
      },
      "current": {
        "on_policy_value_p0": 0.010936157271904,
        "best_response_value_p0": 0.144794352122859,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.133858194850955,
        "unilateral_improvement_p1": 0.344269490605237,
        "nash_conv": 0.478127685456192,
        "exploitability": 0.239063842728096
      }
    },
    {
      "iteration": 19900,
      "average": {
        "on_policy_value_p0": -0.055542082931249,
        "best_response_value_p0": -0.053025693567042,
        "best_response_value_p1": 0.056205737208572,
        "unilateral_improvement_p0": 0.002516389364207,
        "unilateral_improvement_p1": 0.000663654277323,
        "nash_conv": 0.00318004364153,
        "exploitability": 0.001590021820765
      },
      "current": {
        "on_policy_value_p0": -0.115992039252852,
        "best_response_value_p0": 0.115992039252852,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.231984078505703,
        "unilateral_improvement_p1": 0.217341294080482,
        "nash_conv": 0.449325372586185,
        "exploitability": 0.224662686293092
      }
    },
    {
      "iteration": 20000,
      "average": {
        "on_policy_value_p0": -0.055536440314122,
        "best_response_value_p0": -0.053997250601598,
        "best_response_value_p1": 0.057262502702602,
        "unilateral_improvement_p0": 0.001539189712524,
        "unilateral_improvement_p1": 0.00172606238848,
        "nash_conv": 0.003265252101004,
        "exploitability": 0.001632626050502
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 20100,
      "average": {
        "on_policy_value_p0": -0.05554373419659,
        "best_response_value_p0": -0.054975155198567,
        "best_response_value_p1": 0.058193906150396,
        "unilateral_improvement_p0": 0.000568578998023,
        "unilateral_improvement_p1": 0.002650171953806,
        "nash_conv": 0.003218750951829,
        "exploitability": 0.001609375475914
      },
      "current": {
        "on_policy_value_p0": -0.040081387780018,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.128685229105654,
        "unilateral_improvement_p0": 0.373414721113351,
        "unilateral_improvement_p1": 0.088603841325636,
        "nash_conv": 0.462018562438987,
        "exploitability": 0.231009281219493
      }
    },
    {
      "iteration": 20200,
      "average": {
        "on_policy_value_p0": -0.055551096470031,
        "best_response_value_p0": -0.054149750021158,
        "best_response_value_p1": 0.057397178616196,
        "unilateral_improvement_p0": 0.001401346448873,
        "unilateral_improvement_p1": 0.001846082146165,
        "nash_conv": 0.003247428595038,
        "exploitability": 0.001623714297519
      },
      "current": {
        "on_policy_value_p0": 0.196695603045947,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.226724539425228,
        "unilateral_improvement_p0": 0.136637730287386,
        "unilateral_improvement_p1": 0.423420142471175,
        "nash_conv": 0.560057872758561,
        "exploitability": 0.280028936379281
      }
    },
    {
      "iteration": 20300,
      "average": {
        "on_policy_value_p0": -0.055548274842149,
        "best_response_value_p0": -0.053342247778133,
        "best_response_value_p1": 0.056544125905774,
        "unilateral_improvement_p0": 0.002206027064016,
        "unilateral_improvement_p1": 0.000995851063626,
        "nash_conv": 0.003201878127641,
        "exploitability": 0.001600939063821
      },
      "current": {
        "on_policy_value_p0": 0.010674969336282,
        "best_response_value_p0": 0.139184939981074,
        "best_response_value_p1": 0.411423031930191,
        "unilateral_improvement_p0": 0.128509970644791,
        "unilateral_improvement_p1": 0.422098001266473,
        "nash_conv": 0.550607971911265,
        "exploitability": 0.275303985955632
      }
    },
    {
      "iteration": 20400,
      "average": {
        "on_policy_value_p0": -0.055541214979693,
        "best_response_value_p0": -0.053316611606314,
        "best_response_value_p1": 0.056362282135169,
        "unilateral_improvement_p0": 0.002224603373379,
        "unilateral_improvement_p1": 0.000821067155476,
        "nash_conv": 0.003045670528855,
        "exploitability": 0.001522835264428
      },
      "current": {
        "on_policy_value_p0": -0.263076601907898,
        "best_response_value_p0": 0.087692200635966,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.350768802543864,
        "unilateral_improvement_p1": 0.236923398092102,
        "nash_conv": 0.587692200635966,
        "exploitability": 0.293846100317983
      }
    },
    {
      "iteration": 20500,
      "average": {
        "on_policy_value_p0": -0.055543233747248,
        "best_response_value_p0": -0.054786402668588,
        "best_response_value_p1": 0.057997669148776,
        "unilateral_improvement_p0": 0.00075683107866,
        "unilateral_improvement_p1": 0.002454435401528,
        "nash_conv": 0.003211266480187,
        "exploitability": 0.001605633240094
      },
      "current": {
        "on_policy_value_p0": -0.165178316526827,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.165178316526827,
        "unilateral_improvement_p0": 0.331844983193493,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.331844983193493,
        "exploitability": 0.165922491596747
      }
    },
    {
      "iteration": 20600,
      "average": {
        "on_policy_value_p0": -0.055552229390367,
        "best_response_value_p0": -0.054742236246433,
        "best_response_value_p1": 0.057918938914926,
        "unilateral_improvement_p0": 0.000809993143935,
        "unilateral_improvement_p1": 0.002366709524559,
        "nash_conv": 0.003176702668494,
        "exploitability": 0.001588351334247
      },
      "current": {
        "on_policy_value_p0": 0.052760164228133,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.113906502438533,
        "unilateral_improvement_p0": 0.113906502438533,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.2805731691052,
        "exploitability": 0.1402865845526
      }
    },
    {
      "iteration": 20700,
      "average": {
        "on_policy_value_p0": -0.055556817158526,
        "best_response_value_p0": -0.054042959282938,
        "best_response_value_p1": 0.057176907814243,
        "unilateral_improvement_p0": 0.001513857875588,
        "unilateral_improvement_p1": 0.001620090655717,
        "nash_conv": 0.003133948531305,
        "exploitability": 0.001566974265653
      },
      "current": {
        "on_policy_value_p0": 0.092015374669123,
        "best_response_value_p0": 0.111327641687425,
        "best_response_value_p1": 0.294708799296729,
        "unilateral_improvement_p0": 0.019312267018302,
        "unilateral_improvement_p1": 0.386724173965852,
        "nash_conv": 0.406036440984154,
        "exploitability": 0.203018220492077
      }
    },
    {
      "iteration": 20800,
      "average": {
        "on_policy_value_p0": -0.055553360159158,
        "best_response_value_p0": -0.053399873917934,
        "best_response_value_p1": 0.056555797359662,
        "unilateral_improvement_p0": 0.002153486241224,
        "unilateral_improvement_p1": 0.001002437200504,
        "nash_conv": 0.003155923441728,
        "exploitability": 0.001577961720864
      },
      "current": {
        "on_policy_value_p0": 0.0463325050612,
        "best_response_value_p0": 0.074001656544266,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.027669151483066,
        "unilateral_improvement_p1": 0.379665838394534,
        "nash_conv": 0.407334989877599,
        "exploitability": 0.2036674949388
      }
    },
    {
      "iteration": 20900,
      "average": {
        "on_policy_value_p0": -0.055546208967047,
        "best_response_value_p0": -0.053091039364957,
        "best_response_value_p1": 0.05628519545842,
        "unilateral_improvement_p0": 0.002455169602091,
        "unilateral_improvement_p1": 0.000738986491373,
        "nash_conv": 0.003194156093464,
        "exploitability": 0.001597078046732
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 21000,
      "average": {
        "on_policy_value_p0": -0.055538225272695,
        "best_response_value_p0": -0.053521952053796,
        "best_response_value_p1": 0.056704209720601,
        "unilateral_improvement_p0": 0.002016273218899,
        "unilateral_improvement_p1": 0.001165984447906,
        "nash_conv": 0.003182257666805,
        "exploitability": 0.001591128833403
      },
      "current": {
        "on_policy_value_p0": -0.297477240355961,
        "best_response_value_p0": 0.297477240355961,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.594954480711922,
        "unilateral_improvement_p1": 0.035856092977372,
        "nash_conv": 0.630810573689294,
        "exploitability": 0.315405286844647
      }
    },
    {
      "iteration": 21100,
      "average": {
        "on_policy_value_p0": -0.055540454663184,
        "best_response_value_p0": -0.054841519751113,
        "best_response_value_p1": 0.057962585407889,
        "unilateral_improvement_p0": 0.000698934912071,
        "unilateral_improvement_p1": 0.002422130744705,
        "nash_conv": 0.003121065656776,
        "exploitability": 0.001560532828388
      },
      "current": {
        "on_policy_value_p0": -0.227101050820093,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.227101050820093,
        "unilateral_improvement_p0": 0.560434384153426,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.560434384153426,
        "exploitability": 0.280217192076713
      }
    },
    {
      "iteration": 21200,
      "average": {
        "on_policy_value_p0": -0.05554970999437,
        "best_response_value_p0": -0.054665153019747,
        "best_response_value_p1": 0.057810212480341,
        "unilateral_improvement_p0": 0.000884556974623,
        "unilateral_improvement_p1": 0.002260502485971,
        "nash_conv": 0.003145059460594,
        "exploitability": 0.001572529730297
      },
      "current": {
        "on_policy_value_p0": 0.112947298297591,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.220386035035743,
        "unilateral_improvement_p1": 0.279613964964257,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 21300,
      "average": {
        "on_policy_value_p0": -0.055551364304992,
        "best_response_value_p0": -0.053743211076998,
        "best_response_value_p1": 0.056604753600143,
        "unilateral_improvement_p0": 0.001808153227995,
        "unilateral_improvement_p1": 0.00105338929515,
        "nash_conv": 0.002861542523145,
        "exploitability": 0.001430771261572
      },
      "current": {
        "on_policy_value_p0": 0.110216526067541,
        "best_response_value_p0": 0.189360853320062,
        "best_response_value_p1": 0.341711345494958,
        "unilateral_improvement_p0": 0.079144327252521,
        "unilateral_improvement_p1": 0.451927871562499,
        "nash_conv": 0.53107219881502,
        "exploitability": 0.26553609940751
      }
    },
    {
      "iteration": 21400,
      "average": {
        "on_policy_value_p0": -0.055545380883584,
        "best_response_value_p0": -0.053179668504732,
        "best_response_value_p1": 0.056345818739041,
        "unilateral_improvement_p0": 0.002365712378852,
        "unilateral_improvement_p1": 0.000800437855457,
        "nash_conv": 0.003166150234309,
        "exploitability": 0.001583075117155
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 21500,
      "average": {
        "on_policy_value_p0": -0.0555408993589,
        "best_response_value_p0": -0.054324848981051,
        "best_response_value_p1": 0.057409110199374,
        "unilateral_improvement_p0": 0.001216050377849,
        "unilateral_improvement_p1": 0.001868210840474,
        "nash_conv": 0.003084261218323,
        "exploitability": 0.001542130609161
      },
      "current": {
        "on_policy_value_p0": -0.374344230912736,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.374344230912736,
        "unilateral_improvement_p0": 0.541010897579403,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.541010897579403,
        "exploitability": 0.270505448789701
      }
    },
    {
      "iteration": 21600,
      "average": {
        "on_policy_value_p0": -0.055548100409061,
        "best_response_value_p0": -0.054852884777363,
        "best_response_value_p1": 0.057974323714235,
        "unilateral_improvement_p0": 0.000695215631698,
        "unilateral_improvement_p1": 0.002426223305174,
        "nash_conv": 0.003121438936872,
        "exploitability": 0.001560719468436
      },
      "current": {
        "on_policy_value_p0": -0.051866443697635,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.218533110364302,
        "unilateral_improvement_p1": 0.114800222969032,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 21700,
      "average": {
        "on_policy_value_p0": -0.055555019447019,
        "best_response_value_p0": -0.054557398776906,
        "best_response_value_p1": 0.057553296045915,
        "unilateral_improvement_p0": 0.000997620670113,
        "unilateral_improvement_p1": 0.001998276598896,
        "nash_conv": 0.002995897269009,
        "exploitability": 0.001497948634505
      },
      "current": {
        "on_policy_value_p0": 0.083025496056783,
        "best_response_value_p0": 0.162016027550867,
        "best_response_value_p1": 0.175352270345165,
        "unilateral_improvement_p0": 0.078990531494084,
        "unilateral_improvement_p1": 0.258377766401947,
        "nash_conv": 0.337368297896032,
        "exploitability": 0.168684148948016
      }
    },
    {
      "iteration": 21800,
      "average": {
        "on_policy_value_p0": -0.055555679409733,
        "best_response_value_p0": -0.053776121440943,
        "best_response_value_p1": 0.056699973737994,
        "unilateral_improvement_p0": 0.00177955796879,
        "unilateral_improvement_p1": 0.00114429432826,
        "nash_conv": 0.002923852297051,
        "exploitability": 0.001461926148525
      },
      "current": {
        "on_policy_value_p0": 0.081503690854085,
        "best_response_value_p0": 0.081503690854085,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.414837024187419,
        "nash_conv": 0.414837024187419,
        "exploitability": 0.207418512093709
      }
    },
    {
      "iteration": 21900,
      "average": {
        "on_policy_value_p0": -0.055550627188542,
        "best_response_value_p0": -0.053316815982879,
        "best_response_value_p1": 0.056441069748323,
        "unilateral_improvement_p0": 0.002233811205663,
        "unilateral_improvement_p1": 0.000890442559781,
        "nash_conv": 0.003124253765444,
        "exploitability": 0.001562126882722
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 22000,
      "average": {
        "on_policy_value_p0": -0.055543318713817,
        "best_response_value_p0": -0.053162747006272,
        "best_response_value_p1": 0.056184519431286,
        "unilateral_improvement_p0": 0.002380571707545,
        "unilateral_improvement_p1": 0.000641200717468,
        "nash_conv": 0.003021772425014,
        "exploitability": 0.001510886212507
      },
      "current": {
        "on_policy_value_p0": -0.116676910263896,
        "best_response_value_p0": 0.116676910263896,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.233353820527791,
        "unilateral_improvement_p1": 0.216656423069438,
        "nash_conv": 0.450010243597229,
        "exploitability": 0.225005121798614
      }
    },
    {
      "iteration": 22100,
      "average": {
        "on_policy_value_p0": -0.055538309952639,
        "best_response_value_p0": -0.054023969049539,
        "best_response_value_p1": 0.057129356380171,
        "unilateral_improvement_p0": 0.0015143409031,
        "unilateral_improvement_p1": 0.001591046427533,
        "nash_conv": 0.003105387330633,
        "exploitability": 0.001552693665316
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 22200,
      "average": {
        "on_policy_value_p0": -0.055544088940757,
        "best_response_value_p0": -0.055049123052568,
        "best_response_value_p1": 0.058073892446937,
        "unilateral_improvement_p0": 0.000494965888188,
        "unilateral_improvement_p1": 0.00252980350618,
        "nash_conv": 0.003024769394368,
        "exploitability": 0.001512384697184
      },
      "current": {
        "on_policy_value_p0": -0.097423834173956,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.108095722946804,
        "unilateral_improvement_p0": 0.430757167507289,
        "unilateral_improvement_p1": 0.010671888772848,
        "nash_conv": 0.441429056280137,
        "exploitability": 0.220714528140068
      }
    },
    {
      "iteration": 22300,
      "average": {
        "on_policy_value_p0": -0.055551397160686,
        "best_response_value_p0": -0.054374966635398,
        "best_response_value_p1": 0.057473692504076,
        "unilateral_improvement_p0": 0.001176430525288,
        "unilateral_improvement_p1": 0.00192229534339,
        "nash_conv": 0.003098725868678,
        "exploitability": 0.001549362934339
      },
      "current": {
        "on_policy_value_p0": 0.185151813119006,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.203636959571346,
        "unilateral_improvement_p0": 0.148181520214327,
        "unilateral_improvement_p1": 0.388788772690352,
        "nash_conv": 0.536970292904679,
        "exploitability": 0.268485146452339
      }
    },
    {
      "iteration": 22400,
      "average": {
        "on_policy_value_p0": -0.055550195343673,
        "best_response_value_p0": -0.053535814671797,
        "best_response_value_p1": 0.056472254711455,
        "unilateral_improvement_p0": 0.002014380671876,
        "unilateral_improvement_p1": 0.000922059367782,
        "nash_conv": 0.002936440039658,
        "exploitability": 0.001468220019829
      },
      "current": {
        "on_policy_value_p0": 0.061094229173958,
        "best_response_value_p0": 0.117902268458516,
        "best_response_value_p1": 0.386383921430883,
        "unilateral_improvement_p0": 0.056808039284559,
        "unilateral_improvement_p1": 0.44747815060484,
        "nash_conv": 0.504286189889399,
        "exploitability": 0.252143094944699
      }
    },
    {
      "iteration": 22500,
      "average": {
        "on_policy_value_p0": -0.055544088690263,
        "best_response_value_p0": -0.053263793549622,
        "best_response_value_p1": 0.0562867513939,
        "unilateral_improvement_p0": 0.002280295140641,
        "unilateral_improvement_p1": 0.000742662703637,
        "nash_conv": 0.003022957844278,
        "exploitability": 0.001511478922139
      },
      "current": {
        "on_policy_value_p0": -0.155583740904443,
        "best_response_value_p0": 0.062944172730372,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.218527913634814,
        "unilateral_improvement_p1": 0.344416259095557,
        "nash_conv": 0.562944172730372,
        "exploitability": 0.281472086365186
      }
    },
    {
      "iteration": 22600,
      "average": {
        "on_policy_value_p0": -0.055542649602688,
        "best_response_value_p0": -0.054681887187051,
        "best_response_value_p1": 0.057657310550298,
        "unilateral_improvement_p0": 0.000860762415637,
        "unilateral_improvement_p1": 0.00211466094761,
        "nash_conv": 0.002975423363247,
        "exploitability": 0.001487711681624
      },
      "current": {
        "on_policy_value_p0": -0.311248955466587,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.311248955466587,
        "unilateral_improvement_p0": 0.477915622133253,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.477915622133253,
        "exploitability": 0.238957811066627
      }
    },
    {
      "iteration": 22700,
      "average": {
        "on_policy_value_p0": -0.055550211269048,
        "best_response_value_p0": -0.054843785615703,
        "best_response_value_p1": 0.057918010253669,
        "unilateral_improvement_p0": 0.000706425653346,
        "unilateral_improvement_p1": 0.002367798984621,
        "nash_conv": 0.003074224637966,
        "exploitability": 0.001537112318983
      },
      "current": {
        "on_policy_value_p0": -0.005578435700619,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.155588604153596,
        "unilateral_improvement_p0": 0.172245102367286,
        "unilateral_improvement_p1": 0.150010168452976,
        "nash_conv": 0.322255270820262,
        "exploitability": 0.161127635410131
      }
    },
    {
      "iteration": 22800,
      "average": {
        "on_policy_value_p0": -0.055556032452386,
        "best_response_value_p0": -0.054407218407769,
        "best_response_value_p1": 0.057387645462394,
        "unilateral_improvement_p0": 0.001148814044617,
        "unilateral_improvement_p1": 0.001831613010008,
        "nash_conv": 0.002980427054625,
        "exploitability": 0.001490213527312
      },
      "current": {
        "on_policy_value_p0": 0.081059736913188,
        "best_response_value_p0": 0.142311910203595,
        "best_response_value_p1": 0.21082898675252,
        "unilateral_improvement_p0": 0.061252173290407,
        "unilateral_improvement_p1": 0.291888723665708,
        "nash_conv": 0.353140896956115,
        "exploitability": 0.176570448478058
      }
    },
    {
      "iteration": 22900,
      "average": {
        "on_policy_value_p0": -0.055555394409424,
        "best_response_value_p0": -0.05371091603019,
        "best_response_value_p1": 0.056606781411852,
        "unilateral_improvement_p0": 0.001844478379234,
        "unilateral_improvement_p1": 0.001051387002428,
        "nash_conv": 0.002895865381662,
        "exploitability": 0.001447932690831
      },
      "current": {
        "on_policy_value_p0": 0.073865933146303,
        "best_response_value_p0": 0.073865933146303,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.407199266479637,
        "nash_conv": 0.407199266479637,
        "exploitability": 0.203599633239818
      }
    },
    {
      "iteration": 23000,
      "average": {
        "on_policy_value_p0": -0.055550082225282,
        "best_response_value_p0": -0.053313608081179,
        "best_response_value_p1": 0.056360664970931,
        "unilateral_improvement_p0": 0.002236474144103,
        "unilateral_improvement_p1": 0.000810582745649,
        "nash_conv": 0.003047056889752,
        "exploitability": 0.001523528444876
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 23100,
      "average": {
        "on_policy_value_p0": -0.055542850299195,
        "best_response_value_p0": -0.053256178045639,
        "best_response_value_p1": 0.056116679408287,
        "unilateral_improvement_p0": 0.002286672253556,
        "unilateral_improvement_p1": 0.000573829109092,
        "nash_conv": 0.002860501362648,
        "exploitability": 0.001430250681324
      },
      "current": {
        "on_policy_value_p0": -0.152994622131199,
        "best_response_value_p0": 0.152994622131199,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.305989244262398,
        "unilateral_improvement_p1": 0.180338711202135,
        "nash_conv": 0.486327955464532,
        "exploitability": 0.243163977732266
      }
    },
    {
      "iteration": 23200,
      "average": {
        "on_policy_value_p0": -0.055539205196883,
        "best_response_value_p0": -0.05419583349567,
        "best_response_value_p1": 0.0572255560079,
        "unilateral_improvement_p0": 0.001343371701212,
        "unilateral_improvement_p1": 0.001686350811017,
        "nash_conv": 0.003029722512229,
        "exploitability": 0.001514861256115
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 23300,
      "average": {
        "on_policy_value_p0": -0.055545371123058,
        "best_response_value_p0": -0.055022115924619,
        "best_response_value_p1": 0.058008239018475,
        "unilateral_improvement_p0": 0.000523255198439,
        "unilateral_improvement_p1": 0.002462867895417,
        "nash_conv": 0.002986123093856,
        "exploitability": 0.001493061546928
      },
      "current": {
        "on_policy_value_p0": -0.044034902816105,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.127075859362348,
        "unilateral_improvement_p0": 0.377368236149439,
        "unilateral_improvement_p1": 0.083040956546242,
        "nash_conv": 0.460409192695681,
        "exploitability": 0.230204596347841
      }
    },
    {
      "iteration": 23400,
      "average": {
        "on_policy_value_p0": -0.055551598855018,
        "best_response_value_p0": -0.054331066917168,
        "best_response_value_p1": 0.057354048707382,
        "unilateral_improvement_p0": 0.00122053193785,
        "unilateral_improvement_p1": 0.001802449852364,
        "nash_conv": 0.003022981790214,
        "exploitability": 0.001511490895107
      },
      "current": {
        "on_policy_value_p0": 0.18696283169908,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.207258996731494,
        "unilateral_improvement_p0": 0.146370501634253,
        "unilateral_improvement_p1": 0.394221828430575,
        "nash_conv": 0.540592330064828,
        "exploitability": 0.270296165032414
      }
    },
    {
      "iteration": 23500,
      "average": {
        "on_policy_value_p0": -0.055550088737722,
        "best_response_value_p0": -0.053581868396588,
        "best_response_value_p1": 0.056464794235707,
        "unilateral_improvement_p0": 0.001968220341133,
        "unilateral_improvement_p1": 0.000914705497985,
        "nash_conv": 0.002882925839118,
        "exploitability": 0.001441462919559
      },
      "current": {
        "on_policy_value_p0": 0.049370349988084,
        "best_response_value_p0": 0.11305476423711,
        "best_response_value_p1": 0.38429309025794,
        "unilateral_improvement_p0": 0.063684414249026,
        "unilateral_improvement_p1": 0.433663440246023,
        "nash_conv": 0.497347854495049,
        "exploitability": 0.248673927247525
      }
    },
    {
      "iteration": 23600,
      "average": {
        "on_policy_value_p0": -0.055544588745651,
        "best_response_value_p0": -0.053328243968677,
        "best_response_value_p1": 0.056292440907097,
        "unilateral_improvement_p0": 0.002216344776974,
        "unilateral_improvement_p1": 0.000747852161446,
        "nash_conv": 0.00296419693842,
        "exploitability": 0.00148209846921
      },
      "current": {
        "on_policy_value_p0": -0.141667899412727,
        "best_response_value_p0": 0.072221400391515,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.213889299804242,
        "unilateral_improvement_p1": 0.358332100587273,
        "nash_conv": 0.572221400391515,
        "exploitability": 0.286110700195758
      }
    },
    {
      "iteration": 23700,
      "average": {
        "on_policy_value_p0": -0.055542979068216,
        "best_response_value_p0": -0.054688971249769,
        "best_response_value_p1": 0.05756930486853,
        "unilateral_improvement_p0": 0.000854007818447,
        "unilateral_improvement_p1": 0.002026325800314,
        "nash_conv": 0.002880333618761,
        "exploitability": 0.00144016680938
      },
      "current": {
        "on_policy_value_p0": -0.322641353474754,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.322641353474754,
        "unilateral_improvement_p0": 0.48930802014142,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.48930802014142,
        "exploitability": 0.24465401007071
      }
    },
    {
      "iteration": 23800,
      "average": {
        "on_policy_value_p0": -0.05554996937617,
        "best_response_value_p0": -0.0548770272303,
        "best_response_value_p1": 0.057874187255688,
        "unilateral_improvement_p0": 0.000672942145871,
        "unilateral_improvement_p1": 0.002324217879517,
        "nash_conv": 0.002997160025388,
        "exploitability": 0.001498580012694
      },
      "current": {
        "on_policy_value_p0": -0.027918895254311,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.165826618533484,
        "unilateral_improvement_p0": 0.194585561920977,
        "unilateral_improvement_p1": 0.137907723279173,
        "nash_conv": 0.33249328520015,
        "exploitability": 0.166246642600075
      }
    },
    {
      "iteration": 23900,
      "average": {
        "on_policy_value_p0": -0.055555697205872,
        "best_response_value_p0": -0.054505275169844,
        "best_response_value_p1": 0.057400070924109,
        "unilateral_improvement_p0": 0.001050422036028,
        "unilateral_improvement_p1": 0.001844373718237,
        "nash_conv": 0.002894795754265,
        "exploitability": 0.001447397877132
      },
      "current": {
        "on_policy_value_p0": 0.081404562876117,
        "best_response_value_p0": 0.148885702083422,
        "best_response_value_p1": 0.198371054918722,
        "unilateral_improvement_p0": 0.067481139207306,
        "unilateral_improvement_p1": 0.279775617794839,
        "nash_conv": 0.347256757002145,
        "exploitability": 0.173628378501072
      }
    },
    {
      "iteration": 24000,
      "average": {
        "on_policy_value_p0": -0.055555707999169,
        "best_response_value_p0": -0.053820237338422,
        "best_response_value_p1": 0.056621987840993,
        "unilateral_improvement_p0": 0.001735470660747,
        "unilateral_improvement_p1": 0.001066279841824,
        "nash_conv": 0.002801750502571,
        "exploitability": 0.001400875251286
      },
      "current": {
        "on_policy_value_p0": 0.078647592059433,
        "best_response_value_p0": 0.078647592059433,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.411980925392766,
        "nash_conv": 0.411980925392766,
        "exploitability": 0.205990462696383
      }
    },
    {
      "iteration": 24100,
      "average": {
        "on_policy_value_p0": -0.055551146225457,
        "best_response_value_p0": -0.053410057483292,
        "best_response_value_p1": 0.056387041833355,
        "unilateral_improvement_p0": 0.002141088742165,
        "unilateral_improvement_p1": 0.000835895607898,
        "nash_conv": 0.002976984350063,
        "exploitability": 0.001488492175031
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 24200,
      "average": {
        "on_policy_value_p0": -0.055544753950537,
        "best_response_value_p0": -0.05326354068932,
        "best_response_value_p1": 0.056154037528259,
        "unilateral_improvement_p0": 0.002281213261217,
        "unilateral_improvement_p1": 0.000609283577722,
        "nash_conv": 0.002890496838939,
        "exploitability": 0.001445248419469
      },
      "current": {
        "on_policy_value_p0": -0.111055797604016,
        "best_response_value_p0": 0.111138767864659,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.222194565468675,
        "unilateral_improvement_p1": 0.222277535729317,
        "nash_conv": 0.444472101197992,
        "exploitability": 0.222236050598996
      }
    },
    {
      "iteration": 24300,
      "average": {
        "on_policy_value_p0": -0.055539988937579,
        "best_response_value_p0": -0.054008583666178,
        "best_response_value_p1": 0.056969617891069,
        "unilateral_improvement_p0": 0.001531405271401,
        "unilateral_improvement_p1": 0.00142962895349,
        "nash_conv": 0.002961034224891,
        "exploitability": 0.001480517112445
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 24400,
      "average": {
        "on_policy_value_p0": -0.05554421776938,
        "best_response_value_p0": -0.055112584544724,
        "best_response_value_p1": 0.057924569887637,
        "unilateral_improvement_p0": 0.000431633224656,
        "unilateral_improvement_p1": 0.002380352118257,
        "nash_conv": 0.002811985342913,
        "exploitability": 0.001405992671456
      },
      "current": {
        "on_policy_value_p0": -0.157928978009947,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.157928978009947,
        "unilateral_improvement_p0": 0.491262311343281,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.491262311343281,
        "exploitability": 0.24563115567164
      }
    },
    {
      "iteration": 24500,
      "average": {
        "on_policy_value_p0": -0.055551316703421,
        "best_response_value_p0": -0.054594656634512,
        "best_response_value_p1": 0.057545077587721,
        "unilateral_improvement_p0": 0.000956660068909,
        "unilateral_improvement_p1": 0.0019937608843,
        "nash_conv": 0.002950420953209,
        "exploitability": 0.001475210476604
      },
      "current": {
        "on_policy_value_p0": 0.136553061847647,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.196780271485686,
        "unilateral_improvement_p1": 0.303219728514314,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 24600,
      "average": {
        "on_policy_value_p0": -0.055551744394744,
        "best_response_value_p0": -0.053784410141912,
        "best_response_value_p1": 0.056427033690918,
        "unilateral_improvement_p0": 0.001767334252832,
        "unilateral_improvement_p1": 0.000875289296173,
        "nash_conv": 0.002642623549005,
        "exploitability": 0.001321311774503
      },
      "current": {
        "on_policy_value_p0": 0.100058727020773,
        "best_response_value_p0": 0.171699509639048,
        "best_response_value_p1": 0.35671843476345,
        "unilateral_improvement_p0": 0.071640782618275,
        "unilateral_improvement_p1": 0.456777161784223,
        "nash_conv": 0.528417944402498,
        "exploitability": 0.264208972201249
      }
    },
    {
      "iteration": 24700,
      "average": {
        "on_policy_value_p0": -0.055546716126218,
        "best_response_value_p0": -0.053334462571992,
        "best_response_value_p1": 0.056279567486641,
        "unilateral_improvement_p0": 0.002212253554226,
        "unilateral_improvement_p1": 0.000732851360422,
        "nash_conv": 0.002945104914649,
        "exploitability": 0.001472552457324
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 24800,
      "average": {
        "on_policy_value_p0": -0.055542757583726,
        "best_response_value_p0": -0.054306318116482,
        "best_response_value_p1": 0.057202827056008,
        "unilateral_improvement_p0": 0.001236439467245,
        "unilateral_improvement_p1": 0.001660069472282,
        "nash_conv": 0.002896508939527,
        "exploitability": 0.001448254469763
      },
      "current": {
        "on_policy_value_p0": -0.393678222460922,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.393678222460922,
        "unilateral_improvement_p0": 0.560344889127589,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.560344889127589,
        "exploitability": 0.280172444563794
      }
    },
    {
      "iteration": 24900,
      "average": {
        "on_policy_value_p0": -0.055548243062922,
        "best_response_value_p0": -0.054931030725482,
        "best_response_value_p1": 0.057804844346571,
        "unilateral_improvement_p0": 0.00061721233744,
        "unilateral_improvement_p1": 0.00225660128365,
        "nash_conv": 0.002873813621089,
        "exploitability": 0.001436906810545
      },
      "current": {
        "on_policy_value_p0": -0.06767372462079,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.234340391287456,
        "unilateral_improvement_p1": 0.098992942045877,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 25000,
      "average": {
        "on_policy_value_p0": -0.055554247887445,
        "best_response_value_p0": -0.054711306602581,
        "best_response_value_p1": 0.057527917127413,
        "unilateral_improvement_p0": 0.000842941284864,
        "unilateral_improvement_p1": 0.001973669239968,
        "nash_conv": 0.002816610524832,
        "exploitability": 0.001408305262416
      },
      "current": {
        "on_policy_value_p0": 0.072602628816007,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.145205257632015,
        "unilateral_improvement_p0": 0.094064037850659,
        "unilateral_improvement_p1": 0.217807886448022,
        "nash_conv": 0.311871924298681,
        "exploitability": 0.155935962149341
      }
    },
    {
      "iteration": 25100,
      "average": {
        "on_policy_value_p0": -0.055556187893073,
        "best_response_value_p0": -0.054056829648137,
        "best_response_value_p1": 0.056847531165427,
        "unilateral_improvement_p0": 0.001499358244937,
        "unilateral_improvement_p1": 0.001291343272353,
        "nash_conv": 0.00279070151729,
        "exploitability": 0.001395350758645
      },
      "current": {
        "on_policy_value_p0": 0.097535102623678,
        "best_response_value_p0": 0.097535102623678,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.430868435957011,
        "nash_conv": 0.430868435957011,
        "exploitability": 0.215434217978505
      }
    },
    {
      "iteration": 25200,
      "average": {
        "on_policy_value_p0": -0.055553008681942,
        "best_response_value_p0": -0.053567563998159,
        "best_response_value_p1": 0.056453007902994,
        "unilateral_improvement_p0": 0.001985444683783,
        "unilateral_improvement_p1": 0.000899999221052,
        "nash_conv": 0.002885443904835,
        "exploitability": 0.001442721952417
      },
      "current": {
        "on_policy_value_p0": 0.036670674180389,
        "best_response_value_p0": 0.093325318305889,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.0566546441255,
        "unilateral_improvement_p1": 0.370004007513722,
        "nash_conv": 0.426658651639222,
        "exploitability": 0.213329325819611
      }
    },
    {
      "iteration": 25300,
      "average": {
        "on_policy_value_p0": -0.055547677979751,
        "best_response_value_p0": -0.053328357530464,
        "best_response_value_p1": 0.056229873484406,
        "unilateral_improvement_p0": 0.002219320449287,
        "unilateral_improvement_p1": 0.000682195504655,
        "nash_conv": 0.002901515953942,
        "exploitability": 0.001450757976971
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 25400,
      "average": {
        "on_policy_value_p0": -0.05554150025317,
        "best_response_value_p0": -0.053617839264283,
        "best_response_value_p1": 0.056483105937582,
        "unilateral_improvement_p0": 0.001923660988887,
        "unilateral_improvement_p1": 0.000941605684412,
        "nash_conv": 0.002865266673299,
        "exploitability": 0.00143263333665
      },
      "current": {
        "on_policy_value_p0": -0.255254900732842,
        "best_response_value_p0": 0.255254900732842,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.510509801465683,
        "unilateral_improvement_p1": 0.078078432600492,
        "nash_conv": 0.588588234066175,
        "exploitability": 0.294294117033087
      }
    },
    {
      "iteration": 25500,
      "average": {
        "on_policy_value_p0": -0.055541760898957,
        "best_response_value_p0": -0.054681873715041,
        "best_response_value_p1": 0.057568608723906,
        "unilateral_improvement_p0": 0.000859887183915,
        "unilateral_improvement_p1": 0.00202684782495,
        "nash_conv": 0.002886735008865,
        "exploitability": 0.001443367504433
      },
      "current": {
        "on_policy_value_p0": -0.321001849401672,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.321001849401672,
        "unilateral_improvement_p0": 0.654335182735005,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.654335182735005,
        "exploitability": 0.327167591367502
      }
    },
    {
      "iteration": 25600,
      "average": {
        "on_policy_value_p0": -0.055548699913965,
        "best_response_value_p0": -0.054980391738194,
        "best_response_value_p1": 0.057775745050448,
        "unilateral_improvement_p0": 0.000568308175771,
        "unilateral_improvement_p1": 0.002227045136483,
        "nash_conv": 0.002795353312254,
        "exploitability": 0.001397676656127
      },
      "current": {
        "on_policy_value_p0": 0.078011122454901,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.255322210878432,
        "unilateral_improvement_p1": 0.244677789121568,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 25700,
      "average": {
        "on_policy_value_p0": -0.055552393667723,
        "best_response_value_p0": -0.054143704680423,
        "best_response_value_p1": 0.056976318166023,
        "unilateral_improvement_p0": 0.0014086889873,
        "unilateral_improvement_p1": 0.0014239244983,
        "nash_conv": 0.0028326134856,
        "exploitability": 0.0014163067428
      },
      "current": {
        "on_policy_value_p0": 0.189852514592959,
        "best_response_value_p0": 0.302830672267112,
        "best_response_value_p1": 0.274043684651693,
        "unilateral_improvement_p0": 0.112978157674153,
        "unilateral_improvement_p1": 0.463896199244652,
        "nash_conv": 0.576874356918806,
        "exploitability": 0.288437178459403
      }
    },
    {
      "iteration": 25800,
      "average": {
        "on_policy_value_p0": -0.05554957537179,
        "best_response_value_p0": -0.053549667297204,
        "best_response_value_p1": 0.056415069469364,
        "unilateral_improvement_p0": 0.001999908074586,
        "unilateral_improvement_p1": 0.000865494097574,
        "nash_conv": 0.00286540217216,
        "exploitability": 0.00143270108608
      },
      "current": {
        "on_policy_value_p0": -0.01044930732979,
        "best_response_value_p0": 0.155448355774356,
        "best_response_value_p1": 0.434228141771177,
        "unilateral_improvement_p0": 0.165897663104147,
        "unilateral_improvement_p1": 0.423778834441386,
        "nash_conv": 0.589676497545533,
        "exploitability": 0.294838248772766
      }
    },
    {
      "iteration": 25900,
      "average": {
        "on_policy_value_p0": -0.05554447189683,
        "best_response_value_p0": -0.053541115485204,
        "best_response_value_p1": 0.056216282389182,
        "unilateral_improvement_p0": 0.002003356411626,
        "unilateral_improvement_p1": 0.000671810492352,
        "nash_conv": 0.002675166903978,
        "exploitability": 0.001337583451989
      },
      "current": {
        "on_policy_value_p0": -0.250727212705089,
        "best_response_value_p0": 0.083575737568363,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.334302950273452,
        "unilateral_improvement_p1": 0.249272787294911,
        "nash_conv": 0.583575737568363,
        "exploitability": 0.291787868784182
      }
    },
    {
      "iteration": 26000,
      "average": {
        "on_policy_value_p0": -0.055544934375495,
        "best_response_value_p0": -0.054804062389832,
        "best_response_value_p1": 0.05762474723704,
        "unilateral_improvement_p0": 0.000740871985663,
        "unilateral_improvement_p1": 0.002079812861545,
        "nash_conv": 0.002820684847209,
        "exploitability": 0.001410342423604
      },
      "current": {
        "on_policy_value_p0": -0.251578517988391,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.251578517988391,
        "unilateral_improvement_p0": 0.418245184655057,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.418245184655057,
        "exploitability": 0.209122592327529
      }
    },
    {
      "iteration": 26100,
      "average": {
        "on_policy_value_p0": -0.055551256950939,
        "best_response_value_p0": -0.054890998014052,
        "best_response_value_p1": 0.057753689708941,
        "unilateral_improvement_p0": 0.000660258936886,
        "unilateral_improvement_p1": 0.002202432758003,
        "nash_conv": 0.002862691694889,
        "exploitability": 0.001431345847445
      },
      "current": {
        "on_policy_value_p0": 0.017257001460971,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.144994713755449,
        "unilateral_improvement_p0": 0.149409665205696,
        "unilateral_improvement_p1": 0.16225171521642,
        "nash_conv": 0.311661380422116,
        "exploitability": 0.155830690211058
      }
    },
    {
      "iteration": 26200,
      "average": {
        "on_policy_value_p0": -0.055555991090132,
        "best_response_value_p0": -0.054498424675338,
        "best_response_value_p1": 0.057277903432724,
        "unilateral_improvement_p0": 0.001057566414794,
        "unilateral_improvement_p1": 0.001721912342592,
        "nash_conv": 0.002779478757386,
        "exploitability": 0.001389739378693
      },
      "current": {
        "on_policy_value_p0": 0.082472242410618,
        "best_response_value_p0": 0.143864163833219,
        "best_response_value_p1": 0.210549490488131,
        "unilateral_improvement_p0": 0.061391921422601,
        "unilateral_improvement_p1": 0.293021732898749,
        "nash_conv": 0.35441365432135,
        "exploitability": 0.177206827160675
      }
    },
    {
      "iteration": 26300,
      "average": {
        "on_policy_value_p0": -0.055555731606918,
        "best_response_value_p0": -0.053880003031299,
        "best_response_value_p1": 0.056555059984007,
        "unilateral_improvement_p0": 0.001675728575619,
        "unilateral_improvement_p1": 0.000999328377089,
        "nash_conv": 0.002675056952708,
        "exploitability": 0.001337528476354
      },
      "current": {
        "on_policy_value_p0": 0.078661007753801,
        "best_response_value_p0": 0.078661007753801,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.411994341087134,
        "nash_conv": 0.411994341087134,
        "exploitability": 0.205997170543567
      }
    },
    {
      "iteration": 26400,
      "average": {
        "on_policy_value_p0": -0.055551646845524,
        "best_response_value_p0": -0.0534973724172,
        "best_response_value_p1": 0.056340836271947,
        "unilateral_improvement_p0": 0.002054274428324,
        "unilateral_improvement_p1": 0.000789189426423,
        "nash_conv": 0.002843463854747,
        "exploitability": 0.001421731927373
      },
      "current": {
        "on_policy_value_p0": 0.002450327702816,
        "best_response_value_p0": 0.161766011261036,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.15931568355822,
        "unilateral_improvement_p1": 0.335783661036149,
        "nash_conv": 0.495099344594369,
        "exploitability": 0.247549672297184
      }
    },
    {
      "iteration": 26500,
      "average": {
        "on_policy_value_p0": -0.055546040708023,
        "best_response_value_p0": -0.053347189206306,
        "best_response_value_p1": 0.056128229342619,
        "unilateral_improvement_p0": 0.002198851501718,
        "unilateral_improvement_p1": 0.000582188634596,
        "nash_conv": 0.002781040136313,
        "exploitability": 0.001390520068157
      },
      "current": {
        "on_policy_value_p0": -0.098282604728704,
        "best_response_value_p0": 0.117525364302314,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.215807969031019,
        "unilateral_improvement_p1": 0.235050728604629,
        "nash_conv": 0.450858697635648,
        "exploitability": 0.225429348817824
      }
    },
    {
      "iteration": 26600,
      "average": {
        "on_policy_value_p0": -0.055541369062633,
        "best_response_value_p0": -0.053967578279193,
        "best_response_value_p1": 0.05679760045876,
        "unilateral_improvement_p0": 0.00157379078344,
        "unilateral_improvement_p1": 0.001256231396127,
        "nash_conv": 0.002830022179567,
        "exploitability": 0.001415011089783
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 26700,
      "average": {
        "on_policy_value_p0": -0.055544116809586,
        "best_response_value_p0": -0.055013891968534,
        "best_response_value_p1": 0.057742612234785,
        "unilateral_improvement_p0": 0.000530224841053,
        "unilateral_improvement_p1": 0.002198495425199,
        "nash_conv": 0.002728720266252,
        "exploitability": 0.001364360133126
      },
      "current": {
        "on_policy_value_p0": -0.196773675290207,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.196773675290207,
        "unilateral_improvement_p0": 0.53010700862354,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.53010700862354,
        "exploitability": 0.26505350431177
      }
    },
    {
      "iteration": 26800,
      "average": {
        "on_policy_value_p0": -0.055550638234351,
        "best_response_value_p0": -0.054789622956307,
        "best_response_value_p1": 0.057574829506068,
        "unilateral_improvement_p0": 0.000761015278045,
        "unilateral_improvement_p1": 0.002024191271716,
        "nash_conv": 0.002785206549761,
        "exploitability": 0.00139260327488
      },
      "current": {
        "on_policy_value_p0": 0.107817562207384,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.225515771125949,
        "unilateral_improvement_p1": 0.274484228874051,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 26900,
      "average": {
        "on_policy_value_p0": -0.055552444606107,
        "best_response_value_p0": -0.054033471888517,
        "best_response_value_p1": 0.05667318853993,
        "unilateral_improvement_p0": 0.001518972717589,
        "unilateral_improvement_p1": 0.001120743933823,
        "nash_conv": 0.002639716651413,
        "exploitability": 0.001319858325706
      },
      "current": {
        "on_policy_value_p0": 0.131776946843663,
        "best_response_value_p0": 0.224562016613461,
        "best_response_value_p1": 0.314429860460405,
        "unilateral_improvement_p0": 0.092785069769797,
        "unilateral_improvement_p1": 0.446206807304069,
        "nash_conv": 0.538991877073866,
        "exploitability": 0.269495938536933
      }
    },
    {
      "iteration": 27000,
      "average": {
        "on_policy_value_p0": -0.055548893786714,
        "best_response_value_p0": -0.053524001322639,
        "best_response_value_p1": 0.056341352921499,
        "unilateral_improvement_p0": 0.002024892464076,
        "unilateral_improvement_p1": 0.000792459134785,
        "nash_conv": 0.002817351598861,
        "exploitability": 0.00140867579943
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 27100,
      "average": {
        "on_policy_value_p0": -0.055544010853259,
        "best_response_value_p0": -0.053789315418275,
        "best_response_value_p1": 0.056570321387962,
        "unilateral_improvement_p0": 0.001754695434985,
        "unilateral_improvement_p1": 0.001026310534703,
        "nash_conv": 0.002781005969687,
        "exploitability": 0.001390502984844
      },
      "current": {
        "on_policy_value_p0": -0.375688237713875,
        "best_response_value_p0": 0.125229412571292,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.500917650285167,
        "unilateral_improvement_p1": 0.124311762286125,
        "nash_conv": 0.625229412571292,
        "exploitability": 0.312614706285646
      }
    },
    {
      "iteration": 27200,
      "average": {
        "on_policy_value_p0": -0.055546452489317,
        "best_response_value_p0": -0.054882947244397,
        "best_response_value_p1": 0.057666377171439,
        "unilateral_improvement_p0": 0.00066350524492,
        "unilateral_improvement_p1": 0.002119924682122,
        "nash_conv": 0.002783429927042,
        "exploitability": 0.001391714963521
      },
      "current": {
        "on_policy_value_p0": -0.124806070987377,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.160201819000049,
        "unilateral_improvement_p0": 0.291472737654044,
        "unilateral_improvement_p1": 0.035395748012671,
        "nash_conv": 0.326868485666715,
        "exploitability": 0.163434242833358
      }
    },
    {
      "iteration": 27300,
      "average": {
        "on_policy_value_p0": -0.055552214307927,
        "best_response_value_p0": -0.054851639449222,
        "best_response_value_p1": 0.05763594696026,
        "unilateral_improvement_p0": 0.000700574858704,
        "unilateral_improvement_p1": 0.002083732652333,
        "nash_conv": 0.002784307511037,
        "exploitability": 0.001392153755519
      },
      "current": {
        "on_policy_value_p0": 0.036701728426161,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.129964938240505,
        "unilateral_improvement_p0": 0.129964938240505,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.296631604907172,
        "exploitability": 0.148315802453586
      }
    },
    {
      "iteration": 27400,
      "average": {
        "on_policy_value_p0": -0.055556009213575,
        "best_response_value_p0": -0.054427872932161,
        "best_response_value_p1": 0.057147304651949,
        "unilateral_improvement_p0": 0.001128136281414,
        "unilateral_improvement_p1": 0.001591295438374,
        "nash_conv": 0.002719431719788,
        "exploitability": 0.001359715859894
      },
      "current": {
        "on_policy_value_p0": 0.075548719997046,
        "best_response_value_p0": 0.128586734438658,
        "best_response_value_p1": 0.22725730445011,
        "unilateral_improvement_p0": 0.053038014441612,
        "unilateral_improvement_p1": 0.302806024447156,
        "nash_conv": 0.355844038888768,
        "exploitability": 0.177922019444384
      }
    },
    {
      "iteration": 27500,
      "average": {
        "on_policy_value_p0": -0.055555223004721,
        "best_response_value_p0": -0.053873112252036,
        "best_response_value_p1": 0.056529540155828,
        "unilateral_improvement_p0": 0.001682110752685,
        "unilateral_improvement_p1": 0.000974317151107,
        "nash_conv": 0.002656427903792,
        "exploitability": 0.001328213951896
      },
      "current": {
        "on_policy_value_p0": 0.070544868188833,
        "best_response_value_p0": 0.070544868188833,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.403878201522167,
        "nash_conv": 0.403878201522167,
        "exploitability": 0.201939100761083
      }
    },
    {
      "iteration": 27600,
      "average": {
        "on_policy_value_p0": -0.055551338600193,
        "best_response_value_p0": -0.053544557749532,
        "best_response_value_p1": 0.05632472298135,
        "unilateral_improvement_p0": 0.002006780850662,
        "unilateral_improvement_p1": 0.000773384381156,
        "nash_conv": 0.002780165231818,
        "exploitability": 0.001390082615909
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 27700,
      "average": {
        "on_policy_value_p0": -0.055546115666245,
        "best_response_value_p0": -0.053416390300528,
        "best_response_value_p1": 0.056121384631236,
        "unilateral_improvement_p0": 0.002129725365717,
        "unilateral_improvement_p1": 0.000575268964991,
        "nash_conv": 0.002704994330708,
        "exploitability": 0.001352497165354
      },
      "current": {
        "on_policy_value_p0": -0.108276042169995,
        "best_response_value_p0": 0.112528645581669,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.220804687751664,
        "unilateral_improvement_p1": 0.225057291163338,
        "nash_conv": 0.445861978915002,
        "exploitability": 0.222930989457501
      }
    },
    {
      "iteration": 27800,
      "average": {
        "on_policy_value_p0": -0.055541963929574,
        "best_response_value_p0": -0.054033106574845,
        "best_response_value_p1": 0.056800766990702,
        "unilateral_improvement_p0": 0.001508857354729,
        "unilateral_improvement_p1": 0.001258803061128,
        "nash_conv": 0.002767660415857,
        "exploitability": 0.001383830207928
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 27900,
      "average": {
        "on_policy_value_p0": -0.055544674638353,
        "best_response_value_p0": -0.055034182656416,
        "best_response_value_p1": 0.05770554473924,
        "unilateral_improvement_p0": 0.000510491981937,
        "unilateral_improvement_p1": 0.002160870100887,
        "nash_conv": 0.002671362082824,
        "exploitability": 0.001335681041412
      },
      "current": {
        "on_policy_value_p0": -0.198222544760758,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.198222544760758,
        "unilateral_improvement_p0": 0.531555878094091,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.531555878094091,
        "exploitability": 0.265777939047045
      }
    },
    {
      "iteration": 28000,
      "average": {
        "on_policy_value_p0": -0.055550891159768,
        "best_response_value_p0": -0.054828295100743,
        "best_response_value_p1": 0.057551524065882,
        "unilateral_improvement_p0": 0.000722596059026,
        "unilateral_improvement_p1": 0.002000632906114,
        "nash_conv": 0.002723228965139,
        "exploitability": 0.00136161448257
      },
      "current": {
        "on_policy_value_p0": 0.10723689302375,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.226096440309584,
        "unilateral_improvement_p1": 0.273903559690416,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 28100,
      "average": {
        "on_policy_value_p0": -0.055552800937019,
        "best_response_value_p0": -0.054091973116378,
        "best_response_value_p1": 0.056683458512302,
        "unilateral_improvement_p0": 0.001460827820641,
        "unilateral_improvement_p1": 0.001130657575283,
        "nash_conv": 0.002591485395924,
        "exploitability": 0.001295742697962
      },
      "current": {
        "on_policy_value_p0": 0.142471770299111,
        "best_response_value_p0": 0.234054733388245,
        "best_response_value_p1": 0.316834073821731,
        "unilateral_improvement_p0": 0.091582963089134,
        "unilateral_improvement_p1": 0.459305844120842,
        "nash_conv": 0.550888807209976,
        "exploitability": 0.275444403604988
      }
    },
    {
      "iteration": 28200,
      "average": {
        "on_policy_value_p0": -0.055549480381451,
        "best_response_value_p0": -0.05355631059793,
        "best_response_value_p1": 0.056312701244556,
        "unilateral_improvement_p0": 0.001993169783522,
        "unilateral_improvement_p1": 0.000763220863105,
        "nash_conv": 0.002756390646626,
        "exploitability": 0.001378195323313
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 28300,
      "average": {
        "on_policy_value_p0": -0.055544737794921,
        "best_response_value_p0": -0.053766454322432,
        "best_response_value_p1": 0.056466458480777,
        "unilateral_improvement_p0": 0.001778283472489,
        "unilateral_improvement_p1": 0.000921720685856,
        "nash_conv": 0.002700004158345,
        "exploitability": 0.001350002079173
      },
      "current": {
        "on_policy_value_p0": -0.349819257991903,
        "best_response_value_p0": 0.116606419330634,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.466425677322537,
        "unilateral_improvement_p1": 0.150180742008097,
        "nash_conv": 0.616606419330634,
        "exploitability": 0.308303209665317
      }
    },
    {
      "iteration": 28400,
      "average": {
        "on_policy_value_p0": -0.055546512145259,
        "best_response_value_p0": -0.054866039881542,
        "best_response_value_p1": 0.05759333496991,
        "unilateral_improvement_p0": 0.000680472263718,
        "unilateral_improvement_p1": 0.002046822824651,
        "nash_conv": 0.002727295088368,
        "exploitability": 0.001363647544184
      },
      "current": {
        "on_policy_value_p0": -0.174271676462223,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.174271676462223,
        "unilateral_improvement_p0": 0.34093834312889,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.34093834312889,
        "exploitability": 0.170469171564445
      }
    },
    {
      "iteration": 28500,
      "average": {
        "on_policy_value_p0": -0.055551914077266,
        "best_response_value_p0": -0.054883210120168,
        "best_response_value_p1": 0.057619189776408,
        "unilateral_improvement_p0": 0.000668703957098,
        "unilateral_improvement_p1": 0.002067275699142,
        "nash_conv": 0.00273597965624,
        "exploitability": 0.00136798982812
      },
      "current": {
        "on_policy_value_p0": 0.025169952556586,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.141496714110081,
        "unilateral_improvement_p0": 0.141496714110081,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.308163380776747,
        "exploitability": 0.154081690388374
      }
    },
    {
      "iteration": 28600,
      "average": {
        "on_policy_value_p0": -0.055555827567082,
        "best_response_value_p0": -0.05452945581057,
        "best_response_value_p1": 0.057181546444545,
        "unilateral_improvement_p0": 0.001026371756512,
        "unilateral_improvement_p1": 0.001625718877464,
        "nash_conv": 0.002652090633976,
        "exploitability": 0.001326045316988
      },
      "current": {
        "on_policy_value_p0": 0.077593676627189,
        "best_response_value_p0": 0.140273781435945,
        "best_response_value_p1": 0.20797312371582,
        "unilateral_improvement_p0": 0.062680104808757,
        "unilateral_improvement_p1": 0.285566800343009,
        "nash_conv": 0.348246905151765,
        "exploitability": 0.174123452575883
      }
    },
    {
      "iteration": 28700,
      "average": {
        "on_policy_value_p0": -0.055555605812436,
        "best_response_value_p0": -0.053966519309496,
        "best_response_value_p1": 0.05652429423656,
        "unilateral_improvement_p0": 0.001589086502939,
        "unilateral_improvement_p1": 0.000968688424124,
        "nash_conv": 0.002557774927063,
        "exploitability": 0.001278887463532
      },
      "current": {
        "on_policy_value_p0": 0.079033587336081,
        "best_response_value_p0": 0.079033587336081,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.412366920669414,
        "nash_conv": 0.412366920669414,
        "exploitability": 0.206183460334707
      }
    },
    {
      "iteration": 28800,
      "average": {
        "on_policy_value_p0": -0.055552142816929,
        "best_response_value_p0": -0.053607343791399,
        "best_response_value_p1": 0.056328029326016,
        "unilateral_improvement_p0": 0.00194479902553,
        "unilateral_improvement_p1": 0.000775886509087,
        "nash_conv": 0.002720685534617,
        "exploitability": 0.001360342767309
      },
      "current": {
        "on_policy_value_p0": 0.009233084703987,
        "best_response_value_p0": 0.148200497258694,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.138967412554707,
        "unilateral_improvement_p1": 0.34256641803732,
        "nash_conv": 0.481533830592027,
        "exploitability": 0.240766915296013
      }
    },
    {
      "iteration": 28900,
      "average": {
        "on_policy_value_p0": -0.055547414262247,
        "best_response_value_p0": -0.053436028296392,
        "best_response_value_p1": 0.056133122650147,
        "unilateral_improvement_p0": 0.002111385965855,
        "unilateral_improvement_p1": 0.000585708387899,
        "nash_conv": 0.002697094353755,
        "exploitability": 0.001348547176877
      },
      "current": {
        "on_policy_value_p0": -0.061821896987048,
        "best_response_value_p0": 0.135755718173143,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.197577615160191,
        "unilateral_improvement_p1": 0.271511436346285,
        "nash_conv": 0.469089051506476,
        "exploitability": 0.234544525753238
      }
    },
    {
      "iteration": 29000,
      "average": {
        "on_policy_value_p0": -0.055542730899264,
        "best_response_value_p0": -0.053878523852627,
        "best_response_value_p1": 0.056588373007155,
        "unilateral_improvement_p0": 0.001664207046637,
        "unilateral_improvement_p1": 0.001045642107891,
        "nash_conv": 0.002709849154528,
        "exploitability": 0.001354924577264
      },
      "current": {
        "on_policy_value_p0": -0.322997485879197,
        "best_response_value_p0": 0.322997485879197,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.645994971758393,
        "unilateral_improvement_p1": 0.010335847454137,
        "nash_conv": 0.65633081921253,
        "exploitability": 0.328165409606265
      }
    },
    {
      "iteration": 29100,
      "average": {
        "on_policy_value_p0": -0.055543906752901,
        "best_response_value_p0": -0.054838321030954,
        "best_response_value_p1": 0.057523846127376,
        "unilateral_improvement_p0": 0.000705585721947,
        "unilateral_improvement_p1": 0.001979939374475,
        "nash_conv": 0.002685525096422,
        "exploitability": 0.001342762548211
      },
      "current": {
        "on_policy_value_p0": -0.263324807866123,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.263324807866123,
        "unilateral_improvement_p0": 0.596658141199456,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.596658141199456,
        "exploitability": 0.298329070599728
      }
    },
    {
      "iteration": 29200,
      "average": {
        "on_policy_value_p0": -0.055549713133351,
        "best_response_value_p0": -0.055005522393713,
        "best_response_value_p1": 0.057612724165912,
        "unilateral_improvement_p0": 0.000544190739638,
        "unilateral_improvement_p1": 0.002063011032561,
        "nash_conv": 0.0026072017722,
        "exploitability": 0.0013036008861
      },
      "current": {
        "on_policy_value_p0": 0.081290818830529,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.252042514502804,
        "unilateral_improvement_p1": 0.247957485497196,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 29300,
      "average": {
        "on_policy_value_p0": -0.055552725274005,
        "best_response_value_p0": -0.054252694852884,
        "best_response_value_p1": 0.056915034813605,
        "unilateral_improvement_p0": 0.001300030421121,
        "unilateral_improvement_p1": 0.001362309539599,
        "nash_conv": 0.00266233996072,
        "exploitability": 0.00133116998036
      },
      "current": {
        "on_policy_value_p0": 0.200315168319472,
        "best_response_value_p0": 0.317337352381659,
        "best_response_value_p1": 0.265955631875624,
        "unilateral_improvement_p0": 0.117022184062188,
        "unilateral_improvement_p1": 0.466270800195096,
        "nash_conv": 0.583292984257284,
        "exploitability": 0.291646492128642
      }
    },
    {
      "iteration": 29400,
      "average": {
        "on_policy_value_p0": -0.055550628103548,
        "best_response_value_p0": -0.053718047153497,
        "best_response_value_p1": 0.056370491849993,
        "unilateral_improvement_p0": 0.001832580950051,
        "unilateral_improvement_p1": 0.000819863746445,
        "nash_conv": 0.002652444696495,
        "exploitability": 0.001326222348248
      },
      "current": {
        "on_policy_value_p0": 0.016623948832887,
        "best_response_value_p0": 0.136234891577754,
        "best_response_value_p1": 0.411520797310122,
        "unilateral_improvement_p0": 0.119610942744867,
        "unilateral_improvement_p1": 0.428144746143009,
        "nash_conv": 0.547755688887876,
        "exploitability": 0.273877844443938
      }
    },
    {
      "iteration": 29500,
      "average": {
        "on_policy_value_p0": -0.055546604987821,
        "best_response_value_p0": -0.0535819704177,
        "best_response_value_p1": 0.056210052624459,
        "unilateral_improvement_p0": 0.001964634570121,
        "unilateral_improvement_p1": 0.000663447636638,
        "nash_conv": 0.002628082206759,
        "exploitability": 0.001314041103379
      },
      "current": {
        "on_policy_value_p0": -0.16402605575861,
        "best_response_value_p0": 0.057315962827593,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.221342018586203,
        "unilateral_improvement_p1": 0.33597394424139,
        "nash_conv": 0.557315962827593,
        "exploitability": 0.278657981413797
      }
    },
    {
      "iteration": 29600,
      "average": {
        "on_policy_value_p0": -0.055545260483605,
        "best_response_value_p0": -0.0547164520056,
        "best_response_value_p1": 0.057287910689616,
        "unilateral_improvement_p0": 0.000828808478005,
        "unilateral_improvement_p1": 0.001742650206011,
        "nash_conv": 0.002571458684016,
        "exploitability": 0.001285729342008
      },
      "current": {
        "on_policy_value_p0": -0.344432091278943,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.344432091278943,
        "unilateral_improvement_p0": 0.511098757945609,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.511098757945609,
        "exploitability": 0.255549378972805
      }
    },
    {
      "iteration": 29700,
      "average": {
        "on_policy_value_p0": -0.055550136746664,
        "best_response_value_p0": -0.054990839538596,
        "best_response_value_p1": 0.057644054751419,
        "unilateral_improvement_p0": 0.000559297208068,
        "unilateral_improvement_p1": 0.002093918004755,
        "nash_conv": 0.002653215212823,
        "exploitability": 0.001326607606411
      },
      "current": {
        "on_policy_value_p0": -0.054659504035157,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.221326170701824,
        "unilateral_improvement_p1": 0.112007162631509,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 29800,
      "average": {
        "on_policy_value_p0": -0.055554817387571,
        "best_response_value_p0": -0.054806306520011,
        "best_response_value_p1": 0.057357481857908,
        "unilateral_improvement_p0": 0.00074851086756,
        "unilateral_improvement_p1": 0.001802664470336,
        "nash_conv": 0.002551175337897,
        "exploitability": 0.001275587668948
      },
      "current": {
        "on_policy_value_p0": 0.079495079027085,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.15899015805417,
        "unilateral_improvement_p0": 0.087171587639582,
        "unilateral_improvement_p1": 0.238485237081255,
        "nash_conv": 0.325656824720837,
        "exploitability": 0.162828412360418
      }
    },
    {
      "iteration": 29900,
      "average": {
        "on_policy_value_p0": -0.055556348872248,
        "best_response_value_p0": -0.05420315289009,
        "best_response_value_p1": 0.056771836207789,
        "unilateral_improvement_p0": 0.001353195982159,
        "unilateral_improvement_p1": 0.00121548733554,
        "nash_conv": 0.002568683317699,
        "exploitability": 0.001284341658849
      },
      "current": {
        "on_policy_value_p0": 0.100808304841751,
        "best_response_value_p0": 0.100808304841751,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.434141638175084,
        "nash_conv": 0.434141638175084,
        "exploitability": 0.217070819087542
      }
    },
    {
      "iteration": 30000,
      "average": {
        "on_policy_value_p0": -0.05555409198573,
        "best_response_value_p0": -0.053771793561171,
        "best_response_value_p1": 0.056393180508202,
        "unilateral_improvement_p0": 0.001782298424559,
        "unilateral_improvement_p1": 0.000839088522472,
        "nash_conv": 0.002621386947031,
        "exploitability": 0.001310693473515
      },
      "current": {
        "on_policy_value_p0": 0.047398544001876,
        "best_response_value_p0": 0.071869578662915,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.02447103466104,
        "unilateral_improvement_p1": 0.380731877335209,
        "nash_conv": 0.405202911996249,
        "exploitability": 0.202601455998124
      }
    },
    {
      "iteration": 30100,
      "average": {
        "on_policy_value_p0": -0.055550096722945,
        "best_response_value_p0": -0.053546175461152,
        "best_response_value_p1": 0.056205827749038,
        "unilateral_improvement_p0": 0.002003921261793,
        "unilateral_improvement_p1": 0.000655731026093,
        "nash_conv": 0.002659652287887,
        "exploitability": 0.001329826143943
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 30200,
      "average": {
        "on_policy_value_p0": -0.055545112518737,
        "best_response_value_p0": -0.053610832452425,
        "best_response_value_p1": 0.05612206636965,
        "unilateral_improvement_p0": 0.001934280066312,
        "unilateral_improvement_p1": 0.000576953850912,
        "nash_conv": 0.002511233917225,
        "exploitability": 0.001255616958612
      },
      "current": {
        "on_policy_value_p0": -0.18519837009823,
        "best_response_value_p0": 0.18519837009823,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.370396740196461,
        "unilateral_improvement_p1": 0.148134963235103,
        "nash_conv": 0.518531703431564,
        "exploitability": 0.259265851715782
      }
    },
    {
      "iteration": 30300,
      "average": {
        "on_policy_value_p0": -0.055543146831908,
        "best_response_value_p0": -0.054387999145083,
        "best_response_value_p1": 0.057036955039498,
        "unilateral_improvement_p0": 0.001155147686825,
        "unilateral_improvement_p1": 0.00149380820759,
        "nash_conv": 0.002648955894415,
        "exploitability": 0.001324477947208
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 30400,
      "average": {
        "on_policy_value_p0": -0.055547141305199,
        "best_response_value_p0": -0.055131580387167,
        "best_response_value_p1": 0.057705970137038,
        "unilateral_improvement_p0": 0.000415560918033,
        "unilateral_improvement_p1": 0.002158828831839,
        "nash_conv": 0.002574389749871,
        "exploitability": 0.001287194874936
      },
      "current": {
        "on_policy_value_p0": -0.107968227196671,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.107968227196671,
        "unilateral_improvement_p0": 0.441301560530004,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.441301560530004,
        "exploitability": 0.220650780265002
      }
    },
    {
      "iteration": 30500,
      "average": {
        "on_policy_value_p0": -0.055552109119846,
        "best_response_value_p0": -0.054689810299214,
        "best_response_value_p1": 0.057331410341144,
        "unilateral_improvement_p0": 0.000862298820632,
        "unilateral_improvement_p1": 0.001779301221298,
        "nash_conv": 0.00264160004193,
        "exploitability": 0.001320800020965
      },
      "current": {
        "on_policy_value_p0": 0.136475605973219,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.196857727360115,
        "unilateral_improvement_p1": 0.303142272639885,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 30600,
      "average": {
        "on_policy_value_p0": -0.055552691082366,
        "best_response_value_p0": -0.05406626593458,
        "best_response_value_p1": 0.056453722921958,
        "unilateral_improvement_p0": 0.001486425147787,
        "unilateral_improvement_p1": 0.000901031839592,
        "nash_conv": 0.002387456987379,
        "exploitability": 0.001193728493689
      },
      "current": {
        "on_policy_value_p0": 0.113192781674194,
        "best_response_value_p0": 0.193758717834451,
        "best_response_value_p1": 0.338868127679487,
        "unilateral_improvement_p0": 0.080565936160257,
        "unilateral_improvement_p1": 0.452060909353681,
        "nash_conv": 0.532626845513938,
        "exploitability": 0.266313422756969
      }
    },
    {
      "iteration": 30700,
      "average": {
        "on_policy_value_p0": -0.055549401650267,
        "best_response_value_p0": -0.053616077973473,
        "best_response_value_p1": 0.056256260367088,
        "unilateral_improvement_p0": 0.001933323676794,
        "unilateral_improvement_p1": 0.000706858716821,
        "nash_conv": 0.002640182393615,
        "exploitability": 0.001320091196808
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 30800,
      "average": {
        "on_policy_value_p0": -0.055545334378125,
        "best_response_value_p0": -0.053939176206849,
        "best_response_value_p1": 0.056563775337824,
        "unilateral_improvement_p0": 0.001606158171276,
        "unilateral_improvement_p1": 0.001018440959699,
        "nash_conv": 0.002624599130975,
        "exploitability": 0.001312299565488
      },
      "current": {
        "on_policy_value_p0": -0.408879695533813,
        "best_response_value_p0": 0.136293231844604,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.545172927378418,
        "unilateral_improvement_p1": 0.091120304466187,
        "nash_conv": 0.636293231844604,
        "exploitability": 0.318146615922302
      }
    },
    {
      "iteration": 30900,
      "average": {
        "on_policy_value_p0": -0.055547430695589,
        "best_response_value_p0": -0.054907337284406,
        "best_response_value_p1": 0.057520388265274,
        "unilateral_improvement_p0": 0.000640093411184,
        "unilateral_improvement_p1": 0.001972957569684,
        "nash_conv": 0.002613050980868,
        "exploitability": 0.001306525490434
      },
      "current": {
        "on_policy_value_p0": -0.1390062729912,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.154778937331035,
        "unilateral_improvement_p0": 0.305672939657867,
        "unilateral_improvement_p1": 0.015772664339835,
        "nash_conv": 0.321445603997702,
        "exploitability": 0.160722801998851
      }
    },
    {
      "iteration": 31000,
      "average": {
        "on_policy_value_p0": -0.055552153768372,
        "best_response_value_p0": -0.054904415337432,
        "best_response_value_p1": 0.05752745057354,
        "unilateral_improvement_p0": 0.00064773843094,
        "unilateral_improvement_p1": 0.001975296805168,
        "nash_conv": 0.002623035236108,
        "exploitability": 0.001311517618054
      },
      "current": {
        "on_policy_value_p0": 0.023971707210342,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.142694959456325,
        "unilateral_improvement_p0": 0.142694959456325,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.309361626122991,
        "exploitability": 0.154680813061496
      }
    },
    {
      "iteration": 31100,
      "average": {
        "on_policy_value_p0": -0.055555632282014,
        "best_response_value_p0": -0.054599727716936,
        "best_response_value_p1": 0.057132474936721,
        "unilateral_improvement_p0": 0.000955904565078,
        "unilateral_improvement_p1": 0.001576842654707,
        "nash_conv": 0.002532747219785,
        "exploitability": 0.001266373609892
      },
      "current": {
        "on_policy_value_p0": 0.075382074004529,
        "best_response_value_p0": 0.141760298946367,
        "best_response_value_p1": 0.200576883449659,
        "unilateral_improvement_p0": 0.066378224941837,
        "unilateral_improvement_p1": 0.275958957454188,
        "nash_conv": 0.342337182396025,
        "exploitability": 0.171168591198013
      }
    },
    {
      "iteration": 31200,
      "average": {
        "on_policy_value_p0": -0.055555679503007,
        "best_response_value_p0": -0.054074800792152,
        "best_response_value_p1": 0.0565117649055,
        "unilateral_improvement_p0": 0.001480878710855,
        "unilateral_improvement_p1": 0.000956085402493,
        "nash_conv": 0.002436964113348,
        "exploitability": 0.001218482056674
      },
      "current": {
        "on_policy_value_p0": 0.081989671177962,
        "best_response_value_p0": 0.081989671177962,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.415323004511295,
        "nash_conv": 0.415323004511295,
        "exploitability": 0.207661502255648
      }
    },
    {
      "iteration": 31300,
      "average": {
        "on_policy_value_p0": -0.055552865820241,
        "best_response_value_p0": -0.053726762975382,
        "best_response_value_p1": 0.056331216135833,
        "unilateral_improvement_p0": 0.00182610284486,
        "unilateral_improvement_p1": 0.000778350315592,
        "nash_conv": 0.002604453160452,
        "exploitability": 0.001302226580226
      },
      "current": {
        "on_policy_value_p0": 0.020182592428268,
        "best_response_value_p0": 0.12630148181013,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.106118889381862,
        "unilateral_improvement_p1": 0.353515925761601,
        "nash_conv": 0.459634815143464,
        "exploitability": 0.229817407571732
      }
    },
    {
      "iteration": 31400,
      "average": {
        "on_policy_value_p0": -0.055548935407888,
        "best_response_value_p0": -0.053549338231233,
        "best_response_value_p1": 0.05615181735833,
        "unilateral_improvement_p0": 0.001999597176656,
        "unilateral_improvement_p1": 0.000602881950442,
        "nash_conv": 0.002602479127097,
        "exploitability": 0.001301239563549
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 31500,
      "average": {
        "on_policy_value_p0": -0.055544416226139,
        "best_response_value_p0": -0.053789833707312,
        "best_response_value_p1": 0.056347568220534,
        "unilateral_improvement_p0": 0.001754582518827,
        "unilateral_improvement_p1": 0.000803151994395,
        "nash_conv": 0.002557734513222,
        "exploitability": 0.001278867256611
      },
      "current": {
        "on_policy_value_p0": -0.241971520902766,
        "best_response_value_p0": 0.241971520902766,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.483943041805532,
        "unilateral_improvement_p1": 0.091361812430567,
        "nash_conv": 0.575304854236099,
        "exploitability": 0.28765242711805
      }
    },
    {
      "iteration": 31600,
      "average": {
        "on_policy_value_p0": -0.055544012871842,
        "best_response_value_p0": -0.054631523242047,
        "best_response_value_p1": 0.057224105451905,
        "unilateral_improvement_p0": 0.000912489629795,
        "unilateral_improvement_p1": 0.001680092580063,
        "nash_conv": 0.002592582209858,
        "exploitability": 0.001296291104929
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 31700,
      "average": {
        "on_policy_value_p0": -0.055548705974883,
        "best_response_value_p0": -0.055097866301281,
        "best_response_value_p1": 0.057657794995362,
        "unilateral_improvement_p0": 0.000450839673601,
        "unilateral_improvement_p1": 0.002109089020479,
        "nash_conv": 0.00255992869408,
        "exploitability": 0.00127996434704
      },
      "current": {
        "on_policy_value_p0": 0.009063398283642,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.14610078720808,
        "unilateral_improvement_p0": 0.324269935049691,
        "unilateral_improvement_p1": 0.155164185491722,
        "nash_conv": 0.479434120541413,
        "exploitability": 0.239717060270707
      }
    },
    {
      "iteration": 31800,
      "average": {
        "on_policy_value_p0": -0.055552679211188,
        "best_response_value_p0": -0.054561734007348,
        "best_response_value_p1": 0.057151256143468,
        "unilateral_improvement_p0": 0.00099094520384,
        "unilateral_improvement_p1": 0.001598576932281,
        "nash_conv": 0.002589522136121,
        "exploitability": 0.00129476106806
      },
      "current": {
        "on_policy_value_p0": 0.187322727885682,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.207978789104697,
        "unilateral_improvement_p0": 0.146010605447652,
        "unilateral_improvement_p1": 0.395301516990378,
        "nash_conv": 0.54131212243803,
        "exploitability": 0.270656061219015
      }
    },
    {
      "iteration": 31900,
      "average": {
        "on_policy_value_p0": -0.05555230454213,
        "best_response_value_p0": -0.053973744503992,
        "best_response_value_p1": 0.056317374955511,
        "unilateral_improvement_p0": 0.001578560038138,
        "unilateral_improvement_p1": 0.000765070413381,
        "nash_conv": 0.002343630451518,
        "exploitability": 0.001171815225759
      },
      "current": {
        "on_policy_value_p0": 0.085591032467967,
        "best_response_value_p0": 0.154752994422671,
        "best_response_value_p1": 0.361676076090591,
        "unilateral_improvement_p0": 0.069161961954705,
        "unilateral_improvement_p1": 0.447267108558557,
        "nash_conv": 0.516429070513262,
        "exploitability": 0.258214535256631
      }
    },
    {
      "iteration": 32000,
      "average": {
        "on_policy_value_p0": -0.055549042122812,
        "best_response_value_p0": -0.053646378579908,
        "best_response_value_p1": 0.05623124777285,
        "unilateral_improvement_p0": 0.001902663542904,
        "unilateral_improvement_p1": 0.000682205650038,
        "nash_conv": 0.002584869192942,
        "exploitability": 0.001292434596471
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 32100,
      "average": {
        "on_policy_value_p0": -0.055545584679077,
        "best_response_value_p0": -0.054134679779576,
        "best_response_value_p1": 0.056721182496542,
        "unilateral_improvement_p0": 0.0014109048995,
        "unilateral_improvement_p1": 0.001175597817465,
        "nash_conv": 0.002586502716965,
        "exploitability": 0.001293251358483
      },
      "current": {
        "on_policy_value_p0": -0.5,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 32200,
      "average": {
        "on_policy_value_p0": -0.055548305431238,
        "best_response_value_p0": -0.054976551941213,
        "best_response_value_p1": 0.057516657567723,
        "unilateral_improvement_p0": 0.000571753490026,
        "unilateral_improvement_p1": 0.001968352136484,
        "nash_conv": 0.00254010562651,
        "exploitability": 0.001270052813255
      },
      "current": {
        "on_policy_value_p0": -0.102831174572845,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.269497841239512,
        "unilateral_improvement_p1": 0.063835492093821,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 32300,
      "average": {
        "on_policy_value_p0": -0.055552840717515,
        "best_response_value_p0": -0.054917860944219,
        "best_response_value_p1": 0.057473161936976,
        "unilateral_improvement_p0": 0.000634979773296,
        "unilateral_improvement_p1": 0.001920321219461,
        "nash_conv": 0.002555300992757,
        "exploitability": 0.001277650496379
      },
      "current": {
        "on_policy_value_p0": 0.038615505420957,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.12805116124571,
        "unilateral_improvement_p0": 0.12805116124571,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.294717827912376,
        "exploitability": 0.147358913956188
      }
    },
    {
      "iteration": 32400,
      "average": {
        "on_policy_value_p0": -0.055555956347333,
        "best_response_value_p0": -0.054565607080143,
        "best_response_value_p1": 0.05706508597493,
        "unilateral_improvement_p0": 0.00099034926719,
        "unilateral_improvement_p1": 0.001509129627597,
        "nash_conv": 0.002499478894787,
        "exploitability": 0.001249739447393
      },
      "current": {
        "on_policy_value_p0": 0.079353238545474,
        "best_response_value_p0": 0.137405771742962,
        "best_response_value_p1": 0.217228266938359,
        "unilateral_improvement_p0": 0.058052533197487,
        "unilateral_improvement_p1": 0.296581505483833,
        "nash_conv": 0.35463403868132,
        "exploitability": 0.17731701934066
      }
    },
    {
      "iteration": 32500,
      "average": {
        "on_policy_value_p0": -0.05555572034687,
        "best_response_value_p0": -0.054069270511102,
        "best_response_value_p1": 0.056463947582777,
        "unilateral_improvement_p0": 0.001486449835769,
        "unilateral_improvement_p1": 0.000908227235906,
        "nash_conv": 0.002394677071675,
        "exploitability": 0.001197338535838
      },
      "current": {
        "on_policy_value_p0": 0.08051831852452,
        "best_response_value_p0": 0.08051831852452,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.413851651857853,
        "nash_conv": 0.413851651857853,
        "exploitability": 0.206925825928926
      }
    },
    {
      "iteration": 32600,
      "average": {
        "on_policy_value_p0": -0.055552894512453,
        "best_response_value_p0": -0.053738693851393,
        "best_response_value_p1": 0.056290745289577,
        "unilateral_improvement_p0": 0.00181420066106,
        "unilateral_improvement_p1": 0.000737850777124,
        "nash_conv": 0.002552051438184,
        "exploitability": 0.001276025719092
      },
      "current": {
        "on_policy_value_p0": 0.01924167478705,
        "best_response_value_p0": 0.128183317092567,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.108941642305517,
        "unilateral_improvement_p1": 0.352575008120383,
        "nash_conv": 0.4615166504259,
        "exploitability": 0.23075832521295
      }
    },
    {
      "iteration": 32700,
      "average": {
        "on_policy_value_p0": -0.055549058315639,
        "best_response_value_p0": -0.053568811237478,
        "best_response_value_p1": 0.05611860233762,
        "unilateral_improvement_p0": 0.001980247078161,
        "unilateral_improvement_p1": 0.000569544021982,
        "nash_conv": 0.002549791100142,
        "exploitability": 0.001274895550071
      },
      "current": {
        "on_policy_value_p0": -0.006259788959851,
        "best_response_value_p0": 0.163536772186741,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.169796561146592,
        "unilateral_improvement_p1": 0.327073544373482,
        "nash_conv": 0.496870105520074,
        "exploitability": 0.248435052760037
      }
    },
    {
      "iteration": 32800,
      "average": {
        "on_policy_value_p0": -0.055544774061142,
        "best_response_value_p0": -0.053835025256592,
        "best_response_value_p1": 0.05635096128622,
        "unilateral_improvement_p0": 0.00170974880455,
        "unilateral_improvement_p1": 0.000806187225078,
        "nash_conv": 0.002515936029628,
        "exploitability": 0.001257968014814
      },
      "current": {
        "on_policy_value_p0": -0.251363397102721,
        "best_response_value_p0": 0.251363397102721,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.502726794205441,
        "unilateral_improvement_p1": 0.081969936230613,
        "nash_conv": 0.584696730436054,
        "exploitability": 0.292348365218027
      }
    },
    {
      "iteration": 32900,
      "average": {
        "on_policy_value_p0": -0.055544465031148,
        "best_response_value_p0": -0.054652210644737,
        "best_response_value_p1": 0.057192852994571,
        "unilateral_improvement_p0": 0.000892254386411,
        "unilateral_improvement_p1": 0.001648387963423,
        "nash_conv": 0.002540642349834,
        "exploitability": 0.001270321174917
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 33000,
      "average": {
        "on_policy_value_p0": -0.055548843598488,
        "best_response_value_p0": -0.055097136698783,
        "best_response_value_p1": 0.057609177188098,
        "unilateral_improvement_p0": 0.000451706899705,
        "unilateral_improvement_p1": 0.00206033358961,
        "nash_conv": 0.002512040489316,
        "exploitability": 0.001256020244658
      },
      "current": {
        "on_policy_value_p0": 0.003468690247538,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.144895094160763,
        "unilateral_improvement_p0": 0.329864643085795,
        "unilateral_improvement_p1": 0.148363784408301,
        "nash_conv": 0.478228427494096,
        "exploitability": 0.239114213747048
      }
    },
    {
      "iteration": 33100,
      "average": {
        "on_policy_value_p0": -0.055552610909277,
        "best_response_value_p0": -0.054598783464851,
        "best_response_value_p1": 0.057140530109702,
        "unilateral_improvement_p0": 0.000953827444426,
        "unilateral_improvement_p1": 0.001587919200425,
        "nash_conv": 0.002541746644851,
        "exploitability": 0.001270873322426
      },
      "current": {
        "on_policy_value_p0": 0.17500266566695,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.183338664667234,
        "unilateral_improvement_p0": 0.158330667666383,
        "unilateral_improvement_p1": 0.358341330334185,
        "nash_conv": 0.516671998000568,
        "exploitability": 0.258335999000284
      }
    },
    {
      "iteration": 33200,
      "average": {
        "on_policy_value_p0": -0.055552448328327,
        "best_response_value_p0": -0.054061437632907,
        "best_response_value_p1": 0.056306707416686,
        "unilateral_improvement_p0": 0.00149101069542,
        "unilateral_improvement_p1": 0.000754259088358,
        "nash_conv": 0.002245269783778,
        "exploitability": 0.001122634891889
      },
      "current": {
        "on_policy_value_p0": 0.089324224086325,
        "best_response_value_p0": 0.165545064120662,
        "best_response_value_p1": 0.347558319931327,
        "unilateral_improvement_p0": 0.076220840034337,
        "unilateral_improvement_p1": 0.436882544017651,
        "nash_conv": 0.513103384051988,
        "exploitability": 0.256551692025994
      }
    },
    {
      "iteration": 33300,
      "average": {
        "on_policy_value_p0": -0.055549561666645,
        "best_response_value_p0": -0.053715960928541,
        "best_response_value_p1": 0.056249810260623,
        "unilateral_improvement_p0": 0.001833600738104,
        "unilateral_improvement_p1": 0.000700248593979,
        "nash_conv": 0.002533849332082,
        "exploitability": 0.001266924666041
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 33400,
      "average": {
        "on_policy_value_p0": -0.055546094275571,
        "best_response_value_p0": -0.05401563739604,
        "best_response_value_p1": 0.056530549836551,
        "unilateral_improvement_p0": 0.001530456879531,
        "unilateral_improvement_p1": 0.00098445556098,
        "nash_conv": 0.002514912440511,
        "exploitability": 0.001257456220255
      },
      "current": {
        "on_policy_value_p0": -0.398917704912397,
        "best_response_value_p0": 0.132972568304132,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.531890273216529,
        "unilateral_improvement_p1": 0.101082295087603,
        "nash_conv": 0.632972568304132,
        "exploitability": 0.316486284152066
      }
    },
    {
      "iteration": 33500,
      "average": {
        "on_policy_value_p0": -0.055547930179875,
        "best_response_value_p0": -0.054941599539861,
        "best_response_value_p1": 0.057447517906032,
        "unilateral_improvement_p0": 0.000606330640015,
        "unilateral_improvement_p1": 0.001899587726157,
        "nash_conv": 0.002505918366171,
        "exploitability": 0.001252959183086
      },
      "current": {
        "on_policy_value_p0": -0.169873717718117,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.169873717718117,
        "unilateral_improvement_p0": 0.336540384384783,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.336540384384783,
        "exploitability": 0.168270192192392
      }
    },
    {
      "iteration": 33600,
      "average": {
        "on_policy_value_p0": -0.055552238256892,
        "best_response_value_p0": -0.054959458732651,
        "best_response_value_p1": 0.057479392624251,
        "unilateral_improvement_p0": 0.000592779524241,
        "unilateral_improvement_p1": 0.001927154367359,
        "nash_conv": 0.0025199338916,
        "exploitability": 0.0012599669458
      },
      "current": {
        "on_policy_value_p0": 0.021884227004763,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.144782439661904,
        "unilateral_improvement_p0": 0.144782439661904,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.311449106328571,
        "exploitability": 0.155724553164285
      }
    },
    {
      "iteration": 33700,
      "average": {
        "on_policy_value_p0": -0.055555573634699,
        "best_response_value_p0": -0.054696069598735,
        "best_response_value_p1": 0.057122967827244,
        "unilateral_improvement_p0": 0.000859504035964,
        "unilateral_improvement_p1": 0.001567394192545,
        "nash_conv": 0.002426898228508,
        "exploitability": 0.001213449114254
      },
      "current": {
        "on_policy_value_p0": 0.081918989424174,
        "best_response_value_p0": 0.152144108803381,
        "best_response_value_p1": 0.19288309457492,
        "unilateral_improvement_p0": 0.070225119379207,
        "unilateral_improvement_p1": 0.274802083999094,
        "nash_conv": 0.345027203378301,
        "exploitability": 0.17251360168915
      }
    },
    {
      "iteration": 33800,
      "average": {
        "on_policy_value_p0": -0.055555987059705,
        "best_response_value_p0": -0.054188348931828,
        "best_response_value_p1": 0.05655156484665,
        "unilateral_improvement_p0": 0.001367638127877,
        "unilateral_improvement_p1": 0.000995577786945,
        "nash_conv": 0.002363215914822,
        "exploitability": 0.001181607957411
      },
      "current": {
        "on_policy_value_p0": 0.089162986881241,
        "best_response_value_p0": 0.089162986881241,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.422496320214575,
        "nash_conv": 0.422496320214575,
        "exploitability": 0.211248160107287
      }
    },
    {
      "iteration": 33900,
      "average": {
        "on_policy_value_p0": -0.055553737690528,
        "best_response_value_p0": -0.053837953099862,
        "best_response_value_p1": 0.05632409673067,
        "unilateral_improvement_p0": 0.001715784590666,
        "unilateral_improvement_p1": 0.000770359040142,
        "nash_conv": 0.002486143630808,
        "exploitability": 0.001243071815404
      },
      "current": {
        "on_policy_value_p0": 0.035851715968803,
        "best_response_value_p0": 0.09496323472906,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.059111518760257,
        "unilateral_improvement_p1": 0.369185049302137,
        "nash_conv": 0.428296568062393,
        "exploitability": 0.214148284031197
      }
    },
    {
      "iteration": 34000,
      "average": {
        "on_policy_value_p0": -0.055550360897211,
        "best_response_value_p0": -0.05365718764837,
        "best_response_value_p1": 0.056158437622638,
        "unilateral_improvement_p0": 0.001893173248842,
        "unilateral_improvement_p1": 0.000608076725426,
        "nash_conv": 0.002501249974268,
        "exploitability": 0.001250624987134
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 34100,
      "average": {
        "on_policy_value_p0": -0.055546208445716,
        "best_response_value_p0": -0.053738618859132,
        "best_response_value_p1": 0.056109382734188,
        "unilateral_improvement_p0": 0.001807589586584,
        "unilateral_improvement_p1": 0.000563174288471,
        "nash_conv": 0.002370763875055,
        "exploitability": 0.001185381937528
      },
      "current": {
        "on_policy_value_p0": -0.18896240924466,
        "best_response_value_p0": 0.18896240924466,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.37792481848932,
        "unilateral_improvement_p1": 0.144370924088674,
        "nash_conv": 0.522295742577993,
        "exploitability": 0.261147871288997
      }
    },
    {
      "iteration": 34200,
      "average": {
        "on_policy_value_p0": -0.055544534051719,
        "best_response_value_p0": -0.054427281882963,
        "best_response_value_p1": 0.05691997908097,
        "unilateral_improvement_p0": 0.001117252168755,
        "unilateral_improvement_p1": 0.001375445029251,
        "nash_conv": 0.002492697198006,
        "exploitability": 0.001246348599003
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 34300,
      "average": {
        "on_policy_value_p0": -0.055547614883752,
        "best_response_value_p0": -0.055178142135021,
        "best_response_value_p1": 0.057560522501465,
        "unilateral_improvement_p0": 0.000369472748732,
        "unilateral_improvement_p1": 0.002012907617712,
        "nash_conv": 0.002382380366444,
        "exploitability": 0.001191190183222
      },
      "current": {
        "on_policy_value_p0": -0.15079637521767,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.15079637521767,
        "unilateral_improvement_p0": 0.484129708551003,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.484129708551003,
        "exploitability": 0.242064854275502
      }
    },
    {
      "iteration": 34400,
      "average": {
        "on_policy_value_p0": -0.055552058007272,
        "best_response_value_p0": -0.054849077254827,
        "best_response_value_p1": 0.057316659821653,
        "unilateral_improvement_p0": 0.000702980752445,
        "unilateral_improvement_p1": 0.001764601814381,
        "nash_conv": 0.002467582566826,
        "exploitability": 0.001233791283413
      },
      "current": {
        "on_policy_value_p0": 0.116021203343187,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.217312129990147,
        "unilateral_improvement_p1": 0.282687870009853,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 34500,
      "average": {
        "on_policy_value_p0": -0.05555336952574,
        "best_response_value_p0": -0.054245498673918,
        "best_response_value_p1": 0.056595699651226,
        "unilateral_improvement_p0": 0.001307870851823,
        "unilateral_improvement_p1": 0.001042330125486,
        "nash_conv": 0.002350200977308,
        "exploitability": 0.001175100488654
      },
      "current": {
        "on_policy_value_p0": 0.148504478336886,
        "best_response_value_p0": 0.241553025706993,
        "best_response_value_p1": 0.313902905259785,
        "unilateral_improvement_p0": 0.093048547370107,
        "unilateral_improvement_p1": 0.462407383596671,
        "nash_conv": 0.555455930966778,
        "exploitability": 0.277727965483389
      }
    },
    {
      "iteration": 34600,
      "average": {
        "on_policy_value_p0": -0.055551032658426,
        "best_response_value_p0": -0.053767865696904,
        "best_response_value_p1": 0.056251043854881,
        "unilateral_improvement_p0": 0.001783166961522,
        "unilateral_improvement_p1": 0.000700011196456,
        "nash_conv": 0.002483178157978,
        "exploitability": 0.001241589078989
      },
      "current": {
        "on_policy_value_p0": -0.013535736758783,
        "best_response_value_p0": 0.164686567685874,
        "best_response_value_p1": 0.465008130559264,
        "unilateral_improvement_p0": 0.178222304444657,
        "unilateral_improvement_p1": 0.451472393800481,
        "nash_conv": 0.629694698245138,
        "exploitability": 0.314847349122569
      }
    },
    {
      "iteration": 34700,
      "average": {
        "on_policy_value_p0": -0.055547470143022,
        "best_response_value_p0": -0.053776680668762,
        "best_response_value_p1": 0.056092348015234,
        "unilateral_improvement_p0": 0.00177078947426,
        "unilateral_improvement_p1": 0.000544877872212,
        "nash_conv": 0.002315667346472,
        "exploitability": 0.001157833673236
      },
      "current": {
        "on_policy_value_p0": -0.244481608491602,
        "best_response_value_p0": 0.081493869497201,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.325975477988802,
        "unilateral_improvement_p1": 0.255518391508398,
        "nash_conv": 0.581493869497201,
        "exploitability": 0.2907469347486
      }
    },
    {
      "iteration": 34800,
      "average": {
        "on_policy_value_p0": -0.055547007567761,
        "best_response_value_p0": -0.054817943978202,
        "best_response_value_p1": 0.05719186227782,
        "unilateral_improvement_p0": 0.00072906358956,
        "unilateral_improvement_p1": 0.001644854710059,
        "nash_conv": 0.002373918299618,
        "exploitability": 0.001186959149809
      },
      "current": {
        "on_policy_value_p0": -0.321271468114679,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.321271468114679,
        "unilateral_improvement_p0": 0.487938134781346,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.487938134781346,
        "exploitability": 0.243969067390673
      }
    },
    {
      "iteration": 34900,
      "average": {
        "on_policy_value_p0": -0.055550763045533,
        "best_response_value_p0": -0.055015120321615,
        "best_response_value_p1": 0.057456848341043,
        "unilateral_improvement_p0": 0.000535642723917,
        "unilateral_improvement_p1": 0.00190608529551,
        "nash_conv": 0.002441728019427,
        "exploitability": 0.001220864009714
      },
      "current": {
        "on_policy_value_p0": -0.05743902718289,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.224105693849557,
        "unilateral_improvement_p1": 0.109227639483776,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 35000,
      "average": {
        "on_policy_value_p0": -0.055554402195232,
        "best_response_value_p0": -0.054857934263554,
        "best_response_value_p1": 0.057250219115845,
        "unilateral_improvement_p0": 0.000696467931678,
        "unilateral_improvement_p1": 0.001695816920613,
        "nash_conv": 0.002392284852291,
        "exploitability": 0.001196142426145
      },
      "current": {
        "on_policy_value_p0": 0.067718411845316,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.135436823690631,
        "unilateral_improvement_p0": 0.098948254821351,
        "unilateral_improvement_p1": 0.203155235535947,
        "nash_conv": 0.302103490357298,
        "exploitability": 0.151051745178649
      }
    },
    {
      "iteration": 35100,
      "average": {
        "on_policy_value_p0": -0.055556081902999,
        "best_response_value_p0": -0.054409884510943,
        "best_response_value_p1": 0.05680316829533,
        "unilateral_improvement_p0": 0.001146197392057,
        "unilateral_improvement_p1": 0.00124708639233,
        "nash_conv": 0.002393283784387,
        "exploitability": 0.001196641892194
      },
      "current": {
        "on_policy_value_p0": 0.082013254958786,
        "best_response_value_p0": 0.110800889258494,
        "best_response_value_p1": 0.275758064733916,
        "unilateral_improvement_p0": 0.028787634299709,
        "unilateral_improvement_p1": 0.357771319692702,
        "nash_conv": 0.38655895399241,
        "exploitability": 0.193279476996205
      }
    },
    {
      "iteration": 35200,
      "average": {
        "on_policy_value_p0": -0.055554905951349,
        "best_response_value_p0": -0.054009111826483,
        "best_response_value_p1": 0.056384857769401,
        "unilateral_improvement_p0": 0.001545794124866,
        "unilateral_improvement_p1": 0.000829951818052,
        "nash_conv": 0.002375745942918,
        "exploitability": 0.001187872971459
      },
      "current": {
        "on_policy_value_p0": 0.062584444895501,
        "best_response_value_p0": 0.062584444895501,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.395917778228834,
        "nash_conv": 0.395917778228834,
        "exploitability": 0.197958889114417
      }
    },
    {
      "iteration": 35300,
      "average": {
        "on_policy_value_p0": -0.055552083209035,
        "best_response_value_p0": -0.053768900641618,
        "best_response_value_p1": 0.056225127294133,
        "unilateral_improvement_p0": 0.001783182567417,
        "unilateral_improvement_p1": 0.000673044085098,
        "nash_conv": 0.002456226652515,
        "exploitability": 0.001228113326257
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 35400,
      "average": {
        "on_policy_value_p0": -0.05554847510792,
        "best_response_value_p0": -0.053651706961816,
        "best_response_value_p1": 0.056066299250928,
        "unilateral_improvement_p0": 0.001896768146104,
        "unilateral_improvement_p1": 0.000517824143008,
        "nash_conv": 0.002414592289112,
        "exploitability": 0.001207296144556
      },
      "current": {
        "on_policy_value_p0": -0.088652650244855,
        "best_response_value_p0": 0.122340341544239,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.210992991789094,
        "unilateral_improvement_p1": 0.244680683088478,
        "nash_conv": 0.455673674877572,
        "exploitability": 0.227836837438786
      }
    },
    {
      "iteration": 35500,
      "average": {
        "on_policy_value_p0": -0.055545094493383,
        "best_response_value_p0": -0.05403897375934,
        "best_response_value_p1": 0.056486355878453,
        "unilateral_improvement_p0": 0.001506120734044,
        "unilateral_improvement_p1": 0.00094126138507,
        "nash_conv": 0.002447382119114,
        "exploitability": 0.001223691059557
      },
      "current": {
        "on_policy_value_p0": -0.318765072839757,
        "best_response_value_p0": 0.318765072839757,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.637530145679513,
        "unilateral_improvement_p1": 0.014568260493577,
        "nash_conv": 0.65209840617309,
        "exploitability": 0.326049203086545
      }
    },
    {
      "iteration": 35600,
      "average": {
        "on_policy_value_p0": -0.055545700731031,
        "best_response_value_p0": -0.054822667786138,
        "best_response_value_p1": 0.057263695765437,
        "unilateral_improvement_p0": 0.000723032944893,
        "unilateral_improvement_p1": 0.001717995034406,
        "nash_conv": 0.002441027979299,
        "exploitability": 0.001220513989649
      },
      "current": {
        "on_policy_value_p0": -0.318098649259426,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.318098649259426,
        "unilateral_improvement_p0": 0.65143198259276,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.65143198259276,
        "exploitability": 0.32571599129638
      }
    },
    {
      "iteration": 35700,
      "average": {
        "on_policy_value_p0": -0.055549932468478,
        "best_response_value_p0": -0.055099010080137,
        "best_response_value_p1": 0.057498381251098,
        "unilateral_improvement_p0": 0.000450922388341,
        "unilateral_improvement_p1": 0.001948448782619,
        "nash_conv": 0.00239937117096,
        "exploitability": 0.00119968558548
      },
      "current": {
        "on_policy_value_p0": 0.06157584597929,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.271757487354044,
        "unilateral_improvement_p1": 0.228242512645956,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 35800,
      "average": {
        "on_policy_value_p0": -0.055553000133125,
        "best_response_value_p0": -0.054581555025256,
        "best_response_value_p1": 0.0570196154789,
        "unilateral_improvement_p0": 0.000971445107869,
        "unilateral_improvement_p1": 0.001466615345775,
        "nash_conv": 0.002438060453644,
        "exploitability": 0.001219030226822
      },
      "current": {
        "on_policy_value_p0": 0.189584544866686,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.212502423066705,
        "unilateral_improvement_p0": 0.143748788466647,
        "unilateral_improvement_p1": 0.402086967933391,
        "nash_conv": 0.545835756400038,
        "exploitability": 0.272917878200019
      }
    },
    {
      "iteration": 35900,
      "average": {
        "on_policy_value_p0": -0.0555525984958,
        "best_response_value_p0": -0.054083551697918,
        "best_response_value_p1": 0.056282957288826,
        "unilateral_improvement_p0": 0.001469046797883,
        "unilateral_improvement_p1": 0.000730358793025,
        "nash_conv": 0.002199405590908,
        "exploitability": 0.001099702795454
      },
      "current": {
        "on_policy_value_p0": 0.083211783609012,
        "best_response_value_p0": 0.154967077640884,
        "best_response_value_p1": 0.356489411936256,
        "unilateral_improvement_p0": 0.071755294031872,
        "unilateral_improvement_p1": 0.439701195545268,
        "nash_conv": 0.51145648957714,
        "exploitability": 0.25572824478857
      }
    },
    {
      "iteration": 36000,
      "average": {
        "on_policy_value_p0": -0.055549985794658,
        "best_response_value_p0": -0.053783295584505,
        "best_response_value_p1": 0.056219453032161,
        "unilateral_improvement_p0": 0.001766690210152,
        "unilateral_improvement_p1": 0.000669467237504,
        "nash_conv": 0.002436157447656,
        "exploitability": 0.001218078723828
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 36100,
      "average": {
        "on_policy_value_p0": -0.055546869825702,
        "best_response_value_p0": -0.05405615555222,
        "best_response_value_p1": 0.05646821289455,
        "unilateral_improvement_p0": 0.001490714273482,
        "unilateral_improvement_p1": 0.000921343068848,
        "nash_conv": 0.00241205734233,
        "exploitability": 0.001206028671165
      },
      "current": {
        "on_policy_value_p0": -0.386452432485366,
        "best_response_value_p0": 0.128817477495122,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.515269909980488,
        "unilateral_improvement_p1": 0.113547567514634,
        "nash_conv": 0.628817477495122,
        "exploitability": 0.314408738747561
      }
    },
    {
      "iteration": 36200,
      "average": {
        "on_policy_value_p0": -0.055548319560873,
        "best_response_value_p0": -0.054952714565854,
        "best_response_value_p1": 0.057357414504921,
        "unilateral_improvement_p0": 0.000595604995019,
        "unilateral_improvement_p1": 0.001809094944048,
        "nash_conv": 0.002404699939068,
        "exploitability": 0.001202349969534
      },
      "current": {
        "on_policy_value_p0": -0.201945381227252,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.201945381227252,
        "unilateral_improvement_p0": 0.368612047893919,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.368612047893919,
        "exploitability": 0.18430602394696
      }
    },
    {
      "iteration": 36300,
      "average": {
        "on_policy_value_p0": -0.055552184430957,
        "best_response_value_p0": -0.055001979734934,
        "best_response_value_p1": 0.057426873485746,
        "unilateral_improvement_p0": 0.000550204696024,
        "unilateral_improvement_p1": 0.001874689054789,
        "nash_conv": 0.002424893750813,
        "exploitability": 0.001212446875406
      },
      "current": {
        "on_policy_value_p0": -0.006829189263361,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.155095335307788,
        "unilateral_improvement_p0": 0.173495855930028,
        "unilateral_improvement_p1": 0.148266146044427,
        "nash_conv": 0.321762001974455,
        "exploitability": 0.160881000987227
      }
    },
    {
      "iteration": 36400,
      "average": {
        "on_policy_value_p0": -0.055555365341478,
        "best_response_value_p0": -0.054807361217052,
        "best_response_value_p1": 0.057119899977371,
        "unilateral_improvement_p0": 0.000748004124426,
        "unilateral_improvement_p1": 0.001564534635893,
        "nash_conv": 0.002312538760319,
        "exploitability": 0.00115626938016
      },
      "current": {
        "on_policy_value_p0": 0.089039751833197,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.178079503666394,
        "unilateral_improvement_p0": 0.07762691483347,
        "unilateral_improvement_p1": 0.26711925549959,
        "nash_conv": 0.34474617033306,
        "exploitability": 0.17237308516653
      }
    },
    {
      "iteration": 36500,
      "average": {
        "on_policy_value_p0": -0.055556170775734,
        "best_response_value_p0": -0.054306126775918,
        "best_response_value_p1": 0.056618697288228,
        "unilateral_improvement_p0": 0.001250043999816,
        "unilateral_improvement_p1": 0.001062526512495,
        "nash_conv": 0.00231257051231,
        "exploitability": 0.001156285256155
      },
      "current": {
        "on_policy_value_p0": 0.098184969465746,
        "best_response_value_p0": 0.098184969465746,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.43151830279908,
        "nash_conv": 0.43151830279908,
        "exploitability": 0.21575915139954
      }
    },
    {
      "iteration": 36600,
      "average": {
        "on_policy_value_p0": -0.055554463716659,
        "best_response_value_p0": -0.053952364274834,
        "best_response_value_p1": 0.056317181125757,
        "unilateral_improvement_p0": 0.001602099441824,
        "unilateral_improvement_p1": 0.000762717409099,
        "nash_conv": 0.002364816850923,
        "exploitability": 0.001182408425461
      },
      "current": {
        "on_policy_value_p0": 0.050334971347135,
        "best_response_value_p0": 0.065996723972397,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.015661752625262,
        "unilateral_improvement_p1": 0.383668304680468,
        "nash_conv": 0.39933005730573,
        "exploitability": 0.199665028652865
      }
    },
    {
      "iteration": 36700,
      "average": {
        "on_policy_value_p0": -0.055551544558092,
        "best_response_value_p0": -0.053755961569299,
        "best_response_value_p1": 0.056163728316149,
        "unilateral_improvement_p0": 0.001795582988793,
        "unilateral_improvement_p1": 0.000612183758058,
        "nash_conv": 0.002407766746851,
        "exploitability": 0.001203883373425
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 36800,
      "average": {
        "on_policy_value_p0": -0.055547883606285,
        "best_response_value_p0": -0.053720699783597,
        "best_response_value_p1": 0.056011109489202,
        "unilateral_improvement_p0": 0.001827183822689,
        "unilateral_improvement_p1": 0.000463225882917,
        "nash_conv": 0.002290409705606,
        "exploitability": 0.001145204852803
      },
      "current": {
        "on_policy_value_p0": -0.140201168115488,
        "best_response_value_p0": 0.140201168115488,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.280402336230977,
        "unilateral_improvement_p1": 0.193132165217845,
        "nash_conv": 0.473534501448822,
        "exploitability": 0.236767250724411
      }
    },
    {
      "iteration": 36900,
      "average": {
        "on_policy_value_p0": -0.055545308976765,
        "best_response_value_p0": -0.054231136726433,
        "best_response_value_p1": 0.056631484426358,
        "unilateral_improvement_p0": 0.001314172250332,
        "unilateral_improvement_p1": 0.001086175449593,
        "nash_conv": 0.002400347699925,
        "exploitability": 0.001200173849962
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 37000,
      "average": {
        "on_policy_value_p0": -0.055546746333492,
        "best_response_value_p0": -0.054985466987533,
        "best_response_value_p1": 0.057348658955692,
        "unilateral_improvement_p0": 0.00056127934596,
        "unilateral_improvement_p1": 0.0018019126222,
        "nash_conv": 0.002363191968159,
        "exploitability": 0.00118159598408
      },
      "current": {
        "on_policy_value_p0": -0.237782370252452,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.237782370252452,
        "unilateral_improvement_p0": 0.571115703585785,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.571115703585785,
        "exploitability": 0.285557851792893
      }
    },
    {
      "iteration": 37100,
      "average": {
        "on_policy_value_p0": -0.055550904857407,
        "best_response_value_p0": -0.055084055557899,
        "best_response_value_p1": 0.057403067766971,
        "unilateral_improvement_p0": 0.000466849299508,
        "unilateral_improvement_p1": 0.001852162909564,
        "nash_conv": 0.002319012209072,
        "exploitability": 0.001159506104536
      },
      "current": {
        "on_policy_value_p0": 0.078203658596506,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.255129674736827,
        "unilateral_improvement_p1": 0.244870325263173,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 37200,
      "average": {
        "on_policy_value_p0": -0.055553363602421,
        "best_response_value_p0": -0.054507252105874,
        "best_response_value_p1": 0.056883698980782,
        "unilateral_improvement_p0": 0.001046111496547,
        "unilateral_improvement_p1": 0.001330335378361,
        "nash_conv": 0.002376446874908,
        "exploitability": 0.001188223437454
      },
      "current": {
        "on_policy_value_p0": 0.206318670029389,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.245970673392112,
        "unilateral_improvement_p0": 0.127014663303944,
        "unilateral_improvement_p1": 0.452289343421501,
        "nash_conv": 0.579304006725445,
        "exploitability": 0.289652003362722
      }
    },
    {
      "iteration": 37300,
      "average": {
        "on_policy_value_p0": -0.055552482201889,
        "best_response_value_p0": -0.054025896571124,
        "best_response_value_p1": 0.056259319600273,
        "unilateral_improvement_p0": 0.001526585630765,
        "unilateral_improvement_p1": 0.000706837398384,
        "nash_conv": 0.002233423029148,
        "exploitability": 0.001116711514574
      },
      "current": {
        "on_policy_value_p0": 0.06963816348285,
        "best_response_value_p0": 0.131822110619247,
        "best_response_value_p1": 0.375632105727206,
        "unilateral_improvement_p0": 0.062183947136397,
        "unilateral_improvement_p1": 0.445270269210056,
        "nash_conv": 0.507454216346453,
        "exploitability": 0.253727108173227
      }
    },
    {
      "iteration": 37400,
      "average": {
        "on_policy_value_p0": -0.055549813338256,
        "best_response_value_p0": -0.053783009857663,
        "best_response_value_p1": 0.056172410171002,
        "unilateral_improvement_p0": 0.001766803480593,
        "unilateral_improvement_p1": 0.000622596832746,
        "nash_conv": 0.002389400313339,
        "exploitability": 0.001194700156669
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 37500,
      "average": {
        "on_policy_value_p0": -0.05554701386129,
        "best_response_value_p0": -0.054200769213492,
        "best_response_value_p1": 0.056591618343096,
        "unilateral_improvement_p0": 0.001346244647798,
        "unilateral_improvement_p1": 0.001044604481806,
        "nash_conv": 0.002390849129604,
        "exploitability": 0.001195424564802
      },
      "current": {
        "on_policy_value_p0": -0.485132470155011,
        "best_response_value_p0": 0.161710823385004,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.646843293540014,
        "unilateral_improvement_p1": 0.014867529844989,
        "nash_conv": 0.661710823385004,
        "exploitability": 0.330855411692502
      }
    },
    {
      "iteration": 37600,
      "average": {
        "on_policy_value_p0": -0.055548887891665,
        "best_response_value_p0": -0.054978291744977,
        "best_response_value_p1": 0.05734300750237,
        "unilateral_improvement_p0": 0.000570596146688,
        "unilateral_improvement_p1": 0.001794119610706,
        "nash_conv": 0.002364715757394,
        "exploitability": 0.001182357878697
      },
      "current": {
        "on_policy_value_p0": -0.142742479374409,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.152668956686197,
        "unilateral_improvement_p0": 0.309409146041075,
        "unilateral_improvement_p1": 0.009926477311788,
        "nash_conv": 0.319335623352863,
        "exploitability": 0.159667811676432
      }
    },
    {
      "iteration": 37700,
      "average": {
        "on_policy_value_p0": -0.055552463465523,
        "best_response_value_p0": -0.054987632406108,
        "best_response_value_p1": 0.057368014485402,
        "unilateral_improvement_p0": 0.000564831059415,
        "unilateral_improvement_p1": 0.00181555101988,
        "nash_conv": 0.002380382079294,
        "exploitability": 0.001190191039647
      },
      "current": {
        "on_policy_value_p0": 0.011931232037856,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.151426583346925,
        "unilateral_improvement_p0": 0.15473543462881,
        "unilateral_improvement_p1": 0.163357815384781,
        "nash_conv": 0.318093250013592,
        "exploitability": 0.159046625006796
      }
    },
    {
      "iteration": 37800,
      "average": {
        "on_policy_value_p0": -0.055555332021738,
        "best_response_value_p0": -0.054795952997704,
        "best_response_value_p1": 0.057066250893807,
        "unilateral_improvement_p0": 0.000759379024034,
        "unilateral_improvement_p1": 0.001510918872069,
        "nash_conv": 0.002270297896103,
        "exploitability": 0.001135148948052
      },
      "current": {
        "on_policy_value_p0": 0.083279390787889,
        "best_response_value_p0": 0.160385902090261,
        "best_response_value_p1": 0.17912031072859,
        "unilateral_improvement_p0": 0.077106511302372,
        "unilateral_improvement_p1": 0.26239970151648,
        "nash_conv": 0.339506212818851,
        "exploitability": 0.169753106409426
      }
    },
    {
      "iteration": 37900,
      "average": {
        "on_policy_value_p0": -0.055555996629802,
        "best_response_value_p0": -0.054325767933489,
        "best_response_value_p1": 0.056583927874244,
        "unilateral_improvement_p0": 0.001230228696313,
        "unilateral_improvement_p1": 0.001027931244442,
        "nash_conv": 0.002258159940755,
        "exploitability": 0.001129079970377
      },
      "current": {
        "on_policy_value_p0": 0.095440932978973,
        "best_response_value_p0": 0.095440932978973,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.428774266312307,
        "nash_conv": 0.428774266312307,
        "exploitability": 0.214387133156153
      }
    },
    {
      "iteration": 38000,
      "average": {
        "on_policy_value_p0": -0.055554401504844,
        "best_response_value_p0": -0.053991120272647,
        "best_response_value_p1": 0.056316901562451,
        "unilateral_improvement_p0": 0.001563281232196,
        "unilateral_improvement_p1": 0.000762500057608,
        "nash_conv": 0.002325781289804,
        "exploitability": 0.001162890644902
      },
      "current": {
        "on_policy_value_p0": 0.048090104423289,
        "best_response_value_p0": 0.070486457820088,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.022396353396798,
        "unilateral_improvement_p1": 0.381423437756623,
        "nash_conv": 0.403819791153421,
        "exploitability": 0.201909895576711
      }
    },
    {
      "iteration": 38100,
      "average": {
        "on_policy_value_p0": -0.055551725984819,
        "best_response_value_p0": -0.053806144192182,
        "best_response_value_p1": 0.056169088172523,
        "unilateral_improvement_p0": 0.001745581792637,
        "unilateral_improvement_p1": 0.000617362187704,
        "nash_conv": 0.002362943980341,
        "exploitability": 0.001181471990171
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 38200,
      "average": {
        "on_policy_value_p0": -0.055548351510144,
        "best_response_value_p0": -0.053751216457834,
        "best_response_value_p1": 0.056022048674688,
        "unilateral_improvement_p0": 0.001797135052311,
        "unilateral_improvement_p1": 0.000473697164544,
        "nash_conv": 0.002270832216855,
        "exploitability": 0.001135416108427
      },
      "current": {
        "on_policy_value_p0": -0.127904039849242,
        "best_response_value_p0": 0.127904039849242,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.255808079698485,
        "unilateral_improvement_p1": 0.205429293484091,
        "nash_conv": 0.461237373182576,
        "exploitability": 0.230618686591288
      }
    },
    {
      "iteration": 38300,
      "average": {
        "on_policy_value_p0": -0.055545744337294,
        "best_response_value_p0": -0.054204098111828,
        "best_response_value_p1": 0.056560137417283,
        "unilateral_improvement_p0": 0.001341646225466,
        "unilateral_improvement_p1": 0.001014393079989,
        "nash_conv": 0.002356039305456,
        "exploitability": 0.001178019652728
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 38400,
      "average": {
        "on_policy_value_p0": -0.055546787662765,
        "best_response_value_p0": -0.054930997161884,
        "best_response_value_p1": 0.057271408249689,
        "unilateral_improvement_p0": 0.000615790500881,
        "unilateral_improvement_p1": 0.001724620586924,
        "nash_conv": 0.002340411087805,
        "exploitability": 0.001170205543903
      },
      "current": {
        "on_policy_value_p0": -0.27284892707111,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.27284892707111,
        "unilateral_improvement_p0": 0.606182260404443,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.606182260404443,
        "exploitability": 0.303091130202222
      }
    },
    {
      "iteration": 38500,
      "average": {
        "on_policy_value_p0": -0.055550721316091,
        "best_response_value_p0": -0.05511964929808,
        "best_response_value_p1": 0.057415301460103,
        "unilateral_improvement_p0": 0.000431072018011,
        "unilateral_improvement_p1": 0.001864580144011,
        "nash_conv": 0.002295652162022,
        "exploitability": 0.001147826081011
      },
      "current": {
        "on_policy_value_p0": 0.069331707729355,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.264001625603979,
        "unilateral_improvement_p1": 0.235998374396021,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 38600,
      "average": {
        "on_policy_value_p0": -0.055553396243483,
        "best_response_value_p0": -0.054607450060014,
        "best_response_value_p1": 0.056947503249002,
        "unilateral_improvement_p0": 0.000945946183469,
        "unilateral_improvement_p1": 0.001394107005519,
        "nash_conv": 0.002340053188988,
        "exploitability": 0.001170026594494
      },
      "current": {
        "on_policy_value_p0": 0.199622530332207,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.232578393997748,
        "unilateral_improvement_p0": 0.133710803001126,
        "unilateral_improvement_p1": 0.432200924329956,
        "nash_conv": 0.565911727331082,
        "exploitability": 0.282955863665541
      }
    },
    {
      "iteration": 38700,
      "average": {
        "on_policy_value_p0": -0.055552975773408,
        "best_response_value_p0": -0.054114561287052,
        "best_response_value_p1": 0.056235447801484,
        "unilateral_improvement_p0": 0.001438414486356,
        "unilateral_improvement_p1": 0.000682472028076,
        "nash_conv": 0.002120886514432,
        "exploitability": 0.001060443257216
      },
      "current": {
        "on_policy_value_p0": 0.090296832283114,
        "best_response_value_p0": 0.158367107403722,
        "best_response_value_p1": 0.363859449758784,
        "unilateral_improvement_p0": 0.068070275120608,
        "unilateral_improvement_p1": 0.454156282041898,
        "nash_conv": 0.522226557162506,
        "exploitability": 0.261113278581253
      }
    },
    {
      "iteration": 38800,
      "average": {
        "on_policy_value_p0": -0.055550505413632,
        "best_response_value_p0": -0.053822788059845,
        "best_response_value_p1": 0.056168831065736,
        "unilateral_improvement_p0": 0.001727717353787,
        "unilateral_improvement_p1": 0.000618325652104,
        "nash_conv": 0.002346043005891,
        "exploitability": 0.001173021502945
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 38900,
      "average": {
        "on_policy_value_p0": -0.055547582328892,
        "best_response_value_p0": -0.054075446705259,
        "best_response_value_p1": 0.056393465204611,
        "unilateral_improvement_p0": 0.001472135623633,
        "unilateral_improvement_p1": 0.000845882875719,
        "nash_conv": 0.002318018499352,
        "exploitability": 0.001159009249676
      },
      "current": {
        "on_policy_value_p0": -0.377243775734941,
        "best_response_value_p0": 0.12574792524498,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.502991700979922,
        "unilateral_improvement_p1": 0.122756224265059,
        "nash_conv": 0.62574792524498,
        "exploitability": 0.31287396262249
      }
    },
    {
      "iteration": 39000,
      "average": {
        "on_policy_value_p0": -0.055548605097742,
        "best_response_value_p0": -0.054937285393767,
        "best_response_value_p1": 0.057248680001732,
        "unilateral_improvement_p0": 0.000611319703975,
        "unilateral_improvement_p1": 0.00170007490399,
        "nash_conv": 0.002311394607965,
        "exploitability": 0.001155697303982
      },
      "current": {
        "on_policy_value_p0": -0.229657543240208,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.229657543240208,
        "unilateral_improvement_p0": 0.396324209906875,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.396324209906875,
        "exploitability": 0.198162104953438
      }
    },
    {
      "iteration": 39100,
      "average": {
        "on_policy_value_p0": -0.05555196611289,
        "best_response_value_p0": -0.055024182033591,
        "best_response_value_p1": 0.057352871188318,
        "unilateral_improvement_p0": 0.0005277840793,
        "unilateral_improvement_p1": 0.001800905075428,
        "nash_conv": 0.002328689154727,
        "exploitability": 0.001164344577364
      },
      "current": {
        "on_policy_value_p0": -0.035986538100254,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.202653204766921,
        "unilateral_improvement_p1": 0.130680128566412,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 39200,
      "average": {
        "on_policy_value_p0": -0.055554914556619,
        "best_response_value_p0": -0.054883814222279,
        "best_response_value_p1": 0.057114403808343,
        "unilateral_improvement_p0": 0.000671100334339,
        "unilateral_improvement_p1": 0.001559489251724,
        "nash_conv": 0.002230589586063,
        "exploitability": 0.001115294793032
      },
      "current": {
        "on_policy_value_p0": 0.076903155864182,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.153806311728364,
        "unilateral_improvement_p0": 0.089763510802484,
        "unilateral_improvement_p1": 0.230709467592546,
        "nash_conv": 0.320472978395031,
        "exploitability": 0.160236489197515
      }
    },
    {
      "iteration": 39300,
      "average": {
        "on_policy_value_p0": -0.055556092859843,
        "best_response_value_p0": -0.054437503537601,
        "best_response_value_p1": 0.056694061037103,
        "unilateral_improvement_p0": 0.001118589322242,
        "unilateral_improvement_p1": 0.001137968177259,
        "nash_conv": 0.002256557499501,
        "exploitability": 0.001128278749751
      },
      "current": {
        "on_policy_value_p0": 0.088921868827756,
        "best_response_value_p0": 0.107389198049567,
        "best_response_value_p1": 0.29639867488971,
        "unilateral_improvement_p0": 0.018467329221812,
        "unilateral_improvement_p1": 0.385320543717466,
        "nash_conv": 0.403787872939278,
        "exploitability": 0.201893936469639
      }
    },
    {
      "iteration": 39400,
      "average": {
        "on_policy_value_p0": -0.055554975633755,
        "best_response_value_p0": -0.054083902674041,
        "best_response_value_p1": 0.05632992753469,
        "unilateral_improvement_p0": 0.001471072959714,
        "unilateral_improvement_p1": 0.000774951900935,
        "nash_conv": 0.002246024860649,
        "exploitability": 0.001123012430325
      },
      "current": {
        "on_policy_value_p0": 0.062138086766283,
        "best_response_value_p0": 0.062138086766283,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.395471420099616,
        "nash_conv": 0.395471420099616,
        "exploitability": 0.197735710049808
      }
    },
    {
      "iteration": 39500,
      "average": {
        "on_policy_value_p0": -0.055552560170679,
        "best_response_value_p0": -0.05386600824468,
        "best_response_value_p1": 0.056187320123209,
        "unilateral_improvement_p0": 0.001686551925999,
        "unilateral_improvement_p1": 0.00063475995253,
        "nash_conv": 0.002321311878529,
        "exploitability": 0.001160655939264
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 39600,
      "average": {
        "on_policy_value_p0": -0.055549535211377,
        "best_response_value_p0": -0.053747409490761,
        "best_response_value_p1": 0.05604543295118,
        "unilateral_improvement_p0": 0.001802125720617,
        "unilateral_improvement_p1": 0.000495897739803,
        "nash_conv": 0.002298023460419,
        "exploitability": 0.00114901173021
      },
      "current": {
        "on_policy_value_p0": -0.068807531975103,
        "best_response_value_p0": 0.132262900679115,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.201070432654218,
        "unilateral_improvement_p1": 0.26452580135823,
        "nash_conv": 0.465596234012448,
        "exploitability": 0.232798117006224
      }
    },
    {
      "iteration": 39700,
      "average": {
        "on_policy_value_p0": -0.055546450488222,
        "best_response_value_p0": -0.054036325235617,
        "best_response_value_p1": 0.056336099582147,
        "unilateral_improvement_p0": 0.001510125252605,
        "unilateral_improvement_p1": 0.000789649093926,
        "nash_conv": 0.00229977434653,
        "exploitability": 0.001149887173265
      },
      "current": {
        "on_policy_value_p0": -0.271652286712368,
        "best_response_value_p0": 0.271652286712368,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.543304573424736,
        "unilateral_improvement_p1": 0.061681046620965,
        "nash_conv": 0.604985620045701,
        "exploitability": 0.302492810022851
      }
    },
    {
      "iteration": 39800,
      "average": {
        "on_policy_value_p0": -0.055546341246341,
        "best_response_value_p0": -0.054723040496377,
        "best_response_value_p1": 0.05703207253127,
        "unilateral_improvement_p0": 0.000823300749964,
        "unilateral_improvement_p1": 0.001485731284929,
        "nash_conv": 0.002309032034893,
        "exploitability": 0.001154516017446
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 39900,
      "average": {
        "on_policy_value_p0": -0.055549524451835,
        "best_response_value_p0": -0.055155936340163,
        "best_response_value_p1": 0.057431350443749,
        "unilateral_improvement_p0": 0.000393588111672,
        "unilateral_improvement_p1": 0.001881825991914,
        "nash_conv": 0.002275414103586,
        "exploitability": 0.001137707051793
      },
      "current": {
        "on_policy_value_p0": -0.05365244099406,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.123554523000948,
        "unilateral_improvement_p0": 0.386985774327393,
        "unilateral_improvement_p1": 0.069902082006888,
        "nash_conv": 0.456887856334281,
        "exploitability": 0.228443928167141
      }
    },
    {
      "iteration": 40000,
      "average": {
        "on_policy_value_p0": -0.055552712569029,
        "best_response_value_p0": -0.05480711110664,
        "best_response_value_p1": 0.057108225542749,
        "unilateral_improvement_p0": 0.000745601462388,
        "unilateral_improvement_p1": 0.00155551297372,
        "nash_conv": 0.002301114436108,
        "exploitability": 0.001150557218054
      },
      "current": {
        "on_policy_value_p0": 0.131374574515026,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.201958758818308,
        "unilateral_improvement_p1": 0.298041241181692,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 40100,
      "average": {
        "on_policy_value_p0": -0.0555533821639,
        "best_response_value_p0": -0.054313815694979,
        "best_response_value_p1": 0.056472745740099,
        "unilateral_improvement_p0": 0.001239566468921,
        "unilateral_improvement_p1": 0.000919363576199,
        "nash_conv": 0.00215893004512,
        "exploitability": 0.00107946502256
      },
      "current": {
        "on_policy_value_p0": 0.127614040281296,
        "best_response_value_p0": 0.221860203273076,
        "best_response_value_p1": 0.31150767401644,
        "unilateral_improvement_p0": 0.09424616299178,
        "unilateral_improvement_p1": 0.439121714297736,
        "nash_conv": 0.533367877289516,
        "exploitability": 0.266683938644758
      }
    },
    {
      "iteration": 40200,
      "average": {
        "on_policy_value_p0": -0.055551588952011,
        "best_response_value_p0": -0.053942771240408,
        "best_response_value_p1": 0.056239596757139,
        "unilateral_improvement_p0": 0.001608817711603,
        "unilateral_improvement_p1": 0.000688007805128,
        "nash_conv": 0.002296825516731,
        "exploitability": 0.001148412758365
      },
      "current": {
        "on_policy_value_p0": -0.02802242260851,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.44395515478298,
        "unilateral_improvement_p0": 0.194689089275177,
        "unilateral_improvement_p1": 0.41593273217447,
        "nash_conv": 0.610621821449647,
        "exploitability": 0.305310910724823
      }
    },
    {
      "iteration": 40300,
      "average": {
        "on_policy_value_p0": -0.055548978209152,
        "best_response_value_p0": -0.053870489503393,
        "best_response_value_p1": 0.05610891801174,
        "unilateral_improvement_p0": 0.001678488705759,
        "unilateral_improvement_p1": 0.000559939802589,
        "nash_conv": 0.002238428508347,
        "exploitability": 0.001119214254174
      },
      "current": {
        "on_policy_value_p0": -0.171794028803625,
        "best_response_value_p0": 0.057264676267875,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.2290587050715,
        "unilateral_improvement_p1": 0.328205971196375,
        "nash_conv": 0.557264676267875,
        "exploitability": 0.278632338133937
      }
    },
    {
      "iteration": 40400,
      "average": {
        "on_policy_value_p0": -0.055547766995843,
        "best_response_value_p0": -0.05466887419921,
        "best_response_value_p1": 0.056918355340886,
        "unilateral_improvement_p0": 0.000878892796633,
        "unilateral_improvement_p1": 0.001370588345043,
        "nash_conv": 0.002249481141676,
        "exploitability": 0.001124740570838
      },
      "current": {
        "on_policy_value_p0": -0.380741683038095,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.380741683038095,
        "unilateral_improvement_p0": 0.547408349704762,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.547408349704762,
        "exploitability": 0.273704174852381
      }
    },
    {
      "iteration": 40500,
      "average": {
        "on_policy_value_p0": -0.055550504551372,
        "best_response_value_p0": -0.055109139149751,
        "best_response_value_p1": 0.057329550860656,
        "unilateral_improvement_p0": 0.000441365401621,
        "unilateral_improvement_p1": 0.001779046309284,
        "nash_conv": 0.002220411710905,
        "exploitability": 0.001110205855453
      },
      "current": {
        "on_policy_value_p0": -0.083309076587374,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.249975743254041,
        "unilateral_improvement_p1": 0.083357590079292,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 40600,
      "average": {
        "on_policy_value_p0": -0.055553719623244,
        "best_response_value_p0": -0.054979441733023,
        "best_response_value_p1": 0.057242706604723,
        "unilateral_improvement_p0": 0.00057427789022,
        "unilateral_improvement_p1": 0.001688986981479,
        "nash_conv": 0.002263264871699,
        "exploitability": 0.00113163243585
      },
      "current": {
        "on_policy_value_p0": 0.048361926999003,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.118304739667664,
        "unilateral_improvement_p0": 0.118304739667664,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.28497140633433,
        "exploitability": 0.142485703167165
      }
    },
    {
      "iteration": 40700,
      "average": {
        "on_policy_value_p0": -0.055555879372082,
        "best_response_value_p0": -0.054679550913597,
        "best_response_value_p1": 0.056908187265339,
        "unilateral_improvement_p0": 0.000876328458485,
        "unilateral_improvement_p1": 0.001352307893257,
        "nash_conv": 0.002228636351742,
        "exploitability": 0.001114318175871
      },
      "current": {
        "on_policy_value_p0": 0.079533234750551,
        "best_response_value_p0": 0.137282999874688,
        "best_response_value_p1": 0.21783380308506,
        "unilateral_improvement_p0": 0.057749765124137,
        "unilateral_improvement_p1": 0.297367037835612,
        "nash_conv": 0.355116802959748,
        "exploitability": 0.177558401479874
      }
    },
    {
      "iteration": 40800,
      "average": {
        "on_policy_value_p0": -0.055555866005098,
        "best_response_value_p0": -0.054277183854804,
        "best_response_value_p1": 0.056410570467663,
        "unilateral_improvement_p0": 0.001278682150294,
        "unilateral_improvement_p1": 0.000854704462564,
        "nash_conv": 0.002133386612859,
        "exploitability": 0.001066693306429
      },
      "current": {
        "on_policy_value_p0": 0.085693760272996,
        "best_response_value_p0": 0.085693760272996,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.419027093606329,
        "nash_conv": 0.419027093606329,
        "exploitability": 0.209513546803165
      }
    },
    {
      "iteration": 40900,
      "average": {
        "on_policy_value_p0": -0.05555408771721,
        "best_response_value_p0": -0.053989852318855,
        "best_response_value_p1": 0.056250214395401,
        "unilateral_improvement_p0": 0.001564235398355,
        "unilateral_improvement_p1": 0.000696126678191,
        "nash_conv": 0.002260362076546,
        "exploitability": 0.001130181038273
      },
      "current": {
        "on_policy_value_p0": 0.037077974430804,
        "best_response_value_p0": 0.092510717805059,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.055432743374255,
        "unilateral_improvement_p1": 0.370411307764137,
        "nash_conv": 0.425844051138392,
        "exploitability": 0.212922025569196
      }
    },
    {
      "iteration": 41000,
      "average": {
        "on_policy_value_p0": -0.05555154022885,
        "best_response_value_p0": -0.053836070942947,
        "best_response_value_p1": 0.056113018750534,
        "unilateral_improvement_p0": 0.001715469285903,
        "unilateral_improvement_p1": 0.000561478521683,
        "nash_conv": 0.002276947807586,
        "exploitability": 0.001138473903793
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 41100,
      "average": {
        "on_policy_value_p0": -0.055548390272369,
        "best_response_value_p0": -0.053843728387636,
        "best_response_value_p1": 0.055976490724376,
        "unilateral_improvement_p0": 0.001704661884732,
        "unilateral_improvement_p1": 0.000428100452007,
        "nash_conv": 0.00213276233674,
        "exploitability": 0.00106638116837
      },
      "current": {
        "on_policy_value_p0": -0.15647555895268,
        "best_response_value_p0": 0.15647555895268,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.31295111790536,
        "unilateral_improvement_p1": 0.176857774380653,
        "nash_conv": 0.489808892286013,
        "exploitability": 0.244904446143007
      }
    },
    {
      "iteration": 41200,
      "average": {
        "on_policy_value_p0": -0.055546380922671,
        "best_response_value_p0": -0.054330376730731,
        "best_response_value_p1": 0.056601327144894,
        "unilateral_improvement_p0": 0.00121600419194,
        "unilateral_improvement_p1": 0.001054946222223,
        "nash_conv": 0.002270950414163,
        "exploitability": 0.001135475207081
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 41300,
      "average": {
        "on_policy_value_p0": -0.055547638564262,
        "best_response_value_p0": -0.055005928683765,
        "best_response_value_p1": 0.057247676547876,
        "unilateral_improvement_p0": 0.000541709880497,
        "unilateral_improvement_p1": 0.001700037983614,
        "nash_conv": 0.002241747864111,
        "exploitability": 0.001120873932056
      },
      "current": {
        "on_policy_value_p0": -0.245146065746246,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.245146065746246,
        "unilateral_improvement_p0": 0.578479399079579,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.578479399079579,
        "exploitability": 0.28923969953979
      }
    },
    {
      "iteration": 41400,
      "average": {
        "on_policy_value_p0": -0.055551201049359,
        "best_response_value_p0": -0.055126674416343,
        "best_response_value_p1": 0.057333575349528,
        "unilateral_improvement_p0": 0.000424526633016,
        "unilateral_improvement_p1": 0.001782374300169,
        "nash_conv": 0.002206900933185,
        "exploitability": 0.001103450466592
      },
      "current": {
        "on_policy_value_p0": 0.072589306870763,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.260744026462571,
        "unilateral_improvement_p1": 0.239255973537429,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 41500,
      "average": {
        "on_policy_value_p0": -0.055553543196111,
        "best_response_value_p0": -0.054637450351468,
        "best_response_value_p1": 0.056894646130207,
        "unilateral_improvement_p0": 0.000916092844643,
        "unilateral_improvement_p1": 0.001341102934097,
        "nash_conv": 0.002257195778739,
        "exploitability": 0.00112859788937
      },
      "current": {
        "on_policy_value_p0": 0.198499706669656,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.230332746672646,
        "unilateral_improvement_p0": 0.134833626663677,
        "unilateral_improvement_p1": 0.428832453342302,
        "nash_conv": 0.563666080005979,
        "exploitability": 0.28183304000299
      }
    },
    {
      "iteration": 41600,
      "average": {
        "on_policy_value_p0": -0.055553201228671,
        "best_response_value_p0": -0.054185702203372,
        "best_response_value_p1": 0.056207699889052,
        "unilateral_improvement_p0": 0.001367499025299,
        "unilateral_improvement_p1": 0.000654498660381,
        "nash_conv": 0.00202199768568,
        "exploitability": 0.00101099884284
      },
      "current": {
        "on_policy_value_p0": 0.093861053553322,
        "best_response_value_p0": 0.164804636830109,
        "best_response_value_p1": 0.358112833446425,
        "unilateral_improvement_p0": 0.070943583276788,
        "unilateral_improvement_p1": 0.451973886999746,
        "nash_conv": 0.522917470276534,
        "exploitability": 0.261458735138267
      }
    },
    {
      "iteration": 41700,
      "average": {
        "on_policy_value_p0": -0.055551027825742,
        "best_response_value_p0": -0.053894037237445,
        "best_response_value_p1": 0.056156690959428,
        "unilateral_improvement_p0": 0.001656990588296,
        "unilateral_improvement_p1": 0.000605663133686,
        "nash_conv": 0.002262653721982,
        "exploitability": 0.001131326860991
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 41800,
      "average": {
        "on_policy_value_p0": -0.055548328025574,
        "best_response_value_p0": -0.054050297162822,
        "best_response_value_p1": 0.056245449986963,
        "unilateral_improvement_p0": 0.001498030862752,
        "unilateral_improvement_p1": 0.000697121961389,
        "nash_conv": 0.002195152824141,
        "exploitability": 0.00109757641207
      },
      "current": {
        "on_policy_value_p0": -0.324206251766669,
        "best_response_value_p0": 0.10806875058889,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.432275002355558,
        "unilateral_improvement_p1": 0.175793748233331,
        "nash_conv": 0.60806875058889,
        "exploitability": 0.304034375294445
      }
    },
    {
      "iteration": 41900,
      "average": {
        "on_policy_value_p0": -0.055548681642109,
        "best_response_value_p0": -0.05491523108466,
        "best_response_value_p1": 0.057112699939041,
        "unilateral_improvement_p0": 0.000633450557449,
        "unilateral_improvement_p1": 0.001564018296932,
        "nash_conv": 0.002197468854381,
        "exploitability": 0.00109873442719
      },
      "current": {
        "on_policy_value_p0": -0.291871099166433,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.291871099166433,
        "unilateral_improvement_p0": 0.4585377658331,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.4585377658331,
        "exploitability": 0.22926888291655
      }
    },
    {
      "iteration": 42000,
      "average": {
        "on_policy_value_p0": -0.055551603230528,
        "best_response_value_p0": -0.055063166094051,
        "best_response_value_p1": 0.057289000057197,
        "unilateral_improvement_p0": 0.000488437136477,
        "unilateral_improvement_p1": 0.001737396826669,
        "nash_conv": 0.002225833963146,
        "exploitability": 0.001112916981573
      },
      "current": {
        "on_policy_value_p0": -0.05616566767837,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.222832334345036,
        "unilateral_improvement_p1": 0.110500998988297,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 42100,
      "average": {
        "on_policy_value_p0": -0.055554396850364,
        "best_response_value_p0": -0.054932374725656,
        "best_response_value_p1": 0.057124549254398,
        "unilateral_improvement_p0": 0.000622022124708,
        "unilateral_improvement_p1": 0.001570152404034,
        "nash_conv": 0.002192174528743,
        "exploitability": 0.001096087264371
      },
      "current": {
        "on_policy_value_p0": 0.062932736615865,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.12586547323173,
        "unilateral_improvement_p0": 0.103733930050802,
        "unilateral_improvement_p1": 0.188798209847594,
        "nash_conv": 0.292532139898396,
        "exploitability": 0.146266069949198
      }
    },
    {
      "iteration": 42200,
      "average": {
        "on_policy_value_p0": -0.055555924295314,
        "best_response_value_p0": -0.05458542957916,
        "best_response_value_p1": 0.056772120116527,
        "unilateral_improvement_p0": 0.000970494716154,
        "unilateral_improvement_p1": 0.001216195821213,
        "nash_conv": 0.002186690537367,
        "exploitability": 0.001093345268684
      },
      "current": {
        "on_policy_value_p0": 0.075941678453871,
        "best_response_value_p0": 0.119591190643504,
        "best_response_value_p1": 0.246034308954068,
        "unilateral_improvement_p0": 0.043649512189632,
        "unilateral_improvement_p1": 0.32197598740794,
        "nash_conv": 0.365625499597572,
        "exploitability": 0.182812749798786
      }
    },
    {
      "iteration": 42300,
      "average": {
        "on_policy_value_p0": -0.055555440561526,
        "best_response_value_p0": -0.054228732015534,
        "best_response_value_p1": 0.056351677350789,
        "unilateral_improvement_p0": 0.001326708545992,
        "unilateral_improvement_p1": 0.000796236789263,
        "nash_conv": 0.002122945335255,
        "exploitability": 0.001061472667628
      },
      "current": {
        "on_policy_value_p0": 0.074401045639911,
        "best_response_value_p0": 0.074401045639911,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.407734378973244,
        "nash_conv": 0.407734378973244,
        "exploitability": 0.203867189486622
      }
    },
    {
      "iteration": 42400,
      "average": {
        "on_policy_value_p0": -0.05555354566765,
        "best_response_value_p0": -0.053982598948969,
        "best_response_value_p1": 0.056218772451377,
        "unilateral_improvement_p0": 0.001570946718681,
        "unilateral_improvement_p1": 0.000665226783727,
        "nash_conv": 0.002236173502408,
        "exploitability": 0.001118086751204
      },
      "current": {
        "on_policy_value_p0": 0.019072002046424,
        "best_response_value_p0": 0.128522662573819,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.109450660527395,
        "unilateral_improvement_p1": 0.352405335379757,
        "nash_conv": 0.461855995907152,
        "exploitability": 0.230927997953576
      }
    },
    {
      "iteration": 42500,
      "average": {
        "on_policy_value_p0": -0.055551087179187,
        "best_response_value_p0": -0.053850807503151,
        "best_response_value_p1": 0.056086492986785,
        "unilateral_improvement_p0": 0.001700279676037,
        "unilateral_improvement_p1": 0.000535405807597,
        "nash_conv": 0.002235685483634,
        "exploitability": 0.001117842741817
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 42600,
      "average": {
        "on_policy_value_p0": -0.055548099711479,
        "best_response_value_p0": -0.053924242919595,
        "best_response_value_p1": 0.056032671543603,
        "unilateral_improvement_p0": 0.001623856791884,
        "unilateral_improvement_p1": 0.000484571832124,
        "nash_conv": 0.002108428624008,
        "exploitability": 0.001054214312004
      },
      "current": {
        "on_policy_value_p0": -0.183365882077593,
        "best_response_value_p0": 0.183365882077593,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.366731764155186,
        "unilateral_improvement_p1": 0.14996745125574,
        "nash_conv": 0.516699215410926,
        "exploitability": 0.258349607705463
      }
    },
    {
      "iteration": 42700,
      "average": {
        "on_policy_value_p0": -0.055546635437463,
        "best_response_value_p0": -0.054452010006199,
        "best_response_value_p1": 0.056682087613368,
        "unilateral_improvement_p0": 0.001094625431264,
        "unilateral_improvement_p1": 0.001135452175904,
        "nash_conv": 0.002230077607169,
        "exploitability": 0.001115038803584
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 42800,
      "average": {
        "on_policy_value_p0": -0.055548271549946,
        "best_response_value_p0": -0.055103601883132,
        "best_response_value_p1": 0.057270938515811,
        "unilateral_improvement_p0": 0.000444669666814,
        "unilateral_improvement_p1": 0.001722666965865,
        "nash_conv": 0.002167336632679,
        "exploitability": 0.00108366831634
      },
      "current": {
        "on_policy_value_p0": -0.208226051781991,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.208226051781991,
        "unilateral_improvement_p0": 0.541559385115324,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.541559385115324,
        "exploitability": 0.270779692557662
      }
    },
    {
      "iteration": 42900,
      "average": {
        "on_policy_value_p0": -0.05555166065232,
        "best_response_value_p0": -0.055116705689187,
        "best_response_value_p1": 0.057263623963721,
        "unilateral_improvement_p0": 0.000434954963133,
        "unilateral_improvement_p1": 0.001711963311401,
        "nash_conv": 0.002146918274535,
        "exploitability": 0.001073459137267
      },
      "current": {
        "on_policy_value_p0": 0.082130897134266,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.251202436199067,
        "unilateral_improvement_p1": 0.248797563800933,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 43000,
      "average": {
        "on_policy_value_p0": -0.055553650516044,
        "best_response_value_p0": -0.054600930404637,
        "best_response_value_p1": 0.056810675183027,
        "unilateral_improvement_p0": 0.000952720111407,
        "unilateral_improvement_p1": 0.001257024666983,
        "nash_conv": 0.00220974477839,
        "exploitability": 0.001104872389195
      },
      "current": {
        "on_policy_value_p0": 0.205897474213258,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.245128281759849,
        "unilateral_improvement_p0": 0.127435859120076,
        "unilateral_improvement_p1": 0.451025755973107,
        "nash_conv": 0.578461615093182,
        "exploitability": 0.289230807546591
      }
    },
    {
      "iteration": 43100,
      "average": {
        "on_policy_value_p0": -0.055553099402213,
        "best_response_value_p0": -0.054178575213706,
        "best_response_value_p1": 0.05621087237866,
        "unilateral_improvement_p0": 0.001374524188507,
        "unilateral_improvement_p1": 0.000657772976447,
        "nash_conv": 0.002032297164954,
        "exploitability": 0.001016148582477
      },
      "current": {
        "on_policy_value_p0": 0.081868796890002,
        "best_response_value_p0": 0.148813494847909,
        "best_response_value_p1": 0.366110604084185,
        "unilateral_improvement_p0": 0.066944697957907,
        "unilateral_improvement_p1": 0.447979400974187,
        "nash_conv": 0.514924098932094,
        "exploitability": 0.257462049466047
      }
    },
    {
      "iteration": 43200,
      "average": {
        "on_policy_value_p0": -0.055551041060782,
        "best_response_value_p0": -0.053929336932469,
        "best_response_value_p1": 0.056151886452682,
        "unilateral_improvement_p0": 0.001621704128312,
        "unilateral_improvement_p1": 0.000600845391901,
        "nash_conv": 0.002222549520213,
        "exploitability": 0.001111274760106
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 43300,
      "average": {
        "on_policy_value_p0": -0.055548534822593,
        "best_response_value_p0": -0.054099519973753,
        "best_response_value_p1": 0.056263523535564,
        "unilateral_improvement_p0": 0.001449014848841,
        "unilateral_improvement_p1": 0.00071498871297,
        "nash_conv": 0.002164003561811,
        "exploitability": 0.001082001780906
      },
      "current": {
        "on_policy_value_p0": -0.332433366517817,
        "best_response_value_p0": 0.110811122172606,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.443244488690422,
        "unilateral_improvement_p1": 0.167566633482183,
        "nash_conv": 0.610811122172606,
        "exploitability": 0.305405561086303
      }
    },
    {
      "iteration": 43400,
      "average": {
        "on_policy_value_p0": -0.055548970474236,
        "best_response_value_p0": -0.054943307752181,
        "best_response_value_p1": 0.057099953409052,
        "unilateral_improvement_p0": 0.000605662722056,
        "unilateral_improvement_p1": 0.001550982934816,
        "nash_conv": 0.002156645656871,
        "exploitability": 0.001078322828436
      },
      "current": {
        "on_policy_value_p0": -0.289592308092714,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.289592308092714,
        "unilateral_improvement_p0": 0.45625897475938,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.45625897475938,
        "exploitability": 0.22812948737969
      }
    },
    {
      "iteration": 43500,
      "average": {
        "on_policy_value_p0": -0.055551805369185,
        "best_response_value_p0": -0.055083900794667,
        "best_response_value_p1": 0.057270702247596,
        "unilateral_improvement_p0": 0.000467904574518,
        "unilateral_improvement_p1": 0.001718896878411,
        "nash_conv": 0.002186801452928,
        "exploitability": 0.001093400726464
      },
      "current": {
        "on_policy_value_p0": -0.056146324009963,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.22281299067663,
        "unilateral_improvement_p1": 0.110520342656704,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 43600,
      "average": {
        "on_policy_value_p0": -0.055554516121285,
        "best_response_value_p0": -0.05495756157266,
        "best_response_value_p1": 0.057107102742918,
        "unilateral_improvement_p0": 0.000596954548625,
        "unilateral_improvement_p1": 0.001552586621634,
        "nash_conv": 0.002149541170258,
        "exploitability": 0.001074770585129
      },
      "current": {
        "on_policy_value_p0": 0.064684876623093,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.129369753246186,
        "unilateral_improvement_p0": 0.101981790043574,
        "unilateral_improvement_p1": 0.194054629869279,
        "nash_conv": 0.296036419912853,
        "exploitability": 0.148018209956426
      }
    },
    {
      "iteration": 43700,
      "average": {
        "on_policy_value_p0": -0.055556021120873,
        "best_response_value_p0": -0.054608380047673,
        "best_response_value_p1": 0.05676404836984,
        "unilateral_improvement_p0": 0.0009476410732,
        "unilateral_improvement_p1": 0.001208027248967,
        "nash_conv": 0.002155668322167,
        "exploitability": 0.001077834161084
      },
      "current": {
        "on_policy_value_p0": 0.07994950998269,
        "best_response_value_p0": 0.122714639044461,
        "best_response_value_p1": 0.24780307520979,
        "unilateral_improvement_p0": 0.042765129061772,
        "unilateral_improvement_p1": 0.32775258519248,
        "nash_conv": 0.370517714254252,
        "exploitability": 0.185258857127126
      }
    },
    {
      "iteration": 43800,
      "average": {
        "on_policy_value_p0": -0.055555592771011,
        "best_response_value_p0": -0.054257159202449,
        "best_response_value_p1": 0.056330249514868,
        "unilateral_improvement_p0": 0.001298433568561,
        "unilateral_improvement_p1": 0.000774656743857,
        "nash_conv": 0.002073090312418,
        "exploitability": 0.001036545156209
      },
      "current": {
        "on_policy_value_p0": 0.077526831385246,
        "best_response_value_p0": 0.077526831385246,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.41086016471858,
        "nash_conv": 0.41086016471858,
        "exploitability": 0.20543008235929
      }
    },
    {
      "iteration": 43900,
      "average": {
        "on_policy_value_p0": -0.055553791287126,
        "best_response_value_p0": -0.054009261618685,
        "best_response_value_p1": 0.056201934595699,
        "unilateral_improvement_p0": 0.00154452966844,
        "unilateral_improvement_p1": 0.000648143308573,
        "nash_conv": 0.002192672977014,
        "exploitability": 0.001096336488507
      },
      "current": {
        "on_policy_value_p0": 0.026279334356572,
        "best_response_value_p0": 0.114107997953523,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.087828663596952,
        "unilateral_improvement_p1": 0.359612667689905,
        "nash_conv": 0.447441331286857,
        "exploitability": 0.223720665643428
      }
    },
    {
      "iteration": 44000,
      "average": {
        "on_policy_value_p0": -0.055551425259711,
        "best_response_value_p0": -0.053876960222084,
        "best_response_value_p1": 0.056074202926163,
        "unilateral_improvement_p0": 0.001674465037627,
        "unilateral_improvement_p1": 0.000522777666452,
        "nash_conv": 0.002197242704079,
        "exploitability": 0.001098621352039
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 44100,
      "average": {
        "on_policy_value_p0": -0.055548550015705,
        "best_response_value_p0": -0.053927043471219,
        "best_response_value_p1": 0.055978031043732,
        "unilateral_improvement_p0": 0.001621506544486,
        "unilateral_improvement_p1": 0.000429481028028,
        "nash_conv": 0.002050987572513,
        "exploitability": 0.001025493786257
      },
      "current": {
        "on_policy_value_p0": -0.173401750342241,
        "best_response_value_p0": 0.173401750342241,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.346803500684482,
        "unilateral_improvement_p1": 0.159931582991092,
        "nash_conv": 0.506735083675574,
        "exploitability": 0.253367541837787
      }
    },
    {
      "iteration": 44200,
      "average": {
        "on_policy_value_p0": -0.055546938099114,
        "best_response_value_p0": -0.054413542966913,
        "best_response_value_p1": 0.056605531727645,
        "unilateral_improvement_p0": 0.001133395132201,
        "unilateral_improvement_p1": 0.001058593628531,
        "nash_conv": 0.002191988760732,
        "exploitability": 0.001095994380366
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 44300,
      "average": {
        "on_policy_value_p0": -0.055548205642577,
        "best_response_value_p0": -0.055043158746522,
        "best_response_value_p1": 0.057197902482372,
        "unilateral_improvement_p0": 0.000505046896055,
        "unilateral_improvement_p1": 0.001649696839795,
        "nash_conv": 0.00215474373585,
        "exploitability": 0.001077371867925
      },
      "current": {
        "on_policy_value_p0": -0.232552303479838,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.232552303479838,
        "unilateral_improvement_p0": 0.565885636813171,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.565885636813171,
        "exploitability": 0.282942818406585
      }
    },
    {
      "iteration": 44400,
      "average": {
        "on_policy_value_p0": -0.055551390367579,
        "best_response_value_p0": -0.055130651461864,
        "best_response_value_p1": 0.057264174417101,
        "unilateral_improvement_p0": 0.000420738905716,
        "unilateral_improvement_p1": 0.001712784049522,
        "nash_conv": 0.002133522955237,
        "exploitability": 0.001066761477619
      },
      "current": {
        "on_policy_value_p0": 0.071121941014489,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.262211392318844,
        "unilateral_improvement_p1": 0.237788607681156,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 44500,
      "average": {
        "on_policy_value_p0": -0.055553504310491,
        "best_response_value_p0": -0.054682708773297,
        "best_response_value_p1": 0.056868296615492,
        "unilateral_improvement_p0": 0.000870795537194,
        "unilateral_improvement_p1": 0.001314792305002,
        "nash_conv": 0.002185587842195,
        "exploitability": 0.001092793921098
      },
      "current": {
        "on_policy_value_p0": 0.188405846058949,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.210145025451231,
        "unilateral_improvement_p0": 0.144927487274384,
        "unilateral_improvement_p1": 0.398550871510181,
        "nash_conv": 0.543478358784565,
        "exploitability": 0.271739179392282
      }
    },
    {
      "iteration": 44600,
      "average": {
        "on_policy_value_p0": -0.055553341113825,
        "best_response_value_p0": -0.054293891567169,
        "best_response_value_p1": 0.056236770382017,
        "unilateral_improvement_p0": 0.001259449546655,
        "unilateral_improvement_p1": 0.000683429268193,
        "nash_conv": 0.001942878814848,
        "exploitability": 0.000971439407424
      },
      "current": {
        "on_policy_value_p0": 0.096270909343562,
        "best_response_value_p0": 0.174957859717673,
        "best_response_value_p1": 0.342626099251778,
        "unilateral_improvement_p0": 0.078686950374111,
        "unilateral_improvement_p1": 0.43889700859534,
        "nash_conv": 0.517583958969451,
        "exploitability": 0.258791979484725
      }
    },
    {
      "iteration": 44700,
      "average": {
        "on_policy_value_p0": -0.055551563211799,
        "best_response_value_p0": -0.053992474139064,
        "best_response_value_p1": 0.056177603039268,
        "unilateral_improvement_p0": 0.001559089072736,
        "unilateral_improvement_p1": 0.000626039827469,
        "nash_conv": 0.002185128900204,
        "exploitability": 0.001092564450102
      },
      "current": {
        "on_policy_value_p0": -0.001087791471478,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.497824417057044,
        "unilateral_improvement_p0": 0.167754458138145,
        "unilateral_improvement_p1": 0.496736625585566,
        "nash_conv": 0.664491083723711,
        "exploitability": 0.332245541861855
      }
    },
    {
      "iteration": 44800,
      "average": {
        "on_policy_value_p0": -0.055549202405771,
        "best_response_value_p0": -0.054014656733078,
        "best_response_value_p1": 0.056052230884973,
        "unilateral_improvement_p0": 0.001534545672693,
        "unilateral_improvement_p1": 0.000503028479202,
        "nash_conv": 0.002037574151894,
        "exploitability": 0.001018787075947
      },
      "current": {
        "on_policy_value_p0": -0.242946510912432,
        "best_response_value_p0": 0.080982170304144,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.323928681216576,
        "unilateral_improvement_p1": 0.257053489087568,
        "nash_conv": 0.580982170304144,
        "exploitability": 0.290491085152072
      }
    },
    {
      "iteration": 44900,
      "average": {
        "on_policy_value_p0": -0.0555487102503,
        "best_response_value_p0": -0.054837357703008,
        "best_response_value_p1": 0.056930763914808,
        "unilateral_improvement_p0": 0.000711352547291,
        "unilateral_improvement_p1": 0.001382053664509,
        "nash_conv": 0.0020934062118,
        "exploitability": 0.0010467031059
      },
      "current": {
        "on_policy_value_p0": -0.350786155486765,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.350786155486765,
        "unilateral_improvement_p0": 0.517452822153432,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.517452822153432,
        "exploitability": 0.258726411076716
      }
    },
    {
      "iteration": 45000,
      "average": {
        "on_policy_value_p0": -0.055551178437516,
        "best_response_value_p0": -0.055124550086694,
        "best_response_value_p1": 0.057236271951275,
        "unilateral_improvement_p0": 0.000426628350822,
        "unilateral_improvement_p1": 0.001685093513759,
        "nash_conv": 0.00211172186458,
        "exploitability": 0.00105586093229
      },
      "current": {
        "on_policy_value_p0": -0.077939888580548,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.244606555247215,
        "unilateral_improvement_p1": 0.088726778086118,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 45100,
      "average": {
        "on_policy_value_p0": -0.055553888563101,
        "best_response_value_p0": -0.055002322702909,
        "best_response_value_p1": 0.057149726576171,
        "unilateral_improvement_p0": 0.000551565860192,
        "unilateral_improvement_p1": 0.00159583801307,
        "nash_conv": 0.002147403873261,
        "exploitability": 0.001073701936631
      },
      "current": {
        "on_policy_value_p0": 0.047729201220468,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.118937465446199,
        "unilateral_improvement_p0": 0.118937465446199,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.285604132112865,
        "exploitability": 0.142802066056433
      }
    },
    {
      "iteration": 45200,
      "average": {
        "on_policy_value_p0": -0.055555750109294,
        "best_response_value_p0": -0.054745654268985,
        "best_response_value_p1": 0.056853197460061,
        "unilateral_improvement_p0": 0.000810095840309,
        "unilateral_improvement_p1": 0.001297447350767,
        "nash_conv": 0.002107543191076,
        "exploitability": 0.001053771595538
      },
      "current": {
        "on_policy_value_p0": 0.077865536437571,
        "best_response_value_p0": 0.139423135477297,
        "best_response_value_p1": 0.210218135253881,
        "unilateral_improvement_p0": 0.061557599039726,
        "unilateral_improvement_p1": 0.288083671691452,
        "nash_conv": 0.349641270731178,
        "exploitability": 0.174820635365589
      }
    },
    {
      "iteration": 45300,
      "average": {
        "on_policy_value_p0": -0.05555587518189,
        "best_response_value_p0": -0.054377076919108,
        "best_response_value_p1": 0.056415896493128,
        "unilateral_improvement_p0": 0.001178798262783,
        "unilateral_improvement_p1": 0.000860021311238,
        "nash_conv": 0.00203881957402,
        "exploitability": 0.00101940978701
      },
      "current": {
        "on_policy_value_p0": 0.089159827862239,
        "best_response_value_p0": 0.089159827862239,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.422493161195572,
        "nash_conv": 0.422493161195572,
        "exploitability": 0.211246580597786
      }
    },
    {
      "iteration": 45400,
      "average": {
        "on_policy_value_p0": -0.05555450528775,
        "best_response_value_p0": -0.054106780179908,
        "best_response_value_p1": 0.056239148868317,
        "unilateral_improvement_p0": 0.001447725107842,
        "unilateral_improvement_p1": 0.000684643580567,
        "nash_conv": 0.00213236868841,
        "exploitability": 0.001066184344205
      },
      "current": {
        "on_policy_value_p0": 0.045034718930992,
        "best_response_value_p0": 0.076597228804682,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.03156250987369,
        "unilateral_improvement_p1": 0.378368052264326,
        "nash_conv": 0.409930562138015,
        "exploitability": 0.204965281069008
      }
    },
    {
      "iteration": 45500,
      "average": {
        "on_policy_value_p0": -0.055552426227238,
        "best_response_value_p0": -0.053954224372589,
        "best_response_value_p1": 0.056115546343331,
        "unilateral_improvement_p0": 0.001598201854649,
        "unilateral_improvement_p1": 0.000563120116093,
        "nash_conv": 0.002161321970742,
        "exploitability": 0.001080660985371
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 45600,
      "average": {
        "on_policy_value_p0": -0.055549855730161,
        "best_response_value_p0": -0.053887424036936,
        "best_response_value_p1": 0.055992485934683,
        "unilateral_improvement_p0": 0.001662431693225,
        "unilateral_improvement_p1": 0.000442630204522,
        "nash_conv": 0.002105061897747,
        "exploitability": 0.001052530948873
      },
      "current": {
        "on_policy_value_p0": -0.108463068378895,
        "best_response_value_p0": 0.112435132477219,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.220898200856114,
        "unilateral_improvement_p1": 0.224870264954438,
        "nash_conv": 0.445768465810553,
        "exploitability": 0.222884232905276
      }
    },
    {
      "iteration": 45700,
      "average": {
        "on_policy_value_p0": -0.055547545115081,
        "best_response_value_p0": -0.054197798846602,
        "best_response_value_p1": 0.056351471607481,
        "unilateral_improvement_p0": 0.001349746268479,
        "unilateral_improvement_p1": 0.000803926492399,
        "nash_conv": 0.002153672760878,
        "exploitability": 0.001076836380439
      },
      "current": {
        "on_policy_value_p0": -0.30440365319783,
        "best_response_value_p0": 0.30440365319783,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.60880730639566,
        "unilateral_improvement_p1": 0.028929680135503,
        "nash_conv": 0.637736986531164,
        "exploitability": 0.318868493265582
      }
    },
    {
      "iteration": 45800,
      "average": {
        "on_policy_value_p0": -0.055547670350466,
        "best_response_value_p0": -0.054804534385492,
        "best_response_value_p1": 0.056956235497712,
        "unilateral_improvement_p0": 0.000743135964974,
        "unilateral_improvement_p1": 0.001408565147246,
        "nash_conv": 0.00215170111222,
        "exploitability": 0.00107585055611
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 45900,
      "average": {
        "on_policy_value_p0": -0.055550295309987,
        "best_response_value_p0": -0.055204492859904,
        "best_response_value_p1": 0.057314828552565,
        "unilateral_improvement_p0": 0.000345802450084,
        "unilateral_improvement_p1": 0.001764533242578,
        "nash_conv": 0.002110335692662,
        "exploitability": 0.001055167846331
      },
      "current": {
        "on_policy_value_p0": -0.071265258105369,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.11586812804862,
        "unilateral_improvement_p0": 0.404598591438703,
        "unilateral_improvement_p1": 0.04460286994325,
        "nash_conv": 0.449201461381953,
        "exploitability": 0.224600730690977
      }
    },
    {
      "iteration": 46000,
      "average": {
        "on_policy_value_p0": -0.055553043196379,
        "best_response_value_p0": -0.054920585008635,
        "best_response_value_p1": 0.057057512148734,
        "unilateral_improvement_p0": 0.000632458187744,
        "unilateral_improvement_p1": 0.001504468952355,
        "nash_conv": 0.002136927140099,
        "exploitability": 0.00106846357005
      },
      "current": {
        "on_policy_value_p0": 0.121356245955957,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.211977087377376,
        "unilateral_improvement_p1": 0.288022912622624,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 46100,
      "average": {
        "on_policy_value_p0": -0.055553965780824,
        "best_response_value_p0": -0.05445608231394,
        "best_response_value_p1": 0.056520261878289,
        "unilateral_improvement_p0": 0.001097883466884,
        "unilateral_improvement_p1": 0.000966296097465,
        "nash_conv": 0.002064179564349,
        "exploitability": 0.001032089782174
      },
      "current": {
        "on_policy_value_p0": 0.165934422872195,
        "best_response_value_p0": 0.26472601254818,
        "best_response_value_p1": 0.302416820648029,
        "unilateral_improvement_p0": 0.098791589675985,
        "unilateral_improvement_p1": 0.468351243520224,
        "nash_conv": 0.56714283319621,
        "exploitability": 0.283571416598105
      }
    },
    {
      "iteration": 46200,
      "average": {
        "on_policy_value_p0": -0.055552680827285,
        "best_response_value_p0": -0.054069678940531,
        "best_response_value_p1": 0.056172686254708,
        "unilateral_improvement_p0": 0.001483001886754,
        "unilateral_improvement_p1": 0.000620005427423,
        "nash_conv": 0.002103007314178,
        "exploitability": 0.001051503657089
      },
      "current": {
        "on_policy_value_p0": 0.041083884974642,
        "best_response_value_p0": 0.125736842372811,
        "best_response_value_p1": 0.418448472773861,
        "unilateral_improvement_p0": 0.084652957398168,
        "unilateral_improvement_p1": 0.459532357748503,
        "nash_conv": 0.544185315146671,
        "exploitability": 0.272092657573336
      }
    },
    {
      "iteration": 46300,
      "average": {
        "on_policy_value_p0": -0.055550547221543,
        "best_response_value_p0": -0.053927886799672,
        "best_response_value_p1": 0.056071310624167,
        "unilateral_improvement_p0": 0.00162266042187,
        "unilateral_improvement_p1": 0.000520763402624,
        "nash_conv": 0.002143423824494,
        "exploitability": 0.001071711912247
      },
      "current": {
        "on_policy_value_p0": -0.045528044299856,
        "best_response_value_p0": 0.13631463713343,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.181842681433285,
        "unilateral_improvement_p1": 0.454471955700144,
        "nash_conv": 0.63631463713343,
        "exploitability": 0.318157318566715
      }
    },
    {
      "iteration": 46400,
      "average": {
        "on_policy_value_p0": -0.055548653520302,
        "best_response_value_p0": -0.054412750671162,
        "best_response_value_p1": 0.056559702232081,
        "unilateral_improvement_p0": 0.00113590284914,
        "unilateral_improvement_p1": 0.001011048711779,
        "nash_conv": 0.002146951560919,
        "exploitability": 0.001073475780459
      },
      "current": {
        "on_policy_value_p0": -0.478924297411823,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.478924297411823,
        "unilateral_improvement_p0": 0.64559096407849,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.64559096407849,
        "exploitability": 0.322795482039245
      }
    },
    {
      "iteration": 46500,
      "average": {
        "on_policy_value_p0": -0.055550099154345,
        "best_response_value_p0": -0.055023322374668,
        "best_response_value_p1": 0.057149385578462,
        "unilateral_improvement_p0": 0.000526776779677,
        "unilateral_improvement_p1": 0.001599286424117,
        "nash_conv": 0.002126063203794,
        "exploitability": 0.001063031601897
      },
      "current": {
        "on_policy_value_p0": -0.157556492857796,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.157556492857796,
        "unilateral_improvement_p0": 0.324223159524463,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.324223159524463,
        "exploitability": 0.162111579762231
      }
    },
    {
      "iteration": 46600,
      "average": {
        "on_policy_value_p0": -0.055552679642992,
        "best_response_value_p0": -0.055057423295646,
        "best_response_value_p1": 0.057193372469546,
        "unilateral_improvement_p0": 0.000495256347347,
        "unilateral_improvement_p1": 0.001640692826553,
        "nash_conv": 0.0021359491739,
        "exploitability": 0.00106797458695
      },
      "current": {
        "on_policy_value_p0": -0.027819642096627,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.194486308763293,
        "unilateral_improvement_p1": 0.13884702457004,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 46700,
      "average": {
        "on_policy_value_p0": -0.055554922523023,
        "best_response_value_p0": -0.054939527314283,
        "best_response_value_p1": 0.056989422417693,
        "unilateral_improvement_p0": 0.00061539520874,
        "unilateral_improvement_p1": 0.001434499894669,
        "nash_conv": 0.002049895103409,
        "exploitability": 0.001024947551705
      },
      "current": {
        "on_policy_value_p0": 0.074378740941951,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.148757481883902,
        "unilateral_improvement_p0": 0.092287925724716,
        "unilateral_improvement_p1": 0.223136222825853,
        "nash_conv": 0.315424148550568,
        "exploitability": 0.157712074275284
      }
    },
    {
      "iteration": 46800,
      "average": {
        "on_policy_value_p0": -0.055555952362837,
        "best_response_value_p0": -0.054577508285221,
        "best_response_value_p1": 0.056649638209653,
        "unilateral_improvement_p0": 0.000978444077616,
        "unilateral_improvement_p1": 0.001093685846816,
        "nash_conv": 0.002072129924432,
        "exploitability": 0.001036064962216
      },
      "current": {
        "on_policy_value_p0": 0.08069512931204,
        "best_response_value_p0": 0.112380387364634,
        "best_response_value_p1": 0.269962817228146,
        "unilateral_improvement_p0": 0.031685258052594,
        "unilateral_improvement_p1": 0.350657946540187,
        "nash_conv": 0.38234320459278,
        "exploitability": 0.19117160229639
      }
    },
    {
      "iteration": 46900,
      "average": {
        "on_policy_value_p0": -0.055555337473231,
        "best_response_value_p0": -0.054266613581607,
        "best_response_value_p1": 0.056295995082458,
        "unilateral_improvement_p0": 0.001288723891625,
        "unilateral_improvement_p1": 0.000740657609227,
        "nash_conv": 0.002029381500851,
        "exploitability": 0.001014690750426
      },
      "current": {
        "on_policy_value_p0": 0.070639829247692,
        "best_response_value_p0": 0.070639829247692,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.403973162581025,
        "nash_conv": 0.403973162581025,
        "exploitability": 0.201986581290512
      }
    },
    {
      "iteration": 47000,
      "average": {
        "on_policy_value_p0": -0.05555364392922,
        "best_response_value_p0": -0.054051573163184,
        "best_response_value_p1": 0.056176216369516,
        "unilateral_improvement_p0": 0.001502070766036,
        "unilateral_improvement_p1": 0.000622572440296,
        "nash_conv": 0.002124643206332,
        "exploitability": 0.001062321603166
      },
      "current": {
        "on_policy_value_p0": 0.015901445080557,
        "best_response_value_p0": 0.134863776505553,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.118962331424997,
        "unilateral_improvement_p1": 0.34923477841389,
        "nash_conv": 0.468197109838887,
        "exploitability": 0.234098554919443
      }
    },
    {
      "iteration": 47100,
      "average": {
        "on_policy_value_p0": -0.055551515124285,
        "best_response_value_p0": -0.053933751173512,
        "best_response_value_p1": 0.056056946271066,
        "unilateral_improvement_p0": 0.001617763950772,
        "unilateral_improvement_p1": 0.000505431146782,
        "nash_conv": 0.002123195097554,
        "exploitability": 0.001061597548777
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 47200,
      "average": {
        "on_policy_value_p0": -0.055548934681491,
        "best_response_value_p0": -0.053991268612854,
        "best_response_value_p1": 0.055978063230093,
        "unilateral_improvement_p0": 0.001557666068637,
        "unilateral_improvement_p1": 0.000429128548602,
        "nash_conv": 0.001986794617239,
        "exploitability": 0.000993397308619
      },
      "current": {
        "on_policy_value_p0": -0.175639959916389,
        "best_response_value_p0": 0.175639959916389,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.351279919832777,
        "unilateral_improvement_p1": 0.157693373416945,
        "nash_conv": 0.508973293249722,
        "exploitability": 0.254486646624861
      }
    },
    {
      "iteration": 47300,
      "average": {
        "on_policy_value_p0": -0.055547491642121,
        "best_response_value_p0": -0.054445836429113,
        "best_response_value_p1": 0.056564438008324,
        "unilateral_improvement_p0": 0.001101655213008,
        "unilateral_improvement_p1": 0.001016946366203,
        "nash_conv": 0.002118601579211,
        "exploitability": 0.001059300789605
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 47400,
      "average": {
        "on_policy_value_p0": -0.055548565250488,
        "best_response_value_p0": -0.055034206675747,
        "best_response_value_p1": 0.057126679112572,
        "unilateral_improvement_p0": 0.000514358574741,
        "unilateral_improvement_p1": 0.001578113862085,
        "nash_conv": 0.002092472436826,
        "exploitability": 0.001046236218413
      },
      "current": {
        "on_policy_value_p0": -0.245993225813351,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.245993225813351,
        "unilateral_improvement_p0": 0.579326559146684,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.579326559146684,
        "exploitability": 0.289663279573342
      }
    },
    {
      "iteration": 47500,
      "average": {
        "on_policy_value_p0": -0.055551443175746,
        "best_response_value_p0": -0.055157271711808,
        "best_response_value_p1": 0.057230685729879,
        "unilateral_improvement_p0": 0.000394171463938,
        "unilateral_improvement_p1": 0.001679242554133,
        "nash_conv": 0.002073414018071,
        "exploitability": 0.001036707009035
      },
      "current": {
        "on_policy_value_p0": 0.064841836715259,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.268491496618075,
        "unilateral_improvement_p1": 0.231508503381925,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 47600,
      "average": {
        "on_policy_value_p0": -0.055553521285671,
        "best_response_value_p0": -0.054767005061562,
        "best_response_value_p1": 0.056883167462753,
        "unilateral_improvement_p0": 0.000786516224108,
        "unilateral_improvement_p1": 0.001329646177082,
        "nash_conv": 0.00211616240119,
        "exploitability": 0.001058081200595
      },
      "current": {
        "on_policy_value_p0": 0.177815789163561,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.188964911660456,
        "unilateral_improvement_p0": 0.155517544169772,
        "unilateral_improvement_p1": 0.366780700824017,
        "nash_conv": 0.522298244993789,
        "exploitability": 0.261149122496895
      }
    },
    {
      "iteration": 47700,
      "average": {
        "on_policy_value_p0": -0.055553654524459,
        "best_response_value_p0": -0.054392297626851,
        "best_response_value_p1": 0.056308208498163,
        "unilateral_improvement_p0": 0.001161356897608,
        "unilateral_improvement_p1": 0.000754553973703,
        "nash_conv": 0.001915910871311,
        "exploitability": 0.000957955435656
      },
      "current": {
        "on_policy_value_p0": 0.112481983315264,
        "best_response_value_p0": 0.197973257324047,
        "best_response_value_p1": 0.329017451982433,
        "unilateral_improvement_p0": 0.085491274008783,
        "unilateral_improvement_p1": 0.441499435297697,
        "nash_conv": 0.526990709306481,
        "exploitability": 0.26349535465324
      }
    },
    {
      "iteration": 47800,
      "average": {
        "on_policy_value_p0": -0.055552162010225,
        "best_response_value_p0": -0.054064748454763,
        "best_response_value_p1": 0.056173536043448,
        "unilateral_improvement_p0": 0.001487413555462,
        "unilateral_improvement_p1": 0.000621374033223,
        "nash_conv": 0.002108787588685,
        "exploitability": 0.001054393794343
      },
      "current": {
        "on_policy_value_p0": -0.020825137041409,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.458349725917182,
        "unilateral_improvement_p0": 0.187491803708076,
        "unilateral_improvement_p1": 0.437524588875773,
        "nash_conv": 0.625016392583849,
        "exploitability": 0.312508196291924
      }
    },
    {
      "iteration": 47900,
      "average": {
        "on_policy_value_p0": -0.05555010903555,
        "best_response_value_p0": -0.054000862405549,
        "best_response_value_p1": 0.056060443728781,
        "unilateral_improvement_p0": 0.001549246630001,
        "unilateral_improvement_p1": 0.000510334693231,
        "nash_conv": 0.002059581323232,
        "exploitability": 0.001029790661616
      },
      "current": {
        "on_policy_value_p0": -0.162335611087831,
        "best_response_value_p0": 0.058442925941446,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.220778537029277,
        "unilateral_improvement_p1": 0.337664388912169,
        "nash_conv": 0.558442925941446,
        "exploitability": 0.279221462970723
      }
    },
    {
      "iteration": 48000,
      "average": {
        "on_policy_value_p0": -0.055548917494765,
        "best_response_value_p0": -0.054633616211487,
        "best_response_value_p1": 0.056721060617923,
        "unilateral_improvement_p0": 0.000915301283278,
        "unilateral_improvement_p1": 0.001172143123158,
        "nash_conv": 0.002087444406436,
        "exploitability": 0.001043722203218
      },
      "current": {
        "on_policy_value_p0": -0.410234311330008,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.410234311330008,
        "unilateral_improvement_p0": 0.576900977996675,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.576900977996675,
        "exploitability": 0.288450488998338
      }
    },
    {
      "iteration": 48100,
      "average": {
        "on_policy_value_p0": -0.055550791867463,
        "best_response_value_p0": -0.055097515977607,
        "best_response_value_p1": 0.057165909730251,
        "unilateral_improvement_p0": 0.000453275889856,
        "unilateral_improvement_p1": 0.001615117862788,
        "nash_conv": 0.002068393752644,
        "exploitability": 0.001034196876322
      },
      "current": {
        "on_policy_value_p0": -0.099726603865462,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.266393270532129,
        "unilateral_improvement_p1": 0.066940062801205,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 48200,
      "average": {
        "on_policy_value_p0": -0.055553318360398,
        "best_response_value_p0": -0.055060264407381,
        "best_response_value_p1": 0.057159849060457,
        "unilateral_improvement_p0": 0.000493053953017,
        "unilateral_improvement_p1": 0.001606530700059,
        "nash_conv": 0.002099584653077,
        "exploitability": 0.001049792326538
      },
      "current": {
        "on_policy_value_p0": 0.024352445443526,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.142314221223141,
        "unilateral_improvement_p0": 0.142314221223141,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.308980887889807,
        "exploitability": 0.154490443944904
      }
    },
    {
      "iteration": 48300,
      "average": {
        "on_policy_value_p0": -0.055555361607333,
        "best_response_value_p0": -0.054919961038646,
        "best_response_value_p1": 0.05691996561728,
        "unilateral_improvement_p0": 0.000635400568687,
        "unilateral_improvement_p1": 0.001364604009947,
        "nash_conv": 0.002000004578634,
        "exploitability": 0.001000002289317
      },
      "current": {
        "on_policy_value_p0": 0.087302240213856,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.174604480427712,
        "unilateral_improvement_p0": 0.07936442645281,
        "unilateral_improvement_p1": 0.261906720641569,
        "nash_conv": 0.341271147094379,
        "exploitability": 0.17063557354719
      }
    },
    {
      "iteration": 48400,
      "average": {
        "on_policy_value_p0": -0.055556065546221,
        "best_response_value_p0": -0.054530836327108,
        "best_response_value_p1": 0.056559900365205,
        "unilateral_improvement_p0": 0.001025229219113,
        "unilateral_improvement_p1": 0.001003834818984,
        "nash_conv": 0.002029064038097,
        "exploitability": 0.001014532019049
      },
      "current": {
        "on_policy_value_p0": 0.097428909087203,
        "best_response_value_p0": 0.105553866293054,
        "best_response_value_p1": 0.317083418921631,
        "unilateral_improvement_p0": 0.008124957205851,
        "unilateral_improvement_p1": 0.414512328008834,
        "nash_conv": 0.422637285214685,
        "exploitability": 0.211318642607342
      }
    },
    {
      "iteration": 48500,
      "average": {
        "on_policy_value_p0": -0.055555236189601,
        "best_response_value_p0": -0.054242315186458,
        "best_response_value_p1": 0.056256236028888,
        "unilateral_improvement_p0": 0.001312921003143,
        "unilateral_improvement_p1": 0.000700999839287,
        "nash_conv": 0.00201392084243,
        "exploitability": 0.001006960421215
      },
      "current": {
        "on_policy_value_p0": 0.065081281162739,
        "best_response_value_p0": 0.065081281162739,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.398414614496072,
        "nash_conv": 0.398414614496072,
        "exploitability": 0.199207307248036
      }
    },
    {
      "iteration": 48600,
      "average": {
        "on_policy_value_p0": -0.0555534998395,
        "best_response_value_p0": -0.054048933603433,
        "best_response_value_p1": 0.056140482456812,
        "unilateral_improvement_p0": 0.001504566236067,
        "unilateral_improvement_p1": 0.000586982617311,
        "nash_conv": 0.002091548853378,
        "exploitability": 0.001045774426689
      },
      "current": {
        "on_policy_value_p0": 0.005139778706637,
        "best_response_value_p0": 0.156387109253393,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.151247330546756,
        "unilateral_improvement_p1": 0.33847311203997,
        "nash_conv": 0.489720442586726,
        "exploitability": 0.244860221293363
      }
    },
    {
      "iteration": 48700,
      "average": {
        "on_policy_value_p0": -0.055551383730835,
        "best_response_value_p0": -0.053937635344274,
        "best_response_value_p1": 0.056025204258748,
        "unilateral_improvement_p0": 0.001613748386561,
        "unilateral_improvement_p1": 0.000473820527913,
        "nash_conv": 0.002087568914474,
        "exploitability": 0.001043784457237
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 48800,
      "average": {
        "on_policy_value_p0": -0.055548906735848,
        "best_response_value_p0": -0.054047764826004,
        "best_response_value_p1": 0.05604243437292,
        "unilateral_improvement_p0": 0.001501141909844,
        "unilateral_improvement_p1": 0.000493527637072,
        "nash_conv": 0.001994669546916,
        "exploitability": 0.000997334773458
      },
      "current": {
        "on_policy_value_p0": -0.197318271745555,
        "best_response_value_p0": 0.197318271745555,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.39463654349111,
        "unilateral_improvement_p1": 0.136015061587778,
        "nash_conv": 0.530651605078889,
        "exploitability": 0.265325802539444
      }
    },
    {
      "iteration": 48900,
      "average": {
        "on_policy_value_p0": -0.055547785891803,
        "best_response_value_p0": -0.054526212369948,
        "best_response_value_p1": 0.056609491426009,
        "unilateral_improvement_p0": 0.001021573521854,
        "unilateral_improvement_p1": 0.001061705534206,
        "nash_conv": 0.00208327905606,
        "exploitability": 0.00104163952803
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 49000,
      "average": {
        "on_policy_value_p0": -0.055549045091794,
        "best_response_value_p0": -0.055095206494364,
        "best_response_value_p1": 0.057135212725582,
        "unilateral_improvement_p0": 0.00045383859743,
        "unilateral_improvement_p1": 0.001586167633788,
        "nash_conv": 0.002040006231218,
        "exploitability": 0.001020003115609
      },
      "current": {
        "on_policy_value_p0": -0.222853349019698,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.222853349019698,
        "unilateral_improvement_p0": 0.556186682353032,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.556186682353032,
        "exploitability": 0.278093341176516
      }
    },
    {
      "iteration": 49100,
      "average": {
        "on_policy_value_p0": -0.055551816565509,
        "best_response_value_p0": -0.055157820686816,
        "best_response_value_p1": 0.057186152742126,
        "unilateral_improvement_p0": 0.000393995878693,
        "unilateral_improvement_p1": 0.001634336176617,
        "nash_conv": 0.00202833205531,
        "exploitability": 0.001014166027655
      },
      "current": {
        "on_policy_value_p0": 0.071129107145391,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.262204226187942,
        "unilateral_improvement_p1": 0.237795773812058,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 49200,
      "average": {
        "on_policy_value_p0": -0.055553703449787,
        "best_response_value_p0": -0.054755181061616,
        "best_response_value_p1": 0.056834096687469,
        "unilateral_improvement_p0": 0.000798522388172,
        "unilateral_improvement_p1": 0.001280393237682,
        "nash_conv": 0.002078915625854,
        "exploitability": 0.001039457812927
      },
      "current": {
        "on_policy_value_p0": 0.185982801296617,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.205298935926568,
        "unilateral_improvement_p0": 0.147350532036716,
        "unilateral_improvement_p1": 0.391281737223185,
        "nash_conv": 0.538632269259901,
        "exploitability": 0.269316134629951
      }
    },
    {
      "iteration": 49300,
      "average": {
        "on_policy_value_p0": -0.05555372918825,
        "best_response_value_p0": -0.05439855667797,
        "best_response_value_p1": 0.056268892278196,
        "unilateral_improvement_p0": 0.00115517251028,
        "unilateral_improvement_p1": 0.000715163089946,
        "nash_conv": 0.001870335600225,
        "exploitability": 0.000935167800113
      },
      "current": {
        "on_policy_value_p0": 0.110883031396566,
        "best_response_value_p0": 0.193456087902827,
        "best_response_value_p1": 0.334853886987478,
        "unilateral_improvement_p0": 0.082573056506261,
        "unilateral_improvement_p1": 0.445736918384044,
        "nash_conv": 0.528309974890305,
        "exploitability": 0.264154987445152
      }
    },
    {
      "iteration": 49400,
      "average": {
        "on_policy_value_p0": -0.05555224181111,
        "best_response_value_p0": -0.054073407483395,
        "best_response_value_p1": 0.056149048590328,
        "unilateral_improvement_p0": 0.001478834327715,
        "unilateral_improvement_p1": 0.000596806779218,
        "nash_conv": 0.002075641106933,
        "exploitability": 0.001037820553467
      },
      "current": {
        "on_policy_value_p0": -0.016664946359258,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.466670107281485,
        "unilateral_improvement_p0": 0.183331613025924,
        "unilateral_improvement_p1": 0.450005160922227,
        "nash_conv": 0.633336773948151,
        "exploitability": 0.316668386974076
      }
    },
    {
      "iteration": 49500,
      "average": {
        "on_policy_value_p0": -0.055550222779335,
        "best_response_value_p0": -0.054025309594617,
        "best_response_value_p1": 0.056038161027388,
        "unilateral_improvement_p0": 0.001524913184717,
        "unilateral_improvement_p1": 0.000487938248054,
        "nash_conv": 0.002012851432771,
        "exploitability": 0.001006425716385
      },
      "current": {
        "on_policy_value_p0": -0.177827771050955,
        "best_response_value_p0": 0.059275923683652,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.237103694734607,
        "unilateral_improvement_p1": 0.322172228949045,
        "nash_conv": 0.559275923683652,
        "exploitability": 0.279637961841826
      }
    },
    {
      "iteration": 49600,
      "average": {
        "on_policy_value_p0": -0.055549128767314,
        "best_response_value_p0": -0.054658263161748,
        "best_response_value_p1": 0.056706168812088,
        "unilateral_improvement_p0": 0.000890865605566,
        "unilateral_improvement_p1": 0.001157040044774,
        "nash_conv": 0.00204790565034,
        "exploitability": 0.00102395282517
      },
      "current": {
        "on_policy_value_p0": -0.402579715806373,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.402579715806373,
        "unilateral_improvement_p0": 0.56924638247304,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.56924638247304,
        "exploitability": 0.28462319123652
      }
    },
    {
      "iteration": 49700,
      "average": {
        "on_policy_value_p0": -0.05555088758439,
        "best_response_value_p0": -0.055091672153225,
        "best_response_value_p1": 0.057127800068802,
        "unilateral_improvement_p0": 0.000459215431165,
        "unilateral_improvement_p1": 0.001576912484412,
        "nash_conv": 0.002036127915577,
        "exploitability": 0.001018063957788
      },
      "current": {
        "on_policy_value_p0": -0.100688495609264,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.267355162275931,
        "unilateral_improvement_p1": 0.065978171057402,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 49800,
      "average": {
        "on_policy_value_p0": -0.055553248914652,
        "best_response_value_p0": -0.055060744716869,
        "best_response_value_p1": 0.057129255625112,
        "unilateral_improvement_p0": 0.000492504197784,
        "unilateral_improvement_p1": 0.00157600671046,
        "nash_conv": 0.002068510908243,
        "exploitability": 0.001034255454122
      },
      "current": {
        "on_policy_value_p0": 0.016412532258803,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.150072961527363,
        "unilateral_improvement_p0": 0.150254134407864,
        "unilateral_improvement_p1": 0.166485493786166,
        "nash_conv": 0.31673962819403,
        "exploitability": 0.158369814097015
      }
    },
    {
      "iteration": 49900,
      "average": {
        "on_policy_value_p0": -0.055555196215185,
        "best_response_value_p0": -0.054950402543087,
        "best_response_value_p1": 0.056907923831079,
        "unilateral_improvement_p0": 0.000604793672098,
        "unilateral_improvement_p1": 0.001352727615894,
        "nash_conv": 0.001957521287992,
        "exploitability": 0.000978760643996
      },
      "current": {
        "on_policy_value_p0": 0.082521836255469,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.165043672510939,
        "unilateral_improvement_p0": 0.084144830411197,
        "unilateral_improvement_p1": 0.247565508766408,
        "nash_conv": 0.331710339177606,
        "exploitability": 0.165855169588803
      }
    },
    {
      "iteration": 50000,
      "average": {
        "on_policy_value_p0": -0.05555595695098,
        "best_response_value_p0": -0.054575375618386,
        "best_response_value_p1": 0.056573033227029,
        "unilateral_improvement_p0": 0.000980581332594,
        "unilateral_improvement_p1": 0.001017076276048,
        "nash_conv": 0.001997657608643,
        "exploitability": 0.000998828804321
      },
      "current": {
        "on_policy_value_p0": 0.086892731187916,
        "best_response_value_p0": 0.107245859523251,
        "best_response_value_p1": 0.292627076662662,
        "unilateral_improvement_p0": 0.020353128335335,
        "unilateral_improvement_p1": 0.379519807850578,
        "nash_conv": 0.399872936185914,
        "exploitability": 0.199936468092957
      }
    },
    {
      "iteration": 50100,
      "average": {
        "on_policy_value_p0": -0.055555287979842,
        "best_response_value_p0": -0.054292218449536,
        "best_response_value_p1": 0.056266300051993,
        "unilateral_improvement_p0": 0.001263069530307,
        "unilateral_improvement_p1": 0.000711012072151,
        "nash_conv": 0.001974081602458,
        "exploitability": 0.000987040801229
      },
      "current": {
        "on_policy_value_p0": 0.067372210506446,
        "best_response_value_p0": 0.067372210506446,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.400705543839779,
        "nash_conv": 0.400705543839779,
        "exploitability": 0.200352771919889
      }
    },
    {
      "iteration": 50200,
      "average": {
        "on_policy_value_p0": -0.055553727651719,
        "best_response_value_p0": -0.054097522191998,
        "best_response_value_p1": 0.056154215788941,
        "unilateral_improvement_p0": 0.001456205459721,
        "unilateral_improvement_p1": 0.000600488137222,
        "nash_conv": 0.002056693596943,
        "exploitability": 0.001028346798472
      },
      "current": {
        "on_policy_value_p0": 0.011891480623964,
        "best_response_value_p0": 0.142883705418738,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.130992224794774,
        "unilateral_improvement_p1": 0.345224813957298,
        "nash_conv": 0.476217038752072,
        "exploitability": 0.238108519376036
      }
    },
    {
      "iteration": 50300,
      "average": {
        "on_policy_value_p0": -0.055551802530491,
        "best_response_value_p0": -0.053988361676148,
        "best_response_value_p1": 0.056042577188963,
        "unilateral_improvement_p0": 0.001563440854343,
        "unilateral_improvement_p1": 0.000490774658471,
        "nash_conv": 0.002054215512815,
        "exploitability": 0.001027107756407
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 50400,
      "average": {
        "on_policy_value_p0": -0.055549466766093,
        "best_response_value_p0": -0.054035481898684,
        "best_response_value_p1": 0.055946537255796,
        "unilateral_improvement_p0": 0.00151398486741,
        "unilateral_improvement_p1": 0.000397070489703,
        "nash_conv": 0.001911055357112,
        "exploitability": 0.000955527678556
      },
      "current": {
        "on_policy_value_p0": -0.170185791638632,
        "best_response_value_p0": 0.170185791638632,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.340371583277264,
        "unilateral_improvement_p1": 0.163147541694701,
        "nash_conv": 0.503519124971965,
        "exploitability": 0.251759562485983
      }
    },
    {
      "iteration": 50500,
      "average": {
        "on_policy_value_p0": -0.055548059177056,
        "best_response_value_p0": -0.054445625193201,
        "best_response_value_p1": 0.056495818040108,
        "unilateral_improvement_p0": 0.001102433983855,
        "unilateral_improvement_p1": 0.000947758863052,
        "nash_conv": 0.002050192846907,
        "exploitability": 0.001025096423454
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 50600,
      "average": {
        "on_policy_value_p0": -0.055548844069245,
        "best_response_value_p0": -0.054996786671739,
        "best_response_value_p1": 0.057037484726348,
        "unilateral_improvement_p0": 0.000552057397506,
        "unilateral_improvement_p1": 0.001488640657102,
        "nash_conv": 0.002040698054609,
        "exploitability": 0.001020349027304
      },
      "current": {
        "on_policy_value_p0": -0.282630372160862,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.282630372160862,
        "unilateral_improvement_p0": 0.615963705494196,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.615963705494196,
        "exploitability": 0.307981852747098
      }
    },
    {
      "iteration": 50700,
      "average": {
        "on_policy_value_p0": -0.05555142554934,
        "best_response_value_p0": -0.055195375276351,
        "best_response_value_p1": 0.057214873470036,
        "unilateral_improvement_p0": 0.000356050272989,
        "unilateral_improvement_p1": 0.001663447920696,
        "nash_conv": 0.002019498193685,
        "exploitability": 0.001009749096843
      },
      "current": {
        "on_policy_value_p0": 0.028919812768086,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.153883297497925,
        "unilateral_improvement_p0": 0.304413520565247,
        "unilateral_improvement_p1": 0.182803110266012,
        "nash_conv": 0.487216630831259,
        "exploitability": 0.243608315415629
      }
    },
    {
      "iteration": 50800,
      "average": {
        "on_policy_value_p0": -0.055553543243919,
        "best_response_value_p0": -0.054872046444918,
        "best_response_value_p1": 0.056917221851521,
        "unilateral_improvement_p0": 0.000681496799001,
        "unilateral_improvement_p1": 0.001363678607602,
        "nash_conv": 0.002045175406604,
        "exploitability": 0.001022587703302
      },
      "current": {
        "on_policy_value_p0": 0.142244037023611,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.191089296309722,
        "unilateral_improvement_p1": 0.308910703690278,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 50900,
      "average": {
        "on_policy_value_p0": -0.055554034853817,
        "best_response_value_p0": -0.054474521775428,
        "best_response_value_p1": 0.056398429129239,
        "unilateral_improvement_p0": 0.001079513078388,
        "unilateral_improvement_p1": 0.000844394275423,
        "nash_conv": 0.001923907353811,
        "exploitability": 0.000961953676906
      },
      "current": {
        "on_policy_value_p0": 0.143079874925901,
        "best_response_value_p0": 0.235376042091543,
        "best_response_value_p1": 0.315407665668717,
        "unilateral_improvement_p0": 0.092296167165642,
        "unilateral_improvement_p1": 0.458487540594618,
        "nash_conv": 0.550783707760259,
        "exploitability": 0.27539185388013
      }
    },
    {
      "iteration": 51000,
      "average": {
        "on_policy_value_p0": -0.055552807742295,
        "best_response_value_p0": -0.054126480255386,
        "best_response_value_p1": 0.05614522153385,
        "unilateral_improvement_p0": 0.00142632748691,
        "unilateral_improvement_p1": 0.000592413791555,
        "nash_conv": 0.002018741278464,
        "exploitability": 0.001009370639232
      },
      "current": {
        "on_policy_value_p0": 0.026643511357904,
        "best_response_value_p0": 0.135692716309767,
        "best_response_value_p1": 0.429391221288209,
        "unilateral_improvement_p0": 0.109049204951863,
        "unilateral_improvement_p1": 0.456034732646112,
        "nash_conv": 0.565083937597975,
        "exploitability": 0.282541968798988
      }
    },
    {
      "iteration": 51100,
      "average": {
        "on_policy_value_p0": -0.05555096957641,
        "best_response_value_p0": -0.054010758596183,
        "best_response_value_p1": 0.056048774833692,
        "unilateral_improvement_p0": 0.001540210980226,
        "unilateral_improvement_p1": 0.000497805257282,
        "nash_conv": 0.002038016237508,
        "exploitability": 0.001019008118754
      },
      "current": {
        "on_policy_value_p0": -0.057493419381451,
        "best_response_value_p0": 0.128337720412366,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.185831139793817,
        "unilateral_improvement_p1": 0.442506580618549,
        "nash_conv": 0.628337720412366,
        "exploitability": 0.314168860206183
      }
    },
    {
      "iteration": 51200,
      "average": {
        "on_policy_value_p0": -0.05554931354458,
        "best_response_value_p0": -0.054442609088656,
        "best_response_value_p1": 0.056486304829618,
        "unilateral_improvement_p0": 0.001106704455924,
        "unilateral_improvement_p1": 0.000936991285037,
        "nash_conv": 0.002043695740962,
        "exploitability": 0.001021847870481
      },
      "current": {
        "on_policy_value_p0": -0.498406004201389,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.498406004201389,
        "unilateral_improvement_p0": 0.665072670868056,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.665072670868056,
        "exploitability": 0.332536335434028
      }
    },
    {
      "iteration": 51300,
      "average": {
        "on_policy_value_p0": -0.055550452888607,
        "best_response_value_p0": -0.05503735200377,
        "best_response_value_p1": 0.057057513969929,
        "unilateral_improvement_p0": 0.000513100884837,
        "unilateral_improvement_p1": 0.001507061081322,
        "nash_conv": 0.002020161966159,
        "exploitability": 0.001010080983079
      },
      "current": {
        "on_policy_value_p0": -0.195902986143657,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.195902986143657,
        "unilateral_improvement_p0": 0.362569652810324,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.362569652810324,
        "exploitability": 0.181284826405162
      }
    },
    {
      "iteration": 51400,
      "average": {
        "on_policy_value_p0": -0.055552692221519,
        "best_response_value_p0": -0.055099850784451,
        "best_response_value_p1": 0.057124258070896,
        "unilateral_improvement_p0": 0.000452841437068,
        "unilateral_improvement_p1": 0.001571565849377,
        "nash_conv": 0.002024407286445,
        "exploitability": 0.001012203643222
      },
      "current": {
        "on_policy_value_p0": -0.042374071855663,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.209040738522329,
        "unilateral_improvement_p1": 0.124292594811004,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 51500,
      "average": {
        "on_policy_value_p0": -0.055554731239566,
        "best_response_value_p0": -0.054992860782927,
        "best_response_value_p1": 0.056965823855947,
        "unilateral_improvement_p0": 0.000561870456638,
        "unilateral_improvement_p1": 0.001411092616381,
        "nash_conv": 0.00197296307302,
        "exploitability": 0.00098648153651
      },
      "current": {
        "on_policy_value_p0": 0.066182580372016,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.132365160744033,
        "unilateral_improvement_p0": 0.10048408629465,
        "unilateral_improvement_p1": 0.198547741116049,
        "nash_conv": 0.299031827410699,
        "exploitability": 0.14951591370535
      }
    },
    {
      "iteration": 51600,
      "average": {
        "on_policy_value_p0": -0.055555885040245,
        "best_response_value_p0": -0.054697998464941,
        "best_response_value_p1": 0.056677318670009,
        "unilateral_improvement_p0": 0.000857886575304,
        "unilateral_improvement_p1": 0.001121433629764,
        "nash_conv": 0.001979320205068,
        "exploitability": 0.000989660102534
      },
      "current": {
        "on_policy_value_p0": 0.077098777916445,
        "best_response_value_p0": 0.123825989495184,
        "best_response_value_p1": 0.239878910175855,
        "unilateral_improvement_p0": 0.046727211578739,
        "unilateral_improvement_p1": 0.316977688092301,
        "nash_conv": 0.36370489967104,
        "exploitability": 0.18185244983552
      }
    },
    {
      "iteration": 51700,
      "average": {
        "on_policy_value_p0": -0.055555657681887,
        "best_response_value_p0": -0.054395207796742,
        "best_response_value_p1": 0.056285579529505,
        "unilateral_improvement_p0": 0.001160449885144,
        "unilateral_improvement_p1": 0.000729921847618,
        "nash_conv": 0.001890371732762,
        "exploitability": 0.000945185866381
      },
      "current": {
        "on_policy_value_p0": 0.081685173432441,
        "best_response_value_p0": 0.081685173432441,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.415018506765774,
        "nash_conv": 0.415018506765774,
        "exploitability": 0.207509253382887
      }
    },
    {
      "iteration": 51800,
      "average": {
        "on_policy_value_p0": -0.055554381665675,
        "best_response_value_p0": -0.054171319813411,
        "best_response_value_p1": 0.056176920109563,
        "unilateral_improvement_p0": 0.001383061852265,
        "unilateral_improvement_p1": 0.000622538443887,
        "nash_conv": 0.002005600296152,
        "exploitability": 0.001002800148076
      },
      "current": {
        "on_policy_value_p0": 0.038281907310561,
        "best_response_value_p0": 0.090102852045544,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.051820944734982,
        "unilateral_improvement_p1": 0.371615240643895,
        "nash_conv": 0.423436185378877,
        "exploitability": 0.211718092689439
      }
    },
    {
      "iteration": 51900,
      "average": {
        "on_policy_value_p0": -0.055552629476528,
        "best_response_value_p0": -0.054045758999132,
        "best_response_value_p1": 0.05606867941571,
        "unilateral_improvement_p0": 0.001506870477396,
        "unilateral_improvement_p1": 0.000516049939182,
        "nash_conv": 0.002022920416578,
        "exploitability": 0.001011460208289
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 52000,
      "average": {
        "on_policy_value_p0": -0.055550487896056,
        "best_response_value_p0": -0.053992845462484,
        "best_response_value_p1": 0.055960855032218,
        "unilateral_improvement_p0": 0.001557642433573,
        "unilateral_improvement_p1": 0.000410367136161,
        "nash_conv": 0.001968009569734,
        "exploitability": 0.000984004784867
      },
      "current": {
        "on_policy_value_p0": -0.110746476964451,
        "best_response_value_p0": 0.111293428184441,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.222039905148892,
        "unilateral_improvement_p1": 0.222586856368882,
        "nash_conv": 0.444626761517775,
        "exploitability": 0.222313380758887
      }
    },
    {
      "iteration": 52100,
      "average": {
        "on_policy_value_p0": -0.055548543578463,
        "best_response_value_p0": -0.054256647552871,
        "best_response_value_p1": 0.056269091583301,
        "unilateral_improvement_p0": 0.001291896025592,
        "unilateral_improvement_p1": 0.000720548004838,
        "nash_conv": 0.00201244403043,
        "exploitability": 0.001006222015215
      },
      "current": {
        "on_policy_value_p0": -0.287567772891656,
        "best_response_value_p0": 0.287567772891656,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.575135545783311,
        "unilateral_improvement_p1": 0.045765560441678,
        "nash_conv": 0.620901106224989,
        "exploitability": 0.310450553112494
      }
    },
    {
      "iteration": 52200,
      "average": {
        "on_policy_value_p0": -0.055548449465177,
        "best_response_value_p0": -0.054784591935282,
        "best_response_value_p1": 0.056799865992784,
        "unilateral_improvement_p0": 0.000763857529895,
        "unilateral_improvement_p1": 0.001251416527607,
        "nash_conv": 0.002015274057501,
        "exploitability": 0.001007637028751
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 52300,
      "average": {
        "on_policy_value_p0": -0.05555035684179,
        "best_response_value_p0": -0.055233146725546,
        "best_response_value_p1": 0.057175374941481,
        "unilateral_improvement_p0": 0.000317210116244,
        "unilateral_improvement_p1": 0.001625018099691,
        "nash_conv": 0.001942228215935,
        "exploitability": 0.000971114107967
      },
      "current": {
        "on_policy_value_p0": -0.14001232985033,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.14001232985033,
        "unilateral_improvement_p0": 0.473345663183663,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.473345663183663,
        "exploitability": 0.236672831591831
      }
    },
    {
      "iteration": 52400,
      "average": {
        "on_policy_value_p0": -0.055552762429609,
        "best_response_value_p0": -0.05506301560101,
        "best_response_value_p1": 0.057033997793466,
        "unilateral_improvement_p0": 0.0004897468286,
        "unilateral_improvement_p1": 0.001481235363857,
        "nash_conv": 0.001970982192457,
        "exploitability": 0.000985491096228
      },
      "current": {
        "on_policy_value_p0": 0.097525377181592,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.235807956151742,
        "unilateral_improvement_p1": 0.264192043848258,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 52500,
      "average": {
        "on_policy_value_p0": -0.05555397794575,
        "best_response_value_p0": -0.05464067334907,
        "best_response_value_p1": 0.056634644829106,
        "unilateral_improvement_p0": 0.000913304596679,
        "unilateral_improvement_p1": 0.001080666883357,
        "nash_conv": 0.001993971480036,
        "exploitability": 0.000996985740018
      },
      "current": {
        "on_policy_value_p0": 0.209782010700235,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.252897354733803,
        "unilateral_improvement_p0": 0.123551322633098,
        "unilateral_improvement_p1": 0.462679365434038,
        "nash_conv": 0.586230688067136,
        "exploitability": 0.293115344033568
      }
    },
    {
      "iteration": 52600,
      "average": {
        "on_policy_value_p0": -0.055553469596276,
        "best_response_value_p0": -0.054322186328177,
        "best_response_value_p1": 0.056157276802577,
        "unilateral_improvement_p0": 0.001231283268099,
        "unilateral_improvement_p1": 0.0006038072063,
        "nash_conv": 0.001835090474399,
        "exploitability": 0.0009175452372
      },
      "current": {
        "on_policy_value_p0": 0.077534311799518,
        "best_response_value_p0": 0.147067431800395,
        "best_response_value_p1": 0.360933759998245,
        "unilateral_improvement_p0": 0.069533120000878,
        "unilateral_improvement_p1": 0.438468071797762,
        "nash_conv": 0.50800119179864,
        "exploitability": 0.25400059589932
      }
    },
    {
      "iteration": 52700,
      "average": {
        "on_policy_value_p0": -0.055552018371902,
        "best_response_value_p0": -0.054110357065496,
        "best_response_value_p1": 0.056121492994244,
        "unilateral_improvement_p0": 0.001441661306407,
        "unilateral_improvement_p1": 0.000569474622342,
        "nash_conv": 0.002011135928748,
        "exploitability": 0.001005567964374
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 52800,
      "average": {
        "on_policy_value_p0": -0.055550165432879,
        "best_response_value_p0": -0.054140200702913,
        "best_response_value_p1": 0.056015202287815,
        "unilateral_improvement_p0": 0.001409964729966,
        "unilateral_improvement_p1": 0.000465036854936,
        "nash_conv": 0.001875001584902,
        "exploitability": 0.000937500792451
      },
      "current": {
        "on_policy_value_p0": -0.243312565570542,
        "best_response_value_p0": 0.081104188523514,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.324416754094056,
        "unilateral_improvement_p1": 0.256687434429458,
        "nash_conv": 0.581104188523514,
        "exploitability": 0.290552094261757
      }
    },
    {
      "iteration": 52900,
      "average": {
        "on_policy_value_p0": -0.055549664738544,
        "best_response_value_p0": -0.054826458978489,
        "best_response_value_p1": 0.056776955974968,
        "unilateral_improvement_p0": 0.000723205760056,
        "unilateral_improvement_p1": 0.001227291236424,
        "nash_conv": 0.001950496996479,
        "exploitability": 0.00097524849824
      },
      "current": {
        "on_policy_value_p0": -0.368871674531987,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.368871674531987,
        "unilateral_improvement_p0": 0.535538341198654,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.535538341198654,
        "exploitability": 0.267769170599327
      }
    },
    {
      "iteration": 53000,
      "average": {
        "on_policy_value_p0": -0.05555149713708,
        "best_response_value_p0": -0.055146389884248,
        "best_response_value_p1": 0.057098486821524,
        "unilateral_improvement_p0": 0.000405107252832,
        "unilateral_improvement_p1": 0.001546989684444,
        "nash_conv": 0.001952096937276,
        "exploitability": 0.000976048468638
      },
      "current": {
        "on_policy_value_p0": -0.090043257483453,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.256709924150119,
        "unilateral_improvement_p1": 0.076623409183214,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 53100,
      "average": {
        "on_policy_value_p0": -0.055553674693819,
        "best_response_value_p0": -0.055075717691058,
        "best_response_value_p1": 0.057071259918578,
        "unilateral_improvement_p0": 0.000477957002761,
        "unilateral_improvement_p1": 0.001517585224759,
        "nash_conv": 0.00199554222752,
        "exploitability": 0.00099777111376
      },
      "current": {
        "on_policy_value_p0": 0.031010353679503,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.135656312987163,
        "unilateral_improvement_p0": 0.135656312987163,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.30232297965383,
        "exploitability": 0.151161489826915
      }
    },
    {
      "iteration": 53200,
      "average": {
        "on_policy_value_p0": -0.055555405428968,
        "best_response_value_p0": -0.054938554482068,
        "best_response_value_p1": 0.05684757576101,
        "unilateral_improvement_p0": 0.0006168509469,
        "unilateral_improvement_p1": 0.001292170332042,
        "nash_conv": 0.001909021278942,
        "exploitability": 0.000954510639471
      },
      "current": {
        "on_policy_value_p0": 0.088663951559536,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.177327903119071,
        "unilateral_improvement_p0": 0.078002715107131,
        "unilateral_improvement_p1": 0.265991854678607,
        "nash_conv": 0.343994569785738,
        "exploitability": 0.171997284892869
      }
    },
    {
      "iteration": 53300,
      "average": {
        "on_policy_value_p0": -0.055556009746635,
        "best_response_value_p0": -0.054587147485251,
        "best_response_value_p1": 0.056520821186066,
        "unilateral_improvement_p0": 0.000968862261384,
        "unilateral_improvement_p1": 0.000964811439431,
        "nash_conv": 0.001933673700815,
        "exploitability": 0.000966836850407
      },
      "current": {
        "on_policy_value_p0": 0.095267180099468,
        "best_response_value_p0": 0.105982406390217,
        "best_response_value_p1": 0.311902880751836,
        "unilateral_improvement_p0": 0.010715226290749,
        "unilateral_improvement_p1": 0.407170060851304,
        "nash_conv": 0.417885287142053,
        "exploitability": 0.208942643571027
      }
    },
    {
      "iteration": 53400,
      "average": {
        "on_policy_value_p0": -0.055555343450076,
        "best_response_value_p0": -0.054322483302475,
        "best_response_value_p1": 0.056233806533292,
        "unilateral_improvement_p0": 0.0012328601476,
        "unilateral_improvement_p1": 0.000678463083216,
        "nash_conv": 0.001911323230817,
        "exploitability": 0.000955661615408
      },
      "current": {
        "on_policy_value_p0": 0.067476332862445,
        "best_response_value_p0": 0.067476332862445,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.400809666195778,
        "nash_conv": 0.400809666195778,
        "exploitability": 0.200404833097889
      }
    },
    {
      "iteration": 53500,
      "average": {
        "on_policy_value_p0": -0.055553894587894,
        "best_response_value_p0": -0.054137678572705,
        "best_response_value_p1": 0.056128696614538,
        "unilateral_improvement_p0": 0.001416216015189,
        "unilateral_improvement_p1": 0.000574802026644,
        "nash_conv": 0.001991018041833,
        "exploitability": 0.000995509020916
      },
      "current": {
        "on_policy_value_p0": 0.014821844914412,
        "best_response_value_p0": 0.137022976837842,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.12220113192343,
        "unilateral_improvement_p1": 0.348155178247745,
        "nash_conv": 0.470356310171176,
        "exploitability": 0.235178155085588
      }
    },
    {
      "iteration": 53600,
      "average": {
        "on_policy_value_p0": -0.055552123234561,
        "best_response_value_p0": -0.054034210654507,
        "best_response_value_p1": 0.056023978896973,
        "unilateral_improvement_p0": 0.001517912580054,
        "unilateral_improvement_p1": 0.000471855662412,
        "nash_conv": 0.001989768242466,
        "exploitability": 0.000994884121233
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 53700,
      "average": {
        "on_policy_value_p0": -0.055549979595615,
        "best_response_value_p0": -0.054067577143269,
        "best_response_value_p1": 0.05591965118953,
        "unilateral_improvement_p0": 0.001482402452346,
        "unilateral_improvement_p1": 0.000369671593915,
        "nash_conv": 0.001852074046261,
        "exploitability": 0.00092603702313
      },
      "current": {
        "on_policy_value_p0": -0.162695719232319,
        "best_response_value_p0": 0.162695719232319,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.325391438464638,
        "unilateral_improvement_p1": 0.170637614101014,
        "nash_conv": 0.496029052565652,
        "exploitability": 0.248014526282826
      }
    },
    {
      "iteration": 53800,
      "average": {
        "on_policy_value_p0": -0.055548565290633,
        "best_response_value_p0": -0.054432521614925,
        "best_response_value_p1": 0.056418753543255,
        "unilateral_improvement_p0": 0.001116043675709,
        "unilateral_improvement_p1": 0.000870188252621,
        "nash_conv": 0.00198623192833,
        "exploitability": 0.000993115964165
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 53900,
      "average": {
        "on_policy_value_p0": -0.055549066165098,
        "best_response_value_p0": -0.05494996282405,
        "best_response_value_p1": 0.056932509720971,
        "unilateral_improvement_p0": 0.000599103341048,
        "unilateral_improvement_p1": 0.001383443555873,
        "nash_conv": 0.00198254689692,
        "exploitability": 0.00099127344846
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 54000,
      "average": {
        "on_policy_value_p0": -0.055551288665993,
        "best_response_value_p0": -0.055220466965364,
        "best_response_value_p1": 0.057175291155551,
        "unilateral_improvement_p0": 0.000330821700629,
        "unilateral_improvement_p1": 0.001624002489558,
        "nash_conv": 0.001954824190187,
        "exploitability": 0.000977412095093
      },
      "current": {
        "on_policy_value_p0": -0.039434635742699,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.127345061766547,
        "unilateral_improvement_p0": 0.372767969076032,
        "unilateral_improvement_p1": 0.087910426023848,
        "nash_conv": 0.46067839509988,
        "exploitability": 0.23033919754994
      }
    },
    {
      "iteration": 54100,
      "average": {
        "on_policy_value_p0": -0.055553397013937,
        "best_response_value_p0": -0.054969612299385,
        "best_response_value_p1": 0.056938513757707,
        "unilateral_improvement_p0": 0.000583784714552,
        "unilateral_improvement_p1": 0.001385116743769,
        "nash_conv": 0.001968901458322,
        "exploitability": 0.000984450729161
      },
      "current": {
        "on_policy_value_p0": 0.120779232030714,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.21255410130262,
        "unilateral_improvement_p1": 0.28744589869738,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 54200,
      "average": {
        "on_policy_value_p0": -0.055554194612385,
        "best_response_value_p0": -0.054566724790432,
        "best_response_value_p1": 0.05649137969073,
        "unilateral_improvement_p0": 0.000987469821953,
        "unilateral_improvement_p1": 0.000937185078344,
        "nash_conv": 0.001924654900298,
        "exploitability": 0.000962327450149
      },
      "current": {
        "on_policy_value_p0": 0.181726199618412,
        "best_response_value_p0": 0.286224344654125,
        "best_response_value_p1": 0.291003709928573,
        "unilateral_improvement_p0": 0.104498145035714,
        "unilateral_improvement_p1": 0.472729909546984,
        "nash_conv": 0.577228054582698,
        "exploitability": 0.288614027291349
      }
    },
    {
      "iteration": 54300,
      "average": {
        "on_policy_value_p0": -0.055553340431654,
        "best_response_value_p0": -0.054239300816631,
        "best_response_value_p1": 0.056130615937041,
        "unilateral_improvement_p0": 0.001314039615022,
        "unilateral_improvement_p1": 0.000577275505388,
        "nash_conv": 0.00189131512041,
        "exploitability": 0.000945657560205
      },
      "current": {
        "on_policy_value_p0": 0.060505140372776,
        "best_response_value_p0": 0.113523880998648,
        "best_response_value_p1": 0.393962518748256,
        "unilateral_improvement_p0": 0.053018740625872,
        "unilateral_improvement_p1": 0.454467659121032,
        "nash_conv": 0.507486399746904,
        "exploitability": 0.253743199873452
      }
    },
    {
      "iteration": 54400,
      "average": {
        "on_policy_value_p0": -0.055551764715689,
        "best_response_value_p0": -0.054083723610842,
        "best_response_value_p1": 0.056062575696118,
        "unilateral_improvement_p0": 0.001468041104848,
        "unilateral_improvement_p1": 0.000510810980429,
        "nash_conv": 0.001978852085276,
        "exploitability": 0.000989426042638
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 54500,
      "average": {
        "on_policy_value_p0": -0.055549955512936,
        "best_response_value_p0": -0.054261890165984,
        "best_response_value_p1": 0.056195804492701,
        "unilateral_improvement_p0": 0.001288065346952,
        "unilateral_improvement_p1": 0.000645848979765,
        "nash_conv": 0.001933914326717,
        "exploitability": 0.000966957163358
      },
      "current": {
        "on_policy_value_p0": -0.341349091753486,
        "best_response_value_p0": 0.113783030584495,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.455132122337982,
        "unilateral_improvement_p1": 0.158650908246514,
        "nash_conv": 0.613783030584495,
        "exploitability": 0.306891515292248
      }
    },
    {
      "iteration": 54600,
      "average": {
        "on_policy_value_p0": -0.055550142591124,
        "best_response_value_p0": -0.054977435767494,
        "best_response_value_p1": 0.056874880193199,
        "unilateral_improvement_p0": 0.000572706823629,
        "unilateral_improvement_p1": 0.001324737602075,
        "nash_conv": 0.001897444425704,
        "exploitability": 0.000948722212852
      },
      "current": {
        "on_policy_value_p0": -0.316573981036204,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.316573981036204,
        "unilateral_improvement_p0": 0.483240647702871,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.483240647702871,
        "exploitability": 0.241620323851435
      }
    },
    {
      "iteration": 54700,
      "average": {
        "on_policy_value_p0": -0.055552037969792,
        "best_response_value_p0": -0.055149470637687,
        "best_response_value_p1": 0.057069884980899,
        "unilateral_improvement_p0": 0.000402567332105,
        "unilateral_improvement_p1": 0.001517847011107,
        "nash_conv": 0.001920414343212,
        "exploitability": 0.000960207171606
      },
      "current": {
        "on_policy_value_p0": -0.074543122524381,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.241209789191048,
        "unilateral_improvement_p1": 0.092123544142285,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 54800,
      "average": {
        "on_policy_value_p0": -0.055554021333373,
        "best_response_value_p0": -0.055048832917545,
        "best_response_value_p1": 0.057004596654898,
        "unilateral_improvement_p0": 0.000505188415829,
        "unilateral_improvement_p1": 0.001450575321525,
        "nash_conv": 0.001955763737353,
        "exploitability": 0.000977881868677
      },
      "current": {
        "on_policy_value_p0": 0.040355609456782,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.126311057209885,
        "unilateral_improvement_p0": 0.126311057209885,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.292977723876551,
        "exploitability": 0.146488861938276
      }
    },
    {
      "iteration": 54900,
      "average": {
        "on_policy_value_p0": -0.055555499737665,
        "best_response_value_p0": -0.054884875777012,
        "best_response_value_p1": 0.056776120491039,
        "unilateral_improvement_p0": 0.000670623960654,
        "unilateral_improvement_p1": 0.001220620753374,
        "nash_conv": 0.001891244714027,
        "exploitability": 0.000945622357014
      },
      "current": {
        "on_policy_value_p0": 0.078063966627882,
        "best_response_value_p0": 0.150667492692161,
        "best_response_value_p1": 0.188126281204774,
        "unilateral_improvement_p0": 0.07260352606428,
        "unilateral_improvement_p1": 0.266190247832655,
        "nash_conv": 0.338793773896935,
        "exploitability": 0.169396886948468
      }
    },
    {
      "iteration": 55000,
      "average": {
        "on_policy_value_p0": -0.055555872323877,
        "best_response_value_p0": -0.054563075240439,
        "best_response_value_p1": 0.056446965142811,
        "unilateral_improvement_p0": 0.000992797083438,
        "unilateral_improvement_p1": 0.000891092818934,
        "nash_conv": 0.001883889902372,
        "exploitability": 0.000941944951186
      },
      "current": {
        "on_policy_value_p0": 0.099057003698348,
        "best_response_value_p0": 0.099057003698348,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.432390337031682,
        "nash_conv": 0.432390337031682,
        "exploitability": 0.216195168515841
      }
    },
    {
      "iteration": 55100,
      "average": {
        "on_policy_value_p0": -0.055555121018655,
        "best_response_value_p0": -0.05431800403463,
        "best_response_value_p1": 0.056217722615883,
        "unilateral_improvement_p0": 0.001237116984025,
        "unilateral_improvement_p1": 0.000662601597228,
        "nash_conv": 0.001899718581253,
        "exploitability": 0.000949859290627
      },
      "current": {
        "on_policy_value_p0": 0.061205626601019,
        "best_response_value_p0": 0.061205626601019,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.394538959934352,
        "nash_conv": 0.394538959934352,
        "exploitability": 0.197269479967176
      }
    },
    {
      "iteration": 55200,
      "average": {
        "on_policy_value_p0": -0.05555371113816,
        "best_response_value_p0": -0.054153726534032,
        "best_response_value_p1": 0.056115878915492,
        "unilateral_improvement_p0": 0.001399984604128,
        "unilateral_improvement_p1": 0.000562167777332,
        "nash_conv": 0.001962152381461,
        "exploitability": 0.00098107619073
      },
      "current": {
        "on_policy_value_p0": 0.002200640729033,
        "best_response_value_p0": 0.162265385208601,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.160064744479568,
        "unilateral_improvement_p1": 0.335533974062366,
        "nash_conv": 0.495598718541934,
        "exploitability": 0.247799359270967
      }
    },
    {
      "iteration": 55300,
      "average": {
        "on_policy_value_p0": -0.055552016875937,
        "best_response_value_p0": -0.054055734873125,
        "best_response_value_p1": 0.056014403546749,
        "unilateral_improvement_p0": 0.001496282002813,
        "unilateral_improvement_p1": 0.000462386670812,
        "nash_conv": 0.001958668673624,
        "exploitability": 0.000979334336812
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 55400,
      "average": {
        "on_policy_value_p0": -0.055549981511854,
        "best_response_value_p0": -0.054107703031749,
        "best_response_value_p1": 0.055932252525248,
        "unilateral_improvement_p0": 0.001442278480104,
        "unilateral_improvement_p1": 0.000382271013395,
        "nash_conv": 0.001824549493499,
        "exploitability": 0.000912274746749
      },
      "current": {
        "on_policy_value_p0": -0.171281298107533,
        "best_response_value_p0": 0.171281298107533,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.342562596215066,
        "unilateral_improvement_p1": 0.162052035225801,
        "nash_conv": 0.504614631440866,
        "exploitability": 0.252307315720433
      }
    },
    {
      "iteration": 55500,
      "average": {
        "on_policy_value_p0": -0.05554872763129,
        "best_response_value_p0": -0.054476720553002,
        "best_response_value_p1": 0.056432074292468,
        "unilateral_improvement_p0": 0.001072007078289,
        "unilateral_improvement_p1": 0.000883346661178,
        "nash_conv": 0.001955353739466,
        "exploitability": 0.000977676869733
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 55600,
      "average": {
        "on_policy_value_p0": -0.055549287926952,
        "best_response_value_p0": -0.05497826122347,
        "best_response_value_p1": 0.056929690417916,
        "unilateral_improvement_p0": 0.000571026703481,
        "unilateral_improvement_p1": 0.001380402490964,
        "nash_conv": 0.001951429194445,
        "exploitability": 0.000975714597223
      },
      "current": {
        "on_policy_value_p0": -0.315795459063064,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.315795459063064,
        "unilateral_improvement_p0": 0.649128792396397,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.649128792396397,
        "exploitability": 0.324564396198198
      }
    },
    {
      "iteration": 55700,
      "average": {
        "on_policy_value_p0": -0.055551439699562,
        "best_response_value_p0": -0.055221383307397,
        "best_response_value_p1": 0.057148106472818,
        "unilateral_improvement_p0": 0.000330056392165,
        "unilateral_improvement_p1": 0.001596666773256,
        "nash_conv": 0.001926723165421,
        "exploitability": 0.00096336158271
      },
      "current": {
        "on_policy_value_p0": -0.030056406906012,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.131020477739765,
        "unilateral_improvement_p0": 0.363389740239346,
        "unilateral_improvement_p1": 0.100964070833753,
        "nash_conv": 0.464353811073098,
        "exploitability": 0.232176905536549
      }
    },
    {
      "iteration": 55800,
      "average": {
        "on_policy_value_p0": -0.055553432714094,
        "best_response_value_p0": -0.05497476954654,
        "best_response_value_p1": 0.0569135806746,
        "unilateral_improvement_p0": 0.000578663167554,
        "unilateral_improvement_p1": 0.001360147960506,
        "nash_conv": 0.00193881112806,
        "exploitability": 0.00096940556403
      },
      "current": {
        "on_policy_value_p0": 0.12117316869842,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.212160164634913,
        "unilateral_improvement_p1": 0.287839835365087,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 55900,
      "average": {
        "on_policy_value_p0": -0.055554191248867,
        "best_response_value_p0": -0.05458469394998,
        "best_response_value_p1": 0.056481736653179,
        "unilateral_improvement_p0": 0.000969497298887,
        "unilateral_improvement_p1": 0.000927545404312,
        "nash_conv": 0.001897042703199,
        "exploitability": 0.0009485213516
      },
      "current": {
        "on_policy_value_p0": 0.178979528477634,
        "best_response_value_p0": 0.28470979980492,
        "best_response_value_p1": 0.288539457345427,
        "unilateral_improvement_p0": 0.105730271327286,
        "unilateral_improvement_p1": 0.467518985823061,
        "nash_conv": 0.573249257150348,
        "exploitability": 0.286624628575174
      }
    },
    {
      "iteration": 56000,
      "average": {
        "on_policy_value_p0": -0.055553408870748,
        "best_response_value_p0": -0.05427552917534,
        "best_response_value_p1": 0.056131943841826,
        "unilateral_improvement_p0": 0.001277879695409,
        "unilateral_improvement_p1": 0.000578534971078,
        "nash_conv": 0.001856414666486,
        "exploitability": 0.000928207333243
      },
      "current": {
        "on_policy_value_p0": 0.059964092453083,
        "best_response_value_p0": 0.115197856710435,
        "best_response_value_p1": 0.389532471485295,
        "unilateral_improvement_p0": 0.055233764257353,
        "unilateral_improvement_p1": 0.449496563938378,
        "nash_conv": 0.50473032819573,
        "exploitability": 0.252365164097865
      }
    },
    {
      "iteration": 56100,
      "average": {
        "on_policy_value_p0": -0.055551957523397,
        "best_response_value_p0": -0.054121712169477,
        "best_response_value_p1": 0.056070232614057,
        "unilateral_improvement_p0": 0.001430245353921,
        "unilateral_improvement_p1": 0.00051827509066,
        "nash_conv": 0.001948520444581,
        "exploitability": 0.00097426022229
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 56200,
      "average": {
        "on_policy_value_p0": -0.055550232426989,
        "best_response_value_p0": -0.054249985834703,
        "best_response_value_p1": 0.056127700593279,
        "unilateral_improvement_p0": 0.001300246592286,
        "unilateral_improvement_p1": 0.00057746816629,
        "nash_conv": 0.001877714758577,
        "exploitability": 0.000938857379288
      },
      "current": {
        "on_policy_value_p0": -0.309691917997358,
        "best_response_value_p0": 0.103230639332453,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.41292255732981,
        "unilateral_improvement_p1": 0.190308082002642,
        "nash_conv": 0.603230639332453,
        "exploitability": 0.301615319666226
      }
    },
    {
      "iteration": 56300,
      "average": {
        "on_policy_value_p0": -0.055550204144943,
        "best_response_value_p0": -0.054968960115275,
        "best_response_value_p1": 0.056817207405503,
        "unilateral_improvement_p0": 0.000581244029668,
        "unilateral_improvement_p1": 0.00126700326056,
        "nash_conv": 0.001848247290228,
        "exploitability": 0.000924123645114
      },
      "current": {
        "on_policy_value_p0": -0.337237797842177,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.337237797842177,
        "unilateral_improvement_p0": 0.503904464508843,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.503904464508843,
        "exploitability": 0.251952232254422
      }
    },
    {
      "iteration": 56400,
      "average": {
        "on_policy_value_p0": -0.055551984050798,
        "best_response_value_p0": -0.05517495717259,
        "best_response_value_p1": 0.057052438636988,
        "unilateral_improvement_p0": 0.000377026878208,
        "unilateral_improvement_p1": 0.00150045458619,
        "nash_conv": 0.001877481464398,
        "exploitability": 0.000938740732199
      },
      "current": {
        "on_policy_value_p0": -0.081629701069236,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.248296367735903,
        "unilateral_improvement_p1": 0.08503696559743,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 56500,
      "average": {
        "on_policy_value_p0": -0.055553936269544,
        "best_response_value_p0": -0.055077302381133,
        "best_response_value_p1": 0.057007764800322,
        "unilateral_improvement_p0": 0.000476633888411,
        "unilateral_improvement_p1": 0.001453828530778,
        "nash_conv": 0.001930462419189,
        "exploitability": 0.000965231209594
      },
      "current": {
        "on_policy_value_p0": 0.035687223818136,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.13097944284853,
        "unilateral_improvement_p0": 0.13097944284853,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.297646109515197,
        "exploitability": 0.148823054757598
      }
    },
    {
      "iteration": 56600,
      "average": {
        "on_policy_value_p0": -0.055555457879849,
        "best_response_value_p0": -0.054938692407812,
        "best_response_value_p1": 0.056793060768565,
        "unilateral_improvement_p0": 0.000616765472037,
        "unilateral_improvement_p1": 0.001237602888716,
        "nash_conv": 0.001854368360753,
        "exploitability": 0.000927184180377
      },
      "current": {
        "on_policy_value_p0": 0.08720864235819,
        "best_response_value_p0": 0.16377479812453,
        "best_response_value_p1": 0.180201021800653,
        "unilateral_improvement_p0": 0.07656615576634,
        "unilateral_improvement_p1": 0.267409664158844,
        "nash_conv": 0.343975819925184,
        "exploitability": 0.171987909962592
      }
    },
    {
      "iteration": 56700,
      "average": {
        "on_policy_value_p0": -0.055555965699594,
        "best_response_value_p0": -0.054611286428436,
        "best_response_value_p1": 0.056484617627692,
        "unilateral_improvement_p0": 0.000944679271158,
        "unilateral_improvement_p1": 0.000928651928098,
        "nash_conv": 0.001873331199256,
        "exploitability": 0.000936665599628
      },
      "current": {
        "on_policy_value_p0": 0.094687046549067,
        "best_response_value_p0": 0.105590268927747,
        "best_response_value_p1": 0.311526888575974,
        "unilateral_improvement_p0": 0.01090322237868,
        "unilateral_improvement_p1": 0.406213935125041,
        "nash_conv": 0.41711715750372,
        "exploitability": 0.20855857875186
      }
    },
    {
      "iteration": 56800,
      "average": {
        "on_policy_value_p0": -0.055555353126181,
        "best_response_value_p0": -0.054362039857798,
        "best_response_value_p1": 0.056212324851127,
        "unilateral_improvement_p0": 0.001193313268383,
        "unilateral_improvement_p1": 0.000656971724946,
        "nash_conv": 0.001850284993329,
        "exploitability": 0.000925142496664
      },
      "current": {
        "on_policy_value_p0": 0.068314630152727,
        "best_response_value_p0": 0.068314630152727,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.401647963486061,
        "nash_conv": 0.401647963486061,
        "exploitability": 0.20082398174303
      }
    },
    {
      "iteration": 56900,
      "average": {
        "on_policy_value_p0": -0.055554038328395,
        "best_response_value_p0": -0.054184740721359,
        "best_response_value_p1": 0.056113533419052,
        "unilateral_improvement_p0": 0.001369297607036,
        "unilateral_improvement_p1": 0.000559495090657,
        "nash_conv": 0.001928792697693,
        "exploitability": 0.000964396348846
      },
      "current": {
        "on_policy_value_p0": 0.018877506411661,
        "best_response_value_p0": 0.128911653843344,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.110034147431683,
        "unilateral_improvement_p1": 0.352210839744995,
        "nash_conv": 0.462244987176677,
        "exploitability": 0.231122493588339
      }
    },
    {
      "iteration": 57000,
      "average": {
        "on_policy_value_p0": -0.055552432194633,
        "best_response_value_p0": -0.054085671118934,
        "best_response_value_p1": 0.05601508862358,
        "unilateral_improvement_p0": 0.001466761075699,
        "unilateral_improvement_p1": 0.000462656428947,
        "nash_conv": 0.001929417504646,
        "exploitability": 0.000964708752323
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 57100,
      "average": {
        "on_policy_value_p0": -0.055550483240159,
        "best_response_value_p0": -0.05409355476445,
        "best_response_value_p1": 0.055916988643504,
        "unilateral_improvement_p0": 0.001456928475708,
        "unilateral_improvement_p1": 0.000366505403346,
        "nash_conv": 0.001823433879054,
        "exploitability": 0.000911716939527
      },
      "current": {
        "on_policy_value_p0": -0.147967491504031,
        "best_response_value_p0": 0.147967491504031,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.295934983008061,
        "unilateral_improvement_p1": 0.185365841829303,
        "nash_conv": 0.481300824837364,
        "exploitability": 0.240650412418682
      }
    },
    {
      "iteration": 57200,
      "average": {
        "on_policy_value_p0": -0.055549022160642,
        "best_response_value_p0": -0.054400617836313,
        "best_response_value_p1": 0.056326914534461,
        "unilateral_improvement_p0": 0.00114840432433,
        "unilateral_improvement_p1": 0.000777892373818,
        "nash_conv": 0.001926296698148,
        "exploitability": 0.000963148349074
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 57300,
      "average": {
        "on_policy_value_p0": -0.055549222726159,
        "best_response_value_p0": -0.054887411406116,
        "best_response_value_p1": 0.05681034632992,
        "unilateral_improvement_p0": 0.000661811320043,
        "unilateral_improvement_p1": 0.00126112360376,
        "nash_conv": 0.001922934923803,
        "exploitability": 0.000961467461902
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 57400,
      "average": {
        "on_policy_value_p0": -0.055551038699132,
        "best_response_value_p0": -0.055240550664328,
        "best_response_value_p1": 0.057113802806798,
        "unilateral_improvement_p0": 0.000310488034804,
        "unilateral_improvement_p1": 0.001562764107666,
        "nash_conv": 0.00187325214247,
        "exploitability": 0.000936626071235
      },
      "current": {
        "on_policy_value_p0": -0.109220630613835,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.109220630613835,
        "unilateral_improvement_p0": 0.442553963947168,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.442553963947168,
        "exploitability": 0.221276981973584
      }
    },
    {
      "iteration": 57500,
      "average": {
        "on_policy_value_p0": -0.05555310155119,
        "best_response_value_p0": -0.055069844476255,
        "best_response_value_p1": 0.056956068575573,
        "unilateral_improvement_p0": 0.000483257074935,
        "unilateral_improvement_p1": 0.001402967024383,
        "nash_conv": 0.001886224099317,
        "exploitability": 0.000943112049659
      },
      "current": {
        "on_policy_value_p0": 0.100896667033265,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.232436666300069,
        "unilateral_improvement_p1": 0.267563333699931,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 57600,
      "average": {
        "on_policy_value_p0": -0.055554138839024,
        "best_response_value_p0": -0.05468488525552,
        "best_response_value_p1": 0.056587554573788,
        "unilateral_improvement_p0": 0.000869253583503,
        "unilateral_improvement_p1": 0.001033415734764,
        "nash_conv": 0.001902669318268,
        "exploitability": 0.000951334659134
      },
      "current": {
        "on_policy_value_p0": 0.210157194559811,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.253647722452956,
        "unilateral_improvement_p0": 0.123176138773522,
        "unilateral_improvement_p1": 0.463804917012767,
        "nash_conv": 0.586981055786289,
        "exploitability": 0.293490527893145
      }
    },
    {
      "iteration": 57700,
      "average": {
        "on_policy_value_p0": -0.055553726650631,
        "best_response_value_p0": -0.054392858817037,
        "best_response_value_p1": 0.056124204939853,
        "unilateral_improvement_p0": 0.001160867833594,
        "unilateral_improvement_p1": 0.000570478289222,
        "nash_conv": 0.001731346122816,
        "exploitability": 0.000865673061408
      },
      "current": {
        "on_policy_value_p0": 0.083836507992207,
        "best_response_value_p0": 0.155420829311537,
        "best_response_value_p1": 0.356831357361341,
        "unilateral_improvement_p0": 0.071584321319329,
        "unilateral_improvement_p1": 0.440667865353549,
        "nash_conv": 0.512252186672878,
        "exploitability": 0.256126093336439
      }
    },
    {
      "iteration": 57800,
      "average": {
        "on_policy_value_p0": -0.055552474610339,
        "best_response_value_p0": -0.054179794603023,
        "best_response_value_p1": 0.056099768735768,
        "unilateral_improvement_p0": 0.001372680007316,
        "unilateral_improvement_p1": 0.00054729412543,
        "nash_conv": 0.001919974132746,
        "exploitability": 0.000959987066373
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 57900,
      "average": {
        "on_policy_value_p0": -0.055550865713153,
        "best_response_value_p0": -0.05416291233845,
        "best_response_value_p1": 0.056002877943478,
        "unilateral_improvement_p0": 0.001387953374703,
        "unilateral_improvement_p1": 0.000452012230324,
        "nash_conv": 0.001839965605028,
        "exploitability": 0.000919982802514
      },
      "current": {
        "on_policy_value_p0": -0.199751574347973,
        "best_response_value_p0": 0.066583858115991,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.266335432463963,
        "unilateral_improvement_p1": 0.300248425652027,
        "nash_conv": 0.566583858115991,
        "exploitability": 0.283291929057995
      }
    },
    {
      "iteration": 58000,
      "average": {
        "on_policy_value_p0": -0.055550053023527,
        "best_response_value_p0": -0.054718761145844,
        "best_response_value_p1": 0.056614811059815,
        "unilateral_improvement_p0": 0.000831291877683,
        "unilateral_improvement_p1": 0.001064758036288,
        "nash_conv": 0.001896049913971,
        "exploitability": 0.000948024956985
      },
      "current": {
        "on_policy_value_p0": -0.407841353166426,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.407841353166426,
        "unilateral_improvement_p0": 0.574508019833092,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.574508019833092,
        "exploitability": 0.287254009916546
      }
    },
    {
      "iteration": 58100,
      "average": {
        "on_policy_value_p0": -0.055551410451634,
        "best_response_value_p0": -0.055116282579858,
        "best_response_value_p1": 0.057007513223052,
        "unilateral_improvement_p0": 0.000435127871776,
        "unilateral_improvement_p1": 0.001456102771418,
        "nash_conv": 0.001891230643194,
        "exploitability": 0.000945615321597
      },
      "current": {
        "on_policy_value_p0": -0.107429241305933,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.274095907972599,
        "unilateral_improvement_p1": 0.059237425360734,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 58200,
      "average": {
        "on_policy_value_p0": -0.055553302796051,
        "best_response_value_p0": -0.05511991263889,
        "best_response_value_p1": 0.057030241780522,
        "unilateral_improvement_p0": 0.000433390157162,
        "unilateral_improvement_p1": 0.001476938984471,
        "nash_conv": 0.001910329141632,
        "exploitability": 0.000955164570816
      },
      "current": {
        "on_policy_value_p0": -0.025570302537128,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166103701683783,
        "unilateral_improvement_p0": 0.192236969203795,
        "unilateral_improvement_p1": 0.140533399146655,
        "nash_conv": 0.33277036835045,
        "exploitability": 0.166385184175225
      }
    },
    {
      "iteration": 58300,
      "average": {
        "on_policy_value_p0": -0.055554967711254,
        "best_response_value_p0": -0.055025367334191,
        "best_response_value_p1": 0.056864930084095,
        "unilateral_improvement_p0": 0.000529600377063,
        "unilateral_improvement_p1": 0.001309962372841,
        "nash_conv": 0.001839562749904,
        "exploitability": 0.000919781374952
      },
      "current": {
        "on_policy_value_p0": 0.071968233814222,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.143936467628445,
        "unilateral_improvement_p0": 0.094698432852444,
        "unilateral_improvement_p1": 0.215904701442667,
        "nash_conv": 0.310603134295111,
        "exploitability": 0.155301567147556
      }
    },
    {
      "iteration": 58400,
      "average": {
        "on_policy_value_p0": -0.055555876108434,
        "best_response_value_p0": -0.054742202421779,
        "best_response_value_p1": 0.056602878632203,
        "unilateral_improvement_p0": 0.000813673686655,
        "unilateral_improvement_p1": 0.001047002523769,
        "nash_conv": 0.001860676210424,
        "exploitability": 0.000930338105212
      },
      "current": {
        "on_policy_value_p0": 0.078399688178967,
        "best_response_value_p0": 0.121790464790901,
        "best_response_value_p1": 0.246551780109465,
        "unilateral_improvement_p0": 0.043390776611934,
        "unilateral_improvement_p1": 0.324951468288432,
        "nash_conv": 0.368342244900366,
        "exploitability": 0.184171122450183
      }
    },
    {
      "iteration": 58500,
      "average": {
        "on_policy_value_p0": -0.055555691411163,
        "best_response_value_p0": -0.054475322538238,
        "best_response_value_p1": 0.056248525470728,
        "unilateral_improvement_p0": 0.001080368872925,
        "unilateral_improvement_p1": 0.000692834059565,
        "nash_conv": 0.00177320293249,
        "exploitability": 0.000886601466245
      },
      "current": {
        "on_policy_value_p0": 0.082489228219006,
        "best_response_value_p0": 0.082489228219006,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.415822561552339,
        "nash_conv": 0.415822561552339,
        "exploitability": 0.20791128077617
      }
    },
    {
      "iteration": 58600,
      "average": {
        "on_policy_value_p0": -0.055554679139618,
        "best_response_value_p0": -0.054273422179448,
        "best_response_value_p1": 0.056152538225898,
        "unilateral_improvement_p0": 0.00128125696017,
        "unilateral_improvement_p1": 0.00059785908628,
        "nash_conv": 0.00187911604645,
        "exploitability": 0.000939558023225
      },
      "current": {
        "on_policy_value_p0": 0.042681799517463,
        "best_response_value_p0": 0.081303067631742,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.038621268114279,
        "unilateral_improvement_p1": 0.376015132850796,
        "nash_conv": 0.414636400965075,
        "exploitability": 0.207318200482537
      }
    },
    {
      "iteration": 58700,
      "average": {
        "on_policy_value_p0": -0.055553266247054,
        "best_response_value_p0": -0.054155023317256,
        "best_response_value_p1": 0.056056878024491,
        "unilateral_improvement_p0": 0.001398242929798,
        "unilateral_improvement_p1": 0.000503611777438,
        "nash_conv": 0.001901854707236,
        "exploitability": 0.000950927353618
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 58800,
      "average": {
        "on_policy_value_p0": -0.055551573832799,
        "best_response_value_p0": -0.054075264466022,
        "best_response_value_p1": 0.05596154319792,
        "unilateral_improvement_p0": 0.001476309366777,
        "unilateral_improvement_p1": 0.000409969365121,
        "nash_conv": 0.001886278731898,
        "exploitability": 0.000943139365949
      },
      "current": {
        "on_policy_value_p0": -0.064479175664725,
        "best_response_value_p0": 0.134427078834304,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.198906254499029,
        "unilateral_improvement_p1": 0.268854157668608,
        "nash_conv": 0.467760412167637,
        "exploitability": 0.233880206083819
      }
    },
    {
      "iteration": 58900,
      "average": {
        "on_policy_value_p0": -0.055549772054635,
        "best_response_value_p0": -0.05423491924662,
        "best_response_value_p1": 0.056086179721083,
        "unilateral_improvement_p0": 0.001314852808015,
        "unilateral_improvement_p1": 0.000536407666449,
        "nash_conv": 0.001851260474463,
        "exploitability": 0.000925630237232
      },
      "current": {
        "on_policy_value_p0": -0.2249792044191,
        "best_response_value_p0": 0.2249792044191,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.4499584088382,
        "unilateral_improvement_p1": 0.108354128914233,
        "nash_conv": 0.558312537752433,
        "exploitability": 0.279156268876217
      }
    },
    {
      "iteration": 59000,
      "average": {
        "on_policy_value_p0": -0.055549140942305,
        "best_response_value_p0": -0.054660384882821,
        "best_response_value_p1": 0.056556090150933,
        "unilateral_improvement_p0": 0.000888756059484,
        "unilateral_improvement_p1": 0.001006949208628,
        "nash_conv": 0.001895705268112,
        "exploitability": 0.000947852634056
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 59100,
      "average": {
        "on_policy_value_p0": -0.055550152104265,
        "best_response_value_p0": -0.055131912714379,
        "best_response_value_p1": 0.056993332009347,
        "unilateral_improvement_p0": 0.000418239389886,
        "unilateral_improvement_p1": 0.001443179905082,
        "nash_conv": 0.001861419294968,
        "exploitability": 0.000930709647484
      },
      "current": {
        "on_policy_value_p0": -0.228571881564475,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.228571881564475,
        "unilateral_improvement_p0": 0.561905214897808,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.561905214897808,
        "exploitability": 0.280952607448904
      }
    },
    {
      "iteration": 59200,
      "average": {
        "on_policy_value_p0": -0.055552273497584,
        "best_response_value_p0": -0.055212342190768,
        "best_response_value_p1": 0.05707022978977,
        "unilateral_improvement_p0": 0.000339931306816,
        "unilateral_improvement_p1": 0.001517956292186,
        "nash_conv": 0.001857887599003,
        "exploitability": 0.000928943799501
      },
      "current": {
        "on_policy_value_p0": 0.063762679725133,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.2695706536082,
        "unilateral_improvement_p1": 0.2304293463918,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 59300,
      "average": {
        "on_policy_value_p0": -0.055553885772443,
        "best_response_value_p0": -0.054910508912745,
        "best_response_value_p1": 0.056803816533136,
        "unilateral_improvement_p0": 0.000643376859698,
        "unilateral_improvement_p1": 0.001249930760693,
        "nash_conv": 0.001893307620391,
        "exploitability": 0.000946653810195
      },
      "current": {
        "on_policy_value_p0": 0.150634910530422,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.16325463631924,
        "unilateral_improvement_p0": 0.182698422802911,
        "unilateral_improvement_p1": 0.313889546849662,
        "nash_conv": 0.496587969652573,
        "exploitability": 0.248293984826287
      }
    },
    {
      "iteration": 59400,
      "average": {
        "on_policy_value_p0": -0.055554287464781,
        "best_response_value_p0": -0.054564818197328,
        "best_response_value_p1": 0.056357267773884,
        "unilateral_improvement_p0": 0.000989469267453,
        "unilateral_improvement_p1": 0.000802980309103,
        "nash_conv": 0.001792449576556,
        "exploitability": 0.000896224788278
      },
      "current": {
        "on_policy_value_p0": 0.150316020061618,
        "best_response_value_p0": 0.24445580762485,
        "best_response_value_p1": 0.311720424873536,
        "unilateral_improvement_p0": 0.094139787563232,
        "unilateral_improvement_p1": 0.462036444935154,
        "nash_conv": 0.556176232498386,
        "exploitability": 0.278088116249193
      }
    },
    {
      "iteration": 59500,
      "average": {
        "on_policy_value_p0": -0.055553372871534,
        "best_response_value_p0": -0.054257319988283,
        "best_response_value_p1": 0.056100893259962,
        "unilateral_improvement_p0": 0.00129605288325,
        "unilateral_improvement_p1": 0.000547520388429,
        "nash_conv": 0.001843573271679,
        "exploitability": 0.00092178663584
      },
      "current": {
        "on_policy_value_p0": 0.045577362674244,
        "best_response_value_p0": 0.122217864581555,
        "best_response_value_p1": 0.413359517008039,
        "unilateral_improvement_p0": 0.076640501907311,
        "unilateral_improvement_p1": 0.458936879682283,
        "nash_conv": 0.535577381589594,
        "exploitability": 0.267788690794797
      }
    },
    {
      "iteration": 59600,
      "average": {
        "on_policy_value_p0": -0.055551949677838,
        "best_response_value_p0": -0.054137219800235,
        "best_response_value_p1": 0.056027082132845,
        "unilateral_improvement_p0": 0.001414729877603,
        "unilateral_improvement_p1": 0.000475132455007,
        "nash_conv": 0.001889862332611,
        "exploitability": 0.000944931166305
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 59700,
      "average": {
        "on_policy_value_p0": -0.055550389739026,
        "best_response_value_p0": -0.054348069923346,
        "best_response_value_p1": 0.056210411524228,
        "unilateral_improvement_p0": 0.001202319815681,
        "unilateral_improvement_p1": 0.000660021785202,
        "nash_conv": 0.001862341600883,
        "exploitability": 0.000931170800441
      },
      "current": {
        "on_policy_value_p0": -0.366128656518569,
        "best_response_value_p0": 0.12204288550619,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.488171542024758,
        "unilateral_improvement_p1": 0.133871343481431,
        "nash_conv": 0.62204288550619,
        "exploitability": 0.311021442753095
      }
    },
    {
      "iteration": 59800,
      "average": {
        "on_policy_value_p0": -0.055550638677221,
        "best_response_value_p0": -0.055005835957087,
        "best_response_value_p1": 0.056821074048234,
        "unilateral_improvement_p0": 0.000544802720135,
        "unilateral_improvement_p1": 0.001270435371012,
        "nash_conv": 0.001815238091147,
        "exploitability": 0.000907619045574
      },
      "current": {
        "on_policy_value_p0": -0.314361236688324,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.314361236688324,
        "unilateral_improvement_p0": 0.481027903354991,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.481027903354991,
        "exploitability": 0.240513951677495
      }
    },
    {
      "iteration": 59900,
      "average": {
        "on_policy_value_p0": -0.055552291536187,
        "best_response_value_p0": -0.055170897380468,
        "best_response_value_p1": 0.057001166093766,
        "unilateral_improvement_p0": 0.000381394155719,
        "unilateral_improvement_p1": 0.001448874557579,
        "nash_conv": 0.001830268713298,
        "exploitability": 0.000915134356649
      },
      "current": {
        "on_policy_value_p0": -0.076952766621643,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.24361943328831,
        "unilateral_improvement_p1": 0.089713900045023,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 60000,
      "average": {
        "on_policy_value_p0": -0.05555403367039,
        "best_response_value_p0": -0.055078945884833,
        "best_response_value_p1": 0.056952623463803,
        "unilateral_improvement_p0": 0.000475087785557,
        "unilateral_improvement_p1": 0.001398589793413,
        "nash_conv": 0.001873677578971,
        "exploitability": 0.000936838789485
      },
      "current": {
        "on_policy_value_p0": 0.034701354193579,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.131965312473087,
        "unilateral_improvement_p0": 0.131965312473087,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.298631979139754,
        "exploitability": 0.149315989569877
      }
    },
    {
      "iteration": 60100,
      "average": {
        "on_policy_value_p0": -0.055555395952105,
        "best_response_value_p0": -0.054959588120056,
        "best_response_value_p1": 0.056752965638836,
        "unilateral_improvement_p0": 0.000595807832049,
        "unilateral_improvement_p1": 0.001197569686731,
        "nash_conv": 0.00179337751878,
        "exploitability": 0.00089668875939
      },
      "current": {
        "on_policy_value_p0": 0.083296803470838,
        "best_response_value_p0": 0.161937831512392,
        "best_response_value_p1": 0.176051277250225,
        "unilateral_improvement_p0": 0.078641028041554,
        "unilateral_improvement_p1": 0.259348080721062,
        "nash_conv": 0.337989108762617,
        "exploitability": 0.168994554381308
      }
    },
    {
      "iteration": 60200,
      "average": {
        "on_policy_value_p0": -0.055555877308509,
        "best_response_value_p0": -0.054651806503932,
        "best_response_value_p1": 0.056468785098823,
        "unilateral_improvement_p0": 0.000904070804578,
        "unilateral_improvement_p1": 0.000912907790314,
        "nash_conv": 0.001816978594891,
        "exploitability": 0.000908489297446
      },
      "current": {
        "on_policy_value_p0": 0.087597722990601,
        "best_response_value_p0": 0.106025611398496,
        "best_response_value_p1": 0.296477556517543,
        "unilateral_improvement_p0": 0.018427888407895,
        "unilateral_improvement_p1": 0.384075279508143,
        "nash_conv": 0.402503167916039,
        "exploitability": 0.201251583958019
      }
    },
    {
      "iteration": 60300,
      "average": {
        "on_policy_value_p0": -0.055555372599163,
        "best_response_value_p0": -0.054415390255429,
        "best_response_value_p1": 0.056206132215164,
        "unilateral_improvement_p0": 0.001139982343735,
        "unilateral_improvement_p1": 0.000650759616,
        "nash_conv": 0.001790741959735,
        "exploitability": 0.000895370979867
      },
      "current": {
        "on_policy_value_p0": 0.069851111610351,
        "best_response_value_p0": 0.069851111610351,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.403184444943684,
        "nash_conv": 0.403184444943684,
        "exploitability": 0.201592222471842
      }
    },
    {
      "iteration": 60400,
      "average": {
        "on_policy_value_p0": -0.05555422556717,
        "best_response_value_p0": -0.0542437358095,
        "best_response_value_p1": 0.056113075704874,
        "unilateral_improvement_p0": 0.00131048975767,
        "unilateral_improvement_p1": 0.000558850137704,
        "nash_conv": 0.001869339895374,
        "exploitability": 0.000934669947687
      },
      "current": {
        "on_policy_value_p0": 0.023799266246386,
        "best_response_value_p0": 0.119068134173895,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.095268867927509,
        "unilateral_improvement_p1": 0.357132599579719,
        "nash_conv": 0.452401467507228,
        "exploitability": 0.226200733753614
      }
    },
    {
      "iteration": 60500,
      "average": {
        "on_policy_value_p0": -0.055552807954024,
        "best_response_value_p0": -0.054147587341201,
        "best_response_value_p1": 0.056020326819412,
        "unilateral_improvement_p0": 0.001405220612823,
        "unilateral_improvement_p1": 0.000467518865388,
        "nash_conv": 0.001872739478211,
        "exploitability": 0.000936369739105
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 60600,
      "average": {
        "on_policy_value_p0": -0.055551081005468,
        "best_response_value_p0": -0.05411728766996,
        "best_response_value_p1": 0.055927884035882,
        "unilateral_improvement_p0": 0.001433793335508,
        "unilateral_improvement_p1": 0.000376803030414,
        "nash_conv": 0.001810596365921,
        "exploitability": 0.000905298182961
      },
      "current": {
        "on_policy_value_p0": -0.120634216394715,
        "best_response_value_p0": 0.120634216394715,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.24126843278943,
        "unilateral_improvement_p1": 0.212699116938618,
        "nash_conv": 0.453967549728049,
        "exploitability": 0.226983774864024
      }
    },
    {
      "iteration": 60700,
      "average": {
        "on_policy_value_p0": -0.055549552501058,
        "best_response_value_p0": -0.054348608225022,
        "best_response_value_p1": 0.056210950184956,
        "unilateral_improvement_p0": 0.001200944276036,
        "unilateral_improvement_p1": 0.000661397683898,
        "nash_conv": 0.001862341959934,
        "exploitability": 0.000931170979967
      },
      "current": {
        "on_policy_value_p0": -0.282903918563083,
        "best_response_value_p0": 0.282903918563083,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.565807837126166,
        "unilateral_improvement_p1": 0.05042941477025,
        "nash_conv": 0.616237251896416,
        "exploitability": 0.308118625948208
      }
    },
    {
      "iteration": 60800,
      "average": {
        "on_policy_value_p0": -0.055549396856361,
        "best_response_value_p0": -0.054799828241593,
        "best_response_value_p1": 0.05666674357829,
        "unilateral_improvement_p0": 0.000749568614768,
        "unilateral_improvement_p1": 0.001117346721929,
        "nash_conv": 0.001866915336697,
        "exploitability": 0.000933457668349
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 60900,
      "average": {
        "on_policy_value_p0": -0.055550753947473,
        "best_response_value_p0": -0.05525719031892,
        "best_response_value_p1": 0.057029956129662,
        "unilateral_improvement_p0": 0.000293563628553,
        "unilateral_improvement_p1": 0.001479202182189,
        "nash_conv": 0.001772765810742,
        "exploitability": 0.000886382905371
      },
      "current": {
        "on_policy_value_p0": -0.173101474015162,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.173101474015162,
        "unilateral_improvement_p0": 0.506434807348495,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.506434807348495,
        "exploitability": 0.253217403674248
      }
    },
    {
      "iteration": 61000,
      "average": {
        "on_policy_value_p0": -0.055552758382535,
        "best_response_value_p0": -0.055185580980948,
        "best_response_value_p1": 0.056987872047073,
        "unilateral_improvement_p0": 0.000367177401587,
        "unilateral_improvement_p1": 0.001435113664538,
        "nash_conv": 0.001802291066125,
        "exploitability": 0.000901145533062
      },
      "current": {
        "on_policy_value_p0": 0.080391655416542,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.252941677916791,
        "unilateral_improvement_p1": 0.247058322083209,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 61100,
      "average": {
        "on_policy_value_p0": -0.055554045692501,
        "best_response_value_p0": -0.054830034925593,
        "best_response_value_p1": 0.056693843561825,
        "unilateral_improvement_p0": 0.000724010766907,
        "unilateral_improvement_p1": 0.001139797869324,
        "nash_conv": 0.001863808636232,
        "exploitability": 0.000931904318116
      },
      "current": {
        "on_policy_value_p0": 0.187018035239327,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.207369403811987,
        "unilateral_improvement_p0": 0.146315298094006,
        "unilateral_improvement_p1": 0.394387439051314,
        "nash_conv": 0.54070273714532,
        "exploitability": 0.27035136857266
      }
    },
    {
      "iteration": 61200,
      "average": {
        "on_policy_value_p0": -0.055554113264316,
        "best_response_value_p0": -0.054537123070012,
        "best_response_value_p1": 0.056242499328839,
        "unilateral_improvement_p0": 0.001016990194304,
        "unilateral_improvement_p1": 0.000688386064523,
        "nash_conv": 0.001705376258827,
        "exploitability": 0.000852688129413
      },
      "current": {
        "on_policy_value_p0": 0.117724969661821,
        "best_response_value_p0": 0.205067595922028,
        "best_response_value_p1": 0.325314747479585,
        "unilateral_improvement_p0": 0.087342626260207,
        "unilateral_improvement_p1": 0.443039717141406,
        "nash_conv": 0.530382343401613,
        "exploitability": 0.265191171700807
      }
    },
    {
      "iteration": 61300,
      "average": {
        "on_policy_value_p0": -0.055553129361601,
        "best_response_value_p0": -0.054258875853533,
        "best_response_value_p1": 0.056107380886715,
        "unilateral_improvement_p0": 0.001294253508067,
        "unilateral_improvement_p1": 0.000554251525114,
        "nash_conv": 0.001848505033181,
        "exploitability": 0.000924252516591
      },
      "current": {
        "on_policy_value_p0": 0.004686396958301,
        "best_response_value_p0": 0.147029421912791,
        "best_response_value_p1": 0.430823814901099,
        "unilateral_improvement_p0": 0.14234302495449,
        "unilateral_improvement_p1": 0.4355102118594,
        "nash_conv": 0.577853236813889,
        "exploitability": 0.288926618406945
      }
    },
    {
      "iteration": 61400,
      "average": {
        "on_policy_value_p0": -0.055551802502138,
        "best_response_value_p0": -0.05416610017741,
        "best_response_value_p1": 0.056027639259648,
        "unilateral_improvement_p0": 0.001385702324728,
        "unilateral_improvement_p1": 0.00047583675751,
        "nash_conv": 0.001861539082238,
        "exploitability": 0.000930769541119
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 61500,
      "average": {
        "on_policy_value_p0": -0.055550428826265,
        "best_response_value_p0": -0.054434891845362,
        "best_response_value_p1": 0.056287158212699,
        "unilateral_improvement_p0": 0.001115536980903,
        "unilateral_improvement_p1": 0.000736729386434,
        "nash_conv": 0.001852266367338,
        "exploitability": 0.000926133183669
      },
      "current": {
        "on_policy_value_p0": -0.410540380645124,
        "best_response_value_p0": 0.136846793548375,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.547387174193498,
        "unilateral_improvement_p1": 0.089459619354876,
        "nash_conv": 0.636846793548374,
        "exploitability": 0.318423396774187
      }
    },
    {
      "iteration": 61600,
      "average": {
        "on_policy_value_p0": -0.055550914501851,
        "best_response_value_p0": -0.055047656726039,
        "best_response_value_p1": 0.056853319847848,
        "unilateral_improvement_p0": 0.000503257775812,
        "unilateral_improvement_p1": 0.001302405345997,
        "nash_conv": 0.001805663121809,
        "exploitability": 0.000902831560905
      },
      "current": {
        "on_policy_value_p0": -0.290386897661114,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.290386897661114,
        "unilateral_improvement_p0": 0.457053564327781,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.457053564327781,
        "exploitability": 0.22852678216389
      }
    },
    {
      "iteration": 61700,
      "average": {
        "on_policy_value_p0": -0.055552585803178,
        "best_response_value_p0": -0.055180666977479,
        "best_response_value_p1": 0.056995567376219,
        "unilateral_improvement_p0": 0.000371918825699,
        "unilateral_improvement_p1": 0.001442981573041,
        "nash_conv": 0.00181490039874,
        "exploitability": 0.00090745019937
      },
      "current": {
        "on_policy_value_p0": -0.070027152500238,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.236693819166904,
        "unilateral_improvement_p1": 0.096639514166429,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 61800,
      "average": {
        "on_policy_value_p0": -0.055554282057155,
        "best_response_value_p0": -0.055091377872337,
        "best_response_value_p1": 0.056927213210613,
        "unilateral_improvement_p0": 0.000462904184818,
        "unilateral_improvement_p1": 0.001372931153458,
        "nash_conv": 0.001835835338276,
        "exploitability": 0.000917917669138
      },
      "current": {
        "on_policy_value_p0": 0.045023961247211,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.121642705419455,
        "unilateral_improvement_p0": 0.121642705419455,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.288309372086122,
        "exploitability": 0.144154686043061
      }
    },
    {
      "iteration": 61900,
      "average": {
        "on_policy_value_p0": -0.055555557775075,
        "best_response_value_p0": -0.054937731623136,
        "best_response_value_p1": 0.0567210999015,
        "unilateral_improvement_p0": 0.000617826151939,
        "unilateral_improvement_p1": 0.001165542126425,
        "nash_conv": 0.001783368278365,
        "exploitability": 0.000891684139182
      },
      "current": {
        "on_policy_value_p0": 0.084515746575184,
        "best_response_value_p0": 0.156573588016716,
        "best_response_value_p1": 0.189217650450268,
        "unilateral_improvement_p0": 0.072057841441533,
        "unilateral_improvement_p1": 0.273733397025452,
        "nash_conv": 0.345791238466984,
        "exploitability": 0.172895619233492
      }
    },
    {
      "iteration": 62000,
      "average": {
        "on_policy_value_p0": -0.055555944294353,
        "best_response_value_p0": -0.054643778030818,
        "best_response_value_p1": 0.056431804594965,
        "unilateral_improvement_p0": 0.000912166263535,
        "unilateral_improvement_p1": 0.000875860300612,
        "nash_conv": 0.001788026564146,
        "exploitability": 0.000894013282073
      },
      "current": {
        "on_policy_value_p0": 0.09898969101218,
        "best_response_value_p0": 0.10421309612052,
        "best_response_value_p1": 0.322886523116654,
        "unilateral_improvement_p0": 0.00522340510834,
        "unilateral_improvement_p1": 0.421876214128834,
        "nash_conv": 0.427099619237174,
        "exploitability": 0.213549809618587
      }
    },
    {
      "iteration": 62100,
      "average": {
        "on_policy_value_p0": -0.055555397044375,
        "best_response_value_p0": -0.054416518064504,
        "best_response_value_p1": 0.056184491699437,
        "unilateral_improvement_p0": 0.001138878979871,
        "unilateral_improvement_p1": 0.000629094655063,
        "nash_conv": 0.001767973634934,
        "exploitability": 0.000883986817467
      },
      "current": {
        "on_policy_value_p0": 0.068681671240444,
        "best_response_value_p0": 0.068681671240444,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.402015004573777,
        "nash_conv": 0.402015004573777,
        "exploitability": 0.201007502286889
      }
    },
    {
      "iteration": 62200,
      "average": {
        "on_policy_value_p0": -0.055554258881068,
        "best_response_value_p0": -0.054251605876942,
        "best_response_value_p1": 0.056094162934648,
        "unilateral_improvement_p0": 0.001302653004126,
        "unilateral_improvement_p1": 0.00053990405358,
        "nash_conv": 0.001842557057706,
        "exploitability": 0.000921278528853
      },
      "current": {
        "on_policy_value_p0": 0.022668269608115,
        "best_response_value_p0": 0.121330127450437,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.098661857842322,
        "unilateral_improvement_p1": 0.356001602941448,
        "nash_conv": 0.45466346078377,
        "exploitability": 0.227331730391885
      }
    },
    {
      "iteration": 62300,
      "average": {
        "on_policy_value_p0": -0.055552871158574,
        "best_response_value_p0": -0.054158790874847,
        "best_response_value_p1": 0.056004124149842,
        "unilateral_improvement_p0": 0.001394080283727,
        "unilateral_improvement_p1": 0.000451252991268,
        "nash_conv": 0.001845333274995,
        "exploitability": 0.000922666637498
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 62400,
      "average": {
        "on_policy_value_p0": -0.055551190146534,
        "best_response_value_p0": -0.054137937441817,
        "best_response_value_p1": 0.055914373950885,
        "unilateral_improvement_p0": 0.001413252704717,
        "unilateral_improvement_p1": 0.000363183804351,
        "nash_conv": 0.001776436509068,
        "exploitability": 0.000888218254534
      },
      "current": {
        "on_policy_value_p0": -0.126609185620342,
        "best_response_value_p0": 0.126609185620342,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.253218371240683,
        "unilateral_improvement_p1": 0.206724147712992,
        "nash_conv": 0.459942518953675,
        "exploitability": 0.229971259476837
      }
    },
    {
      "iteration": 62500,
      "average": {
        "on_policy_value_p0": -0.055549737216062,
        "best_response_value_p0": -0.054370610605997,
        "best_response_value_p1": 0.056207737641148,
        "unilateral_improvement_p0": 0.001179126610065,
        "unilateral_improvement_p1": 0.000658000425086,
        "nash_conv": 0.001837127035151,
        "exploitability": 0.000918563517575
      },
      "current": {
        "on_policy_value_p0": -0.289142186816165,
        "best_response_value_p0": 0.289142186816165,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.57828437363233,
        "unilateral_improvement_p1": 0.044191146517169,
        "nash_conv": 0.622475520149498,
        "exploitability": 0.311237760074749
      }
    },
    {
      "iteration": 62600,
      "average": {
        "on_policy_value_p0": -0.055549599524953,
        "best_response_value_p0": -0.054810623403716,
        "best_response_value_p1": 0.056650430286023,
        "unilateral_improvement_p0": 0.000738976121237,
        "unilateral_improvement_p1": 0.001100830761069,
        "nash_conv": 0.001839806882307,
        "exploitability": 0.000919903441153
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 62700,
      "average": {
        "on_policy_value_p0": -0.055550889682875,
        "best_response_value_p0": -0.055254838252089,
        "best_response_value_p1": 0.057006706219591,
        "unilateral_improvement_p0": 0.000296051430786,
        "unilateral_improvement_p1": 0.001455816536717,
        "nash_conv": 0.001751867967502,
        "exploitability": 0.000875933983751
      },
      "current": {
        "on_policy_value_p0": -0.176592865386329,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.176592865386329,
        "unilateral_improvement_p0": 0.509926198719662,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.509926198719662,
        "exploitability": 0.254963099359831
      }
    },
    {
      "iteration": 62800,
      "average": {
        "on_policy_value_p0": -0.055552819834309,
        "best_response_value_p0": -0.055195671306079,
        "best_response_value_p1": 0.056975352208462,
        "unilateral_improvement_p0": 0.00035714852823,
        "unilateral_improvement_p1": 0.001422532374153,
        "nash_conv": 0.001779680902383,
        "exploitability": 0.000889840451191
      },
      "current": {
        "on_policy_value_p0": 0.078524545241704,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.254808788091629,
        "unilateral_improvement_p1": 0.24519121190837,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 62900,
      "average": {
        "on_policy_value_p0": -0.055554093669994,
        "best_response_value_p0": -0.054857710133872,
        "best_response_value_p1": 0.056695433140348,
        "unilateral_improvement_p0": 0.000696383536122,
        "unilateral_improvement_p1": 0.001141339470354,
        "nash_conv": 0.001837723006476,
        "exploitability": 0.000918861503238
      },
      "current": {
        "on_policy_value_p0": 0.184349721384016,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.202032776101364,
        "unilateral_improvement_p0": 0.148983611949318,
        "unilateral_improvement_p1": 0.38638249748538,
        "nash_conv": 0.535366109434698,
        "exploitability": 0.267683054717349
      }
    },
    {
      "iteration": 63000,
      "average": {
        "on_policy_value_p0": -0.055554226829127,
        "best_response_value_p0": -0.054559808225311,
        "best_response_value_p1": 0.056260302307472,
        "unilateral_improvement_p0": 0.000994418603817,
        "unilateral_improvement_p1": 0.000706075478344,
        "nash_conv": 0.001700494082161,
        "exploitability": 0.000850247041081
      },
      "current": {
        "on_policy_value_p0": 0.126955709809277,
        "best_response_value_p0": 0.216296040452499,
        "best_response_value_p1": 0.321319338713557,
        "unilateral_improvement_p0": 0.089340330643222,
        "unilateral_improvement_p1": 0.448275048522834,
        "nash_conv": 0.537615379166056,
        "exploitability": 0.268807689583028
      }
    },
    {
      "iteration": 63100,
      "average": {
        "on_policy_value_p0": -0.055553303728141,
        "best_response_value_p0": -0.054280795331675,
        "best_response_value_p1": 0.056092030047229,
        "unilateral_improvement_p0": 0.001272508396466,
        "unilateral_improvement_p1": 0.000538726319088,
        "nash_conv": 0.001811234715554,
        "exploitability": 0.000905617357777
      },
      "current": {
        "on_policy_value_p0": 0.022568056956287,
        "best_response_value_p0": 0.136095964137929,
        "best_response_value_p1": 0.422853303797625,
        "unilateral_improvement_p0": 0.113527907181643,
        "unilateral_improvement_p1": 0.445421360753912,
        "nash_conv": 0.558949267935554,
        "exploitability": 0.279474633967777
      }
    },
    {
      "iteration": 63200,
      "average": {
        "on_policy_value_p0": -0.055552018956311,
        "best_response_value_p0": -0.054183309963444,
        "best_response_value_p1": 0.056018152094596,
        "unilateral_improvement_p0": 0.001368708992867,
        "unilateral_improvement_p1": 0.000466133138284,
        "nash_conv": 0.001834842131152,
        "exploitability": 0.000917421065576
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 63300,
      "average": {
        "on_policy_value_p0": -0.055550633205692,
        "best_response_value_p0": -0.05440439616832,
        "best_response_value_p1": 0.056217424929648,
        "unilateral_improvement_p0": 0.001146237037373,
        "unilateral_improvement_p1": 0.000666791723956,
        "nash_conv": 0.001813028761329,
        "exploitability": 0.000906514380664
      },
      "current": {
        "on_policy_value_p0": -0.376033031222323,
        "best_response_value_p0": 0.125344343740774,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.501377374963098,
        "unilateral_improvement_p1": 0.123966968777677,
        "nash_conv": 0.625344343740774,
        "exploitability": 0.312672171870387
      }
    },
    {
      "iteration": 63400,
      "average": {
        "on_policy_value_p0": -0.05555091150781,
        "best_response_value_p0": -0.055027968414389,
        "best_response_value_p1": 0.056790106884283,
        "unilateral_improvement_p0": 0.000522943093421,
        "unilateral_improvement_p1": 0.001239195376473,
        "nash_conv": 0.001762138469894,
        "exploitability": 0.000881069234947
      },
      "current": {
        "on_policy_value_p0": -0.314626001352576,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.314626001352576,
        "unilateral_improvement_p0": 0.481292668019243,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.481292668019243,
        "exploitability": 0.240646334009621
      }
    },
    {
      "iteration": 63500,
      "average": {
        "on_policy_value_p0": -0.055552438787995,
        "best_response_value_p0": -0.055187899180251,
        "best_response_value_p1": 0.056962870305419,
        "unilateral_improvement_p0": 0.000364539607745,
        "unilateral_improvement_p1": 0.001410431517423,
        "nash_conv": 0.001774971125168,
        "exploitability": 0.000887485562584
      },
      "current": {
        "on_policy_value_p0": -0.078236952152531,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.244903618819198,
        "unilateral_improvement_p1": 0.088429714514136,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 63600,
      "average": {
        "on_policy_value_p0": -0.055554054868797,
        "best_response_value_p0": -0.055101125753866,
        "best_response_value_p1": 0.056922029189322,
        "unilateral_improvement_p0": 0.000452929114931,
        "unilateral_improvement_p1": 0.001367974320525,
        "nash_conv": 0.001820903435456,
        "exploitability": 0.000910451717728
      },
      "current": {
        "on_policy_value_p0": 0.032903085503276,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.13376358116339,
        "unilateral_improvement_p0": 0.13376358116339,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.300430247830057,
        "exploitability": 0.150215123915029
      }
    },
    {
      "iteration": 63700,
      "average": {
        "on_policy_value_p0": -0.055555344874885,
        "best_response_value_p0": -0.055001888329452,
        "best_response_value_p1": 0.056736881940634,
        "unilateral_improvement_p0": 0.000553456545433,
        "unilateral_improvement_p1": 0.00118153706575,
        "nash_conv": 0.001734993611183,
        "exploitability": 0.000867496805591
      },
      "current": {
        "on_policy_value_p0": 0.085538214770818,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.171076429541636,
        "unilateral_improvement_p0": 0.081128451895848,
        "unilateral_improvement_p1": 0.256614644312455,
        "nash_conv": 0.337743096208303,
        "exploitability": 0.168871548104152
      }
    },
    {
      "iteration": 63800,
      "average": {
        "on_policy_value_p0": -0.055555869278746,
        "best_response_value_p0": -0.054704715678752,
        "best_response_value_p1": 0.056475070604177,
        "unilateral_improvement_p0": 0.000851153599995,
        "unilateral_improvement_p1": 0.000919201325431,
        "nash_conv": 0.001770354925425,
        "exploitability": 0.000885177462713
      },
      "current": {
        "on_policy_value_p0": 0.083937884556843,
        "best_response_value_p0": 0.109634620826278,
        "best_response_value_p1": 0.281939860794462,
        "unilateral_improvement_p0": 0.025696736269435,
        "unilateral_improvement_p1": 0.365877745351305,
        "nash_conv": 0.391574481620741,
        "exploitability": 0.19578724081037
      }
    },
    {
      "iteration": 63900,
      "average": {
        "on_policy_value_p0": -0.055555498686147,
        "best_response_value_p0": -0.054475590702186,
        "best_response_value_p1": 0.056201134525745,
        "unilateral_improvement_p0": 0.001079907983961,
        "unilateral_improvement_p1": 0.000645635839598,
        "nash_conv": 0.001725543823559,
        "exploitability": 0.00086277191178
      },
      "current": {
        "on_policy_value_p0": 0.074168411764552,
        "best_response_value_p0": 0.074168411764552,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.407501745097885,
        "nash_conv": 0.407501745097885,
        "exploitability": 0.203750872548943
      }
    },
    {
      "iteration": 64000,
      "average": {
        "on_policy_value_p0": -0.055554514809137,
        "best_response_value_p0": -0.05430419310803,
        "best_response_value_p1": 0.056113320253049,
        "unilateral_improvement_p0": 0.001250321701107,
        "unilateral_improvement_p1": 0.000558805443913,
        "nash_conv": 0.00180912714502,
        "exploitability": 0.00090456357251
      },
      "current": {
        "on_policy_value_p0": 0.032752983993817,
        "best_response_value_p0": 0.101160698679032,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.068407714685215,
        "unilateral_improvement_p1": 0.366086317327151,
        "nash_conv": 0.434494032012365,
        "exploitability": 0.217247016006183
      }
    },
    {
      "iteration": 64100,
      "average": {
        "on_policy_value_p0": -0.055553256040015,
        "best_response_value_p0": -0.054206343195281,
        "best_response_value_p1": 0.056025779971844,
        "unilateral_improvement_p0": 0.001346912844735,
        "unilateral_improvement_p1": 0.000472523931829,
        "nash_conv": 0.001819436776564,
        "exploitability": 0.000909718388282
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 64200,
      "average": {
        "on_policy_value_p0": -0.055551742910482,
        "best_response_value_p0": -0.054143312533141,
        "best_response_value_p1": 0.055938512401796,
        "unilateral_improvement_p0": 0.001408430377341,
        "unilateral_improvement_p1": 0.000386769491314,
        "nash_conv": 0.001795199868655,
        "exploitability": 0.000897599934328
      },
      "current": {
        "on_policy_value_p0": -0.081944682162776,
        "best_response_value_p0": 0.125694325585279,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.207639007748055,
        "unilateral_improvement_p1": 0.251388651170557,
        "nash_conv": 0.459027658918612,
        "exploitability": 0.229513829459306
      }
    },
    {
      "iteration": 64300,
      "average": {
        "on_policy_value_p0": -0.055550192028421,
        "best_response_value_p0": -0.054304408364096,
        "best_response_value_p1": 0.056082397903039,
        "unilateral_improvement_p0": 0.001245783664325,
        "unilateral_improvement_p1": 0.000532205874619,
        "nash_conv": 0.001777989538944,
        "exploitability": 0.000888994769472
      },
      "current": {
        "on_policy_value_p0": -0.231511451045757,
        "best_response_value_p0": 0.231511451045757,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.463022902091515,
        "unilateral_improvement_p1": 0.101821882287576,
        "nash_conv": 0.564844784379091,
        "exploitability": 0.282422392189545
      }
    },
    {
      "iteration": 64400,
      "average": {
        "on_policy_value_p0": -0.055549668042163,
        "best_response_value_p0": -0.054698718913033,
        "best_response_value_p1": 0.05651291177793,
        "unilateral_improvement_p0": 0.00085094912913,
        "unilateral_improvement_p1": 0.000963243735767,
        "nash_conv": 0.001814192864898,
        "exploitability": 0.000907096432449
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 64500,
      "average": {
        "on_policy_value_p0": -0.055550522687879,
        "best_response_value_p0": -0.055130710563297,
        "best_response_value_p1": 0.056918971193794,
        "unilateral_improvement_p0": 0.000419812124582,
        "unilateral_improvement_p1": 0.001368448505915,
        "nash_conv": 0.001788260630497,
        "exploitability": 0.000894130315248
      },
      "current": {
        "on_policy_value_p0": -0.238421518451759,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.238421518451759,
        "unilateral_improvement_p0": 0.571754851785092,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.571754851785092,
        "exploitability": 0.285877425892546
      }
    },
    {
      "iteration": 64600,
      "average": {
        "on_policy_value_p0": -0.055552372567075,
        "best_response_value_p0": -0.055234042217533,
        "best_response_value_p1": 0.057020577715221,
        "unilateral_improvement_p0": 0.000318330349542,
        "unilateral_improvement_p1": 0.001468205148146,
        "nash_conv": 0.001786535497688,
        "exploitability": 0.000893267748844
      },
      "current": {
        "on_policy_value_p0": 0.045415545163359,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.16092231293634,
        "unilateral_improvement_p0": 0.287917788169975,
        "unilateral_improvement_p1": 0.206337858099698,
        "nash_conv": 0.494255646269673,
        "exploitability": 0.247127823134837
      }
    },
    {
      "iteration": 64700,
      "average": {
        "on_policy_value_p0": -0.055553874621426,
        "best_response_value_p0": -0.054984064373364,
        "best_response_value_p1": 0.056790476710064,
        "unilateral_improvement_p0": 0.000569810248062,
        "unilateral_improvement_p1": 0.001236602088638,
        "nash_conv": 0.0018064123367,
        "exploitability": 0.00090320616835
      },
      "current": {
        "on_policy_value_p0": 0.131445276855158,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.201888056478175,
        "unilateral_improvement_p1": 0.298111943521825,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 64800,
      "average": {
        "on_policy_value_p0": -0.055554418471824,
        "best_response_value_p0": -0.054649362811552,
        "best_response_value_p1": 0.056403609598608,
        "unilateral_improvement_p0": 0.000905055660272,
        "unilateral_improvement_p1": 0.000849191126785,
        "nash_conv": 0.001754246787056,
        "exploitability": 0.000877123393528
      },
      "current": {
        "on_policy_value_p0": 0.177454916098288,
        "best_response_value_p0": 0.280365626294028,
        "best_response_value_p1": 0.294178579608521,
        "unilateral_improvement_p0": 0.10291071019574,
        "unilateral_improvement_p1": 0.471633495706809,
        "nash_conv": 0.574544205902549,
        "exploitability": 0.287272102951274
      }
    },
    {
      "iteration": 64900,
      "average": {
        "on_policy_value_p0": -0.055553786869269,
        "best_response_value_p0": -0.054372428512725,
        "best_response_value_p1": 0.056083477140306,
        "unilateral_improvement_p0": 0.001181358356544,
        "unilateral_improvement_p1": 0.000529690271036,
        "nash_conv": 0.001711048627581,
        "exploitability": 0.00085552431379
      },
      "current": {
        "on_policy_value_p0": 0.067896764294407,
        "best_response_value_p0": 0.124141082658582,
        "best_response_value_p1": 0.38751136327165,
        "unilateral_improvement_p0": 0.056244318364175,
        "unilateral_improvement_p1": 0.455408127566057,
        "nash_conv": 0.511652445930232,
        "exploitability": 0.255826222965116
      }
    },
    {
      "iteration": 65000,
      "average": {
        "on_policy_value_p0": -0.055552615945368,
        "best_response_value_p0": -0.054224856674166,
        "best_response_value_p1": 0.056034449419107,
        "unilateral_improvement_p0": 0.001327759271203,
        "unilateral_improvement_p1": 0.000481833473739,
        "nash_conv": 0.001809592744941,
        "exploitability": 0.000904796372471
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 65100,
      "average": {
        "on_policy_value_p0": -0.055551191446483,
        "best_response_value_p0": -0.054276847065032,
        "best_response_value_p1": 0.055963203312111,
        "unilateral_improvement_p0": 0.001274344381451,
        "unilateral_improvement_p1": 0.000412011865628,
        "nash_conv": 0.00168635624708,
        "exploitability": 0.00084317812354
      },
      "current": {
        "on_policy_value_p0": -0.255896949389162,
        "best_response_value_p0": 0.085298983129721,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.341195932518882,
        "unilateral_improvement_p1": 0.244103050610838,
        "nash_conv": 0.585298983129721,
        "exploitability": 0.29264949156486
      }
    },
    {
      "iteration": 65200,
      "average": {
        "on_policy_value_p0": -0.055550736943745,
        "best_response_value_p0": -0.054835423856869,
        "best_response_value_p1": 0.056604814223981,
        "unilateral_improvement_p0": 0.000715313086876,
        "unilateral_improvement_p1": 0.001054077280236,
        "nash_conv": 0.001769390367112,
        "exploitability": 0.000884695183556
      },
      "current": {
        "on_policy_value_p0": -0.383213520597012,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.383213520597012,
        "unilateral_improvement_p0": 0.549880187263679,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.549880187263679,
        "exploitability": 0.274940093631839
      }
    },
    {
      "iteration": 65300,
      "average": {
        "on_policy_value_p0": -0.055551940220368,
        "best_response_value_p0": -0.055142186365366,
        "best_response_value_p1": 0.056921764956887,
        "unilateral_improvement_p0": 0.000409753855002,
        "unilateral_improvement_p1": 0.001369824736519,
        "nash_conv": 0.001779578591521,
        "exploitability": 0.000889789295761
      },
      "current": {
        "on_policy_value_p0": -0.104721856613981,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.271388523280648,
        "unilateral_improvement_p1": 0.061944810052686,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 65400,
      "average": {
        "on_policy_value_p0": -0.055553513318416,
        "best_response_value_p0": -0.055141065938649,
        "best_response_value_p1": 0.05694155202203,
        "unilateral_improvement_p0": 0.000412447379767,
        "unilateral_improvement_p1": 0.001388038703614,
        "nash_conv": 0.001800486083382,
        "exploitability": 0.000900243041691
      },
      "current": {
        "on_policy_value_p0": -0.028782515699093,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.19544918236576,
        "unilateral_improvement_p1": 0.137884150967574,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 65500,
      "average": {
        "on_policy_value_p0": -0.055554915092965,
        "best_response_value_p0": -0.055056881105154,
        "best_response_value_p1": 0.056804306509727,
        "unilateral_improvement_p0": 0.000498033987811,
        "unilateral_improvement_p1": 0.001249391416762,
        "nash_conv": 0.001747425404574,
        "exploitability": 0.000873712702287
      },
      "current": {
        "on_policy_value_p0": 0.066459294247176,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.132918588494352,
        "unilateral_improvement_p0": 0.100207372419491,
        "unilateral_improvement_p1": 0.199377882741528,
        "nash_conv": 0.299585255161019,
        "exploitability": 0.149792627580509
      }
    },
    {
      "iteration": 65600,
      "average": {
        "on_policy_value_p0": -0.055555768402963,
        "best_response_value_p0": -0.054829824742508,
        "best_response_value_p1": 0.056581860987853,
        "unilateral_improvement_p0": 0.000725943660454,
        "unilateral_improvement_p1": 0.00102609258489,
        "nash_conv": 0.001752036245344,
        "exploitability": 0.000876018122672
      },
      "current": {
        "on_policy_value_p0": 0.076038326868806,
        "best_response_value_p0": 0.129541923260342,
        "best_response_value_p1": 0.226326140550262,
        "unilateral_improvement_p0": 0.053503596391536,
        "unilateral_improvement_p1": 0.302364467419068,
        "nash_conv": 0.355868063810604,
        "exploitability": 0.177934031905302
      }
    },
    {
      "iteration": 65700,
      "average": {
        "on_policy_value_p0": -0.055555764071696,
        "best_response_value_p0": -0.054580938460455,
        "best_response_value_p1": 0.056274698878816,
        "unilateral_improvement_p0": 0.000974825611241,
        "unilateral_improvement_p1": 0.00071893480712,
        "nash_conv": 0.001693760418361,
        "exploitability": 0.00084688020918
      },
      "current": {
        "on_policy_value_p0": 0.090033339857149,
        "best_response_value_p0": 0.090033339857149,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.423366673190483,
        "nash_conv": 0.423366673190483,
        "exploitability": 0.211683336595241
      }
    },
    {
      "iteration": 65800,
      "average": {
        "on_policy_value_p0": -0.055555035886018,
        "best_response_value_p0": -0.054386974570995,
        "best_response_value_p1": 0.056138973732489,
        "unilateral_improvement_p0": 0.001168061315024,
        "unilateral_improvement_p1": 0.00058393784647,
        "nash_conv": 0.001751999161494,
        "exploitability": 0.000875999580747
      },
      "current": {
        "on_policy_value_p0": 0.054799416164096,
        "best_response_value_p0": 0.057067834338475,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.00226841817438,
        "unilateral_improvement_p1": 0.388132749497429,
        "nash_conv": 0.390401167671809,
        "exploitability": 0.195200583835904
      }
    },
    {
      "iteration": 65900,
      "average": {
        "on_policy_value_p0": -0.055553898200312,
        "best_response_value_p0": -0.054258826459692,
        "best_response_value_p1": 0.056053785608465,
        "unilateral_improvement_p0": 0.00129507174062,
        "unilateral_improvement_p1": 0.000499887408153,
        "nash_conv": 0.001794959148773,
        "exploitability": 0.000897479574387
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 66000,
      "average": {
        "on_policy_value_p0": -0.055552567666857,
        "best_response_value_p0": -0.054176616116571,
        "best_response_value_p1": 0.055968855630271,
        "unilateral_improvement_p0": 0.001375951550285,
        "unilateral_improvement_p1": 0.000416287963414,
        "nash_conv": 0.0017922395137,
        "exploitability": 0.00089611975685
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 66100,
      "average": {
        "on_policy_value_p0": -0.055550985512961,
        "best_response_value_p0": -0.054215738878494,
        "best_response_value_p1": 0.055884182626293,
        "unilateral_improvement_p0": 0.001335246634467,
        "unilateral_improvement_p1": 0.000333197113331,
        "nash_conv": 0.001668443747798,
        "exploitability": 0.000834221873899
      },
      "current": {
        "on_policy_value_p0": -0.162738665028824,
        "best_response_value_p0": 0.162738665028823,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.325477330057647,
        "unilateral_improvement_p1": 0.17059466830451,
        "nash_conv": 0.496071998362157,
        "exploitability": 0.248035999181078
      }
    },
    {
      "iteration": 66200,
      "average": {
        "on_policy_value_p0": -0.055549882173885,
        "best_response_value_p0": -0.05449857076402,
        "best_response_value_p1": 0.056288532743495,
        "unilateral_improvement_p0": 0.001051311409864,
        "unilateral_improvement_p1": 0.00073865056961,
        "nash_conv": 0.001789961979474,
        "exploitability": 0.000894980989737
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 66300,
      "average": {
        "on_policy_value_p0": -0.055550043245241,
        "best_response_value_p0": -0.054919136016765,
        "best_response_value_p1": 0.056706398204414,
        "unilateral_improvement_p0": 0.000630907228476,
        "unilateral_improvement_p1": 0.001156354959173,
        "nash_conv": 0.001787262187649,
        "exploitability": 0.000893631093824
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 66400,
      "average": {
        "on_policy_value_p0": -0.055551437342057,
        "best_response_value_p0": -0.055272253607783,
        "best_response_value_p1": 0.05699159240121,
        "unilateral_improvement_p0": 0.000279183734274,
        "unilateral_improvement_p1": 0.001440155059153,
        "nash_conv": 0.001719338793427,
        "exploitability": 0.000859669396713
      },
      "current": {
        "on_policy_value_p0": -0.143076465147088,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.143076465147088,
        "unilateral_improvement_p0": 0.476409798480421,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.476409798480421,
        "exploitability": 0.238204899240211
      }
    },
    {
      "iteration": 66500,
      "average": {
        "on_policy_value_p0": -0.055553161477429,
        "best_response_value_p0": -0.055167459380053,
        "best_response_value_p1": 0.056900199540629,
        "unilateral_improvement_p0": 0.000385702097376,
        "unilateral_improvement_p1": 0.0013470380632,
        "nash_conv": 0.001732740160576,
        "exploitability": 0.000866370080288
      },
      "current": {
        "on_policy_value_p0": 0.088057510367541,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.245275822965793,
        "unilateral_improvement_p1": 0.254724177034207,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 66600,
      "average": {
        "on_policy_value_p0": -0.055554203603082,
        "best_response_value_p0": -0.054834375106709,
        "best_response_value_p1": 0.056617083897223,
        "unilateral_improvement_p0": 0.000719828496373,
        "unilateral_improvement_p1": 0.001062880294141,
        "nash_conv": 0.001782708790514,
        "exploitability": 0.000891354395257
      },
      "current": {
        "on_policy_value_p0": 0.191821598203767,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.216976529740867,
        "unilateral_improvement_p0": 0.141511735129567,
        "unilateral_improvement_p1": 0.408798127944633,
        "nash_conv": 0.5503098630742,
        "exploitability": 0.2751549315371
      }
    },
    {
      "iteration": 66700,
      "average": {
        "on_policy_value_p0": -0.055554200781677,
        "best_response_value_p0": -0.054574556323908,
        "best_response_value_p1": 0.056199908620923,
        "unilateral_improvement_p0": 0.000979644457769,
        "unilateral_improvement_p1": 0.000645707839245,
        "nash_conv": 0.001625352297015,
        "exploitability": 0.000812676148507
      },
      "current": {
        "on_policy_value_p0": 0.113832581011235,
        "best_response_value_p0": 0.200588594162791,
        "best_response_value_p1": 0.326487973696888,
        "unilateral_improvement_p0": 0.086756013151556,
        "unilateral_improvement_p1": 0.440320554708123,
        "nash_conv": 0.527076567859679,
        "exploitability": 0.26353828392984
      }
    },
    {
      "iteration": 66800,
      "average": {
        "on_policy_value_p0": -0.055553333404388,
        "best_response_value_p0": -0.054318569850685,
        "best_response_value_p1": 0.056086460880687,
        "unilateral_improvement_p0": 0.001234763553704,
        "unilateral_improvement_p1": 0.000533127476298,
        "nash_conv": 0.001767891030002,
        "exploitability": 0.000883945515001
      },
      "current": {
        "on_policy_value_p0": 0.005895305400962,
        "best_response_value_p0": 0.145391183992874,
        "best_response_value_p1": 0.426688680106755,
        "unilateral_improvement_p0": 0.139495878591913,
        "unilateral_improvement_p1": 0.432583985507716,
        "nash_conv": 0.572079864099629,
        "exploitability": 0.286039932049815
      }
    },
    {
      "iteration": 66900,
      "average": {
        "on_policy_value_p0": -0.055552185621881,
        "best_response_value_p0": -0.054232371354086,
        "best_response_value_p1": 0.056015411307081,
        "unilateral_improvement_p0": 0.001319814267795,
        "unilateral_improvement_p1": 0.0004632256852,
        "nash_conv": 0.001783039952995,
        "exploitability": 0.000891519976498
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 67000,
      "average": {
        "on_policy_value_p0": -0.055550925271407,
        "best_response_value_p0": -0.054429274335,
        "best_response_value_p1": 0.056183714010341,
        "unilateral_improvement_p0": 0.001121650936407,
        "unilateral_improvement_p1": 0.000632788738934,
        "nash_conv": 0.001754439675341,
        "exploitability": 0.00087721983767
      },
      "current": {
        "on_policy_value_p0": -0.36089185809698,
        "best_response_value_p0": 0.120297286032327,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.481189144129307,
        "unilateral_improvement_p1": 0.13910814190302,
        "nash_conv": 0.620297286032327,
        "exploitability": 0.310148643016163
      }
    },
    {
      "iteration": 67100,
      "average": {
        "on_policy_value_p0": -0.055551118064214,
        "best_response_value_p0": -0.055043809845204,
        "best_response_value_p1": 0.056740042398139,
        "unilateral_improvement_p0": 0.000507308219011,
        "unilateral_improvement_p1": 0.001188924333924,
        "nash_conv": 0.001696232552935,
        "exploitability": 0.000848116276467
      },
      "current": {
        "on_policy_value_p0": -0.327069449993757,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.327069449993757,
        "unilateral_improvement_p0": 0.493736116660424,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.493736116660424,
        "exploitability": 0.246868058330212
      }
    },
    {
      "iteration": 67200,
      "average": {
        "on_policy_value_p0": -0.055552517557036,
        "best_response_value_p0": -0.055209915286902,
        "best_response_value_p1": 0.056928927917638,
        "unilateral_improvement_p0": 0.000342602270134,
        "unilateral_improvement_p1": 0.001376410360602,
        "nash_conv": 0.001719012630736,
        "exploitability": 0.000859506315368
      },
      "current": {
        "on_policy_value_p0": -0.083075748459957,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.249742415126624,
        "unilateral_improvement_p1": 0.083590918206709,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 67300,
      "average": {
        "on_policy_value_p0": -0.055554039678156,
        "best_response_value_p0": -0.055130620962244,
        "best_response_value_p1": 0.056902931976633,
        "unilateral_improvement_p0": 0.000423418715912,
        "unilateral_improvement_p1": 0.001348892298477,
        "nash_conv": 0.001772311014389,
        "exploitability": 0.000886155507194
      },
      "current": {
        "on_policy_value_p0": 0.029074237258721,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.137592429407946,
        "unilateral_improvement_p0": 0.137592429407946,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.304259096074612,
        "exploitability": 0.152129548037306
      }
    },
    {
      "iteration": 67400,
      "average": {
        "on_policy_value_p0": -0.055555292429159,
        "best_response_value_p0": -0.055048824788709,
        "best_response_value_p1": 0.056733057389351,
        "unilateral_improvement_p0": 0.000506467640449,
        "unilateral_improvement_p1": 0.001177764960192,
        "nash_conv": 0.001684232600642,
        "exploitability": 0.000842116300321
      },
      "current": {
        "on_policy_value_p0": 0.081953528017954,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.163907056035907,
        "unilateral_improvement_p0": 0.084713138648713,
        "unilateral_improvement_p1": 0.245860584053861,
        "nash_conv": 0.330573722702574,
        "exploitability": 0.165286861351287
      }
    },
    {
      "iteration": 67500,
      "average": {
        "on_policy_value_p0": -0.055555891352033,
        "best_response_value_p0": -0.054763375186105,
        "best_response_value_p1": 0.056493088182561,
        "unilateral_improvement_p0": 0.000792516165928,
        "unilateral_improvement_p1": 0.000937196830528,
        "nash_conv": 0.001729712996456,
        "exploitability": 0.000864856498228
      },
      "current": {
        "on_policy_value_p0": 0.082678985796997,
        "best_response_value_p0": 0.116487220415303,
        "best_response_value_p1": 0.265716864096721,
        "unilateral_improvement_p0": 0.033808234618306,
        "unilateral_improvement_p1": 0.348395849893718,
        "nash_conv": 0.382204084512024,
        "exploitability": 0.191102042256012
      }
    },
    {
      "iteration": 67600,
      "average": {
        "on_policy_value_p0": -0.055555660294383,
        "best_response_value_p0": -0.05453707534566,
        "best_response_value_p1": 0.056191535608507,
        "unilateral_improvement_p0": 0.001018584948723,
        "unilateral_improvement_p1": 0.000635875314124,
        "nash_conv": 0.001654460262848,
        "exploitability": 0.000827230131424
      },
      "current": {
        "on_policy_value_p0": 0.080896715945047,
        "best_response_value_p0": 0.080896715945047,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.41423004927838,
        "nash_conv": 0.41423004927838,
        "exploitability": 0.20711502463919
      }
    },
    {
      "iteration": 67700,
      "average": {
        "on_policy_value_p0": -0.055554825858527,
        "best_response_value_p0": -0.054362691202667,
        "best_response_value_p1": 0.056108534817358,
        "unilateral_improvement_p0": 0.001192134655861,
        "unilateral_improvement_p1": 0.000553708958831,
        "nash_conv": 0.001745843614692,
        "exploitability": 0.000872921807346
      },
      "current": {
        "on_policy_value_p0": 0.043931740030246,
        "best_response_value_p0": 0.078803186606174,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.034871446575928,
        "unilateral_improvement_p1": 0.377265073363579,
        "nash_conv": 0.412136519939508,
        "exploitability": 0.206068259969754
      }
    },
    {
      "iteration": 67800,
      "average": {
        "on_policy_value_p0": -0.055553685554735,
        "best_response_value_p0": -0.054256608568309,
        "best_response_value_p1": 0.056025778866301,
        "unilateral_improvement_p0": 0.001297076986426,
        "unilateral_improvement_p1": 0.000472093311566,
        "nash_conv": 0.001769170297992,
        "exploitability": 0.000884585148996
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 67900,
      "average": {
        "on_policy_value_p0": -0.055552350085979,
        "best_response_value_p0": -0.054178184501447,
        "best_response_value_p1": 0.055943266673568,
        "unilateral_improvement_p0": 0.001374165584532,
        "unilateral_improvement_p1": 0.000390916587589,
        "nash_conv": 0.001765082172121,
        "exploitability": 0.00088254108606
      },
      "current": {
        "on_policy_value_p0": -0.027013264977113,
        "best_response_value_p0": 0.15316003417811,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.180173299155223,
        "unilateral_improvement_p1": 0.30632006835622,
        "nash_conv": 0.486493367511443,
        "exploitability": 0.243246683755722
      }
    },
    {
      "iteration": 68000,
      "average": {
        "on_policy_value_p0": -0.055550835736702,
        "best_response_value_p0": -0.054274101355589,
        "best_response_value_p1": 0.055955189519883,
        "unilateral_improvement_p0": 0.001276734381113,
        "unilateral_improvement_p1": 0.00040435378318,
        "nash_conv": 0.001681088164294,
        "exploitability": 0.000840544082147
      },
      "current": {
        "on_policy_value_p0": -0.192334209583673,
        "best_response_value_p0": 0.192334209583673,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.384668419167346,
        "unilateral_improvement_p1": 0.140999123749661,
        "nash_conv": 0.525667542917006,
        "exploitability": 0.262833771458503
      }
    },
    {
      "iteration": 68100,
      "average": {
        "on_policy_value_p0": -0.05555000534746,
        "best_response_value_p0": -0.054598069358365,
        "best_response_value_p1": 0.056362499569535,
        "unilateral_improvement_p0": 0.000951935989095,
        "unilateral_improvement_p1": 0.000812494222075,
        "nash_conv": 0.00176443021117,
        "exploitability": 0.000882215105585
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 68200,
      "average": {
        "on_policy_value_p0": -0.055550398411427,
        "best_response_value_p0": -0.055006772091467,
        "best_response_value_p1": 0.056768615161563,
        "unilateral_improvement_p0": 0.00054362631996,
        "unilateral_improvement_p1": 0.001218216750136,
        "nash_conv": 0.001761843070096,
        "exploitability": 0.000880921535048
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 68300,
      "average": {
        "on_policy_value_p0": -0.055551906228963,
        "best_response_value_p0": -0.055270529123701,
        "best_response_value_p1": 0.056993333604018,
        "unilateral_improvement_p0": 0.000281377105262,
        "unilateral_improvement_p1": 0.001441427375054,
        "nash_conv": 0.001722804480317,
        "exploitability": 0.000861402240158
      },
      "current": {
        "on_policy_value_p0": -0.089855039090639,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.110358559391649,
        "unilateral_improvement_p0": 0.423188372423973,
        "unilateral_improvement_p1": 0.020503520301009,
        "nash_conv": 0.443691892724982,
        "exploitability": 0.221845946362491
      }
    },
    {
      "iteration": 68400,
      "average": {
        "on_policy_value_p0": -0.055553514359444,
        "best_response_value_p0": -0.055125377849504,
        "best_response_value_p1": 0.056852384433949,
        "unilateral_improvement_p0": 0.00042813650994,
        "unilateral_improvement_p1": 0.001298870074505,
        "nash_conv": 0.001727006584445,
        "exploitability": 0.000863503292222
      },
      "current": {
        "on_policy_value_p0": 0.099844764165454,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.23348856916788,
        "unilateral_improvement_p1": 0.26651143083212,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 68500,
      "average": {
        "on_policy_value_p0": -0.05555439974447,
        "best_response_value_p0": -0.05480159384291,
        "best_response_value_p1": 0.056548367153195,
        "unilateral_improvement_p0": 0.00075280590156,
        "unilateral_improvement_p1": 0.000993967408725,
        "nash_conv": 0.001746773310285,
        "exploitability": 0.000873386655143
      },
      "current": {
        "on_policy_value_p0": 0.207339327244553,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.248011987822439,
        "unilateral_improvement_p0": 0.12599400608878,
        "unilateral_improvement_p1": 0.455351315066992,
        "nash_conv": 0.581345321155772,
        "exploitability": 0.290672660577886
      }
    },
    {
      "iteration": 68600,
      "average": {
        "on_policy_value_p0": -0.055554241954439,
        "best_response_value_p0": -0.054543109394994,
        "best_response_value_p1": 0.056126786958971,
        "unilateral_improvement_p0": 0.001011132559445,
        "unilateral_improvement_p1": 0.000572545004532,
        "nash_conv": 0.001583677563978,
        "exploitability": 0.000791838781989
      },
      "current": {
        "on_policy_value_p0": 0.107264625896758,
        "best_response_value_p0": 0.184920469630798,
        "best_response_value_p1": 0.344688312531919,
        "unilateral_improvement_p0": 0.077655843734041,
        "unilateral_improvement_p1": 0.451952938428677,
        "nash_conv": 0.529608782162717,
        "exploitability": 0.264804391081359
      }
    },
    {
      "iteration": 68700,
      "average": {
        "on_policy_value_p0": -0.05555331002823,
        "best_response_value_p0": -0.054299540351202,
        "best_response_value_p1": 0.056055466619723,
        "unilateral_improvement_p0": 0.001253769677028,
        "unilateral_improvement_p1": 0.000502156591493,
        "nash_conv": 0.001755926268521,
        "exploitability": 0.000877963134261
      },
      "current": {
        "on_policy_value_p0": -0.008706429622154,
        "best_response_value_p0": 0.159802599135139,
        "best_response_value_p1": 0.455130870629582,
        "unilateral_improvement_p0": 0.168509028757294,
        "unilateral_improvement_p1": 0.446424441007428,
        "nash_conv": 0.614933469764721,
        "exploitability": 0.307466734882361
      }
    },
    {
      "iteration": 68800,
      "average": {
        "on_policy_value_p0": -0.055552109172063,
        "best_response_value_p0": -0.054225540823729,
        "best_response_value_p1": 0.055978064674244,
        "unilateral_improvement_p0": 0.001326568348335,
        "unilateral_improvement_p1": 0.000425955502181,
        "nash_conv": 0.001752523850515,
        "exploitability": 0.000876261925258
      },
      "current": {
        "on_policy_value_p0": -0.072040066450447,
        "best_response_value_p0": 0.118639955699702,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.190680022150149,
        "unilateral_improvement_p1": 0.427959933549553,
        "nash_conv": 0.618639955699702,
        "exploitability": 0.309319977849851
      }
    },
    {
      "iteration": 68900,
      "average": {
        "on_policy_value_p0": -0.055550960497938,
        "best_response_value_p0": -0.054513869253085,
        "best_response_value_p1": 0.056269512953545,
        "unilateral_improvement_p0": 0.001037091244853,
        "unilateral_improvement_p1": 0.000718552455607,
        "nash_conv": 0.001755643700461,
        "exploitability": 0.00087782185023
      },
      "current": {
        "on_policy_value_p0": -0.441116675178882,
        "best_response_value_p0": 0.147038891726294,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.588155566905176,
        "unilateral_improvement_p1": 0.058883324821118,
        "nash_conv": 0.647038891726294,
        "exploitability": 0.323519445863147
      }
    },
    {
      "iteration": 69000,
      "average": {
        "on_policy_value_p0": -0.05555138148588,
        "best_response_value_p0": -0.055059321645537,
        "best_response_value_p1": 0.05676579968571,
        "unilateral_improvement_p0": 0.000492059840343,
        "unilateral_improvement_p1": 0.00121441819983,
        "nash_conv": 0.001706478040173,
        "exploitability": 0.000853239020086
      },
      "current": {
        "on_policy_value_p0": -0.29472139058453,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.29472139058453,
        "unilateral_improvement_p0": 0.461388057251196,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.461388057251196,
        "exploitability": 0.230694028625598
      }
    },
    {
      "iteration": 69100,
      "average": {
        "on_policy_value_p0": -0.055552742315834,
        "best_response_value_p0": -0.055197011529437,
        "best_response_value_p1": 0.056901946429259,
        "unilateral_improvement_p0": 0.000355730786397,
        "unilateral_improvement_p1": 0.001349204113425,
        "nash_conv": 0.001704934899822,
        "exploitability": 0.000852467449911
      },
      "current": {
        "on_policy_value_p0": -0.075953155910492,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.242619822577158,
        "unilateral_improvement_p1": 0.090713510756175,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 69200,
      "average": {
        "on_policy_value_p0": -0.055554153868442,
        "best_response_value_p0": -0.055117247061908,
        "best_response_value_p1": 0.056863817075658,
        "unilateral_improvement_p0": 0.000436906806534,
        "unilateral_improvement_p1": 0.001309663207216,
        "nash_conv": 0.00174657001375,
        "exploitability": 0.000873285006875
      },
      "current": {
        "on_policy_value_p0": 0.030771005526608,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.135895661140058,
        "unilateral_improvement_p0": 0.135895661140058,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.302562327806725,
        "exploitability": 0.151281163903362
      }
    },
    {
      "iteration": 69300,
      "average": {
        "on_policy_value_p0": -0.055555298504212,
        "best_response_value_p0": -0.055037712794863,
        "best_response_value_p1": 0.056697385968721,
        "unilateral_improvement_p0": 0.000517585709349,
        "unilateral_improvement_p1": 0.001142087464509,
        "nash_conv": 0.001659673173858,
        "exploitability": 0.000829836586929
      },
      "current": {
        "on_policy_value_p0": 0.082355550342498,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.164711100684996,
        "unilateral_improvement_p0": 0.084311116324169,
        "unilateral_improvement_p1": 0.247066651027493,
        "nash_conv": 0.331377767351662,
        "exploitability": 0.165688883675831
      }
    },
    {
      "iteration": 69400,
      "average": {
        "on_policy_value_p0": -0.055555817605659,
        "best_response_value_p0": -0.054764100149239,
        "best_response_value_p1": 0.056463905200796,
        "unilateral_improvement_p0": 0.00079171745642,
        "unilateral_improvement_p1": 0.000908087595137,
        "nash_conv": 0.001699805051557,
        "exploitability": 0.000849902525778
      },
      "current": {
        "on_policy_value_p0": 0.079312103635262,
        "best_response_value_p0": 0.113438622471806,
        "best_response_value_p1": 0.265080295660247,
        "unilateral_improvement_p0": 0.034126518836543,
        "unilateral_improvement_p1": 0.34439239929551,
        "nash_conv": 0.378518918132053,
        "exploitability": 0.189259459066026
      }
    },
    {
      "iteration": 69500,
      "average": {
        "on_policy_value_p0": -0.055555574799264,
        "best_response_value_p0": -0.054547457582009,
        "best_response_value_p1": 0.056186091874654,
        "unilateral_improvement_p0": 0.001008117217255,
        "unilateral_improvement_p1": 0.00063051707539,
        "nash_conv": 0.001638634292645,
        "exploitability": 0.000819317146322
      },
      "current": {
        "on_policy_value_p0": 0.07886504964256,
        "best_response_value_p0": 0.07886504964256,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.412198382975893,
        "nash_conv": 0.412198382975893,
        "exploitability": 0.206099191487947
      }
    },
    {
      "iteration": 69600,
      "average": {
        "on_policy_value_p0": -0.055554774068124,
        "best_response_value_p0": -0.054380695666082,
        "best_response_value_p1": 0.056105364731156,
        "unilateral_improvement_p0": 0.001174078402042,
        "unilateral_improvement_p1": 0.000550590663032,
        "nash_conv": 0.001724669065074,
        "exploitability": 0.000862334532537
      },
      "current": {
        "on_policy_value_p0": 0.041783401741831,
        "best_response_value_p0": 0.083099863183005,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.041316461441174,
        "unilateral_improvement_p1": 0.375116735075164,
        "nash_conv": 0.416433196516338,
        "exploitability": 0.208216598258169
      }
    },
    {
      "iteration": 69700,
      "average": {
        "on_policy_value_p0": -0.055553697236799,
        "best_response_value_p0": -0.054280086772142,
        "best_response_value_p1": 0.056024869229391,
        "unilateral_improvement_p0": 0.001273610464657,
        "unilateral_improvement_p1": 0.000471171992592,
        "nash_conv": 0.001744782457249,
        "exploitability": 0.000872391228624
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 69800,
      "average": {
        "on_policy_value_p0": -0.055552434603072,
        "best_response_value_p0": -0.054202838195524,
        "best_response_value_p1": 0.055944604373762,
        "unilateral_improvement_p0": 0.001349596407549,
        "unilateral_improvement_p1": 0.00039216977069,
        "nash_conv": 0.001741766178239,
        "exploitability": 0.000870883089119
      },
      "current": {
        "on_policy_value_p0": -0.017296728568912,
        "best_response_value_p0": 0.158018302382211,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.175315030951123,
        "unilateral_improvement_p1": 0.316036604764422,
        "nash_conv": 0.491351635715544,
        "exploitability": 0.245675817857772
      }
    },
    {
      "iteration": 69900,
      "average": {
        "on_policy_value_p0": -0.055550987618082,
        "best_response_value_p0": -0.0542870716534,
        "best_response_value_p1": 0.05593557537426,
        "unilateral_improvement_p0": 0.001263915964682,
        "unilateral_improvement_p1": 0.000384587756178,
        "nash_conv": 0.00164850372086,
        "exploitability": 0.00082425186043
      },
      "current": {
        "on_policy_value_p0": -0.186189464210618,
        "best_response_value_p0": 0.186189464210618,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.372378928421235,
        "unilateral_improvement_p1": 0.147143869122716,
        "nash_conv": 0.519522797543951,
        "exploitability": 0.259761398771976
      }
    },
    {
      "iteration": 70000,
      "average": {
        "on_policy_value_p0": -0.05555014564493,
        "best_response_value_p0": -0.054591583394595,
        "best_response_value_p1": 0.056331857885629,
        "unilateral_improvement_p0": 0.000958562250334,
        "unilateral_improvement_p1": 0.0007817122407,
        "nash_conv": 0.001740274491034,
        "exploitability": 0.000870137245517
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 70100,
      "average": {
        "on_policy_value_p0": -0.055550459010553,
        "best_response_value_p0": -0.054989217845293,
        "best_response_value_p1": 0.056727009776424,
        "unilateral_improvement_p0": 0.000561241165261,
        "unilateral_improvement_p1": 0.001176550765871,
        "nash_conv": 0.001737791931131,
        "exploitability": 0.000868895965566
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 70200,
      "average": {
        "on_policy_value_p0": -0.055551849506487,
        "best_response_value_p0": -0.055278760095166,
        "best_response_value_p1": 0.056967388707382,
        "unilateral_improvement_p0": 0.000273089411321,
        "unilateral_improvement_p1": 0.001415539200895,
        "nash_conv": 0.001688628612216,
        "exploitability": 0.000844314306108
      },
      "current": {
        "on_policy_value_p0": -0.114693375880369,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.114693375880369,
        "unilateral_improvement_p0": 0.448026709213702,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.448026709213702,
        "exploitability": 0.224013354606851
      }
    },
    {
      "iteration": 70300,
      "average": {
        "on_policy_value_p0": -0.055553424928953,
        "best_response_value_p0": -0.055158905698505,
        "best_response_value_p1": 0.056853006436906,
        "unilateral_improvement_p0": 0.000394519230449,
        "unilateral_improvement_p1": 0.001299581507952,
        "nash_conv": 0.001694100738401,
        "exploitability": 0.0008470503692
      },
      "current": {
        "on_policy_value_p0": 0.093604280657145,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.239729052676188,
        "unilateral_improvement_p1": 0.260270947323812,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 70400,
      "average": {
        "on_policy_value_p0": -0.055554351182009,
        "best_response_value_p0": -0.05484381255594,
        "best_response_value_p1": 0.056573054057855,
        "unilateral_improvement_p0": 0.000710538626069,
        "unilateral_improvement_p1": 0.001018702875846,
        "nash_conv": 0.001729241501915,
        "exploitability": 0.000864620750958
      },
      "current": {
        "on_policy_value_p0": 0.199563252432039,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.232459838197411,
        "unilateral_improvement_p0": 0.133770080901294,
        "unilateral_improvement_p1": 0.43202309062945,
        "nash_conv": 0.565793171530744,
        "exploitability": 0.282896585765372
      }
    },
    {
      "iteration": 70500,
      "average": {
        "on_policy_value_p0": -0.055554316252972,
        "best_response_value_p0": -0.054602330901543,
        "best_response_value_p1": 0.056171590668335,
        "unilateral_improvement_p0": 0.000951985351428,
        "unilateral_improvement_p1": 0.000617274415363,
        "nash_conv": 0.001569259766791,
        "exploitability": 0.000784629883396
      },
      "current": {
        "on_policy_value_p0": 0.115426443537013,
        "best_response_value_p0": 0.198848980963647,
        "best_response_value_p1": 0.333154925146732,
        "unilateral_improvement_p0": 0.083422537426634,
        "unilateral_improvement_p1": 0.448581368683745,
        "nash_conv": 0.532003906110379,
        "exploitability": 0.266001953055189
      }
    },
    {
      "iteration": 70600,
      "average": {
        "on_policy_value_p0": -0.0555534931806,
        "best_response_value_p0": -0.054341754647038,
        "best_response_value_p1": 0.056062584833673,
        "unilateral_improvement_p0": 0.001211738533562,
        "unilateral_improvement_p1": 0.000509091653073,
        "nash_conv": 0.001720830186635,
        "exploitability": 0.000860415093318
      },
      "current": {
        "on_policy_value_p0": 0.012948766928495,
        "best_response_value_p0": 0.143501340793418,
        "best_response_value_p1": 0.433236230363993,
        "unilateral_improvement_p0": 0.130552573864922,
        "unilateral_improvement_p1": 0.446184997292488,
        "nash_conv": 0.576737571157411,
        "exploitability": 0.288368785578705
      }
    },
    {
      "iteration": 70700,
      "average": {
        "on_policy_value_p0": -0.055552403122721,
        "best_response_value_p0": -0.054259035724467,
        "best_response_value_p1": 0.055993244756348,
        "unilateral_improvement_p0": 0.001293367398253,
        "unilateral_improvement_p1": 0.000440841633627,
        "nash_conv": 0.00173420903188,
        "exploitability": 0.00086710451594
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 70800,
      "average": {
        "on_policy_value_p0": -0.055551212221995,
        "best_response_value_p0": -0.054444510422959,
        "best_response_value_p1": 0.056147561748304,
        "unilateral_improvement_p0": 0.001106701799035,
        "unilateral_improvement_p1": 0.000596349526309,
        "nash_conv": 0.001703051325344,
        "exploitability": 0.000851525662672
      },
      "current": {
        "on_policy_value_p0": -0.354650946455357,
        "best_response_value_p0": 0.118216982151786,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.472867928607142,
        "unilateral_improvement_p1": 0.145349053544643,
        "nash_conv": 0.618216982151786,
        "exploitability": 0.309108491075893
      }
    },
    {
      "iteration": 70900,
      "average": {
        "on_policy_value_p0": -0.055551316802874,
        "best_response_value_p0": -0.055039868128571,
        "best_response_value_p1": 0.056682968533825,
        "unilateral_improvement_p0": 0.000511448674303,
        "unilateral_improvement_p1": 0.001131651730951,
        "nash_conv": 0.001643100405255,
        "exploitability": 0.000821550202627
      },
      "current": {
        "on_policy_value_p0": -0.335482217595153,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.335482217595153,
        "unilateral_improvement_p0": 0.50214888426182,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.50214888426182,
        "exploitability": 0.25107444213091
      }
    },
    {
      "iteration": 71000,
      "average": {
        "on_policy_value_p0": -0.055552562151323,
        "best_response_value_p0": -0.055201820385481,
        "best_response_value_p1": 0.056883987029264,
        "unilateral_improvement_p0": 0.000350741765842,
        "unilateral_improvement_p1": 0.001331424877941,
        "nash_conv": 0.001682166643783,
        "exploitability": 0.000841083321891
      },
      "current": {
        "on_policy_value_p0": -0.088757616456088,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.255424283122755,
        "unilateral_improvement_p1": 0.077909050210579,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 71100,
      "average": {
        "on_policy_value_p0": -0.055553960411872,
        "best_response_value_p0": -0.055147591732733,
        "best_response_value_p1": 0.056876304379323,
        "unilateral_improvement_p0": 0.000406368679139,
        "unilateral_improvement_p1": 0.001322343967451,
        "nash_conv": 0.00172871264659,
        "exploitability": 0.000864356323295
      },
      "current": {
        "on_policy_value_p0": 0.015621263480564,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.148915071201597,
        "unilateral_improvement_p0": 0.151045403186103,
        "unilateral_improvement_p1": 0.164536334682161,
        "nash_conv": 0.315581737868264,
        "exploitability": 0.157790868934132
      }
    },
    {
      "iteration": 71200,
      "average": {
        "on_policy_value_p0": -0.055555155160758,
        "best_response_value_p0": -0.055070137249961,
        "best_response_value_p1": 0.056727246219014,
        "unilateral_improvement_p0": 0.000485017910796,
        "unilateral_improvement_p1": 0.001172091058256,
        "nash_conv": 0.001657108969052,
        "exploitability": 0.000828554484526
      },
      "current": {
        "on_policy_value_p0": 0.074665521372888,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.149331042745777,
        "unilateral_improvement_p0": 0.092001145293778,
        "unilateral_improvement_p1": 0.223996564118665,
        "nash_conv": 0.315997709412443,
        "exploitability": 0.157998854706222
      }
    },
    {
      "iteration": 71300,
      "average": {
        "on_policy_value_p0": -0.055555820891499,
        "best_response_value_p0": -0.054828506197717,
        "best_response_value_p1": 0.056512327355756,
        "unilateral_improvement_p0": 0.000727314693781,
        "unilateral_improvement_p1": 0.000956506464257,
        "nash_conv": 0.001683821158038,
        "exploitability": 0.000841910579019
      },
      "current": {
        "on_policy_value_p0": 0.078520408753983,
        "best_response_value_p0": 0.124396774917162,
        "best_response_value_p1": 0.241580601006976,
        "unilateral_improvement_p0": 0.045876366163179,
        "unilateral_improvement_p1": 0.320101009760959,
        "nash_conv": 0.365977375924138,
        "exploitability": 0.182988687962069
      }
    },
    {
      "iteration": 71400,
      "average": {
        "on_policy_value_p0": -0.05555574251011,
        "best_response_value_p0": -0.054604218647351,
        "best_response_value_p1": 0.056221071085341,
        "unilateral_improvement_p0": 0.000951523862759,
        "unilateral_improvement_p1": 0.00066532857523,
        "nash_conv": 0.001616852437989,
        "exploitability": 0.000808426218995
      },
      "current": {
        "on_policy_value_p0": 0.087832166672077,
        "best_response_value_p0": 0.087832166672077,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.421165500005411,
        "nash_conv": 0.421165500005411,
        "exploitability": 0.210582750002705
      }
    },
    {
      "iteration": 71500,
      "average": {
        "on_policy_value_p0": -0.055555068395303,
        "best_response_value_p0": -0.054427858449206,
        "best_response_value_p1": 0.056110032556395,
        "unilateral_improvement_p0": 0.001127209946096,
        "unilateral_improvement_p1": 0.000554964161093,
        "nash_conv": 0.001682174107189,
        "exploitability": 0.000841087053594
      },
      "current": {
        "on_policy_value_p0": 0.053846357410781,
        "best_response_value_p0": 0.058973951845105,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.005127594434324,
        "unilateral_improvement_p1": 0.387179690744114,
        "nash_conv": 0.392307285178438,
        "exploitability": 0.196153642589219
      }
    },
    {
      "iteration": 71600,
      "average": {
        "on_policy_value_p0": -0.055554053093262,
        "best_response_value_p0": -0.054309971907079,
        "best_response_value_p1": 0.056031666589139,
        "unilateral_improvement_p0": 0.001244081186183,
        "unilateral_improvement_p1": 0.000477613495877,
        "nash_conv": 0.00172169468206,
        "exploitability": 0.00086084734103
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 71700,
      "average": {
        "on_policy_value_p0": -0.055552877528488,
        "best_response_value_p0": -0.054234225781685,
        "best_response_value_p1": 0.055953519215933,
        "unilateral_improvement_p0": 0.001318651746802,
        "unilateral_improvement_p1": 0.000400641687445,
        "nash_conv": 0.001719293434247,
        "exploitability": 0.000859646717124
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 71800,
      "average": {
        "on_policy_value_p0": -0.055551477541098,
        "best_response_value_p0": -0.054257001813937,
        "best_response_value_p1": 0.055875589523432,
        "unilateral_improvement_p0": 0.001294475727161,
        "unilateral_improvement_p1": 0.000324111982334,
        "nash_conv": 0.001618587709495,
        "exploitability": 0.000809293854747
      },
      "current": {
        "on_policy_value_p0": -0.15211270974234,
        "best_response_value_p0": 0.15211270974234,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.30422541948468,
        "unilateral_improvement_p1": 0.181220623590994,
        "nash_conv": 0.485446043075673,
        "exploitability": 0.242723021537837
      }
    },
    {
      "iteration": 71900,
      "average": {
        "on_policy_value_p0": -0.055550405626265,
        "best_response_value_p0": -0.054492139707303,
        "best_response_value_p1": 0.05620919683282,
        "unilateral_improvement_p0": 0.001058265918963,
        "unilateral_improvement_p1": 0.000658791206555,
        "nash_conv": 0.001717057125518,
        "exploitability": 0.000858528562759
      },
      "current": {
        "on_policy_value_p0": -0.322054749557654,
        "best_response_value_p0": 0.322054749557654,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.644109499115308,
        "unilateral_improvement_p1": 0.011278583775679,
        "nash_conv": 0.655388082890987,
        "exploitability": 0.327694041445494
      }
    },
    {
      "iteration": 72000,
      "average": {
        "on_policy_value_p0": -0.05555038651223,
        "best_response_value_p0": -0.054879073745217,
        "best_response_value_p1": 0.056594091466848,
        "unilateral_improvement_p0": 0.000671312767013,
        "unilateral_improvement_p1": 0.001043704954618,
        "nash_conv": 0.001715017721631,
        "exploitability": 0.000857508860815
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 72100,
      "average": {
        "on_policy_value_p0": -0.055551444771437,
        "best_response_value_p0": -0.055265279375714,
        "best_response_value_p1": 0.056902754877343,
        "unilateral_improvement_p0": 0.000286165395723,
        "unilateral_improvement_p1": 0.001351310105906,
        "nash_conv": 0.001637475501629,
        "exploitability": 0.000818737750815
      },
      "current": {
        "on_policy_value_p0": -0.179966096481036,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.179966096481036,
        "unilateral_improvement_p0": 0.51329942981437,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.51329942981437,
        "exploitability": 0.256649714907185
      }
    },
    {
      "iteration": 72200,
      "average": {
        "on_policy_value_p0": -0.055553016495564,
        "best_response_value_p0": -0.055224695949176,
        "best_response_value_p1": 0.056894865957593,
        "unilateral_improvement_p0": 0.000328320546388,
        "unilateral_improvement_p1": 0.001341849462029,
        "nash_conv": 0.001670170008417,
        "exploitability": 0.000835085004208
      },
      "current": {
        "on_policy_value_p0": 0.071857217249975,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.261476116083359,
        "unilateral_improvement_p1": 0.238523883916641,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 72300,
      "average": {
        "on_policy_value_p0": -0.055554140997076,
        "best_response_value_p0": -0.054955692848224,
        "best_response_value_p1": 0.056669838189633,
        "unilateral_improvement_p0": 0.000598448148853,
        "unilateral_improvement_p1": 0.001115697192557,
        "nash_conv": 0.001714145341409,
        "exploitability": 0.000857072670705
      },
      "current": {
        "on_policy_value_p0": 0.149518560597498,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.183814772735835,
        "unilateral_improvement_p1": 0.316185227264165,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 72400,
      "average": {
        "on_policy_value_p0": -0.055554448015378,
        "best_response_value_p0": -0.054668116304705,
        "best_response_value_p1": 0.056312083509728,
        "unilateral_improvement_p0": 0.000886331710673,
        "unilateral_improvement_p1": 0.00075763549435,
        "nash_conv": 0.001643967205023,
        "exploitability": 0.000821983602511
      },
      "current": {
        "on_policy_value_p0": 0.153099284600262,
        "best_response_value_p0": 0.2542541716304,
        "best_response_value_p1": 0.297690225939725,
        "unilateral_improvement_p0": 0.101154887030138,
        "unilateral_improvement_p1": 0.450789510539987,
        "nash_conv": 0.551944397570125,
        "exploitability": 0.275972198785062
      }
    },
    {
      "iteration": 72500,
      "average": {
        "on_policy_value_p0": -0.055553864463791,
        "best_response_value_p0": -0.054437005377509,
        "best_response_value_p1": 0.056065522298214,
        "unilateral_improvement_p0": 0.001116859086282,
        "unilateral_improvement_p1": 0.000511657834423,
        "nash_conv": 0.001628516920705,
        "exploitability": 0.000814258460353
      },
      "current": {
        "on_policy_value_p0": 0.056869817696288,
        "best_response_value_p0": 0.114019529641925,
        "best_response_value_p1": 0.385700576108726,
        "unilateral_improvement_p0": 0.057149711945637,
        "unilateral_improvement_p1": 0.442570393805014,
        "nash_conv": 0.499720105750651,
        "exploitability": 0.249860052875326
      }
    },
    {
      "iteration": 72600,
      "average": {
        "on_policy_value_p0": -0.055552912992935,
        "best_response_value_p0": -0.0543132451408,
        "best_response_value_p1": 0.056024971322599,
        "unilateral_improvement_p0": 0.001239667852135,
        "unilateral_improvement_p1": 0.000472058329664,
        "nash_conv": 0.001711726181799,
        "exploitability": 0.0008558630909
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 72700,
      "average": {
        "on_policy_value_p0": -0.055551741316459,
        "best_response_value_p0": -0.054327776283947,
        "best_response_value_p1": 0.055947908088317,
        "unilateral_improvement_p0": 0.001223965032511,
        "unilateral_improvement_p1": 0.000396166771858,
        "nash_conv": 0.00162013180437,
        "exploitability": 0.000810065902185
      },
      "current": {
        "on_policy_value_p0": -0.221179363492709,
        "best_response_value_p0": 0.07372645449757,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.294905817990279,
        "unilateral_improvement_p1": 0.278820636507291,
        "nash_conv": 0.57372645449757,
        "exploitability": 0.286863227248785
      }
    },
    {
      "iteration": 72800,
      "average": {
        "on_policy_value_p0": -0.055551160073915,
        "best_response_value_p0": -0.054778507992367,
        "best_response_value_p1": 0.056475353674106,
        "unilateral_improvement_p0": 0.000772652081549,
        "unilateral_improvement_p1": 0.00092419360019,
        "nash_conv": 0.001696845681739,
        "exploitability": 0.000848422840869
      },
      "current": {
        "on_policy_value_p0": -0.419415772919689,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.419415772919689,
        "unilateral_improvement_p0": 0.586082439586356,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.586082439586356,
        "exploitability": 0.293041219793178
      }
    },
    {
      "iteration": 72900,
      "average": {
        "on_policy_value_p0": -0.055552033557053,
        "best_response_value_p0": -0.05513782810158,
        "best_response_value_p1": 0.05683010420162,
        "unilateral_improvement_p0": 0.000414205455473,
        "unilateral_improvement_p1": 0.001278070644567,
        "nash_conv": 0.00169227610004,
        "exploitability": 0.00084613805002
      },
      "current": {
        "on_policy_value_p0": -0.176885163313231,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.176885163313231,
        "unilateral_improvement_p0": 0.343551829979898,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.343551829979898,
        "exploitability": 0.171775914989949
      }
    },
    {
      "iteration": 73000,
      "average": {
        "on_policy_value_p0": -0.0555533710563,
        "best_response_value_p0": -0.055188639139208,
        "best_response_value_p1": 0.056877346399172,
        "unilateral_improvement_p0": 0.000364731917092,
        "unilateral_improvement_p1": 0.001323975342872,
        "nash_conv": 0.001688707259964,
        "exploitability": 0.000844353629982
      },
      "current": {
        "on_policy_value_p0": -0.052397411475133,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.2190640781418,
        "unilateral_improvement_p1": 0.114269255191534,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 73100,
      "average": {
        "on_policy_value_p0": -0.055554643727109,
        "best_response_value_p0": -0.055113141684845,
        "best_response_value_p1": 0.056790619870733,
        "unilateral_improvement_p0": 0.000441502042264,
        "unilateral_improvement_p1": 0.001235976143624,
        "nash_conv": 0.001677478185888,
        "exploitability": 0.000838739092944
      },
      "current": {
        "on_policy_value_p0": 0.052614572233836,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.114052094432831,
        "unilateral_improvement_p0": 0.114052094432831,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.280718761099498,
        "exploitability": 0.140359380549749
      }
    },
    {
      "iteration": 73200,
      "average": {
        "on_policy_value_p0": -0.055555573477314,
        "best_response_value_p0": -0.054967968888129,
        "best_response_value_p1": 0.056610395367405,
        "unilateral_improvement_p0": 0.000587604589184,
        "unilateral_improvement_p1": 0.001054821890091,
        "nash_conv": 0.001642426479276,
        "exploitability": 0.000821213239638
      },
      "current": {
        "on_policy_value_p0": 0.079824424619767,
        "best_response_value_p0": 0.149534120102749,
        "best_response_value_p1": 0.19391394236737,
        "unilateral_improvement_p0": 0.069709695482982,
        "unilateral_improvement_p1": 0.273738366987137,
        "nash_conv": 0.343448062470119,
        "exploitability": 0.171724031235059
      }
    },
    {
      "iteration": 73300,
      "average": {
        "on_policy_value_p0": -0.055555854754895,
        "best_response_value_p0": -0.054723196629691,
        "best_response_value_p1": 0.05636548945698,
        "unilateral_improvement_p0": 0.000832658125203,
        "unilateral_improvement_p1": 0.000809634702085,
        "nash_conv": 0.001642292827289,
        "exploitability": 0.000821146413644
      },
      "current": {
        "on_policy_value_p0": 0.095098867890195,
        "best_response_value_p0": 0.103826607691993,
        "best_response_value_p1": 0.315877853729738,
        "unilateral_improvement_p0": 0.008727739801798,
        "unilateral_improvement_p1": 0.410976721619932,
        "nash_conv": 0.41970446142173,
        "exploitability": 0.209852230710865
      }
    },
    {
      "iteration": 73400,
      "average": {
        "on_policy_value_p0": -0.055555467890968,
        "best_response_value_p0": -0.05452938682694,
        "best_response_value_p1": 0.056147530745766,
        "unilateral_improvement_p0": 0.001026081064028,
        "unilateral_improvement_p1": 0.000592062854798,
        "nash_conv": 0.001618143918826,
        "exploitability": 0.000809071959413
      },
      "current": {
        "on_policy_value_p0": 0.071230223322128,
        "best_response_value_p0": 0.071230223322128,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.404563556655461,
        "nash_conv": 0.404563556655461,
        "exploitability": 0.20228177832773
      }
    },
    {
      "iteration": 73500,
      "average": {
        "on_policy_value_p0": -0.055554639826104,
        "best_response_value_p0": -0.054382798029796,
        "best_response_value_p1": 0.056071139547473,
        "unilateral_improvement_p0": 0.001171841796309,
        "unilateral_improvement_p1": 0.000516499721369,
        "nash_conv": 0.001688341517677,
        "exploitability": 0.000844170758839
      },
      "current": {
        "on_policy_value_p0": 0.031842228006208,
        "best_response_value_p0": 0.10298221065425,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.071139982648042,
        "unilateral_improvement_p1": 0.365175561339542,
        "nash_conv": 0.436315543987583,
        "exploitability": 0.218157771993792
      }
    },
    {
      "iteration": 73600,
      "average": {
        "on_policy_value_p0": -0.055553611070918,
        "best_response_value_p0": -0.05429744733014,
        "best_response_value_p1": 0.055994955933958,
        "unilateral_improvement_p0": 0.001256163740779,
        "unilateral_improvement_p1": 0.00044134486304,
        "nash_conv": 0.001697508603818,
        "exploitability": 0.000848754301909
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 73700,
      "average": {
        "on_policy_value_p0": -0.055552395187533,
        "best_response_value_p0": -0.054233825725065,
        "best_response_value_p1": 0.055918979060236,
        "unilateral_improvement_p0": 0.001318569462468,
        "unilateral_improvement_p1": 0.000366583872703,
        "nash_conv": 0.001685153335171,
        "exploitability": 0.000842576667586
      },
      "current": {
        "on_policy_value_p0": -0.061931472513007,
        "best_response_value_p0": 0.135700930410163,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.19763240292317,
        "unilateral_improvement_p1": 0.271401860820327,
        "nash_conv": 0.469034263743497,
        "exploitability": 0.234517131871748
      }
    },
    {
      "iteration": 73800,
      "average": {
        "on_policy_value_p0": -0.055551081725159,
        "best_response_value_p0": -0.054347773533362,
        "best_response_value_p1": 0.055979683072242,
        "unilateral_improvement_p0": 0.001203308191797,
        "unilateral_improvement_p1": 0.000428601347083,
        "nash_conv": 0.00163190953888,
        "exploitability": 0.00081595476944
      },
      "current": {
        "on_policy_value_p0": -0.205993677543751,
        "best_response_value_p0": 0.205993677543751,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.411987355087502,
        "unilateral_improvement_p1": 0.127339655789582,
        "nash_conv": 0.539327010877084,
        "exploitability": 0.269663505438542
      }
    },
    {
      "iteration": 73900,
      "average": {
        "on_policy_value_p0": -0.055550431184415,
        "best_response_value_p0": -0.054661565748333,
        "best_response_value_p1": 0.056354992477195,
        "unilateral_improvement_p0": 0.000888865436083,
        "unilateral_improvement_p1": 0.00080456129278,
        "nash_conv": 0.001693426728862,
        "exploitability": 0.000846713364431
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 74000,
      "average": {
        "on_policy_value_p0": -0.055550824687228,
        "best_response_value_p0": -0.055038149218042,
        "best_response_value_p1": 0.056729287532405,
        "unilateral_improvement_p0": 0.000512675469186,
        "unilateral_improvement_p1": 0.001178462845177,
        "nash_conv": 0.001691138314363,
        "exploitability": 0.000845569157182
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 74100,
      "average": {
        "on_policy_value_p0": -0.055552166942887,
        "best_response_value_p0": -0.055285645734219,
        "best_response_value_p1": 0.056936371648003,
        "unilateral_improvement_p0": 0.000266521208668,
        "unilateral_improvement_p1": 0.001384204705116,
        "nash_conv": 0.001650725913784,
        "exploitability": 0.000825362956892
      },
      "current": {
        "on_policy_value_p0": -0.095790490633655,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.108236335414534,
        "unilateral_improvement_p0": 0.429123823966989,
        "unilateral_improvement_p1": 0.012445844780879,
        "nash_conv": 0.441569668747868,
        "exploitability": 0.220784834373934
      }
    },
    {
      "iteration": 74200,
      "average": {
        "on_policy_value_p0": -0.055553612822872,
        "best_response_value_p0": -0.055161950808198,
        "best_response_value_p1": 0.056814596209358,
        "unilateral_improvement_p0": 0.000391662014674,
        "unilateral_improvement_p1": 0.001260983386487,
        "nash_conv": 0.00165264540116,
        "exploitability": 0.00082632270058
      },
      "current": {
        "on_policy_value_p0": 0.096225408861681,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.237107924471653,
        "unilateral_improvement_p1": 0.262892075528347,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 74300,
      "average": {
        "on_policy_value_p0": -0.055554461227953,
        "best_response_value_p0": -0.054863392776604,
        "best_response_value_p1": 0.056544369546117,
        "unilateral_improvement_p0": 0.000691068451349,
        "unilateral_improvement_p1": 0.000989908318164,
        "nash_conv": 0.001680976769513,
        "exploitability": 0.000840488384757
      },
      "current": {
        "on_policy_value_p0": 0.202434496948038,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.238202327229409,
        "unilateral_improvement_p0": 0.130898836385296,
        "unilateral_improvement_p1": 0.440636824177446,
        "nash_conv": 0.571535660562742,
        "exploitability": 0.285767830281371
      }
    },
    {
      "iteration": 74400,
      "average": {
        "on_policy_value_p0": -0.055554435043143,
        "best_response_value_p0": -0.054628207618399,
        "best_response_value_p1": 0.056161940280852,
        "unilateral_improvement_p0": 0.000926227424744,
        "unilateral_improvement_p1": 0.000607505237709,
        "nash_conv": 0.001533732662453,
        "exploitability": 0.000766866331227
      },
      "current": {
        "on_policy_value_p0": 0.120798408108623,
        "best_response_value_p0": 0.203531369437566,
        "best_response_value_p1": 0.334534077342113,
        "unilateral_improvement_p0": 0.082732961328944,
        "unilateral_improvement_p1": 0.455332485450735,
        "nash_conv": 0.538065446779679,
        "exploitability": 0.269032723389839
      }
    },
    {
      "iteration": 74500,
      "average": {
        "on_policy_value_p0": -0.055553668856695,
        "best_response_value_p0": -0.054369994482465,
        "best_response_value_p1": 0.056039942571289,
        "unilateral_improvement_p0": 0.001183674374231,
        "unilateral_improvement_p1": 0.000486273714594,
        "nash_conv": 0.001669948088825,
        "exploitability": 0.000834974044412
      },
      "current": {
        "on_policy_value_p0": 0.026019694653092,
        "best_response_value_p0": 0.136611645819331,
        "best_response_value_p1": 0.431819305916842,
        "unilateral_improvement_p0": 0.110591951166239,
        "unilateral_improvement_p1": 0.457839000569934,
        "nash_conv": 0.568430951736173,
        "exploitability": 0.284215475868087
      }
    },
    {
      "iteration": 74600,
      "average": {
        "on_policy_value_p0": -0.055552639330212,
        "best_response_value_p0": -0.054286879744917,
        "best_response_value_p1": 0.055975000024322,
        "unilateral_improvement_p0": 0.001265759585295,
        "unilateral_improvement_p1": 0.00042236069411,
        "nash_conv": 0.001688120279405,
        "exploitability": 0.000844060139703
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 74700,
      "average": {
        "on_policy_value_p0": -0.055551498682291,
        "best_response_value_p0": -0.054443146011651,
        "best_response_value_p1": 0.056088858738896,
        "unilateral_improvement_p0": 0.001108352670641,
        "unilateral_improvement_p1": 0.000537360056604,
        "nash_conv": 0.001645712727245,
        "exploitability": 0.000822856363622
      },
      "current": {
        "on_policy_value_p0": -0.33468944533318,
        "best_response_value_p0": 0.111563148444393,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.446252593777574,
        "unilateral_improvement_p1": 0.16531055466682,
        "nash_conv": 0.611563148444394,
        "exploitability": 0.305781574222197
      }
    },
    {
      "iteration": 74800,
      "average": {
        "on_policy_value_p0": -0.055551463514587,
        "best_response_value_p0": -0.054994546470073,
        "best_response_value_p1": 0.056610963222955,
        "unilateral_improvement_p0": 0.000556917044514,
        "unilateral_improvement_p1": 0.001059499708367,
        "nash_conv": 0.001616416752881,
        "exploitability": 0.000808208376441
      },
      "current": {
        "on_policy_value_p0": -0.348949933032107,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.348949933032107,
        "unilateral_improvement_p0": 0.515616599698774,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.515616599698774,
        "exploitability": 0.257808299849387
      }
    },
    {
      "iteration": 74900,
      "average": {
        "on_policy_value_p0": -0.055552538236787,
        "best_response_value_p0": -0.055184116184932,
        "best_response_value_p1": 0.056834723947953,
        "unilateral_improvement_p0": 0.000368422051854,
        "unilateral_improvement_p1": 0.001282185711167,
        "nash_conv": 0.001650607763021,
        "exploitability": 0.00082530388151
      },
      "current": {
        "on_policy_value_p0": -0.097185402881109,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.263852069547776,
        "unilateral_improvement_p1": 0.069481263785558,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 75000,
      "average": {
        "on_policy_value_p0": -0.055553813443054,
        "best_response_value_p0": -0.05516274901401,
        "best_response_value_p1": 0.056844844586326,
        "unilateral_improvement_p0": 0.000391064429044,
        "unilateral_improvement_p1": 0.001291031143272,
        "nash_conv": 0.001682095572316,
        "exploitability": 0.000841047786158
      },
      "current": {
        "on_policy_value_p0": -0.024040721153243,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.19070738781991,
        "unilateral_improvement_p1": 0.142625945513423,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 75100,
      "average": {
        "on_policy_value_p0": -0.055554947159466,
        "best_response_value_p0": -0.05508929661852,
        "best_response_value_p1": 0.05672474325512,
        "unilateral_improvement_p0": 0.000465650540946,
        "unilateral_improvement_p1": 0.001169796095655,
        "nash_conv": 0.001635446636601,
        "exploitability": 0.0008177233183
      },
      "current": {
        "on_policy_value_p0": 0.064302515071957,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.128605030143913,
        "unilateral_improvement_p0": 0.10236415159471,
        "unilateral_improvement_p1": 0.19290754521587,
        "nash_conv": 0.29527169681058,
        "exploitability": 0.14763584840529
      }
    },
    {
      "iteration": 75200,
      "average": {
        "on_policy_value_p0": -0.05555567871624,
        "best_response_value_p0": -0.054904712561526,
        "best_response_value_p1": 0.05653562581779,
        "unilateral_improvement_p0": 0.000650966154714,
        "unilateral_improvement_p1": 0.00097994710155,
        "nash_conv": 0.001630913256264,
        "exploitability": 0.000815456628132
      },
      "current": {
        "on_policy_value_p0": 0.074552498839923,
        "best_response_value_p0": 0.133809640608903,
        "best_response_value_p1": 0.214819049795373,
        "unilateral_improvement_p0": 0.05925714176898,
        "unilateral_improvement_p1": 0.289371548635295,
        "nash_conv": 0.348628690404276,
        "exploitability": 0.174314345202138
      }
    },
    {
      "iteration": 75300,
      "average": {
        "on_policy_value_p0": -0.05555576952656,
        "best_response_value_p0": -0.054681404560077,
        "best_response_value_p1": 0.056278624510506,
        "unilateral_improvement_p0": 0.000874364966483,
        "unilateral_improvement_p1": 0.000722854983946,
        "nash_conv": 0.001597219950429,
        "exploitability": 0.000798609975215
      },
      "current": {
        "on_policy_value_p0": 0.095132585350175,
        "best_response_value_p0": 0.095132585350175,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.428465918683508,
        "nash_conv": 0.428465918683508,
        "exploitability": 0.214232959341754
      }
    },
    {
      "iteration": 75400,
      "average": {
        "on_policy_value_p0": -0.055555274488979,
        "best_response_value_p0": -0.054503613481822,
        "best_response_value_p1": 0.056121888711242,
        "unilateral_improvement_p0": 0.001051661007157,
        "unilateral_improvement_p1": 0.000566614222263,
        "nash_conv": 0.00161827522942,
        "exploitability": 0.00080913761471
      },
      "current": {
        "on_policy_value_p0": 0.062994522555126,
        "best_response_value_p0": 0.062994522555126,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.396327855888459,
        "nash_conv": 0.396327855888459,
        "exploitability": 0.198163927944229
      }
    },
    {
      "iteration": 75500,
      "average": {
        "on_policy_value_p0": -0.055554411771401,
        "best_response_value_p0": -0.054374095286664,
        "best_response_value_p1": 0.056047555083811,
        "unilateral_improvement_p0": 0.001180316484736,
        "unilateral_improvement_p1": 0.00049314331241,
        "nash_conv": 0.001673459797146,
        "exploitability": 0.000836729898573
      },
      "current": {
        "on_policy_value_p0": 0.018549769031454,
        "best_response_value_p0": 0.129567128603759,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.111017359572305,
        "unilateral_improvement_p1": 0.351883102364787,
        "nash_conv": 0.462900461937092,
        "exploitability": 0.231450230968546
      }
    },
    {
      "iteration": 75600,
      "average": {
        "on_policy_value_p0": -0.055553395211561,
        "best_response_value_p0": -0.054298835367387,
        "best_response_value_p1": 0.055973418106187,
        "unilateral_improvement_p0": 0.001254559844173,
        "unilateral_improvement_p1": 0.000420022894627,
        "nash_conv": 0.0016745827388,
        "exploitability": 0.0008372913694
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 75700,
      "average": {
        "on_policy_value_p0": -0.055552176375041,
        "best_response_value_p0": -0.05425852907555,
        "best_response_value_p1": 0.055899476999046,
        "unilateral_improvement_p0": 0.001293647299491,
        "unilateral_improvement_p1": 0.000347300624005,
        "nash_conv": 0.001640947923496,
        "exploitability": 0.000820473961748
      },
      "current": {
        "on_policy_value_p0": -0.098420050824481,
        "best_response_value_p0": 0.117456641254426,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.215876692078907,
        "unilateral_improvement_p1": 0.234913282508852,
        "nash_conv": 0.450789974587759,
        "exploitability": 0.22539498729388
      }
    },
    {
      "iteration": 75800,
      "average": {
        "on_policy_value_p0": -0.055550975637893,
        "best_response_value_p0": -0.054405168120227,
        "best_response_value_p1": 0.05604382877423,
        "unilateral_improvement_p0": 0.001145807517666,
        "unilateral_improvement_p1": 0.000492853136337,
        "nash_conv": 0.001638660654002,
        "exploitability": 0.000819330327001
      },
      "current": {
        "on_policy_value_p0": -0.233409220590224,
        "best_response_value_p0": 0.233409220590224,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.466818441180448,
        "unilateral_improvement_p1": 0.099924112743109,
        "nash_conv": 0.566742553923557,
        "exploitability": 0.283371276961779
      }
    },
    {
      "iteration": 75900,
      "average": {
        "on_policy_value_p0": -0.055550531284824,
        "best_response_value_p0": -0.054738467153831,
        "best_response_value_p1": 0.056409164089854,
        "unilateral_improvement_p0": 0.000812064130993,
        "unilateral_improvement_p1": 0.00085863280503,
        "nash_conv": 0.001670696936023,
        "exploitability": 0.000835348468011
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 76000,
      "average": {
        "on_policy_value_p0": -0.055551078141527,
        "best_response_value_p0": -0.055105039346172,
        "best_response_value_p1": 0.056766908299737,
        "unilateral_improvement_p0": 0.000446038795354,
        "unilateral_improvement_p1": 0.00121583015821,
        "nash_conv": 0.001661868953564,
        "exploitability": 0.000830934476782
      },
      "current": {
        "on_policy_value_p0": -0.273983453105312,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.273983453105312,
        "unilateral_improvement_p0": 0.607316786438645,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.607316786438645,
        "exploitability": 0.303658393219322
      }
    },
    {
      "iteration": 76100,
      "average": {
        "on_policy_value_p0": -0.055552449963434,
        "best_response_value_p0": -0.055270478371053,
        "best_response_value_p1": 0.056916377156769,
        "unilateral_improvement_p0": 0.000281971592382,
        "unilateral_improvement_p1": 0.001363927193335,
        "nash_conv": 0.001645898785717,
        "exploitability": 0.000822949392858
      },
      "current": {
        "on_policy_value_p0": -0.046138639505604,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.125693488855449,
        "unilateral_improvement_p0": 0.379471972838937,
        "unilateral_improvement_p1": 0.079554849349846,
        "nash_conv": 0.459026822188783,
        "exploitability": 0.229513411094391
      }
    },
    {
      "iteration": 76200,
      "average": {
        "on_policy_value_p0": -0.055553759999589,
        "best_response_value_p0": -0.055122037607845,
        "best_response_value_p1": 0.056763565418158,
        "unilateral_improvement_p0": 0.000431722391745,
        "unilateral_improvement_p1": 0.001209805418568,
        "nash_conv": 0.001641527810313,
        "exploitability": 0.000820763905157
      },
      "current": {
        "on_policy_value_p0": 0.104442039561544,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.228891293771789,
        "unilateral_improvement_p1": 0.27110870622821,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 76300,
      "average": {
        "on_policy_value_p0": -0.055554466133799,
        "best_response_value_p0": -0.05483135778573,
        "best_response_value_p1": 0.056487344069051,
        "unilateral_improvement_p0": 0.000723108348069,
        "unilateral_improvement_p1": 0.000932877935252,
        "nash_conv": 0.001655986283321,
        "exploitability": 0.000827993141661
      },
      "current": {
        "on_policy_value_p0": 0.20568841353513,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.244710160403592,
        "unilateral_improvement_p0": 0.127644919798204,
        "unilateral_improvement_p1": 0.450398573938722,
        "nash_conv": 0.578043493736926,
        "exploitability": 0.289021746868463
      }
    },
    {
      "iteration": 76400,
      "average": {
        "on_policy_value_p0": -0.055554336282258,
        "best_response_value_p0": -0.054618471694631,
        "best_response_value_p1": 0.056112124939764,
        "unilateral_improvement_p0": 0.000935864587627,
        "unilateral_improvement_p1": 0.000557788657506,
        "nash_conv": 0.001493653245133,
        "exploitability": 0.000746826622566
      },
      "current": {
        "on_policy_value_p0": 0.105209502175487,
        "best_response_value_p0": 0.186312221753472,
        "best_response_value_p1": 0.33779456084403,
        "unilateral_improvement_p0": 0.081102719577985,
        "unilateral_improvement_p1": 0.443004063019516,
        "nash_conv": 0.524106782597502,
        "exploitability": 0.262053391298751
      }
    },
    {
      "iteration": 76500,
      "average": {
        "on_policy_value_p0": -0.055553591989042,
        "best_response_value_p0": -0.054391011302864,
        "best_response_value_p1": 0.056047397906009,
        "unilateral_improvement_p0": 0.001162580686178,
        "unilateral_improvement_p1": 0.000493805916967,
        "nash_conv": 0.001656386603144,
        "exploitability": 0.000828193301572
      },
      "current": {
        "on_policy_value_p0": 0.000141915545391,
        "best_response_value_p0": 0.150406791208584,
        "best_response_value_p1": 0.435244329258451,
        "unilateral_improvement_p0": 0.150264875663192,
        "unilateral_improvement_p1": 0.435386244803842,
        "nash_conv": 0.585651120467034,
        "exploitability": 0.292825560233517
      }
    },
    {
      "iteration": 76600,
      "average": {
        "on_policy_value_p0": -0.05555264285587,
        "best_response_value_p0": -0.054317382764406,
        "best_response_value_p1": 0.055983122414716,
        "unilateral_improvement_p0": 0.001235260091464,
        "unilateral_improvement_p1": 0.000430479558846,
        "nash_conv": 0.00166573965031,
        "exploitability": 0.000832869825155
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 76700,
      "average": {
        "on_policy_value_p0": -0.055551582188221,
        "best_response_value_p0": -0.054470107584132,
        "best_response_value_p1": 0.056092766086085,
        "unilateral_improvement_p0": 0.001081474604089,
        "unilateral_improvement_p1": 0.000541183897864,
        "nash_conv": 0.001622658501953,
        "exploitability": 0.000811329250977
      },
      "current": {
        "on_policy_value_p0": -0.332829241938864,
        "best_response_value_p0": 0.110943080646288,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.443772322585151,
        "unilateral_improvement_p1": 0.167170758061136,
        "nash_conv": 0.610943080646288,
        "exploitability": 0.305471540323144
      }
    },
    {
      "iteration": 76800,
      "average": {
        "on_policy_value_p0": -0.055551568076996,
        "best_response_value_p0": -0.055005316380918,
        "best_response_value_p1": 0.056604474019236,
        "unilateral_improvement_p0": 0.000546251696078,
        "unilateral_improvement_p1": 0.001052905942241,
        "nash_conv": 0.001599157638318,
        "exploitability": 0.000799578819159
      },
      "current": {
        "on_policy_value_p0": -0.352525588182051,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.352525588182051,
        "unilateral_improvement_p0": 0.519192254848718,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.519192254848718,
        "exploitability": 0.259596127424359
      }
    },
    {
      "iteration": 76900,
      "average": {
        "on_policy_value_p0": -0.055552627243398,
        "best_response_value_p0": -0.055199757062909,
        "best_response_value_p1": 0.056828391789316,
        "unilateral_improvement_p0": 0.000352870180489,
        "unilateral_improvement_p1": 0.001275764545918,
        "nash_conv": 0.001628634726407,
        "exploitability": 0.000814317363203
      },
      "current": {
        "on_policy_value_p0": -0.097083729325673,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.26375039599234,
        "unilateral_improvement_p1": 0.069582937340994,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 77000,
      "average": {
        "on_policy_value_p0": -0.055553880521623,
        "best_response_value_p0": -0.055179205681565,
        "best_response_value_p1": 0.056838738177181,
        "unilateral_improvement_p0": 0.000374674840058,
        "unilateral_improvement_p1": 0.001284857655558,
        "nash_conv": 0.001659532495616,
        "exploitability": 0.000829766247808
      },
      "current": {
        "on_policy_value_p0": -0.022404209417994,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.16374682040596,
        "unilateral_improvement_p0": 0.189070876084661,
        "unilateral_improvement_p1": 0.141342610987966,
        "nash_conv": 0.330413487072627,
        "exploitability": 0.165206743536313
      }
    },
    {
      "iteration": 77100,
      "average": {
        "on_policy_value_p0": -0.055555000551873,
        "best_response_value_p0": -0.055107637321409,
        "best_response_value_p1": 0.056717595258034,
        "unilateral_improvement_p0": 0.000447363230465,
        "unilateral_improvement_p1": 0.001162594706161,
        "nash_conv": 0.001609957936625,
        "exploitability": 0.000804978968313
      },
      "current": {
        "on_policy_value_p0": 0.066387508432265,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.13277501686453,
        "unilateral_improvement_p0": 0.100279158234401,
        "unilateral_improvement_p1": 0.199162525296796,
        "nash_conv": 0.299441683531197,
        "exploitability": 0.149720841765599
      }
    },
    {
      "iteration": 77200,
      "average": {
        "on_policy_value_p0": -0.055555727199253,
        "best_response_value_p0": -0.054917184207147,
        "best_response_value_p1": 0.056531144295957,
        "unilateral_improvement_p0": 0.000638542992106,
        "unilateral_improvement_p1": 0.000975417096704,
        "nash_conv": 0.00161396008881,
        "exploitability": 0.000806980044405
      },
      "current": {
        "on_policy_value_p0": 0.078436966907454,
        "best_response_value_p0": 0.136103783920075,
        "best_response_value_p1": 0.217999699308091,
        "unilateral_improvement_p0": 0.057666817012621,
        "unilateral_improvement_p1": 0.296436666215545,
        "nash_conv": 0.354103483228166,
        "exploitability": 0.177051741614083
      }
    },
    {
      "iteration": 77300,
      "average": {
        "on_policy_value_p0": -0.05555582809638,
        "best_response_value_p0": -0.054697064203418,
        "best_response_value_p1": 0.05627871326899,
        "unilateral_improvement_p0": 0.000858763892962,
        "unilateral_improvement_p1": 0.00072288517261,
        "nash_conv": 0.001581649065572,
        "exploitability": 0.000790824532786
      },
      "current": {
        "on_policy_value_p0": 0.097016998084868,
        "best_response_value_p0": 0.097016998084868,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.430350331418202,
        "nash_conv": 0.430350331418202,
        "exploitability": 0.215175165709101
      }
    },
    {
      "iteration": 77400,
      "average": {
        "on_policy_value_p0": -0.055555362214676,
        "best_response_value_p0": -0.054521179050548,
        "best_response_value_p1": 0.056112342674935,
        "unilateral_improvement_p0": 0.001034183164128,
        "unilateral_improvement_p1": 0.000556980460259,
        "nash_conv": 0.001591163624387,
        "exploitability": 0.000795581812193
      },
      "current": {
        "on_policy_value_p0": 0.065402927842351,
        "best_response_value_p0": 0.065402927842351,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.398736261175684,
        "nash_conv": 0.398736261175684,
        "exploitability": 0.199368130587842
      }
    },
    {
      "iteration": 77500,
      "average": {
        "on_policy_value_p0": -0.055554533197219,
        "best_response_value_p0": -0.054390628255101,
        "best_response_value_p1": 0.056039939652129,
        "unilateral_improvement_p0": 0.001163904942119,
        "unilateral_improvement_p1": 0.00048540645491,
        "nash_conv": 0.001649311397029,
        "exploitability": 0.000824655698514
      },
      "current": {
        "on_policy_value_p0": 0.023671062040381,
        "best_response_value_p0": 0.119324542585905,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.095653480545524,
        "unilateral_improvement_p1": 0.357004395373714,
        "nash_conv": 0.452657875919238,
        "exploitability": 0.226328937959619
      }
    },
    {
      "iteration": 77600,
      "average": {
        "on_policy_value_p0": -0.055553549049689,
        "best_response_value_p0": -0.054314890771604,
        "best_response_value_p1": 0.055967723235052,
        "unilateral_improvement_p0": 0.001238658278086,
        "unilateral_improvement_p1": 0.000414174185363,
        "nash_conv": 0.001652832463449,
        "exploitability": 0.000826416231724
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 77700,
      "average": {
        "on_policy_value_p0": -0.055552379791272,
        "best_response_value_p0": -0.054268782610402,
        "best_response_value_p1": 0.055895692703219,
        "unilateral_improvement_p0": 0.00128359718087,
        "unilateral_improvement_p1": 0.000343312911947,
        "nash_conv": 0.001626910092817,
        "exploitability": 0.000813455046409
      },
      "current": {
        "on_policy_value_p0": -0.088652222282552,
        "best_response_value_p0": 0.122340555525391,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.210992777807943,
        "unilateral_improvement_p1": 0.244681111050781,
        "nash_conv": 0.455673888858724,
        "exploitability": 0.227836944429362
      }
    },
    {
      "iteration": 77800,
      "average": {
        "on_policy_value_p0": -0.055551191805452,
        "best_response_value_p0": -0.054400146919477,
        "best_response_value_p1": 0.056007538948018,
        "unilateral_improvement_p0": 0.001151044885974,
        "unilateral_improvement_p1": 0.000456347142567,
        "nash_conv": 0.001607392028541,
        "exploitability": 0.00080369601427
      },
      "current": {
        "on_policy_value_p0": -0.22242192099871,
        "best_response_value_p0": 0.22242192099871,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.44484384199742,
        "unilateral_improvement_p1": 0.110911412334623,
        "nash_conv": 0.555755254332044,
        "exploitability": 0.277877627166022
      }
    },
    {
      "iteration": 77900,
      "average": {
        "on_policy_value_p0": -0.055550678573746,
        "best_response_value_p0": -0.054714404924209,
        "best_response_value_p1": 0.056363541251465,
        "unilateral_improvement_p0": 0.000836273649537,
        "unilateral_improvement_p1": 0.000812862677719,
        "nash_conv": 0.001649136327256,
        "exploitability": 0.000824568163628
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 78000,
      "average": {
        "on_policy_value_p0": -0.055551105729954,
        "best_response_value_p0": -0.055071608678579,
        "best_response_value_p1": 0.056717981888428,
        "unilateral_improvement_p0": 0.000479497051375,
        "unilateral_improvement_p1": 0.001166876158474,
        "nash_conv": 0.001646373209849,
        "exploitability": 0.000823186604924
      },
      "current": {
        "on_policy_value_p0": -0.310950356422093,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.310950356422093,
        "unilateral_improvement_p0": 0.644283689755426,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.644283689755426,
        "exploitability": 0.322141844877713
      }
    },
    {
      "iteration": 78100,
      "average": {
        "on_policy_value_p0": -0.055552369038372,
        "best_response_value_p0": -0.055284729488384,
        "best_response_value_p1": 0.056898198391756,
        "unilateral_improvement_p0": 0.000267639549989,
        "unilateral_improvement_p1": 0.001345829353383,
        "nash_conv": 0.001613468903372,
        "exploitability": 0.000806734451686
      },
      "current": {
        "on_policy_value_p0": -0.085320554670686,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.1123219589755,
        "unilateral_improvement_p0": 0.41865388800402,
        "unilateral_improvement_p1": 0.027001404304814,
        "nash_conv": 0.445655292308834,
        "exploitability": 0.222827646154417
      }
    },
    {
      "iteration": 78200,
      "average": {
        "on_policy_value_p0": -0.055553686693693,
        "best_response_value_p0": -0.055167664515165,
        "best_response_value_p1": 0.056776635863414,
        "unilateral_improvement_p0": 0.000386022178527,
        "unilateral_improvement_p1": 0.001222949169721,
        "nash_conv": 0.001608971348248,
        "exploitability": 0.000804485674124
      },
      "current": {
        "on_policy_value_p0": 0.095797111727266,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.237536221606067,
        "unilateral_improvement_p1": 0.262463778393932,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 78300,
      "average": {
        "on_policy_value_p0": -0.055554465622102,
        "best_response_value_p0": -0.054884351193095,
        "best_response_value_p1": 0.056524961163178,
        "unilateral_improvement_p0": 0.000670114429008,
        "unilateral_improvement_p1": 0.000970495541075,
        "nash_conv": 0.001640609970083,
        "exploitability": 0.000820304985041
      },
      "current": {
        "on_policy_value_p0": 0.197266567527635,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.227866468388604,
        "unilateral_improvement_p0": 0.136066765805698,
        "unilateral_improvement_p1": 0.425133035916239,
        "nash_conv": 0.561199801721937,
        "exploitability": 0.280599900860968
      }
    },
    {
      "iteration": 78400,
      "average": {
        "on_policy_value_p0": -0.055554471603081,
        "best_response_value_p0": -0.054657769375908,
        "best_response_value_p1": 0.056167683144839,
        "unilateral_improvement_p0": 0.000896702227172,
        "unilateral_improvement_p1": 0.000613211541759,
        "nash_conv": 0.001509913768931,
        "exploitability": 0.000754956884466
      },
      "current": {
        "on_policy_value_p0": 0.122287244476648,
        "best_response_value_p0": 0.209350548084456,
        "best_response_value_p1": 0.325873392784383,
        "unilateral_improvement_p0": 0.087063303607808,
        "unilateral_improvement_p1": 0.448160637261032,
        "nash_conv": 0.53522394086884,
        "exploitability": 0.26761197043442
      }
    },
    {
      "iteration": 78500,
      "average": {
        "on_policy_value_p0": -0.055553804697486,
        "best_response_value_p0": -0.054423075193271,
        "best_response_value_p1": 0.05603720289418,
        "unilateral_improvement_p0": 0.001130729504215,
        "unilateral_improvement_p1": 0.000483398196694,
        "nash_conv": 0.001614127700909,
        "exploitability": 0.000807063850454
      },
      "current": {
        "on_policy_value_p0": 0.031636448992295,
        "best_response_value_p0": 0.129829169768762,
        "best_response_value_p1": 0.415922910392973,
        "unilateral_improvement_p0": 0.098192720776467,
        "unilateral_improvement_p1": 0.447559359385268,
        "nash_conv": 0.545752080161735,
        "exploitability": 0.272876040080868
      }
    },
    {
      "iteration": 78600,
      "average": {
        "on_policy_value_p0": -0.055552898048483,
        "best_response_value_p0": -0.054337810454044,
        "best_response_value_p1": 0.055982305478294,
        "unilateral_improvement_p0": 0.001215087594439,
        "unilateral_improvement_p1": 0.00042940742981,
        "nash_conv": 0.001644495024249,
        "exploitability": 0.000822247512125
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 78700,
      "average": {
        "on_policy_value_p0": -0.055551833441291,
        "best_response_value_p0": -0.054423220570132,
        "best_response_value_p1": 0.055982843260509,
        "unilateral_improvement_p0": 0.001128612871159,
        "unilateral_improvement_p1": 0.000431009819218,
        "nash_conv": 0.001559622690377,
        "exploitability": 0.000779811345189
      },
      "current": {
        "on_policy_value_p0": -0.28153056084091,
        "best_response_value_p0": 0.093843520280304,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.375374081121214,
        "unilateral_improvement_p1": 0.21846943915909,
        "nash_conv": 0.593843520280304,
        "exploitability": 0.296921760140152
      }
    },
    {
      "iteration": 78800,
      "average": {
        "on_policy_value_p0": -0.055551556021869,
        "best_response_value_p0": -0.054902042414596,
        "best_response_value_p1": 0.056510368461728,
        "unilateral_improvement_p0": 0.000649513607273,
        "unilateral_improvement_p1": 0.000958812439859,
        "nash_conv": 0.001608326047132,
        "exploitability": 0.000804163023566
      },
      "current": {
        "on_policy_value_p0": -0.383202305436089,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.383202305436089,
        "unilateral_improvement_p0": 0.549868972102756,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.549868972102756,
        "exploitability": 0.274934486051378
      }
    },
    {
      "iteration": 78900,
      "average": {
        "on_policy_value_p0": -0.0555524404377,
        "best_response_value_p0": -0.0551670315763,
        "best_response_value_p1": 0.056791659515615,
        "unilateral_improvement_p0": 0.0003854088614,
        "unilateral_improvement_p1": 0.001239219077915,
        "nash_conv": 0.001624627939315,
        "exploitability": 0.000812313969658
      },
      "current": {
        "on_policy_value_p0": -0.122133428623302,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.161293658291966,
        "unilateral_improvement_p0": 0.288800095289969,
        "unilateral_improvement_p1": 0.039160229668664,
        "nash_conv": 0.327960324958632,
        "exploitability": 0.163980162479316
      }
    },
    {
      "iteration": 79000,
      "average": {
        "on_policy_value_p0": -0.055553622331007,
        "best_response_value_p0": -0.055192971383854,
        "best_response_value_p1": 0.056820912954163,
        "unilateral_improvement_p0": 0.000360650947153,
        "unilateral_improvement_p1": 0.001267290623156,
        "nash_conv": 0.001627941570309,
        "exploitability": 0.000813970785155
      },
      "current": {
        "on_policy_value_p0": -0.046238642025182,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.212905308691849,
        "unilateral_improvement_p1": 0.120428024641484,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 79100,
      "average": {
        "on_policy_value_p0": -0.055554730689781,
        "best_response_value_p0": -0.055123195187415,
        "best_response_value_p1": 0.056734760224001,
        "unilateral_improvement_p0": 0.000431535502365,
        "unilateral_improvement_p1": 0.001180029534221,
        "nash_conv": 0.001611565036586,
        "exploitability": 0.000805782518293
      },
      "current": {
        "on_policy_value_p0": 0.053066951592684,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.113599715073983,
        "unilateral_improvement_p0": 0.113599715073983,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.280266381740649,
        "exploitability": 0.140133190870325
      }
    },
    {
      "iteration": 79200,
      "average": {
        "on_policy_value_p0": -0.055555544705979,
        "best_response_value_p0": -0.054993030150499,
        "best_response_value_p1": 0.056568806814233,
        "unilateral_improvement_p0": 0.00056251455548,
        "unilateral_improvement_p1": 0.001013262108254,
        "nash_conv": 0.001575776663734,
        "exploitability": 0.000787888331867
      },
      "current": {
        "on_policy_value_p0": 0.078252678142356,
        "best_response_value_p0": 0.149265845365745,
        "best_response_value_p1": 0.191306998886557,
        "unilateral_improvement_p0": 0.071013167223388,
        "unilateral_improvement_p1": 0.269559677028913,
        "nash_conv": 0.340572844252301,
        "exploitability": 0.170286422126151
      }
    },
    {
      "iteration": 79300,
      "average": {
        "on_policy_value_p0": -0.055555810408172,
        "best_response_value_p0": -0.054765982299621,
        "best_response_value_p1": 0.05634631540646,
        "unilateral_improvement_p0": 0.000789828108551,
        "unilateral_improvement_p1": 0.000790504998289,
        "nash_conv": 0.00158033310684,
        "exploitability": 0.00079016655342
      },
      "current": {
        "on_policy_value_p0": 0.089773124590869,
        "best_response_value_p0": 0.105049322999797,
        "best_response_value_p1": 0.302780936515477,
        "unilateral_improvement_p0": 0.015276198408928,
        "unilateral_improvement_p1": 0.392554061106345,
        "nash_conv": 0.407830259515274,
        "exploitability": 0.203915129757637
      }
    },
    {
      "iteration": 79400,
      "average": {
        "on_policy_value_p0": -0.055555502107335,
        "best_response_value_p0": -0.054584546695573,
        "best_response_value_p1": 0.056133064979365,
        "unilateral_improvement_p0": 0.000970955411762,
        "unilateral_improvement_p1": 0.00057756287203,
        "nash_conv": 0.001548518283792,
        "exploitability": 0.000774259141896
      },
      "current": {
        "on_policy_value_p0": 0.07364190305717,
        "best_response_value_p0": 0.07364190305717,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.406975236390504,
        "nash_conv": 0.406975236390504,
        "exploitability": 0.203487618195252
      }
    },
    {
      "iteration": 79500,
      "average": {
        "on_policy_value_p0": -0.055554801795613,
        "best_response_value_p0": -0.054444390775916,
        "best_response_value_p1": 0.056062457350461,
        "unilateral_improvement_p0": 0.001110411019697,
        "unilateral_improvement_p1": 0.000507655554848,
        "nash_conv": 0.001618066574545,
        "exploitability": 0.000809033287272
      },
      "current": {
        "on_policy_value_p0": 0.037406814437458,
        "best_response_value_p0": 0.091853037791751,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.054446223354293,
        "unilateral_improvement_p1": 0.370740147770791,
        "nash_conv": 0.425186371125085,
        "exploitability": 0.212593185562542
      }
    },
    {
      "iteration": 79600,
      "average": {
        "on_policy_value_p0": -0.055553910871946,
        "best_response_value_p0": -0.054359839746078,
        "best_response_value_p1": 0.055992027127659,
        "unilateral_improvement_p0": 0.001194071125868,
        "unilateral_improvement_p1": 0.000438116255713,
        "nash_conv": 0.001632187381581,
        "exploitability": 0.000816093690791
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 79700,
      "average": {
        "on_policy_value_p0": -0.055552876058765,
        "best_response_value_p0": -0.054291672228862,
        "best_response_value_p1": 0.055921773643183,
        "unilateral_improvement_p0": 0.001261203829903,
        "unilateral_improvement_p1": 0.000368897584418,
        "nash_conv": 0.001630101414321,
        "exploitability": 0.000815050707161
      },
      "current": {
        "on_policy_value_p0": -0.00636456462845,
        "best_response_value_p0": 0.163484384352442,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.169848948980892,
        "unilateral_improvement_p1": 0.326968768704884,
        "nash_conv": 0.496817717685775,
        "exploitability": 0.248408858842888
      }
    },
    {
      "iteration": 79800,
      "average": {
        "on_policy_value_p0": -0.055551680922711,
        "best_response_value_p0": -0.054352283439887,
        "best_response_value_p1": 0.055874699898158,
        "unilateral_improvement_p0": 0.001199397482823,
        "unilateral_improvement_p1": 0.000323018975447,
        "nash_conv": 0.00152241645827,
        "exploitability": 0.000761208229135
      },
      "current": {
        "on_policy_value_p0": -0.173383484279331,
        "best_response_value_p0": 0.173383484279331,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.346766968558662,
        "unilateral_improvement_p1": 0.159949849054002,
        "nash_conv": 0.506716817612664,
        "exploitability": 0.253358408806332
      }
    },
    {
      "iteration": 79900,
      "average": {
        "on_policy_value_p0": -0.055550877759785,
        "best_response_value_p0": -0.054593295018493,
        "best_response_value_p1": 0.056221957261656,
        "unilateral_improvement_p0": 0.000957582741293,
        "unilateral_improvement_p1": 0.000671079501871,
        "nash_conv": 0.001628662243164,
        "exploitability": 0.000814331121582
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 80000,
      "average": {
        "on_policy_value_p0": -0.055550954113484,
        "best_response_value_p0": -0.054941720066386,
        "best_response_value_p1": 0.056568346481746,
        "unilateral_improvement_p0": 0.000609234047098,
        "unilateral_improvement_p1": 0.001017392368262,
        "nash_conv": 0.00162662641536,
        "exploitability": 0.00081331320768
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 80100,
      "average": {
        "on_policy_value_p0": -0.055551902933314,
        "best_response_value_p0": -0.055289275139128,
        "best_response_value_p1": 0.056841498402532,
        "unilateral_improvement_p0": 0.000262627794186,
        "unilateral_improvement_p1": 0.001289595469218,
        "nash_conv": 0.001552223263404,
        "exploitability": 0.000776111631702
      },
      "current": {
        "on_policy_value_p0": -0.179109969307806,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.179109969307806,
        "unilateral_improvement_p0": 0.512443302641139,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.512443302641139,
        "exploitability": 0.25622165132057
      }
    },
    {
      "iteration": 80200,
      "average": {
        "on_policy_value_p0": -0.05555326755253,
        "best_response_value_p0": -0.055252274711981,
        "best_response_value_p1": 0.056837978221814,
        "unilateral_improvement_p0": 0.000300992840549,
        "unilateral_improvement_p1": 0.001284710669283,
        "nash_conv": 0.001585703509833,
        "exploitability": 0.000792851754916
      },
      "current": {
        "on_policy_value_p0": 0.070921041279078,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.262412292054256,
        "unilateral_improvement_p1": 0.237587707945744,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 80300,
      "average": {
        "on_policy_value_p0": -0.055554275880491,
        "best_response_value_p0": -0.055015165170378,
        "best_response_value_p1": 0.056639405071268,
        "unilateral_improvement_p0": 0.000539110710113,
        "unilateral_improvement_p1": 0.001085129190777,
        "nash_conv": 0.00162423990089,
        "exploitability": 0.000812119950445
      },
      "current": {
        "on_policy_value_p0": 0.141561774176317,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.191771559157016,
        "unilateral_improvement_p1": 0.308228440842984,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 80400,
      "average": {
        "on_policy_value_p0": -0.05555463889056,
        "best_response_value_p0": -0.054745385821951,
        "best_response_value_p1": 0.05632103138667,
        "unilateral_improvement_p0": 0.000809253068609,
        "unilateral_improvement_p1": 0.000766392496109,
        "nash_conv": 0.001575645564719,
        "exploitability": 0.000787822782359
      },
      "current": {
        "on_policy_value_p0": 0.179281453089658,
        "best_response_value_p0": 0.282787566982469,
        "best_response_value_p1": 0.292987772214377,
        "unilateral_improvement_p0": 0.103506113892812,
        "unilateral_improvement_p1": 0.472269225304034,
        "nash_conv": 0.575775339196846,
        "exploitability": 0.287887669598423
      }
    },
    {
      "iteration": 80500,
      "average": {
        "on_policy_value_p0": -0.055554219592376,
        "best_response_value_p0": -0.054523810542919,
        "best_response_value_p1": 0.05602917801417,
        "unilateral_improvement_p0": 0.001030409049457,
        "unilateral_improvement_p1": 0.000474958421794,
        "nash_conv": 0.001505367471251,
        "exploitability": 0.000752683735626
      },
      "current": {
        "on_policy_value_p0": 0.078274835793072,
        "best_response_value_p0": 0.140303368654209,
        "best_response_value_p1": 0.375942934277727,
        "unilateral_improvement_p0": 0.062028532861136,
        "unilateral_improvement_p1": 0.454217770070799,
        "nash_conv": 0.516246302931936,
        "exploitability": 0.258123151465968
      }
    },
    {
      "iteration": 80600,
      "average": {
        "on_policy_value_p0": -0.05555340826909,
        "best_response_value_p0": -0.054378152548156,
        "best_response_value_p1": 0.056002466667305,
        "unilateral_improvement_p0": 0.001175255720935,
        "unilateral_improvement_p1": 0.000449058398214,
        "nash_conv": 0.001624314119149,
        "exploitability": 0.000812157059574
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 80700,
      "average": {
        "on_policy_value_p0": -0.055552411001111,
        "best_response_value_p0": -0.054342619597529,
        "best_response_value_p1": 0.055933070797828,
        "unilateral_improvement_p0": 0.001209791403582,
        "unilateral_improvement_p1": 0.000380659796717,
        "nash_conv": 0.001590451200299,
        "exploitability": 0.000795225600149
      },
      "current": {
        "on_policy_value_p0": -0.15123069823008,
        "best_response_value_p0": 0.065846201179947,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.217076899410027,
        "unilateral_improvement_p1": 0.34876930176992,
        "nash_conv": 0.565846201179947,
        "exploitability": 0.282923100589973
      }
    },
    {
      "iteration": 80800,
      "average": {
        "on_policy_value_p0": -0.055551610701365,
        "best_response_value_p0": -0.054651426461381,
        "best_response_value_p1": 0.056275578130073,
        "unilateral_improvement_p0": 0.000900184239984,
        "unilateral_improvement_p1": 0.000723967428708,
        "nash_conv": 0.001624151668692,
        "exploitability": 0.000812075834346
      },
      "current": {
        "on_policy_value_p0": -0.5,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 80900,
      "average": {
        "on_policy_value_p0": -0.055552031759906,
        "best_response_value_p0": -0.055105252136317,
        "best_response_value_p1": 0.056683209799973,
        "unilateral_improvement_p0": 0.000446779623589,
        "unilateral_improvement_p1": 0.001131178040067,
        "nash_conv": 0.001577957663656,
        "exploitability": 0.000788978831828
      },
      "current": {
        "on_policy_value_p0": -0.286558768149877,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.286558768149877,
        "unilateral_improvement_p0": 0.453225434816543,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.453225434816543,
        "exploitability": 0.226612717408272
      }
    },
    {
      "iteration": 81000,
      "average": {
        "on_policy_value_p0": -0.055553112693186,
        "best_response_value_p0": -0.055231221543109,
        "best_response_value_p1": 0.056800315245083,
        "unilateral_improvement_p0": 0.000321891150077,
        "unilateral_improvement_p1": 0.001247202551897,
        "nash_conv": 0.001569093701974,
        "exploitability": 0.000784546850987
      },
      "current": {
        "on_policy_value_p0": -0.079018156766505,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.245684823433172,
        "unilateral_improvement_p1": 0.087648509900161,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 81100,
      "average": {
        "on_policy_value_p0": -0.055554241917552,
        "best_response_value_p0": -0.055163118927151,
        "best_response_value_p1": 0.056779485179313,
        "unilateral_improvement_p0": 0.000391122990401,
        "unilateral_improvement_p1": 0.001225243261761,
        "nash_conv": 0.001616366252162,
        "exploitability": 0.000808183126081
      },
      "current": {
        "on_policy_value_p0": 0.022786876501108,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.143879790165559,
        "unilateral_improvement_p0": 0.143879790165559,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.310546456832226,
        "exploitability": 0.155273228416113
      }
    },
    {
      "iteration": 81200,
      "average": {
        "on_policy_value_p0": -0.055555200635357,
        "best_response_value_p0": -0.055095184051626,
        "best_response_value_p1": 0.056646698248839,
        "unilateral_improvement_p0": 0.000460016583731,
        "unilateral_improvement_p1": 0.001091497613482,
        "nash_conv": 0.001551514197213,
        "exploitability": 0.000775757098606
      },
      "current": {
        "on_policy_value_p0": 0.074477067237665,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.148954134475329,
        "unilateral_improvement_p0": 0.092189599429002,
        "unilateral_improvement_p1": 0.223431201712994,
        "nash_conv": 0.315620801141996,
        "exploitability": 0.157810400570998
      }
    },
    {
      "iteration": 81300,
      "average": {
        "on_policy_value_p0": -0.055555750944284,
        "best_response_value_p0": -0.054886548104776,
        "best_response_value_p1": 0.056460395077379,
        "unilateral_improvement_p0": 0.000669202839508,
        "unilateral_improvement_p1": 0.000904644133095,
        "nash_conv": 0.001573846972603,
        "exploitability": 0.000786923486302
      },
      "current": {
        "on_policy_value_p0": 0.07629895673314,
        "best_response_value_p0": 0.125903817466217,
        "best_response_value_p1": 0.234123611867178,
        "unilateral_improvement_p0": 0.049604860733077,
        "unilateral_improvement_p1": 0.310422568600318,
        "nash_conv": 0.360027429333396,
        "exploitability": 0.180013714666698
      }
    },
    {
      "iteration": 81400,
      "average": {
        "on_policy_value_p0": -0.055555734937561,
        "best_response_value_p0": -0.054686872403907,
        "best_response_value_p1": 0.056210505237645,
        "unilateral_improvement_p0": 0.000868862533654,
        "unilateral_improvement_p1": 0.000654770300084,
        "nash_conv": 0.001523632833737,
        "exploitability": 0.000761816416869
      },
      "current": {
        "on_policy_value_p0": 0.091002584790128,
        "best_response_value_p0": 0.091002584790128,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.424335918123461,
        "nash_conv": 0.424335918123461,
        "exploitability": 0.212167959061731
      }
    },
    {
      "iteration": 81500,
      "average": {
        "on_policy_value_p0": -0.055555240216267,
        "best_response_value_p0": -0.054526706859243,
        "best_response_value_p1": 0.056089899998646,
        "unilateral_improvement_p0": 0.001028533357025,
        "unilateral_improvement_p1": 0.000534659782379,
        "nash_conv": 0.001563193139404,
        "exploitability": 0.000781596569702
      },
      "current": {
        "on_policy_value_p0": 0.059860416033718,
        "best_response_value_p0": 0.059860416033718,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.393193749367052,
        "nash_conv": 0.393193749367052,
        "exploitability": 0.196596874683526
      }
    },
    {
      "iteration": 81600,
      "average": {
        "on_policy_value_p0": -0.055554446927431,
        "best_response_value_p0": -0.054410687327787,
        "best_response_value_p1": 0.0560211623761,
        "unilateral_improvement_p0": 0.001143759599644,
        "unilateral_improvement_p1": 0.000466715448668,
        "nash_conv": 0.001610475048312,
        "exploitability": 0.000805237524156
      },
      "current": {
        "on_policy_value_p0": 0.015141456388261,
        "best_response_value_p0": 0.136383753890144,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.121242297501883,
        "unilateral_improvement_p1": 0.348474789721595,
        "nash_conv": 0.469717087223477,
        "exploitability": 0.234858543611739
      }
    },
    {
      "iteration": 81700,
      "average": {
        "on_policy_value_p0": -0.055553529054183,
        "best_response_value_p0": -0.054342018612429,
        "best_response_value_p1": 0.055952593021907,
        "unilateral_improvement_p0": 0.001211510441754,
        "unilateral_improvement_p1": 0.000399063967724,
        "nash_conv": 0.001610574409478,
        "exploitability": 0.000805287204739
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 81800,
      "average": {
        "on_policy_value_p0": -0.055552434518427,
        "best_response_value_p0": -0.054306232705218,
        "best_response_value_p1": 0.055884191318946,
        "unilateral_improvement_p0": 0.001246201813209,
        "unilateral_improvement_p1": 0.000331756800519,
        "nash_conv": 0.001577958613728,
        "exploitability": 0.000788979306864
      },
      "current": {
        "on_policy_value_p0": -0.098908080396113,
        "best_response_value_p0": 0.11721262646861,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.216120706864723,
        "unilateral_improvement_p1": 0.23442525293722,
        "nash_conv": 0.450545959801943,
        "exploitability": 0.225272979900972
      }
    },
    {
      "iteration": 81900,
      "average": {
        "on_policy_value_p0": -0.05555135316649,
        "best_response_value_p0": -0.054439309967006,
        "best_response_value_p1": 0.05601096148948,
        "unilateral_improvement_p0": 0.001112043199483,
        "unilateral_improvement_p1": 0.000459608322991,
        "nash_conv": 0.001571651522474,
        "exploitability": 0.000785825761237
      },
      "current": {
        "on_policy_value_p0": -0.228026202084406,
        "best_response_value_p0": 0.228026202084406,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.456052404168812,
        "unilateral_improvement_p1": 0.105307131248927,
        "nash_conv": 0.561359535417739,
        "exploitability": 0.28067976770887
      }
    },
    {
      "iteration": 82000,
      "average": {
        "on_policy_value_p0": -0.055550903997594,
        "best_response_value_p0": -0.054741944557164,
        "best_response_value_p1": 0.056349159503924,
        "unilateral_improvement_p0": 0.00080895944043,
        "unilateral_improvement_p1": 0.000798255506331,
        "nash_conv": 0.00160721494676,
        "exploitability": 0.00080360747338
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 82100,
      "average": {
        "on_policy_value_p0": -0.05555130363958,
        "best_response_value_p0": -0.055081276333992,
        "best_response_value_p1": 0.056685871961618,
        "unilateral_improvement_p0": 0.000470027305588,
        "unilateral_improvement_p1": 0.001134568322038,
        "nash_conv": 0.001604595627626,
        "exploitability": 0.000802297813813
      },
      "current": {
        "on_policy_value_p0": -0.31058454714137,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.31058454714137,
        "unilateral_improvement_p0": 0.643917880474703,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.643917880474703,
        "exploitability": 0.321958940237352
      }
    },
    {
      "iteration": 82200,
      "average": {
        "on_policy_value_p0": -0.055552461452045,
        "best_response_value_p0": -0.055290371240928,
        "best_response_value_p1": 0.056860568716392,
        "unilateral_improvement_p0": 0.000262090211117,
        "unilateral_improvement_p1": 0.001308107264346,
        "nash_conv": 0.001570197475463,
        "exploitability": 0.000785098737732
      },
      "current": {
        "on_policy_value_p0": -0.094616103891765,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.109994936383849,
        "unilateral_improvement_p0": 0.427949437225098,
        "unilateral_improvement_p1": 0.015378832492084,
        "nash_conv": 0.443328269717182,
        "exploitability": 0.221664134858591
      }
    },
    {
      "iteration": 82300,
      "average": {
        "on_policy_value_p0": -0.055553689858184,
        "best_response_value_p0": -0.055191393658692,
        "best_response_value_p1": 0.056753762182096,
        "unilateral_improvement_p0": 0.000362296199492,
        "unilateral_improvement_p1": 0.001200072323912,
        "nash_conv": 0.001562368523404,
        "exploitability": 0.000781184261702
      },
      "current": {
        "on_policy_value_p0": 0.091803123583818,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.241530209749516,
        "unilateral_improvement_p1": 0.258469790250484,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 82400,
      "average": {
        "on_policy_value_p0": -0.055554447253008,
        "best_response_value_p0": -0.054922148439851,
        "best_response_value_p1": 0.056525102948184,
        "unilateral_improvement_p0": 0.000632298813157,
        "unilateral_improvement_p1": 0.000970655695176,
        "nash_conv": 0.001602954508332,
        "exploitability": 0.000801477254166
      },
      "current": {
        "on_policy_value_p0": 0.188706536076997,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.210746405487328,
        "unilateral_improvement_p0": 0.144626797256336,
        "unilateral_improvement_p1": 0.399452941564325,
        "nash_conv": 0.544079738820661,
        "exploitability": 0.272039869410331
      }
    },
    {
      "iteration": 82500,
      "average": {
        "on_policy_value_p0": -0.055554526586666,
        "best_response_value_p0": -0.0546945357402,
        "best_response_value_p1": 0.056193562825549,
        "unilateral_improvement_p0": 0.000859990846466,
        "unilateral_improvement_p1": 0.000639036238884,
        "nash_conv": 0.00149902708535,
        "exploitability": 0.000749513542675
      },
      "current": {
        "on_policy_value_p0": 0.129499212879468,
        "best_response_value_p0": 0.222452869662134,
        "best_response_value_p1": 0.314092686434668,
        "unilateral_improvement_p0": 0.092953656782666,
        "unilateral_improvement_p1": 0.443591899314136,
        "nash_conv": 0.536545556096802,
        "exploitability": 0.268272778048401
      }
    },
    {
      "iteration": 82600,
      "average": {
        "on_policy_value_p0": -0.055553971807658,
        "best_response_value_p0": -0.054484267728865,
        "best_response_value_p1": 0.056035383161719,
        "unilateral_improvement_p0": 0.001069704078793,
        "unilateral_improvement_p1": 0.000481411354061,
        "nash_conv": 0.001551115432854,
        "exploitability": 0.000775557716427
      },
      "current": {
        "on_policy_value_p0": 0.042989947277447,
        "best_response_value_p0": 0.119681508132547,
        "best_response_value_p1": 0.398039260418414,
        "unilateral_improvement_p0": 0.0766915608551,
        "unilateral_improvement_p1": 0.441029207695861,
        "nash_conv": 0.517720768550961,
        "exploitability": 0.258860384275481
      }
    },
    {
      "iteration": 82700,
      "average": {
        "on_policy_value_p0": -0.055553181655016,
        "best_response_value_p0": -0.054390217305424,
        "best_response_value_p1": 0.055993381105133,
        "unilateral_improvement_p0": 0.001162964349593,
        "unilateral_improvement_p1": 0.000440199450116,
        "nash_conv": 0.001603163799709,
        "exploitability": 0.000801581899854
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 82800,
      "average": {
        "on_policy_value_p0": -0.055552217336599,
        "best_response_value_p0": -0.054405316670047,
        "best_response_value_p1": 0.055925756248725,
        "unilateral_improvement_p0": 0.001146900666551,
        "unilateral_improvement_p1": 0.000373538912126,
        "nash_conv": 0.001520439578678,
        "exploitability": 0.000760219789339
      },
      "current": {
        "on_policy_value_p0": -0.218016986751854,
        "best_response_value_p0": 0.072672328917285,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.290689315669138,
        "unilateral_improvement_p1": 0.281983013248146,
        "nash_conv": 0.572672328917285,
        "exploitability": 0.286336164458642
      }
    },
    {
      "iteration": 82900,
      "average": {
        "on_policy_value_p0": -0.05555168634023,
        "best_response_value_p0": -0.054787671936166,
        "best_response_value_p1": 0.056384014222069,
        "unilateral_improvement_p0": 0.000764014404064,
        "unilateral_improvement_p1": 0.000832327881839,
        "nash_conv": 0.001596342285903,
        "exploitability": 0.000798171142952
      },
      "current": {
        "on_policy_value_p0": -0.439436627979718,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.439436627979718,
        "unilateral_improvement_p0": 0.606103294646385,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.606103294646385,
        "exploitability": 0.303051647323192
      }
    },
    {
      "iteration": 83000,
      "average": {
        "on_policy_value_p0": -0.055552321380831,
        "best_response_value_p0": -0.055148362934001,
        "best_response_value_p1": 0.056726743893039,
        "unilateral_improvement_p0": 0.00040395844683,
        "unilateral_improvement_p1": 0.001174422512208,
        "nash_conv": 0.001578380959038,
        "exploitability": 0.000789190479519
      },
      "current": {
        "on_policy_value_p0": -0.223661146972411,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.223661146972411,
        "unilateral_improvement_p0": 0.390327813639078,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.390327813639078,
        "exploitability": 0.195163906819539
      }
    },
    {
      "iteration": 83100,
      "average": {
        "on_policy_value_p0": -0.055553415680812,
        "best_response_value_p0": -0.055228127371769,
        "best_response_value_p1": 0.056795648934152,
        "unilateral_improvement_p0": 0.000325288309042,
        "unilateral_improvement_p1": 0.00124223325334,
        "nash_conv": 0.001567521562383,
        "exploitability": 0.000783760781191
      },
      "current": {
        "on_policy_value_p0": -0.066093764338132,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.232760431004798,
        "unilateral_improvement_p1": 0.100572902328535,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 83200,
      "average": {
        "on_policy_value_p0": -0.055554502502742,
        "best_response_value_p0": -0.055161747410987,
        "best_response_value_p1": 0.05674782253043,
        "unilateral_improvement_p0": 0.000392755091755,
        "unilateral_improvement_p1": 0.001193320027688,
        "nash_conv": 0.001586075119443,
        "exploitability": 0.000793037559722
      },
      "current": {
        "on_policy_value_p0": 0.039267868571058,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.127398798095608,
        "unilateral_improvement_p0": 0.127398798095608,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.294065464762275,
        "exploitability": 0.147032732381137
      }
    },
    {
      "iteration": 83300,
      "average": {
        "on_policy_value_p0": -0.055555380777314,
        "best_response_value_p0": -0.055092053031123,
        "best_response_value_p1": 0.056604299120986,
        "unilateral_improvement_p0": 0.000463327746191,
        "unilateral_improvement_p1": 0.001048918343672,
        "nash_conv": 0.001512246089863,
        "exploitability": 0.000756123044932
      },
      "current": {
        "on_policy_value_p0": 0.084020794866464,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.168041589732929,
        "unilateral_improvement_p0": 0.082645871800202,
        "unilateral_improvement_p1": 0.252062384599393,
        "nash_conv": 0.334708256399595,
        "exploitability": 0.167354128199798
      }
    },
    {
      "iteration": 83400,
      "average": {
        "on_policy_value_p0": -0.055555815509427,
        "best_response_value_p0": -0.054854972786186,
        "best_response_value_p1": 0.056410609672226,
        "unilateral_improvement_p0": 0.000700842723241,
        "unilateral_improvement_p1": 0.0008547941628,
        "nash_conv": 0.001555636886041,
        "exploitability": 0.00077781844302
      },
      "current": {
        "on_policy_value_p0": 0.081012532414647,
        "best_response_value_p0": 0.118404848268664,
        "best_response_value_p1": 0.258548701625299,
        "unilateral_improvement_p0": 0.037392315854017,
        "unilateral_improvement_p1": 0.339561234039946,
        "nash_conv": 0.376953549893963,
        "exploitability": 0.188476774946982
      }
    },
    {
      "iteration": 83500,
      "average": {
        "on_policy_value_p0": -0.055555702203429,
        "best_response_value_p0": -0.054667614376196,
        "best_response_value_p1": 0.056155414282221,
        "unilateral_improvement_p0": 0.000888087827233,
        "unilateral_improvement_p1": 0.000599712078792,
        "nash_conv": 0.001487799906025,
        "exploitability": 0.000743899953012
      },
      "current": {
        "on_policy_value_p0": 0.085777481181618,
        "best_response_value_p0": 0.085777481181618,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.419110814514951,
        "nash_conv": 0.419110814514951,
        "exploitability": 0.209555407257475
      }
    },
    {
      "iteration": 83600,
      "average": {
        "on_policy_value_p0": -0.055555165850653,
        "best_response_value_p0": -0.054517713571977,
        "best_response_value_p1": 0.05607193140602,
        "unilateral_improvement_p0": 0.001037452278676,
        "unilateral_improvement_p1": 0.000516765555367,
        "nash_conv": 0.001554217834043,
        "exploitability": 0.000777108917021
      },
      "current": {
        "on_policy_value_p0": 0.054321391353434,
        "best_response_value_p0": 0.0580238839598,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.003702492606366,
        "unilateral_improvement_p1": 0.387654724686767,
        "nash_conv": 0.391357217293133,
        "exploitability": 0.195678608646566
      }
    },
    {
      "iteration": 83700,
      "average": {
        "on_policy_value_p0": -0.05555437492506,
        "best_response_value_p0": -0.054413184357993,
        "best_response_value_p1": 0.056004939851175,
        "unilateral_improvement_p0": 0.001141190567067,
        "unilateral_improvement_p1": 0.000450564926115,
        "nash_conv": 0.001591755493182,
        "exploitability": 0.000795877746591
      },
      "current": {
        "on_policy_value_p0": 0.004708265222808,
        "best_response_value_p0": 0.15725013622105,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.152541870998242,
        "unilateral_improvement_p1": 0.338041598556142,
        "nash_conv": 0.490583469554383,
        "exploitability": 0.245291734777192
      }
    },
    {
      "iteration": 83800,
      "average": {
        "on_policy_value_p0": -0.055553471983748,
        "best_response_value_p0": -0.054348057066242,
        "best_response_value_p1": 0.055938108180709,
        "unilateral_improvement_p0": 0.001205414917507,
        "unilateral_improvement_p1": 0.000384636196961,
        "nash_conv": 0.001590051114467,
        "exploitability": 0.000795025557234
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 83900,
      "average": {
        "on_policy_value_p0": -0.05555239374857,
        "best_response_value_p0": -0.054329688095729,
        "best_response_value_p1": 0.055871435822925,
        "unilateral_improvement_p0": 0.001222705652842,
        "unilateral_improvement_p1": 0.000319042074355,
        "nash_conv": 0.001541747727196,
        "exploitability": 0.000770873863598
      },
      "current": {
        "on_policy_value_p0": -0.116841775708526,
        "best_response_value_p0": 0.116841775708526,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.233683551417052,
        "unilateral_improvement_p1": 0.216491557624807,
        "nash_conv": 0.450175109041859,
        "exploitability": 0.22508755452093
      }
    },
    {
      "iteration": 84000,
      "average": {
        "on_policy_value_p0": -0.055551392026814,
        "best_response_value_p0": -0.054477606336495,
        "best_response_value_p1": 0.056042490048972,
        "unilateral_improvement_p0": 0.001073785690319,
        "unilateral_improvement_p1": 0.000491098022159,
        "nash_conv": 0.001564883712477,
        "exploitability": 0.000782441856239
      },
      "current": {
        "on_policy_value_p0": -0.245171893909504,
        "best_response_value_p0": 0.245171893909504,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.490343787819008,
        "unilateral_improvement_p1": 0.088161439423829,
        "nash_conv": 0.578505227242837,
        "exploitability": 0.289252613621419
      }
    },
    {
      "iteration": 84100,
      "average": {
        "on_policy_value_p0": -0.055551047452539,
        "best_response_value_p0": -0.054785355588261,
        "best_response_value_p1": 0.056372205677135,
        "unilateral_improvement_p0": 0.000765691864278,
        "unilateral_improvement_p1": 0.000821158224596,
        "nash_conv": 0.001586850088873,
        "exploitability": 0.000793425044437
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 84200,
      "average": {
        "on_policy_value_p0": -0.055551507547913,
        "best_response_value_p0": -0.055116172663968,
        "best_response_value_p1": 0.056697763892936,
        "unilateral_improvement_p0": 0.000435334883945,
        "unilateral_improvement_p1": 0.001146256345023,
        "nash_conv": 0.001581591228968,
        "exploitability": 0.000790795614484
      },
      "current": {
        "on_policy_value_p0": -0.287635570974057,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.287635570974057,
        "unilateral_improvement_p0": 0.620968904307391,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.620968904307391,
        "exploitability": 0.310484452153695
      }
    },
    {
      "iteration": 84300,
      "average": {
        "on_policy_value_p0": -0.055552667256471,
        "best_response_value_p0": -0.055295958139693,
        "best_response_value_p1": 0.056851304840791,
        "unilateral_improvement_p0": 0.000256709116778,
        "unilateral_improvement_p1": 0.00129863758432,
        "nash_conv": 0.001555346701098,
        "exploitability": 0.000777673350549
      },
      "current": {
        "on_policy_value_p0": -0.074459236415648,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.1153046418651,
        "unilateral_improvement_p0": 0.407792569748982,
        "unilateral_improvement_p1": 0.040845405449452,
        "nash_conv": 0.448637975198434,
        "exploitability": 0.224318987599217
      }
    },
    {
      "iteration": 84400,
      "average": {
        "on_policy_value_p0": -0.055553844188552,
        "best_response_value_p0": -0.055184434052544,
        "best_response_value_p1": 0.056733482779446,
        "unilateral_improvement_p0": 0.000369410136008,
        "unilateral_improvement_p1": 0.001179638590894,
        "nash_conv": 0.001549048726902,
        "exploitability": 0.000774524363451
      },
      "current": {
        "on_policy_value_p0": 0.096257500778573,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.23707583255476,
        "unilateral_improvement_p1": 0.26292416744524,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 84500,
      "average": {
        "on_policy_value_p0": -0.055554556680022,
        "best_response_value_p0": -0.05492188837122,
        "best_response_value_p1": 0.056500875120774,
        "unilateral_improvement_p0": 0.000632668308802,
        "unilateral_improvement_p1": 0.000946318440751,
        "nash_conv": 0.001578986749554,
        "exploitability": 0.000789493374777
      },
      "current": {
        "on_policy_value_p0": 0.197243302708193,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.22781993874972,
        "unilateral_improvement_p0": 0.13609003062514,
        "unilateral_improvement_p1": 0.425063241457914,
        "nash_conv": 0.561153272083053,
        "exploitability": 0.280576636041527
      }
    },
    {
      "iteration": 84600,
      "average": {
        "on_policy_value_p0": -0.055554605692582,
        "best_response_value_p0": -0.05470245799387,
        "best_response_value_p1": 0.056170817133833,
        "unilateral_improvement_p0": 0.000852147698712,
        "unilateral_improvement_p1": 0.000616211441251,
        "nash_conv": 0.001468359139963,
        "exploitability": 0.000734179569981
      },
      "current": {
        "on_policy_value_p0": 0.131728661811724,
        "best_response_value_p0": 0.220339495360237,
        "best_response_value_p1": 0.322778332902973,
        "unilateral_improvement_p0": 0.088610833548514,
        "unilateral_improvement_p1": 0.454506994714697,
        "nash_conv": 0.54311782826321,
        "exploitability": 0.271558914131605
      }
    },
    {
      "iteration": 84700,
      "average": {
        "on_policy_value_p0": -0.055554038138152,
        "best_response_value_p0": -0.054477830758977,
        "best_response_value_p1": 0.056015330764852,
        "unilateral_improvement_p0": 0.001076207379175,
        "unilateral_improvement_p1": 0.0004612926267,
        "nash_conv": 0.001537500005875,
        "exploitability": 0.000768750002938
      },
      "current": {
        "on_policy_value_p0": 0.0473963806042,
        "best_response_value_p0": 0.119886239721033,
        "best_response_value_p1": 0.407671053425868,
        "unilateral_improvement_p0": 0.072489859116833,
        "unilateral_improvement_p1": 0.455067434030068,
        "nash_conv": 0.527557293146901,
        "exploitability": 0.263778646573451
      }
    },
    {
      "iteration": 84800,
      "average": {
        "on_policy_value_p0": -0.055553229316204,
        "best_response_value_p0": -0.054386073247862,
        "best_response_value_p1": 0.055969136558771,
        "unilateral_improvement_p0": 0.001167156068342,
        "unilateral_improvement_p1": 0.000415907242567,
        "nash_conv": 0.001583063310909,
        "exploitability": 0.000791531655454
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 84900,
      "average": {
        "on_policy_value_p0": -0.05555226340248,
        "best_response_value_p0": -0.054422190641601,
        "best_response_value_p1": 0.055903212958583,
        "unilateral_improvement_p0": 0.001130072760879,
        "unilateral_improvement_p1": 0.000350949556103,
        "nash_conv": 0.001481022316982,
        "exploitability": 0.000740511158491
      },
      "current": {
        "on_policy_value_p0": -0.238621200307876,
        "best_response_value_p0": 0.079540400102625,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.318161600410501,
        "unilateral_improvement_p1": 0.261378799692124,
        "nash_conv": 0.579540400102625,
        "exploitability": 0.289770200051313
      }
    },
    {
      "iteration": 85000,
      "average": {
        "on_policy_value_p0": -0.055551795578569,
        "best_response_value_p0": -0.054817644930198,
        "best_response_value_p1": 0.056388762344348,
        "unilateral_improvement_p0": 0.000734150648371,
        "unilateral_improvement_p1": 0.00083696676578,
        "nash_conv": 0.001571117414151,
        "exploitability": 0.000785558707075
      },
      "current": {
        "on_policy_value_p0": -0.423269559078056,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.423269559078056,
        "unilateral_improvement_p0": 0.589936225744723,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.589936225744723,
        "exploitability": 0.294968112872362
      }
    },
    {
      "iteration": 85100,
      "average": {
        "on_policy_value_p0": -0.055552422920239,
        "best_response_value_p0": -0.055146329323281,
        "best_response_value_p1": 0.056709719813296,
        "unilateral_improvement_p0": 0.000406093596958,
        "unilateral_improvement_p1": 0.001157296893058,
        "nash_conv": 0.001563390490016,
        "exploitability": 0.000781695245008
      },
      "current": {
        "on_policy_value_p0": -0.210692552494674,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.210692552494674,
        "unilateral_improvement_p0": 0.37735921916134,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.37735921916134,
        "exploitability": 0.18867960958067
      }
    },
    {
      "iteration": 85200,
      "average": {
        "on_policy_value_p0": -0.055553454745705,
        "best_response_value_p0": -0.055222482758006,
        "best_response_value_p1": 0.056770847774863,
        "unilateral_improvement_p0": 0.000330971987699,
        "unilateral_improvement_p1": 0.001217393029159,
        "nash_conv": 0.001548365016858,
        "exploitability": 0.000774182508429
      },
      "current": {
        "on_policy_value_p0": -0.06579302137767,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.232459688044337,
        "unilateral_improvement_p1": 0.100873645288997,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 85300,
      "average": {
        "on_policy_value_p0": -0.05555447702124,
        "best_response_value_p0": -0.0551577436223,
        "best_response_value_p1": 0.056727532281505,
        "unilateral_improvement_p0": 0.00039673339894,
        "unilateral_improvement_p1": 0.001173055260265,
        "nash_conv": 0.001569788659205,
        "exploitability": 0.000784894329602
      },
      "current": {
        "on_policy_value_p0": 0.03455231364376,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.132114353022906,
        "unilateral_improvement_p0": 0.132114353022906,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.298781019689573,
        "exploitability": 0.149390509844787
      }
    },
    {
      "iteration": 85400,
      "average": {
        "on_policy_value_p0": -0.055555314899568,
        "best_response_value_p0": -0.055093156100495,
        "best_response_value_p1": 0.056592100882819,
        "unilateral_improvement_p0": 0.000462158799073,
        "unilateral_improvement_p1": 0.001036785983251,
        "nash_conv": 0.001498944782324,
        "exploitability": 0.000749472391162
      },
      "current": {
        "on_policy_value_p0": 0.080391696703999,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.160783393407999,
        "unilateral_improvement_p0": 0.086274969962667,
        "unilateral_improvement_p1": 0.241175090111998,
        "nash_conv": 0.327450060074665,
        "exploitability": 0.163725030037333
      }
    },
    {
      "iteration": 85500,
      "average": {
        "on_policy_value_p0": -0.055555747462923,
        "best_response_value_p0": -0.054875508362182,
        "best_response_value_p1": 0.056408264385047,
        "unilateral_improvement_p0": 0.000680239100742,
        "unilateral_improvement_p1": 0.000852516922124,
        "nash_conv": 0.001532756022866,
        "exploitability": 0.000766378011433
      },
      "current": {
        "on_policy_value_p0": 0.076115515356083,
        "best_response_value_p0": 0.119361095378886,
        "best_response_value_p1": 0.246842173287729,
        "unilateral_improvement_p0": 0.043245580022802,
        "unilateral_improvement_p1": 0.322957688643812,
        "nash_conv": 0.366203268666614,
        "exploitability": 0.183101634333307
      }
    },
    {
      "iteration": 85600,
      "average": {
        "on_policy_value_p0": -0.055555672460676,
        "best_response_value_p0": -0.054691542516341,
        "best_response_value_p1": 0.056164474527124,
        "unilateral_improvement_p0": 0.000864129944335,
        "unilateral_improvement_p1": 0.000608802066447,
        "nash_conv": 0.001472932010782,
        "exploitability": 0.000736466005391
      },
      "current": {
        "on_policy_value_p0": 0.086895343625879,
        "best_response_value_p0": 0.086895343625879,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.420228676959212,
        "nash_conv": 0.420228676959212,
        "exploitability": 0.210114338479606
      }
    },
    {
      "iteration": 85700,
      "average": {
        "on_policy_value_p0": -0.05555518562471,
        "best_response_value_p0": -0.054543690814763,
        "best_response_value_p1": 0.05607546848659,
        "unilateral_improvement_p0": 0.001011494809947,
        "unilateral_improvement_p1": 0.00052028286188,
        "nash_conv": 0.001531777671827,
        "exploitability": 0.000765888835914
      },
      "current": {
        "on_policy_value_p0": 0.056068398484469,
        "best_response_value_p0": 0.056068398484469,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.389401731817803,
        "nash_conv": 0.389401731817803,
        "exploitability": 0.194700865908901
      }
    },
    {
      "iteration": 85800,
      "average": {
        "on_policy_value_p0": -0.055554447151776,
        "best_response_value_p0": -0.054438498218344,
        "best_response_value_p1": 0.056010112462714,
        "unilateral_improvement_p0": 0.001115948933431,
        "unilateral_improvement_p1": 0.000455665310938,
        "nash_conv": 0.001571614244369,
        "exploitability": 0.000785807122185
      },
      "current": {
        "on_policy_value_p0": 0.009325013003103,
        "best_response_value_p0": 0.14801664066046,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.138691627657357,
        "unilateral_improvement_p1": 0.342658346336436,
        "nash_conv": 0.481349973993794,
        "exploitability": 0.240674986996897
      }
    },
    {
      "iteration": 85900,
      "average": {
        "on_policy_value_p0": -0.055553601622784,
        "best_response_value_p0": -0.054374387297453,
        "best_response_value_p1": 0.055944908606529,
        "unilateral_improvement_p0": 0.001179214325331,
        "unilateral_improvement_p1": 0.000391306983746,
        "nash_conv": 0.001570521309077,
        "exploitability": 0.000785260654538
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 86000,
      "average": {
        "on_policy_value_p0": -0.055552593251477,
        "best_response_value_p0": -0.054339771710193,
        "best_response_value_p1": 0.05587985638722,
        "unilateral_improvement_p0": 0.001212821541284,
        "unilateral_improvement_p1": 0.000327263135743,
        "nash_conv": 0.001540084677027,
        "exploitability": 0.000770042338514
      },
      "current": {
        "on_policy_value_p0": -0.097198331046111,
        "best_response_value_p0": 0.118067501143611,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.215265832189722,
        "unilateral_improvement_p1": 0.236135002287223,
        "nash_conv": 0.451400834476945,
        "exploitability": 0.225700417238472
      }
    },
    {
      "iteration": 86100,
      "average": {
        "on_policy_value_p0": -0.055551589224793,
        "best_response_value_p0": -0.054462819026937,
        "best_response_value_p1": 0.055990860085944,
        "unilateral_improvement_p0": 0.001088770197856,
        "unilateral_improvement_p1": 0.000439270861151,
        "nash_conv": 0.001528041059007,
        "exploitability": 0.000764020529503
      },
      "current": {
        "on_policy_value_p0": -0.222869853805394,
        "best_response_value_p0": 0.222869853805394,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.445739707610788,
        "unilateral_improvement_p1": 0.110463479527939,
        "nash_conv": 0.556203187138727,
        "exploitability": 0.278101593569364
      }
    },
    {
      "iteration": 86200,
      "average": {
        "on_policy_value_p0": -0.055551140956004,
        "best_response_value_p0": -0.05474513833334,
        "best_response_value_p1": 0.056312603094352,
        "unilateral_improvement_p0": 0.000806002622665,
        "unilateral_improvement_p1": 0.000761462138348,
        "nash_conv": 0.001567464761013,
        "exploitability": 0.000783732380506
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 86300,
      "average": {
        "on_policy_value_p0": -0.055551462382238,
        "best_response_value_p0": -0.055067952000778,
        "best_response_value_p1": 0.05663360046427,
        "unilateral_improvement_p0": 0.00048351038146,
        "unilateral_improvement_p1": 0.001082138082032,
        "nash_conv": 0.001565648463492,
        "exploitability": 0.000782824231746
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 86400,
      "average": {
        "on_policy_value_p0": -0.055552495422989,
        "best_response_value_p0": -0.055309083268282,
        "best_response_value_p1": 0.056824999611373,
        "unilateral_improvement_p0": 0.000243412154707,
        "unilateral_improvement_p1": 0.001272504188385,
        "nash_conv": 0.001515916343092,
        "exploitability": 0.000757958171546
      },
      "current": {
        "on_policy_value_p0": -0.127071762911276,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.127071762911276,
        "unilateral_improvement_p0": 0.460405096244609,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.460405096244609,
        "exploitability": 0.230202548122305
      }
    },
    {
      "iteration": 86500,
      "average": {
        "on_policy_value_p0": -0.055553676800906,
        "best_response_value_p0": -0.05523716651171,
        "best_response_value_p1": 0.056750182911999,
        "unilateral_improvement_p0": 0.000316510289197,
        "unilateral_improvement_p1": 0.001196506111092,
        "nash_conv": 0.001513016400289,
        "exploitability": 0.000756508200144
      },
      "current": {
        "on_policy_value_p0": 0.084861781450236,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.248471551883097,
        "unilateral_improvement_p1": 0.251528448116903,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 86600,
      "average": {
        "on_policy_value_p0": -0.055554464047586,
        "best_response_value_p0": -0.054980926519588,
        "best_response_value_p1": 0.0565464351859,
        "unilateral_improvement_p0": 0.000573537527998,
        "unilateral_improvement_p1": 0.000991971138314,
        "nash_conv": 0.001565508666312,
        "exploitability": 0.000782754333156
      },
      "current": {
        "on_policy_value_p0": 0.181005971138927,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.195345275611187,
        "unilateral_improvement_p0": 0.152327362194407,
        "unilateral_improvement_p1": 0.376351246750113,
        "nash_conv": 0.52867860894452,
        "exploitability": 0.26433930447226
      }
    },
    {
      "iteration": 86700,
      "average": {
        "on_policy_value_p0": -0.055554658475127,
        "best_response_value_p0": -0.054744551105314,
        "best_response_value_p1": 0.056237469923321,
        "unilateral_improvement_p0": 0.000810107369813,
        "unilateral_improvement_p1": 0.000682811448194,
        "nash_conv": 0.001492918818008,
        "exploitability": 0.000746459409004
      },
      "current": {
        "on_policy_value_p0": 0.152308438408325,
        "best_response_value_p0": 0.250149489417469,
        "best_response_value_p1": 0.304317897981713,
        "unilateral_improvement_p0": 0.097841051009144,
        "unilateral_improvement_p1": 0.456626336390038,
        "nash_conv": 0.554467387399182,
        "exploitability": 0.277233693699591
      }
    },
    {
      "iteration": 86800,
      "average": {
        "on_policy_value_p0": -0.055554214767103,
        "best_response_value_p0": -0.054540373299188,
        "best_response_value_p1": 0.05601750414712,
        "unilateral_improvement_p0": 0.001013841467915,
        "unilateral_improvement_p1": 0.000463289380017,
        "nash_conv": 0.001477130847932,
        "exploitability": 0.000738565423966
      },
      "current": {
        "on_policy_value_p0": 0.065441831520923,
        "best_response_value_p0": 0.123189067557144,
        "best_response_value_p1": 0.384505527927558,
        "unilateral_improvement_p0": 0.057747236036221,
        "unilateral_improvement_p1": 0.449947359448482,
        "nash_conv": 0.507694595484703,
        "exploitability": 0.253847297742351
      }
    },
    {
      "iteration": 86900,
      "average": {
        "on_policy_value_p0": -0.055553487971489,
        "best_response_value_p0": -0.05442349473634,
        "best_response_value_p1": 0.055987422512099,
        "unilateral_improvement_p0": 0.001129993235149,
        "unilateral_improvement_p1": 0.000433934540609,
        "nash_conv": 0.001563927775759,
        "exploitability": 0.000781963887879
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 87000,
      "average": {
        "on_policy_value_p0": -0.055552602543325,
        "best_response_value_p0": -0.054395128578228,
        "best_response_value_p1": 0.055923069152889,
        "unilateral_improvement_p0": 0.001157473965097,
        "unilateral_improvement_p1": 0.000370466609564,
        "nash_conv": 0.001527940574662,
        "exploitability": 0.000763970287331
      },
      "current": {
        "on_policy_value_p0": -0.157646349087263,
        "best_response_value_p0": 0.061569100608492,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.219215449695754,
        "unilateral_improvement_p1": 0.342353650912737,
        "nash_conv": 0.561569100608492,
        "exploitability": 0.280784550304246
      }
    },
    {
      "iteration": 87100,
      "average": {
        "on_policy_value_p0": -0.055551895432557,
        "best_response_value_p0": -0.054680803359959,
        "best_response_value_p1": 0.056244709175167,
        "unilateral_improvement_p0": 0.000871092072598,
        "unilateral_improvement_p1": 0.00069281374261,
        "nash_conv": 0.001563905815208,
        "exploitability": 0.000781952907604
      },
      "current": {
        "on_policy_value_p0": -0.499077460461371,
        "best_response_value_p0": 0.166359153487124,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.665436613948495,
        "unilateral_improvement_p1": 0.000922539538629,
        "nash_conv": 0.666359153487124,
        "exploitability": 0.333179576743562
      }
    },
    {
      "iteration": 87200,
      "average": {
        "on_policy_value_p0": -0.055552249686775,
        "best_response_value_p0": -0.05512051459115,
        "best_response_value_p1": 0.056631335327265,
        "unilateral_improvement_p0": 0.000431735095625,
        "unilateral_improvement_p1": 0.00107908564049,
        "nash_conv": 0.001510820736115,
        "exploitability": 0.000755410368057
      },
      "current": {
        "on_policy_value_p0": -0.302297738089169,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.302297738089169,
        "unilateral_improvement_p0": 0.468964404755836,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.468964404755836,
        "exploitability": 0.234482202377918
      }
    },
    {
      "iteration": 87300,
      "average": {
        "on_policy_value_p0": -0.055553211595033,
        "best_response_value_p0": -0.055248287961226,
        "best_response_value_p1": 0.056755835515451,
        "unilateral_improvement_p0": 0.000304923633806,
        "unilateral_improvement_p1": 0.001202623920418,
        "nash_conv": 0.001507547554225,
        "exploitability": 0.000753773777112
      },
      "current": {
        "on_policy_value_p0": -0.083659499177259,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.250326165843926,
        "unilateral_improvement_p1": 0.083007167489407,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 87400,
      "average": {
        "on_policy_value_p0": -0.055554236862396,
        "best_response_value_p0": -0.055189015939143,
        "best_response_value_p1": 0.056747593826299,
        "unilateral_improvement_p0": 0.000365220923253,
        "unilateral_improvement_p1": 0.001193356963903,
        "nash_conv": 0.001558577887156,
        "exploitability": 0.000779288943578
      },
      "current": {
        "on_policy_value_p0": 0.010837443853456,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.150789921867984,
        "unilateral_improvement_p0": 0.155829222813211,
        "unilateral_improvement_p1": 0.16162736572144,
        "nash_conv": 0.317456588534651,
        "exploitability": 0.158728294267325
      }
    },
    {
      "iteration": 87500,
      "average": {
        "on_policy_value_p0": -0.055555131082062,
        "best_response_value_p0": -0.055125942778071,
        "best_response_value_p1": 0.05663078586869,
        "unilateral_improvement_p0": 0.000429188303991,
        "unilateral_improvement_p1": 0.001075654786628,
        "nash_conv": 0.001504843090619,
        "exploitability": 0.000752421545309
      },
      "current": {
        "on_policy_value_p0": 0.06942195039881,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.13884390079762,
        "unilateral_improvement_p0": 0.097244716267857,
        "unilateral_improvement_p1": 0.20826585119643,
        "nash_conv": 0.305510567464287,
        "exploitability": 0.152755283732143
      }
    },
    {
      "iteration": 87600,
      "average": {
        "on_policy_value_p0": -0.055555703542654,
        "best_response_value_p0": -0.054950594797498,
        "best_response_value_p1": 0.056464659157698,
        "unilateral_improvement_p0": 0.000605108745156,
        "unilateral_improvement_p1": 0.000908955615044,
        "nash_conv": 0.0015140643602,
        "exploitability": 0.0007570321801
      },
      "current": {
        "on_policy_value_p0": 0.077337802519724,
        "best_response_value_p0": 0.134589247583213,
        "best_response_value_p1": 0.218830443206355,
        "unilateral_improvement_p0": 0.057251445063489,
        "unilateral_improvement_p1": 0.29616824572608,
        "nash_conv": 0.353419690789569,
        "exploitability": 0.176709845394784
      }
    },
    {
      "iteration": 87700,
      "average": {
        "on_policy_value_p0": -0.05555579010969,
        "best_response_value_p0": -0.054756607232206,
        "best_response_value_p1": 0.056243683914927,
        "unilateral_improvement_p0": 0.000799182877483,
        "unilateral_improvement_p1": 0.000687893805238,
        "nash_conv": 0.001487076682721,
        "exploitability": 0.00074353834136
      },
      "current": {
        "on_policy_value_p0": 0.098053413288858,
        "best_response_value_p0": 0.098053413288858,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.431386746622191,
        "nash_conv": 0.431386746622191,
        "exploitability": 0.215693373311096
      }
    },
    {
      "iteration": 87800,
      "average": {
        "on_policy_value_p0": -0.055555429492032,
        "best_response_value_p0": -0.054599240517153,
        "best_response_value_p1": 0.056085170523777,
        "unilateral_improvement_p0": 0.000956188974878,
        "unilateral_improvement_p1": 0.000529741031745,
        "nash_conv": 0.001485930006623,
        "exploitability": 0.000742965003312
      },
      "current": {
        "on_policy_value_p0": 0.068475392126228,
        "best_response_value_p0": 0.068475392126228,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.401808725459561,
        "nash_conv": 0.401808725459561,
        "exploitability": 0.200904362729781
      }
    },
    {
      "iteration": 87900,
      "average": {
        "on_policy_value_p0": -0.055554772391466,
        "best_response_value_p0": -0.054478269631779,
        "best_response_value_p1": 0.056021364869029,
        "unilateral_improvement_p0": 0.001076502759687,
        "unilateral_improvement_p1": 0.000466592477564,
        "nash_conv": 0.00154309523725,
        "exploitability": 0.000771547618625
      },
      "current": {
        "on_policy_value_p0": 0.031993691688227,
        "best_response_value_p0": 0.102679283290213,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.070685591601986,
        "unilateral_improvement_p1": 0.36532702502156,
        "nash_conv": 0.436012616623547,
        "exploitability": 0.218006308311773
      }
    },
    {
      "iteration": 88000,
      "average": {
        "on_policy_value_p0": -0.055553978251984,
        "best_response_value_p0": -0.054405774681537,
        "best_response_value_p1": 0.055957704227133,
        "unilateral_improvement_p0": 0.001148203570447,
        "unilateral_improvement_p1": 0.000403725975149,
        "nash_conv": 0.001551929545596,
        "exploitability": 0.000775964772798
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 88100,
      "average": {
        "on_policy_value_p0": -0.055553060844731,
        "best_response_value_p0": -0.054345720151701,
        "best_response_value_p1": 0.055894188104288,
        "unilateral_improvement_p0": 0.001207340693029,
        "unilateral_improvement_p1": 0.000341127259557,
        "nash_conv": 0.001548467952586,
        "exploitability": 0.000774233976293
      },
      "current": {
        "on_policy_value_p0": -0.030088043874249,
        "best_response_value_p0": 0.151622644729542,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.181710688603791,
        "unilateral_improvement_p1": 0.303245289459085,
        "nash_conv": 0.484955978062876,
        "exploitability": 0.242477989031438
      }
    },
    {
      "iteration": 88200,
      "average": {
        "on_policy_value_p0": -0.055552021581436,
        "best_response_value_p0": -0.054411453791679,
        "best_response_value_p1": 0.055865200350291,
        "unilateral_improvement_p0": 0.001140567789757,
        "unilateral_improvement_p1": 0.000313178768855,
        "nash_conv": 0.001453746558612,
        "exploitability": 0.000726873279306
      },
      "current": {
        "on_policy_value_p0": -0.177231075869547,
        "best_response_value_p0": 0.177231075869547,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.354462151739094,
        "unilateral_improvement_p1": 0.156102257463786,
        "nash_conv": 0.51056440920288,
        "exploitability": 0.25528220460144
      }
    },
    {
      "iteration": 88300,
      "average": {
        "on_policy_value_p0": -0.055551322162284,
        "best_response_value_p0": -0.054630427031558,
        "best_response_value_p1": 0.056179433796478,
        "unilateral_improvement_p0": 0.000920895130726,
        "unilateral_improvement_p1": 0.000628111634194,
        "nash_conv": 0.00154900676492,
        "exploitability": 0.00077450338246
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 88400,
      "average": {
        "on_policy_value_p0": -0.055551346801143,
        "best_response_value_p0": -0.054945701812442,
        "best_response_value_p1": 0.056492956307267,
        "unilateral_improvement_p0": 0.000605644988701,
        "unilateral_improvement_p1": 0.000941609506124,
        "nash_conv": 0.001547254494825,
        "exploitability": 0.000773627247412
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 88500,
      "average": {
        "on_policy_value_p0": -0.05555209174044,
        "best_response_value_p0": -0.055260264107945,
        "best_response_value_p1": 0.056755550645234,
        "unilateral_improvement_p0": 0.000291827632495,
        "unilateral_improvement_p1": 0.001203458904794,
        "nash_conv": 0.001495286537289,
        "exploitability": 0.000747643268644
      },
      "current": {
        "on_policy_value_p0": -0.197182547244676,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.197182547244676,
        "unilateral_improvement_p0": 0.53051588057801,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.53051588057801,
        "exploitability": 0.265257940289005
      }
    },
    {
      "iteration": 88600,
      "average": {
        "on_policy_value_p0": -0.055553256475766,
        "best_response_value_p0": -0.055271373907393,
        "best_response_value_p1": 0.056794420603371,
        "unilateral_improvement_p0": 0.000281882568373,
        "unilateral_improvement_p1": 0.001241164127605,
        "nash_conv": 0.001523046695978,
        "exploitability": 0.000761523347989
      },
      "current": {
        "on_policy_value_p0": 0.058644874353888,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.274688458979445,
        "unilateral_improvement_p1": 0.225311541020555,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 88700,
      "average": {
        "on_policy_value_p0": -0.055554198619184,
        "best_response_value_p0": -0.0550960203394,
        "best_response_value_p1": 0.056630842925921,
        "unilateral_improvement_p0": 0.000458178279784,
        "unilateral_improvement_p1": 0.001076644306738,
        "nash_conv": 0.001534822586522,
        "exploitability": 0.000767411293261
      },
      "current": {
        "on_policy_value_p0": 0.119928636830794,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.21340469650254,
        "unilateral_improvement_p1": 0.28659530349746,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 88800,
      "average": {
        "on_policy_value_p0": -0.055554655968623,
        "best_response_value_p0": -0.054846287583763,
        "best_response_value_p1": 0.056373695286763,
        "unilateral_improvement_p0": 0.00070836838486,
        "unilateral_improvement_p1": 0.00081903931814,
        "nash_conv": 0.001527407703,
        "exploitability": 0.0007637038515
      },
      "current": {
        "on_policy_value_p0": 0.213225385569672,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.259784104472678,
        "unilateral_improvement_p0": 0.120107947763661,
        "unilateral_improvement_p1": 0.47300949004235,
        "nash_conv": 0.593117437806011,
        "exploitability": 0.296558718903005
      }
    },
    {
      "iteration": 88900,
      "average": {
        "on_policy_value_p0": -0.055554478309763,
        "best_response_value_p0": -0.054667736529667,
        "best_response_value_p1": 0.056046320576793,
        "unilateral_improvement_p0": 0.000886741780096,
        "unilateral_improvement_p1": 0.00049184226703,
        "nash_conv": 0.001378584047127,
        "exploitability": 0.000689292023563
      },
      "current": {
        "on_policy_value_p0": 0.099935019707719,
        "best_response_value_p0": 0.17858031868974,
        "best_response_value_p1": 0.342709402035956,
        "unilateral_improvement_p0": 0.078645298982022,
        "unilateral_improvement_p1": 0.442644421743675,
        "nash_conv": 0.521289720725697,
        "exploitability": 0.260644860362848
      }
    },
    {
      "iteration": 89000,
      "average": {
        "on_policy_value_p0": -0.055553871687132,
        "best_response_value_p0": -0.054475772984797,
        "best_response_value_p1": 0.056009779177849,
        "unilateral_improvement_p0": 0.001078098702335,
        "unilateral_improvement_p1": 0.000455907490717,
        "nash_conv": 0.001534006193052,
        "exploitability": 0.000767003096526
      },
      "current": {
        "on_policy_value_p0": 0.002250226894187,
        "best_response_value_p0": 0.14890890563608,
        "best_response_value_p1": 0.433469409666026,
        "unilateral_improvement_p0": 0.146658678741893,
        "unilateral_improvement_p1": 0.435719636560212,
        "nash_conv": 0.582378315302106,
        "exploitability": 0.291189157651053
      }
    },
    {
      "iteration": 89100,
      "average": {
        "on_policy_value_p0": -0.055553118830497,
        "best_response_value_p0": -0.054411722939032,
        "best_response_value_p1": 0.055955687147202,
        "unilateral_improvement_p0": 0.001141395891465,
        "unilateral_improvement_p1": 0.000402568316705,
        "nash_conv": 0.00154396420817,
        "exploitability": 0.000771982104085
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 89200,
      "average": {
        "on_policy_value_p0": -0.05555224589034,
        "best_response_value_p0": -0.054502939906594,
        "best_response_value_p1": 0.055972557345029,
        "unilateral_improvement_p0": 0.001049305983746,
        "unilateral_improvement_p1": 0.000420311454688,
        "nash_conv": 0.001469617438434,
        "exploitability": 0.000734808719217
      },
      "current": {
        "on_policy_value_p0": -0.287375712341336,
        "best_response_value_p0": 0.095791904113779,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.383167616455115,
        "unilateral_improvement_p1": 0.212624287658664,
        "nash_conv": 0.595791904113779,
        "exploitability": 0.297895952056889
      }
    },
    {
      "iteration": 89300,
      "average": {
        "on_policy_value_p0": -0.05555201368496,
        "best_response_value_p0": -0.054925799196315,
        "best_response_value_p1": 0.056441497883763,
        "unilateral_improvement_p0": 0.000626214488645,
        "unilateral_improvement_p1": 0.000889484198804,
        "nash_conv": 0.001515698687448,
        "exploitability": 0.000757849343724
      },
      "current": {
        "on_policy_value_p0": -0.391008740582248,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.391008740582248,
        "unilateral_improvement_p0": 0.557675407248915,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.557675407248915,
        "exploitability": 0.278837703624458
      }
    },
    {
      "iteration": 89400,
      "average": {
        "on_policy_value_p0": -0.055552719493437,
        "best_response_value_p0": -0.055183132108259,
        "best_response_value_p1": 0.0567102880372,
        "unilateral_improvement_p0": 0.000369587385177,
        "unilateral_improvement_p1": 0.001157568543764,
        "nash_conv": 0.001527155928941,
        "exploitability": 0.00076357796447
      },
      "current": {
        "on_policy_value_p0": -0.164252522213854,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.164252522213854,
        "unilateral_improvement_p0": 0.330919188880521,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.330919188880521,
        "exploitability": 0.16545959444026
      }
    },
    {
      "iteration": 89500,
      "average": {
        "on_policy_value_p0": -0.055553702887921,
        "best_response_value_p0": -0.055228582271075,
        "best_response_value_p1": 0.05674844952792,
        "unilateral_improvement_p0": 0.000325120616847,
        "unilateral_improvement_p1": 0.001194746639998,
        "nash_conv": 0.001519867256845,
        "exploitability": 0.000759933628423
      },
      "current": {
        "on_policy_value_p0": -0.057146521535099,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.223813188201765,
        "unilateral_improvement_p1": 0.109520145131568,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 89600,
      "average": {
        "on_policy_value_p0": -0.055554655600191,
        "best_response_value_p0": -0.055166943228362,
        "best_response_value_p1": 0.056691109278385,
        "unilateral_improvement_p0": 0.000387712371829,
        "unilateral_improvement_p1": 0.001136453678194,
        "nash_conv": 0.001524166050023,
        "exploitability": 0.000762083025012
      },
      "current": {
        "on_policy_value_p0": 0.043678790540235,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.122987876126431,
        "unilateral_improvement_p0": 0.122987876126431,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.289654542793098,
        "exploitability": 0.144827271396549
      }
    },
    {
      "iteration": 89700,
      "average": {
        "on_policy_value_p0": -0.055555415242451,
        "best_response_value_p0": -0.055093815485603,
        "best_response_value_p1": 0.056554814272808,
        "unilateral_improvement_p0": 0.000461599756849,
        "unilateral_improvement_p1": 0.000999399030357,
        "nash_conv": 0.001460998787206,
        "exploitability": 0.000730499393603
      },
      "current": {
        "on_policy_value_p0": 0.085721714698605,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.17144342939721,
        "unilateral_improvement_p0": 0.080944951968062,
        "unilateral_improvement_p1": 0.257165144095815,
        "nash_conv": 0.338110096063877,
        "exploitability": 0.169055048031938
      }
    },
    {
      "iteration": 89800,
      "average": {
        "on_policy_value_p0": -0.055555781702889,
        "best_response_value_p0": -0.054876796014084,
        "best_response_value_p1": 0.056373783391382,
        "unilateral_improvement_p0": 0.000678985688805,
        "unilateral_improvement_p1": 0.000818001688492,
        "nash_conv": 0.001496987377298,
        "exploitability": 0.000748493688649
      },
      "current": {
        "on_policy_value_p0": 0.079969851952628,
        "best_response_value_p0": 0.116969158345232,
        "best_response_value_p1": 0.259334720548126,
        "unilateral_improvement_p0": 0.036999306392604,
        "unilateral_improvement_p1": 0.339304572500754,
        "nash_conv": 0.376303878893357,
        "exploitability": 0.188151939446679
      }
    },
    {
      "iteration": 89900,
      "average": {
        "on_policy_value_p0": -0.055555680187933,
        "best_response_value_p0": -0.054703496787557,
        "best_response_value_p1": 0.056136917567741,
        "unilateral_improvement_p0": 0.000852183400376,
        "unilateral_improvement_p1": 0.000581237379808,
        "nash_conv": 0.001433420780184,
        "exploitability": 0.000716710390092
      },
      "current": {
        "on_policy_value_p0": 0.085735131112745,
        "best_response_value_p0": 0.085735131112745,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.419068464446078,
        "nash_conv": 0.419068464446078,
        "exploitability": 0.209534232223039
      }
    },
    {
      "iteration": 90000,
      "average": {
        "on_policy_value_p0": -0.055555213162813,
        "best_response_value_p0": -0.054563618791352,
        "best_response_value_p1": 0.056059090109674,
        "unilateral_improvement_p0": 0.000991594371462,
        "unilateral_improvement_p1": 0.00050387694686,
        "nash_conv": 0.001495471318322,
        "exploitability": 0.000747735659161
      },
      "current": {
        "on_policy_value_p0": 0.055554233687062,
        "best_response_value_p0": 0.055558199292542,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.00000396560548,
        "unilateral_improvement_p1": 0.388887567020396,
        "nash_conv": 0.388891532625875,
        "exploitability": 0.194445766312938
      }
    },
    {
      "iteration": 90100,
      "average": {
        "on_policy_value_p0": -0.055554519877486,
        "best_response_value_p0": -0.054463502125298,
        "best_response_value_p1": 0.055996871363715,
        "unilateral_improvement_p0": 0.001091017752188,
        "unilateral_improvement_p1": 0.000442351486229,
        "nash_conv": 0.001533369238416,
        "exploitability": 0.000766684619208
      },
      "current": {
        "on_policy_value_p0": 0.009977278811495,
        "best_response_value_p0": 0.146712109043677,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.136734830232182,
        "unilateral_improvement_p1": 0.343310612144828,
        "nash_conv": 0.48004544237701,
        "exploitability": 0.240022721188505
      }
    },
    {
      "iteration": 90200,
      "average": {
        "on_policy_value_p0": -0.055553730586331,
        "best_response_value_p0": -0.05440229734331,
        "best_response_value_p1": 0.055934790575064,
        "unilateral_improvement_p0": 0.001151433243021,
        "unilateral_improvement_p1": 0.000381059988732,
        "nash_conv": 0.001532493231754,
        "exploitability": 0.000766246615877
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 90300,
      "average": {
        "on_policy_value_p0": -0.055552794339485,
        "best_response_value_p0": -0.054366212358558,
        "best_response_value_p1": 0.055872847285391,
        "unilateral_improvement_p0": 0.001186581980927,
        "unilateral_improvement_p1": 0.000320052945906,
        "nash_conv": 0.001506634926833,
        "exploitability": 0.000753317463417
      },
      "current": {
        "on_policy_value_p0": -0.091793542988574,
        "best_response_value_p0": 0.12076989517238,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.212563438160954,
        "unilateral_improvement_p1": 0.241539790344759,
        "nash_conv": 0.454103228505713,
        "exploitability": 0.227051614252857
      }
    },
    {
      "iteration": 90400,
      "average": {
        "on_policy_value_p0": -0.055551845253252,
        "best_response_value_p0": -0.054476800196765,
        "best_response_value_p1": 0.055960170239806,
        "unilateral_improvement_p0": 0.001075045056487,
        "unilateral_improvement_p1": 0.000408324986554,
        "nash_conv": 0.001483370043041,
        "exploitability": 0.000741685021521
      },
      "current": {
        "on_policy_value_p0": -0.214824853954446,
        "best_response_value_p0": 0.214824853954446,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.429649707908892,
        "unilateral_improvement_p1": 0.118508479378887,
        "nash_conv": 0.548158187287779,
        "exploitability": 0.27407909364389
      }
    },
    {
      "iteration": 90500,
      "average": {
        "on_policy_value_p0": -0.055551374063045,
        "best_response_value_p0": -0.054736949182056,
        "best_response_value_p1": 0.056266659922783,
        "unilateral_improvement_p0": 0.000814424880989,
        "unilateral_improvement_p1": 0.000715285859738,
        "nash_conv": 0.001529710740727,
        "exploitability": 0.000764855370364
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 90600,
      "average": {
        "on_policy_value_p0": -0.055551603172325,
        "best_response_value_p0": -0.055044450709816,
        "best_response_value_p1": 0.056572473028093,
        "unilateral_improvement_p0": 0.000507152462509,
        "unilateral_improvement_p1": 0.001020869855768,
        "nash_conv": 0.001528022318277,
        "exploitability": 0.000764011159138
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 90700,
      "average": {
        "on_policy_value_p0": -0.055552499699066,
        "best_response_value_p0": -0.05532701298345,
        "best_response_value_p1": 0.056782700231581,
        "unilateral_improvement_p0": 0.000225486715616,
        "unilateral_improvement_p1": 0.001230200532515,
        "nash_conv": 0.001455687248131,
        "exploitability": 0.000727843624066
      },
      "current": {
        "on_policy_value_p0": -0.156636320266906,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.156636320266906,
        "unilateral_improvement_p0": 0.489969653600239,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.489969653600239,
        "exploitability": 0.24498482680012
      }
    },
    {
      "iteration": 90800,
      "average": {
        "on_policy_value_p0": -0.055553628796331,
        "best_response_value_p0": -0.055266080149768,
        "best_response_value_p1": 0.056749024324187,
        "unilateral_improvement_p0": 0.000287548646562,
        "unilateral_improvement_p1": 0.001195395527856,
        "nash_conv": 0.001482944174418,
        "exploitability": 0.000741472087209
      },
      "current": {
        "on_policy_value_p0": 0.075879531175731,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.257453802157602,
        "unilateral_improvement_p1": 0.242546197842398,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 90900,
      "average": {
        "on_policy_value_p0": -0.055554445777623,
        "best_response_value_p0": -0.055043256287532,
        "best_response_value_p1": 0.056569926939298,
        "unilateral_improvement_p0": 0.000511189490091,
        "unilateral_improvement_p1": 0.001015481161675,
        "nash_conv": 0.001526670651765,
        "exploitability": 0.000763335325883
      },
      "current": {
        "on_policy_value_p0": 0.143913403876423,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166639301139271,
        "unilateral_improvement_p0": 0.18941992945691,
        "unilateral_improvement_p1": 0.310552705015695,
        "nash_conv": 0.499972634472605,
        "exploitability": 0.249986317236302
      }
    },
    {
      "iteration": 91000,
      "average": {
        "on_policy_value_p0": -0.055554756847673,
        "best_response_value_p0": -0.054802871716833,
        "best_response_value_p1": 0.056288259142945,
        "unilateral_improvement_p0": 0.00075188513084,
        "unilateral_improvement_p1": 0.000733502295272,
        "nash_conv": 0.001485387426112,
        "exploitability": 0.000742693713056
      },
      "current": {
        "on_policy_value_p0": 0.188242174580748,
        "best_response_value_p0": 0.293250989392669,
        "best_response_value_p1": 0.289982370376159,
        "unilateral_improvement_p0": 0.10500881481192,
        "unilateral_improvement_p1": 0.478224544956908,
        "nash_conv": 0.583233359768828,
        "exploitability": 0.291616679884414
      }
    },
    {
      "iteration": 91100,
      "average": {
        "on_policy_value_p0": -0.055554442330607,
        "best_response_value_p0": -0.054607769657627,
        "best_response_value_p1": 0.055997180683189,
        "unilateral_improvement_p0": 0.00094667267298,
        "unilateral_improvement_p1": 0.000442738352582,
        "nash_conv": 0.001389411025562,
        "exploitability": 0.000694705512781
      },
      "current": {
        "on_policy_value_p0": 0.087257679712749,
        "best_response_value_p0": 0.153375406852981,
        "best_response_value_p1": 0.367764545719537,
        "unilateral_improvement_p0": 0.066117727140232,
        "unilateral_improvement_p1": 0.455022225432286,
        "nash_conv": 0.521139952572518,
        "exploitability": 0.260569976286259
      }
    },
    {
      "iteration": 91200,
      "average": {
        "on_policy_value_p0": -0.055553785044153,
        "best_response_value_p0": -0.05445641180651,
        "best_response_value_p1": 0.055982269974603,
        "unilateral_improvement_p0": 0.001097373237643,
        "unilateral_improvement_p1": 0.00042848493045,
        "nash_conv": 0.001525858168093,
        "exploitability": 0.000762929084046
      },
      "current": {
        "on_policy_value_p0": -0.011052111050062,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.477895777899876,
        "unilateral_improvement_p0": 0.177718777716729,
        "unilateral_improvement_p1": 0.466843666849814,
        "nash_conv": 0.644562444566543,
        "exploitability": 0.322281222283271
      }
    },
    {
      "iteration": 91300,
      "average": {
        "on_policy_value_p0": -0.055552986399312,
        "best_response_value_p0": -0.054400331403176,
        "best_response_value_p1": 0.055921732651478,
        "unilateral_improvement_p0": 0.001152654996136,
        "unilateral_improvement_p1": 0.000368746252166,
        "nash_conv": 0.001521401248302,
        "exploitability": 0.000760700624151
      },
      "current": {
        "on_policy_value_p0": -0.06367769266737,
        "best_response_value_p0": 0.124214871555087,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.187892564222457,
        "unilateral_improvement_p1": 0.43632230733263,
        "nash_conv": 0.624214871555087,
        "exploitability": 0.312107435777543
      }
    },
    {
      "iteration": 91400,
      "average": {
        "on_policy_value_p0": -0.055552176508951,
        "best_response_value_p0": -0.054586540322551,
        "best_response_value_p1": 0.056091166975406,
        "unilateral_improvement_p0": 0.000965636186399,
        "unilateral_improvement_p1": 0.000538990466456,
        "nash_conv": 0.001504626652855,
        "exploitability": 0.000752313326428
      },
      "current": {
        "on_policy_value_p0": -0.37007316302539,
        "best_response_value_p0": 0.123357721008463,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.493430884033853,
        "unilateral_improvement_p1": 0.12992683697461,
        "nash_conv": 0.623357721008463,
        "exploitability": 0.311678860504232
      }
    },
    {
      "iteration": 91500,
      "average": {
        "on_policy_value_p0": -0.05555221427384,
        "best_response_value_p0": -0.055051279124957,
        "best_response_value_p1": 0.056511148608874,
        "unilateral_improvement_p0": 0.000500935148882,
        "unilateral_improvement_p1": 0.000958934335034,
        "nash_conv": 0.001459869483916,
        "exploitability": 0.000729934741958
      },
      "current": {
        "on_policy_value_p0": -0.34842908740454,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.34842908740454,
        "unilateral_improvement_p0": 0.515095754071206,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.515095754071206,
        "exploitability": 0.257547877035603
      }
    },
    {
      "iteration": 91600,
      "average": {
        "on_policy_value_p0": -0.055552997954112,
        "best_response_value_p0": -0.055208575104749,
        "best_response_value_p1": 0.056707246162833,
        "unilateral_improvement_p0": 0.000344422849363,
        "unilateral_improvement_p1": 0.00115424820872,
        "nash_conv": 0.001498671058083,
        "exploitability": 0.000749335529042
      },
      "current": {
        "on_policy_value_p0": -0.103166235956096,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.269832902622762,
        "unilateral_improvement_p1": 0.063500430710571,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 91700,
      "average": {
        "on_policy_value_p0": -0.055553941692607,
        "best_response_value_p0": -0.055213461193042,
        "best_response_value_p1": 0.056726350713641,
        "unilateral_improvement_p0": 0.000340480499565,
        "unilateral_improvement_p1": 0.001172409021034,
        "nash_conv": 0.001512889520599,
        "exploitability": 0.0007564447603
      },
      "current": {
        "on_policy_value_p0": -0.042253183566021,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.208919850232688,
        "unilateral_improvement_p1": 0.124413483100646,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 91800,
      "average": {
        "on_policy_value_p0": -0.055554820246423,
        "best_response_value_p0": -0.055153315810479,
        "best_response_value_p1": 0.0566513130488,
        "unilateral_improvement_p0": 0.000401504435944,
        "unilateral_improvement_p1": 0.001096492802376,
        "nash_conv": 0.00149799723832,
        "exploitability": 0.00074899861916
      },
      "current": {
        "on_policy_value_p0": 0.050882151599714,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.115784515066953,
        "unilateral_improvement_p0": 0.115784515066953,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.28245118173362,
        "exploitability": 0.14122559086681
      }
    },
    {
      "iteration": 91900,
      "average": {
        "on_policy_value_p0": -0.055555491589029,
        "best_response_value_p0": -0.055058012169921,
        "best_response_value_p1": 0.056512043512715,
        "unilateral_improvement_p0": 0.000497479419108,
        "unilateral_improvement_p1": 0.000956551923686,
        "nash_conv": 0.001454031342794,
        "exploitability": 0.000727015671397
      },
      "current": {
        "on_policy_value_p0": 0.080732918326917,
        "best_response_value_p0": 0.156550807144167,
        "best_response_value_p1": 0.181697555698833,
        "unilateral_improvement_p0": 0.07581788881725,
        "unilateral_improvement_p1": 0.26243047402575,
        "nash_conv": 0.338248362843,
        "exploitability": 0.1691241814215
      }
    },
    {
      "iteration": 92000,
      "average": {
        "on_policy_value_p0": -0.055555766532735,
        "best_response_value_p0": -0.05485520083493,
        "best_response_value_p1": 0.056329142531368,
        "unilateral_improvement_p0": 0.000700565697805,
        "unilateral_improvement_p1": 0.000773375998633,
        "nash_conv": 0.001473941696437,
        "exploitability": 0.000736970848219
      },
      "current": {
        "on_policy_value_p0": 0.081649579175135,
        "best_response_value_p0": 0.111308100763282,
        "best_response_value_p1": 0.27401629015704,
        "unilateral_improvement_p0": 0.029658521588146,
        "unilateral_improvement_p1": 0.355665869332176,
        "nash_conv": 0.385324390920322,
        "exploitability": 0.192662195460161
      }
    },
    {
      "iteration": 92100,
      "average": {
        "on_policy_value_p0": -0.055555604862893,
        "best_response_value_p0": -0.054691285288777,
        "best_response_value_p1": 0.056105818666782,
        "unilateral_improvement_p0": 0.000864319574116,
        "unilateral_improvement_p1": 0.00055021380389,
        "nash_conv": 0.001414533378006,
        "exploitability": 0.000707266689003
      },
      "current": {
        "on_policy_value_p0": 0.081414904780935,
        "best_response_value_p0": 0.081414904780935,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.414748238114268,
        "nash_conv": 0.414748238114268,
        "exploitability": 0.207374119057134
      }
    },
    {
      "iteration": 92200,
      "average": {
        "on_policy_value_p0": -0.055555115331328,
        "best_response_value_p0": -0.054559527882379,
        "best_response_value_p1": 0.05604496636888,
        "unilateral_improvement_p0": 0.000995587448949,
        "unilateral_improvement_p1": 0.000489851037551,
        "nash_conv": 0.001485438486501,
        "exploitability": 0.00074271924325
      },
      "current": {
        "on_policy_value_p0": 0.050780242987275,
        "best_response_value_p0": 0.065106180692117,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.014325937704842,
        "unilateral_improvement_p1": 0.384113576320608,
        "nash_conv": 0.39843951402545,
        "exploitability": 0.199219757012725
      }
    },
    {
      "iteration": 92300,
      "average": {
        "on_policy_value_p0": -0.055554431544705,
        "best_response_value_p0": -0.05446861148169,
        "best_response_value_p1": 0.055984245928611,
        "unilateral_improvement_p0": 0.001085820063015,
        "unilateral_improvement_p1": 0.000429814383905,
        "nash_conv": 0.001515634446921,
        "exploitability": 0.00075781722346
      },
      "current": {
        "on_policy_value_p0": 0.000697826859428,
        "best_response_value_p0": 0.16527101294781,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.164573186088382,
        "unilateral_improvement_p1": 0.334031160192761,
        "nash_conv": 0.498604346281144,
        "exploitability": 0.249302173140572
      }
    },
    {
      "iteration": 92400,
      "average": {
        "on_policy_value_p0": -0.055553657256041,
        "best_response_value_p0": -0.054409655215726,
        "best_response_value_p1": 0.055923656917866,
        "unilateral_improvement_p0": 0.001144002040315,
        "unilateral_improvement_p1": 0.000369999661825,
        "nash_conv": 0.00151400170214,
        "exploitability": 0.00075700085107
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 92500,
      "average": {
        "on_policy_value_p0": -0.055552737709976,
        "best_response_value_p0": -0.054385367389709,
        "best_response_value_p1": 0.055863198910388,
        "unilateral_improvement_p0": 0.001167370320268,
        "unilateral_improvement_p1": 0.000310461200411,
        "nash_conv": 0.001477831520679,
        "exploitability": 0.000738915760339
      },
      "current": {
        "on_policy_value_p0": -0.106092044851455,
        "best_response_value_p0": 0.113620644240939,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.219712689092394,
        "unilateral_improvement_p1": 0.227241288481878,
        "nash_conv": 0.446953977574272,
        "exploitability": 0.223476988787136
      }
    },
    {
      "iteration": 92600,
      "average": {
        "on_policy_value_p0": -0.055551844669692,
        "best_response_value_p0": -0.054505607551706,
        "best_response_value_p1": 0.055981921323071,
        "unilateral_improvement_p0": 0.001046237117986,
        "unilateral_improvement_p1": 0.000430076653379,
        "nash_conv": 0.001476313771365,
        "exploitability": 0.000738156885682
      },
      "current": {
        "on_policy_value_p0": -0.226371458821819,
        "best_response_value_p0": 0.226371458821819,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.452742917643639,
        "unilateral_improvement_p1": 0.106961874511514,
        "nash_conv": 0.559704792155153,
        "exploitability": 0.279852396077576
      }
    },
    {
      "iteration": 92700,
      "average": {
        "on_policy_value_p0": -0.055551444803808,
        "best_response_value_p0": -0.054769771587302,
        "best_response_value_p1": 0.056281113784786,
        "unilateral_improvement_p0": 0.000781673216506,
        "unilateral_improvement_p1": 0.000729668980979,
        "nash_conv": 0.001511342197484,
        "exploitability": 0.000755671098742
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 92800,
      "average": {
        "on_policy_value_p0": -0.055551711383264,
        "best_response_value_p0": -0.055069947839183,
        "best_response_value_p1": 0.056579661435162,
        "unilateral_improvement_p0": 0.000481763544081,
        "unilateral_improvement_p1": 0.001027950051898,
        "nash_conv": 0.001509713595979,
        "exploitability": 0.00075485679799
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 92900,
      "average": {
        "on_policy_value_p0": -0.055552604624706,
        "best_response_value_p0": -0.055320154657771,
        "best_response_value_p1": 0.056769983728528,
        "unilateral_improvement_p0": 0.000232449966936,
        "unilateral_improvement_p1": 0.001217379103822,
        "nash_conv": 0.001449829070758,
        "exploitability": 0.000724914535379
      },
      "current": {
        "on_policy_value_p0": -0.14591305603068,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.14591305603068,
        "unilateral_improvement_p0": 0.479246389364014,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.479246389364014,
        "exploitability": 0.239623194682007
      }
    },
    {
      "iteration": 93000,
      "average": {
        "on_policy_value_p0": -0.055553676640632,
        "best_response_value_p0": -0.055260670620505,
        "best_response_value_p1": 0.056723552526159,
        "unilateral_improvement_p0": 0.000293006020127,
        "unilateral_improvement_p1": 0.001169875885527,
        "nash_conv": 0.001462881905654,
        "exploitability": 0.000731440952827
      },
      "current": {
        "on_policy_value_p0": 0.077484116556354,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.255849216776979,
        "unilateral_improvement_p1": 0.244150783223021,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 93100,
      "average": {
        "on_policy_value_p0": -0.055554440433201,
        "best_response_value_p0": -0.055038307863083,
        "best_response_value_p1": 0.056547128129708,
        "unilateral_improvement_p0": 0.000516132570119,
        "unilateral_improvement_p1": 0.000992687696507,
        "nash_conv": 0.001508820266625,
        "exploitability": 0.000754410133313
      },
      "current": {
        "on_policy_value_p0": 0.145644117828531,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.187689215504802,
        "unilateral_improvement_p1": 0.312310784495198,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 93200,
      "average": {
        "on_policy_value_p0": -0.055554718215831,
        "best_response_value_p0": -0.054805419531432,
        "best_response_value_p1": 0.056274899989201,
        "unilateral_improvement_p0": 0.000749298684399,
        "unilateral_improvement_p1": 0.000720181773369,
        "nash_conv": 0.001469480457769,
        "exploitability": 0.000734740228884
      },
      "current": {
        "on_policy_value_p0": 0.177643654183835,
        "best_response_value_p0": 0.284973502811602,
        "best_response_value_p1": 0.285340302744465,
        "unilateral_improvement_p0": 0.107329848627768,
        "unilateral_improvement_p1": 0.462983956928299,
        "nash_conv": 0.570313805556067,
        "exploitability": 0.285156902778033
      }
    },
    {
      "iteration": 93300,
      "average": {
        "on_policy_value_p0": -0.055554417698943,
        "best_response_value_p0": -0.054630882707832,
        "best_response_value_p1": 0.056003118868609,
        "unilateral_improvement_p0": 0.000923534991111,
        "unilateral_improvement_p1": 0.000448701169666,
        "nash_conv": 0.001372236160777,
        "exploitability": 0.000686118080389
      },
      "current": {
        "on_policy_value_p0": 0.081605636227794,
        "best_response_value_p0": 0.150389221857765,
        "best_response_value_p1": 0.362432828740058,
        "unilateral_improvement_p0": 0.068783585629971,
        "unilateral_improvement_p1": 0.444038464967852,
        "nash_conv": 0.512822050597823,
        "exploitability": 0.256411025298911
      }
    },
    {
      "iteration": 93400,
      "average": {
        "on_policy_value_p0": -0.055553813873354,
        "best_response_value_p0": -0.054486287410639,
        "best_response_value_p1": 0.05599273483747,
        "unilateral_improvement_p0": 0.001067526462715,
        "unilateral_improvement_p1": 0.000438920964116,
        "nash_conv": 0.001506447426831,
        "exploitability": 0.000753223713416
      },
      "current": {
        "on_policy_value_p0": -0.017541851193993,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.464916297612013,
        "unilateral_improvement_p0": 0.18420851786066,
        "unilateral_improvement_p1": 0.44737444641802,
        "nash_conv": 0.63158296427868,
        "exploitability": 0.31579148213934
      }
    },
    {
      "iteration": 93500,
      "average": {
        "on_policy_value_p0": -0.055553083897416,
        "best_response_value_p0": -0.0544280134497,
        "best_response_value_p1": 0.055934876687428,
        "unilateral_improvement_p0": 0.001125070447715,
        "unilateral_improvement_p1": 0.000381792790012,
        "nash_conv": 0.001506863237728,
        "exploitability": 0.000753431618864
      },
      "current": {
        "on_policy_value_p0": -0.007134487979043,
        "best_response_value_p0": 0.161910341347304,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.169044829326348,
        "unilateral_improvement_p1": 0.492865512020957,
        "nash_conv": 0.661910341347304,
        "exploitability": 0.330955170673652
      }
    },
    {
      "iteration": 93600,
      "average": {
        "on_policy_value_p0": -0.055552297092851,
        "best_response_value_p0": -0.05457552252628,
        "best_response_value_p1": 0.056044954231553,
        "unilateral_improvement_p0": 0.000976774566571,
        "unilateral_improvement_p1": 0.000492657138702,
        "nash_conv": 0.001469431705273,
        "exploitability": 0.000734715852637
      },
      "current": {
        "on_policy_value_p0": -0.335272395305943,
        "best_response_value_p0": 0.111757465101981,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.447029860407924,
        "unilateral_improvement_p1": 0.164727604694057,
        "nash_conv": 0.611757465101981,
        "exploitability": 0.30587873255099
      }
    },
    {
      "iteration": 93700,
      "average": {
        "on_policy_value_p0": -0.055552243033059,
        "best_response_value_p0": -0.055011784809937,
        "best_response_value_p1": 0.056472929861461,
        "unilateral_improvement_p0": 0.000540458223123,
        "unilateral_improvement_p1": 0.000920686828402,
        "nash_conv": 0.001461145051524,
        "exploitability": 0.000730572525762
      },
      "current": {
        "on_policy_value_p0": -0.367384688409948,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.367384688409948,
        "unilateral_improvement_p0": 0.534051355076614,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.534051355076614,
        "exploitability": 0.267025677538307
      }
    },
    {
      "iteration": 93800,
      "average": {
        "on_policy_value_p0": -0.055552974505381,
        "best_response_value_p0": -0.055209614133382,
        "best_response_value_p1": 0.05669672837132,
        "unilateral_improvement_p0": 0.000343360372,
        "unilateral_improvement_p1": 0.001143753865939,
        "nash_conv": 0.001487114237938,
        "exploitability": 0.000743557118969
      },
      "current": {
        "on_policy_value_p0": -0.117377318241812,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.162626660919704,
        "unilateral_improvement_p0": 0.284043984908479,
        "unilateral_improvement_p1": 0.045249342677892,
        "nash_conv": 0.329293327586371,
        "exploitability": 0.164646663793186
      }
    },
    {
      "iteration": 93900,
      "average": {
        "on_policy_value_p0": -0.055553899552981,
        "best_response_value_p0": -0.055232641555101,
        "best_response_value_p1": 0.05672216360564,
        "unilateral_improvement_p0": 0.00032125799788,
        "unilateral_improvement_p1": 0.001168264052659,
        "nash_conv": 0.001489522050539,
        "exploitability": 0.000744761025269
      },
      "current": {
        "on_policy_value_p0": -0.050284660033193,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.216951326699859,
        "unilateral_improvement_p1": 0.116382006633474,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 94000,
      "average": {
        "on_policy_value_p0": -0.055554781582107,
        "best_response_value_p0": -0.055173883425788,
        "best_response_value_p1": 0.056656868350323,
        "unilateral_improvement_p0": 0.000380898156319,
        "unilateral_improvement_p1": 0.001102086768216,
        "nash_conv": 0.001482984924535,
        "exploitability": 0.000741492462268
      },
      "current": {
        "on_policy_value_p0": 0.048526577113495,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.118140089553172,
        "unilateral_improvement_p0": 0.118140089553172,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.284806756219839,
        "exploitability": 0.142403378109919
      }
    },
    {
      "iteration": 94100,
      "average": {
        "on_policy_value_p0": -0.055555473613211,
        "best_response_value_p0": -0.055090692365982,
        "best_response_value_p1": 0.056523217664062,
        "unilateral_improvement_p0": 0.000464781247229,
        "unilateral_improvement_p1": 0.000967744050852,
        "nash_conv": 0.001432525298081,
        "exploitability": 0.00071626264904
      },
      "current": {
        "on_policy_value_p0": 0.088507949279225,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.177015898558451,
        "unilateral_improvement_p0": 0.078158717387441,
        "unilateral_improvement_p1": 0.265523847837677,
        "nash_conv": 0.343682565225118,
        "exploitability": 0.171841282612559
      }
    },
    {
      "iteration": 94200,
      "average": {
        "on_policy_value_p0": -0.055555796123085,
        "best_response_value_p0": -0.054885173828976,
        "best_response_value_p1": 0.056347834763342,
        "unilateral_improvement_p0": 0.000670622294109,
        "unilateral_improvement_p1": 0.000792038640257,
        "nash_conv": 0.001462660934366,
        "exploitability": 0.000731330467183
      },
      "current": {
        "on_policy_value_p0": 0.082396006855555,
        "best_response_value_p0": 0.116530534596025,
        "best_response_value_p1": 0.265064277852393,
        "unilateral_improvement_p0": 0.03413452774047,
        "unilateral_improvement_p1": 0.347460284707948,
        "nash_conv": 0.381594812448418,
        "exploitability": 0.190797406224209
      }
    },
    {
      "iteration": 94300,
      "average": {
        "on_policy_value_p0": -0.055555688160557,
        "best_response_value_p0": -0.054719979970992,
        "best_response_value_p1": 0.056120387476107,
        "unilateral_improvement_p0": 0.000835708189565,
        "unilateral_improvement_p1": 0.00056469931555,
        "nash_conv": 0.001400407505115,
        "exploitability": 0.000700203752558
      },
      "current": {
        "on_policy_value_p0": 0.08606933528182,
        "best_response_value_p0": 0.08606933528182,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.419402668615154,
        "nash_conv": 0.419402668615154,
        "exploitability": 0.209701334307577
      }
    },
    {
      "iteration": 94400,
      "average": {
        "on_policy_value_p0": -0.055555248001261,
        "best_response_value_p0": -0.054585850540412,
        "best_response_value_p1": 0.056043748875106,
        "unilateral_improvement_p0": 0.000969397460849,
        "unilateral_improvement_p1": 0.000488500873845,
        "nash_conv": 0.001457898334694,
        "exploitability": 0.000728949167347
      },
      "current": {
        "on_policy_value_p0": 0.056736278271898,
        "best_response_value_p0": 0.056736278271898,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.390069611605231,
        "nash_conv": 0.390069611605231,
        "exploitability": 0.195034805802616
      }
    },
    {
      "iteration": 94500,
      "average": {
        "on_policy_value_p0": -0.055554596911152,
        "best_response_value_p0": -0.054488088693362,
        "best_response_value_p1": 0.055984443320741,
        "unilateral_improvement_p0": 0.00106650821779,
        "unilateral_improvement_p1": 0.000429846409589,
        "nash_conv": 0.001496354627379,
        "exploitability": 0.000748177313689
      },
      "current": {
        "on_policy_value_p0": 0.013844949388561,
        "best_response_value_p0": 0.138976767889544,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.125131818500983,
        "unilateral_improvement_p1": 0.347178282721894,
        "nash_conv": 0.472310101222878,
        "exploitability": 0.236155050611439
      }
    },
    {
      "iteration": 94600,
      "average": {
        "on_policy_value_p0": -0.055553857673502,
        "best_response_value_p0": -0.054428904072414,
        "best_response_value_p1": 0.055925263148098,
        "unilateral_improvement_p0": 0.001124953601088,
        "unilateral_improvement_p1": 0.000371405474596,
        "nash_conv": 0.001496359075684,
        "exploitability": 0.000748179537842
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 94700,
      "average": {
        "on_policy_value_p0": -0.055552988084358,
        "best_response_value_p0": -0.054389424687435,
        "best_response_value_p1": 0.05586620795998,
        "unilateral_improvement_p0": 0.001163563396923,
        "unilateral_improvement_p1": 0.000313219875622,
        "nash_conv": 0.001476783272545,
        "exploitability": 0.000738391636273
      },
      "current": {
        "on_policy_value_p0": -0.082396968055821,
        "best_response_value_p0": 0.125468182638756,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.207865150694577,
        "unilateral_improvement_p1": 0.250936365277512,
        "nash_conv": 0.458801515972089,
        "exploitability": 0.229400757986045
      }
    },
    {
      "iteration": 94800,
      "average": {
        "on_policy_value_p0": -0.055552083395525,
        "best_response_value_p0": -0.054485518270017,
        "best_response_value_p1": 0.055922756217463,
        "unilateral_improvement_p0": 0.001066565125508,
        "unilateral_improvement_p1": 0.000370672821938,
        "nash_conv": 0.001437237947446,
        "exploitability": 0.000718618973723
      },
      "current": {
        "on_policy_value_p0": -0.204278442741363,
        "best_response_value_p0": 0.204278442741364,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.408556885482727,
        "unilateral_improvement_p1": 0.12905489059197,
        "nash_conv": 0.537611776074697,
        "exploitability": 0.268805888037348
      }
    },
    {
      "iteration": 94900,
      "average": {
        "on_policy_value_p0": -0.055551577995597,
        "best_response_value_p0": -0.054721256372218,
        "best_response_value_p1": 0.056215075055309,
        "unilateral_improvement_p0": 0.000830321623379,
        "unilateral_improvement_p1": 0.000663497059713,
        "nash_conv": 0.001493818683092,
        "exploitability": 0.000746909341546
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 95000,
      "average": {
        "on_policy_value_p0": -0.055551710367058,
        "best_response_value_p0": -0.055014532242703,
        "best_response_value_p1": 0.056506778485076,
        "unilateral_improvement_p0": 0.000537178124356,
        "unilateral_improvement_p1": 0.000955068118017,
        "nash_conv": 0.001492246242373,
        "exploitability": 0.000746123121187
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 95100,
      "average": {
        "on_policy_value_p0": -0.055552462455793,
        "best_response_value_p0": -0.055307191339538,
        "best_response_value_p1": 0.056731722751665,
        "unilateral_improvement_p0": 0.000245271116255,
        "unilateral_improvement_p1": 0.001179260295873,
        "nash_conv": 0.001424531412128,
        "exploitability": 0.000712265706064
      },
      "current": {
        "on_policy_value_p0": -0.179455445809477,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.179455445809477,
        "unilateral_improvement_p0": 0.51278877914281,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.51278877914281,
        "exploitability": 0.256394389571405
      }
    },
    {
      "iteration": 95200,
      "average": {
        "on_policy_value_p0": -0.055553516816173,
        "best_response_value_p0": -0.055279223705183,
        "best_response_value_p1": 0.056742281204006,
        "unilateral_improvement_p0": 0.000274293110989,
        "unilateral_improvement_p1": 0.001188764387834,
        "nash_conv": 0.001463057498823,
        "exploitability": 0.000731528749411
      },
      "current": {
        "on_policy_value_p0": 0.064329281992516,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.269004051340817,
        "unilateral_improvement_p1": 0.230995948659183,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 95300,
      "average": {
        "on_policy_value_p0": -0.055554344831936,
        "best_response_value_p0": -0.05510236692451,
        "best_response_value_p1": 0.056585447711632,
        "unilateral_improvement_p0": 0.000451977907426,
        "unilateral_improvement_p1": 0.001031102879696,
        "nash_conv": 0.001483080787121,
        "exploitability": 0.000741540393561
      },
      "current": {
        "on_policy_value_p0": 0.124111790760576,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.209221542572757,
        "unilateral_improvement_p1": 0.290778457427243,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 95400,
      "average": {
        "on_policy_value_p0": -0.055554744302714,
        "best_response_value_p0": -0.05486990462515,
        "best_response_value_p1": 0.056340882960346,
        "unilateral_improvement_p0": 0.000684839677564,
        "unilateral_improvement_p1": 0.000786138657633,
        "nash_conv": 0.001470978335197,
        "exploitability": 0.000735489167598
      },
      "current": {
        "on_policy_value_p0": 0.21548603143979,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.264305396212914,
        "unilateral_improvement_p0": 0.117847301893543,
        "unilateral_improvement_p1": 0.479791427652705,
        "nash_conv": 0.597638729546248,
        "exploitability": 0.298819364773124
      }
    },
    {
      "iteration": 95500,
      "average": {
        "on_policy_value_p0": -0.055554584362123,
        "best_response_value_p0": -0.054699249175409,
        "best_response_value_p1": 0.056034908568169,
        "unilateral_improvement_p0": 0.000855335186714,
        "unilateral_improvement_p1": 0.000480324206046,
        "nash_conv": 0.00133565939276,
        "exploitability": 0.00066782969638
      },
      "current": {
        "on_policy_value_p0": 0.103798681676477,
        "best_response_value_p0": 0.182036330513759,
        "best_response_value_p1": 0.343524702325437,
        "unilateral_improvement_p0": 0.078237648837282,
        "unilateral_improvement_p1": 0.447323384001914,
        "nash_conv": 0.525561032839196,
        "exploitability": 0.262780516419598
      }
    },
    {
      "iteration": 95600,
      "average": {
        "on_policy_value_p0": -0.055554037653913,
        "best_response_value_p0": -0.054513364522542,
        "best_response_value_p1": 0.055989240060132,
        "unilateral_improvement_p0": 0.00104067313137,
        "unilateral_improvement_p1": 0.000435202406219,
        "nash_conv": 0.001475875537589,
        "exploitability": 0.000737937768795
      },
      "current": {
        "on_policy_value_p0": 0.015165217545556,
        "best_response_value_p0": 0.141716506232708,
        "best_response_value_p1": 0.430529793355277,
        "unilateral_improvement_p0": 0.126551288687152,
        "unilateral_improvement_p1": 0.445695010900833,
        "nash_conv": 0.572246299587985,
        "exploitability": 0.286123149793993
      }
    },
    {
      "iteration": 95700,
      "average": {
        "on_policy_value_p0": -0.055553354394471,
        "best_response_value_p0": -0.05445051353528,
        "best_response_value_p1": 0.055940101142939,
        "unilateral_improvement_p0": 0.001102840859191,
        "unilateral_improvement_p1": 0.000386746748468,
        "nash_conv": 0.001489587607659,
        "exploitability": 0.00074479380383
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 95800,
      "average": {
        "on_policy_value_p0": -0.055552553380847,
        "best_response_value_p0": -0.054512907883011,
        "best_response_value_p1": 0.05590781708806,
        "unilateral_improvement_p0": 0.001039645497836,
        "unilateral_improvement_p1": 0.000355263707212,
        "nash_conv": 0.001394909205048,
        "exploitability": 0.000697454602524
      },
      "current": {
        "on_policy_value_p0": -0.262582514569029,
        "best_response_value_p0": 0.087527504856343,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.350110019425371,
        "unilateral_improvement_p1": 0.237417485430971,
        "nash_conv": 0.587527504856343,
        "exploitability": 0.293763752428171
      }
    },
    {
      "iteration": 95900,
      "average": {
        "on_policy_value_p0": -0.055552225661528,
        "best_response_value_p0": -0.054881158213334,
        "best_response_value_p1": 0.056356843407462,
        "unilateral_improvement_p0": 0.000671067448194,
        "unilateral_improvement_p1": 0.000804617745934,
        "nash_conv": 0.001475685194128,
        "exploitability": 0.000737842597064
      },
      "current": {
        "on_policy_value_p0": -0.416037188037152,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.416037188037152,
        "unilateral_improvement_p0": 0.582703854703819,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.582703854703819,
        "exploitability": 0.291351927351909
      }
    },
    {
      "iteration": 96000,
      "average": {
        "on_policy_value_p0": -0.055552762130658,
        "best_response_value_p0": -0.055171446894279,
        "best_response_value_p1": 0.056640925585392,
        "unilateral_improvement_p0": 0.00038131523638,
        "unilateral_improvement_p1": 0.001088163454733,
        "nash_conv": 0.001469478691113,
        "exploitability": 0.000734739345557
      },
      "current": {
        "on_policy_value_p0": -0.217706073304923,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.217706073304923,
        "unilateral_improvement_p0": 0.38437273997159,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.38437273997159,
        "exploitability": 0.192186369985795
      }
    },
    {
      "iteration": 96100,
      "average": {
        "on_policy_value_p0": -0.055553625854315,
        "best_response_value_p0": -0.055249262448098,
        "best_response_value_p1": 0.056701203298229,
        "unilateral_improvement_p0": 0.000304363406217,
        "unilateral_improvement_p1": 0.001147577443914,
        "nash_conv": 0.001451940850132,
        "exploitability": 0.000725970425066
      },
      "current": {
        "on_policy_value_p0": -0.070269945039792,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.236936611706459,
        "unilateral_improvement_p1": 0.096396721626875,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 96200,
      "average": {
        "on_policy_value_p0": -0.055554493926933,
        "best_response_value_p0": -0.055191830782352,
        "best_response_value_p1": 0.056673371290241,
        "unilateral_improvement_p0": 0.000362663144581,
        "unilateral_improvement_p1": 0.001118877363308,
        "nash_conv": 0.001481540507889,
        "exploitability": 0.000740770253945
      },
      "current": {
        "on_policy_value_p0": 0.027400602886345,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.139266063780321,
        "unilateral_improvement_p0": 0.139266063780321,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.305932730446988,
        "exploitability": 0.152966365223494
      }
    },
    {
      "iteration": 96300,
      "average": {
        "on_policy_value_p0": -0.055555233145019,
        "best_response_value_p0": -0.055134518393171,
        "best_response_value_p1": 0.056560103555806,
        "unilateral_improvement_p0": 0.000420714751848,
        "unilateral_improvement_p1": 0.001004870410787,
        "nash_conv": 0.001425585162635,
        "exploitability": 0.000712792581318
      },
      "current": {
        "on_policy_value_p0": 0.07370894027912,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.147417880558239,
        "unilateral_improvement_p0": 0.092957726387547,
        "unilateral_improvement_p1": 0.221126820837359,
        "nash_conv": 0.314084547224906,
        "exploitability": 0.157042273612453
      }
    },
    {
      "iteration": 96400,
      "average": {
        "on_policy_value_p0": -0.055555683864413,
        "best_response_value_p0": -0.054963505579804,
        "best_response_value_p1": 0.056405717868524,
        "unilateral_improvement_p0": 0.00059217828461,
        "unilateral_improvement_p1": 0.00085003400411,
        "nash_conv": 0.00144221228872,
        "exploitability": 0.00072110614436
      },
      "current": {
        "on_policy_value_p0": 0.074803448686548,
        "best_response_value_p0": 0.129313625403334,
        "best_response_value_p1": 0.224312979899761,
        "unilateral_improvement_p0": 0.054510176716786,
        "unilateral_improvement_p1": 0.29911642858631,
        "nash_conv": 0.353626605303096,
        "exploitability": 0.176813302651548
      }
    },
    {
      "iteration": 96500,
      "average": {
        "on_policy_value_p0": -0.055555736470346,
        "best_response_value_p0": -0.054790714023472,
        "best_response_value_p1": 0.05620271414147,
        "unilateral_improvement_p0": 0.000765022446874,
        "unilateral_improvement_p1": 0.000646977671124,
        "nash_conv": 0.001412000117998,
        "exploitability": 0.000706000058999
      },
      "current": {
        "on_policy_value_p0": 0.095946537399441,
        "best_response_value_p0": 0.095946537399441,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.429279870732775,
        "nash_conv": 0.429279870732775,
        "exploitability": 0.214639935366387
      }
    },
    {
      "iteration": 96600,
      "average": {
        "on_policy_value_p0": -0.055555426203372,
        "best_response_value_p0": -0.054649061412944,
        "best_response_value_p1": 0.056067029309107,
        "unilateral_improvement_p0": 0.000906364790428,
        "unilateral_improvement_p1": 0.000511603105735,
        "nash_conv": 0.001417967896163,
        "exploitability": 0.000708983948081
      },
      "current": {
        "on_policy_value_p0": 0.067787151869598,
        "best_response_value_p0": 0.067787151869598,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.401120485202932,
        "nash_conv": 0.401120485202932,
        "exploitability": 0.200560242601466
      }
    },
    {
      "iteration": 96700,
      "average": {
        "on_policy_value_p0": -0.05555487304082,
        "best_response_value_p0": -0.054538959489372,
        "best_response_value_p1": 0.056009048927196,
        "unilateral_improvement_p0": 0.001015913551447,
        "unilateral_improvement_p1": 0.000454175886376,
        "nash_conv": 0.001470089437823,
        "exploitability": 0.000735044718912
      },
      "current": {
        "on_policy_value_p0": 0.033131582007632,
        "best_response_value_p0": 0.100403502651403,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.067271920643771,
        "unilateral_improvement_p1": 0.366464915340965,
        "nash_conv": 0.433736835984736,
        "exploitability": 0.216868417992368
      }
    },
    {
      "iteration": 96800,
      "average": {
        "on_policy_value_p0": -0.055554205171689,
        "best_response_value_p0": -0.054471660749922,
        "best_response_value_p1": 0.055951188339461,
        "unilateral_improvement_p0": 0.001082544421767,
        "unilateral_improvement_p1": 0.000396983167772,
        "nash_conv": 0.001479527589539,
        "exploitability": 0.00073976379477
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 96900,
      "average": {
        "on_policy_value_p0": -0.055553439242647,
        "best_response_value_p0": -0.054415446445742,
        "best_response_value_p1": 0.055893447175025,
        "unilateral_improvement_p0": 0.001137992796905,
        "unilateral_improvement_p1": 0.000340007932377,
        "nash_conv": 0.001478000729283,
        "exploitability": 0.000739000364641
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 97000,
      "average": {
        "on_policy_value_p0": -0.055552545930426,
        "best_response_value_p0": -0.054441760971188,
        "best_response_value_p1": 0.055835825064536,
        "unilateral_improvement_p0": 0.001110784959238,
        "unilateral_improvement_p1": 0.00028327913411,
        "nash_conv": 0.001394064093347,
        "exploitability": 0.000697032046674
      },
      "current": {
        "on_policy_value_p0": -0.15038995362699,
        "best_response_value_p0": 0.15038995362699,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.30077990725398,
        "unilateral_improvement_p1": 0.182943379706343,
        "nash_conv": 0.483723286960323,
        "exploitability": 0.241861643480162
      }
    },
    {
      "iteration": 97100,
      "average": {
        "on_policy_value_p0": -0.055551824718989,
        "best_response_value_p0": -0.054600788191057,
        "best_response_value_p1": 0.056071295327608,
        "unilateral_improvement_p0": 0.000951036527933,
        "unilateral_improvement_p1": 0.000519470608619,
        "nash_conv": 0.001470507136552,
        "exploitability": 0.000735253568276
      },
      "current": {
        "on_policy_value_p0": -0.280672190203569,
        "best_response_value_p0": 0.280672190203569,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.561344380407138,
        "unilateral_improvement_p1": 0.052661143129765,
        "nash_conv": 0.614005523536902,
        "exploitability": 0.307002761768451
      }
    },
    {
      "iteration": 97200,
      "average": {
        "on_policy_value_p0": -0.055551662642502,
        "best_response_value_p0": -0.054880967757287,
        "best_response_value_p1": 0.056356544337902,
        "unilateral_improvement_p0": 0.000670694885215,
        "unilateral_improvement_p1": 0.0008048816954,
        "nash_conv": 0.001475576580615,
        "exploitability": 0.000737788290308
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 97300,
      "average": {
        "on_policy_value_p0": -0.055552098036601,
        "best_response_value_p0": -0.055167146961373,
        "best_response_value_p1": 0.056634367381463,
        "unilateral_improvement_p0": 0.000384951075228,
        "unilateral_improvement_p1": 0.001082269344862,
        "nash_conv": 0.00146722042009,
        "exploitability": 0.000733610210045
      },
      "current": {
        "on_policy_value_p0": -0.270273240712847,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.270273240712847,
        "unilateral_improvement_p0": 0.603606574046181,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.603606574046181,
        "exploitability": 0.30180328702309
      }
    },
    {
      "iteration": 97400,
      "average": {
        "on_policy_value_p0": -0.055553045888485,
        "best_response_value_p0": -0.0553156397243,
        "best_response_value_p1": 0.05676151416047,
        "unilateral_improvement_p0": 0.000237406164185,
        "unilateral_improvement_p1": 0.001208468271985,
        "nash_conv": 0.00144587443617,
        "exploitability": 0.000722937218085
      },
      "current": {
        "on_policy_value_p0": -0.076156865285986,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.114608423165999,
        "unilateral_improvement_p0": 0.409490198619319,
        "unilateral_improvement_p1": 0.038451557880013,
        "nash_conv": 0.447941756499332,
        "exploitability": 0.223970878249666
      }
    },
    {
      "iteration": 97500,
      "average": {
        "on_policy_value_p0": -0.055554007462315,
        "best_response_value_p0": -0.055229595860882,
        "best_response_value_p1": 0.056664519001585,
        "unilateral_improvement_p0": 0.000324411601433,
        "unilateral_improvement_p1": 0.00111051153927,
        "nash_conv": 0.001434923140703,
        "exploitability": 0.000717461570352
      },
      "current": {
        "on_policy_value_p0": 0.091912641237378,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.241420692095955,
        "unilateral_improvement_p1": 0.258579307904045,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 97600,
      "average": {
        "on_policy_value_p0": -0.055554632911077,
        "best_response_value_p0": -0.055002243132883,
        "best_response_value_p1": 0.056474724176626,
        "unilateral_improvement_p0": 0.000552389778194,
        "unilateral_improvement_p1": 0.000920091265549,
        "nash_conv": 0.001472481043743,
        "exploitability": 0.000736240521871
      },
      "current": {
        "on_policy_value_p0": 0.188006887529019,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.209347108391372,
        "unilateral_improvement_p0": 0.145326445804314,
        "unilateral_improvement_p1": 0.397353995920392,
        "nash_conv": 0.542680441724706,
        "exploitability": 0.271340220862353
      }
    },
    {
      "iteration": 97700,
      "average": {
        "on_policy_value_p0": -0.055554782120209,
        "best_response_value_p0": -0.054792848802494,
        "best_response_value_p1": 0.056197237973426,
        "unilateral_improvement_p0": 0.000761933317715,
        "unilateral_improvement_p1": 0.000642455853217,
        "nash_conv": 0.001404389170932,
        "exploitability": 0.000702194585466
      },
      "current": {
        "on_policy_value_p0": 0.154790814224995,
        "best_response_value_p0": 0.251442010164105,
        "best_response_value_p1": 0.30669760812178,
        "unilateral_improvement_p0": 0.09665119593911,
        "unilateral_improvement_p1": 0.461488422346775,
        "nash_conv": 0.558139618285885,
        "exploitability": 0.279069809142942
      }
    },
    {
      "iteration": 97800,
      "average": {
        "on_policy_value_p0": -0.055554416253899,
        "best_response_value_p0": -0.054605525535315,
        "best_response_value_p1": 0.055985741663056,
        "unilateral_improvement_p0": 0.000948890718584,
        "unilateral_improvement_p1": 0.000431325409158,
        "nash_conv": 0.001380216127742,
        "exploitability": 0.000690108063871
      },
      "current": {
        "on_policy_value_p0": 0.072524211702311,
        "best_response_value_p0": 0.131525531076975,
        "best_response_value_p1": 0.381997361250671,
        "unilateral_improvement_p0": 0.059001319374664,
        "unilateral_improvement_p1": 0.454521572952982,
        "nash_conv": 0.513522892327646,
        "exploitability": 0.256761446163823
      }
    },
    {
      "iteration": 97900,
      "average": {
        "on_policy_value_p0": -0.055553806617116,
        "best_response_value_p0": -0.054489640142445,
        "best_response_value_p1": 0.055962735482087,
        "unilateral_improvement_p0": 0.001064166474671,
        "unilateral_improvement_p1": 0.000408928864971,
        "nash_conv": 0.001473095339642,
        "exploitability": 0.000736547669821
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.5,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 98000,
      "average": {
        "on_policy_value_p0": -0.055553074736838,
        "best_response_value_p0": -0.054446016556261,
        "best_response_value_p1": 0.055905630649963,
        "unilateral_improvement_p0": 0.001107058180576,
        "unilateral_improvement_p1": 0.000352555913125,
        "nash_conv": 0.001459614093702,
        "exploitability": 0.000729807046851
      },
      "current": {
        "on_policy_value_p0": -0.106512552533456,
        "best_response_value_p0": 0.095658298311029,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.202170850844485,
        "unilateral_improvement_p1": 0.393487447466544,
        "nash_conv": 0.595658298311029,
        "exploitability": 0.297829149155515
      }
    },
    {
      "iteration": 98100,
      "average": {
        "on_policy_value_p0": -0.055552377852477,
        "best_response_value_p0": -0.054643829374152,
        "best_response_value_p1": 0.056103608880914,
        "unilateral_improvement_p0": 0.000908548478325,
        "unilateral_improvement_p1": 0.000551231028437,
        "nash_conv": 0.001459779506762,
        "exploitability": 0.000729889753381
      },
      "current": {
        "on_policy_value_p0": -0.392645197197389,
        "best_response_value_p0": 0.130881732399129,
        "best_response_value_p1": 0.5,
        "unilateral_improvement_p0": 0.523526929596518,
        "unilateral_improvement_p1": 0.107354802802611,
        "nash_conv": 0.630881732399129,
        "exploitability": 0.315440866199565
      }
    },
    {
      "iteration": 98200,
      "average": {
        "on_policy_value_p0": -0.055552459849371,
        "best_response_value_p0": -0.055083867961372,
        "best_response_value_p1": 0.05648890978541,
        "unilateral_improvement_p0": 0.000468591887998,
        "unilateral_improvement_p1": 0.000936449936039,
        "nash_conv": 0.001405041824037,
        "exploitability": 0.000702520912019
      },
      "current": {
        "on_policy_value_p0": -0.344836982118935,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.344836982118935,
        "unilateral_improvement_p0": 0.511503648785601,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.511503648785601,
        "exploitability": 0.255751824392801
      }
    },
    {
      "iteration": 98300,
      "average": {
        "on_policy_value_p0": -0.0555531766225,
        "best_response_value_p0": -0.055223320999472,
        "best_response_value_p1": 0.056669646690787,
        "unilateral_improvement_p0": 0.000329855623028,
        "unilateral_improvement_p1": 0.001116470068288,
        "nash_conv": 0.001446325691315,
        "exploitability": 0.000723162845658
      },
      "current": {
        "on_policy_value_p0": -0.103170812560675,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.269837479227342,
        "unilateral_improvement_p1": 0.063495854105991,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 98400,
      "average": {
        "on_policy_value_p0": -0.055554030565993,
        "best_response_value_p0": -0.0552299463086,
        "best_response_value_p1": 0.056688579622812,
        "unilateral_improvement_p0": 0.000324084257393,
        "unilateral_improvement_p1": 0.001134549056819,
        "nash_conv": 0.001458633314212,
        "exploitability": 0.000729316657106
      },
      "current": {
        "on_policy_value_p0": -0.044879579397616,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.211546246064283,
        "unilateral_improvement_p1": 0.12178708726905,
        "nash_conv": 0.333333333333333,
        "exploitability": 0.166666666666667
      }
    },
    {
      "iteration": 98500,
      "average": {
        "on_policy_value_p0": -0.055554832650561,
        "best_response_value_p0": -0.05517387529712,
        "best_response_value_p1": 0.056622322823684,
        "unilateral_improvement_p0": 0.000380957353441,
        "unilateral_improvement_p1": 0.001067490173123,
        "nash_conv": 0.001448447526564,
        "exploitability": 0.000724223763282
      },
      "current": {
        "on_policy_value_p0": 0.048598126428135,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.118068540238531,
        "unilateral_improvement_p0": 0.118068540238531,
        "unilateral_improvement_p1": 0.166666666666667,
        "nash_conv": 0.284735206905198,
        "exploitability": 0.142367603452599
      }
    },
    {
      "iteration": 98600,
      "average": {
        "on_policy_value_p0": -0.055555462572008,
        "best_response_value_p0": -0.055097684219364,
        "best_response_value_p1": 0.056495191740446,
        "unilateral_improvement_p0": 0.000457778352644,
        "unilateral_improvement_p1": 0.000939729168438,
        "nash_conv": 0.001397507521082,
        "exploitability": 0.000698753760541
      },
      "current": {
        "on_policy_value_p0": 0.085655061745006,
        "best_response_value_p0": 0.164622550158957,
        "best_response_value_p1": 0.17539835650543,
        "unilateral_improvement_p0": 0.078967488413952,
        "unilateral_improvement_p1": 0.261053418250435,
        "nash_conv": 0.340020906664387,
        "exploitability": 0.170010453332194
      }
    },
    {
      "iteration": 98700,
      "average": {
        "on_policy_value_p0": -0.055555757217749,
        "best_response_value_p0": -0.054902467950163,
        "best_response_value_p1": 0.056329326245803,
        "unilateral_improvement_p0": 0.000653289267586,
        "unilateral_improvement_p1": 0.000773569028054,
        "nash_conv": 0.00142685829564,
        "exploitability": 0.00071342914782
      },
      "current": {
        "on_policy_value_p0": 0.079665094979889,
        "best_response_value_p0": 0.116347681425997,
        "best_response_value_p1": 0.259968160441118,
        "unilateral_improvement_p0": 0.036682586446107,
        "unilateral_improvement_p1": 0.339633255421007,
        "nash_conv": 0.376315841867115,
        "exploitability": 0.188157920933557
      }
    },
    {
      "iteration": 98800,
      "average": {
        "on_policy_value_p0": -0.055555665742068,
        "best_response_value_p0": -0.054744599353629,
        "best_response_value_p1": 0.056114108395201,
        "unilateral_improvement_p0": 0.000811066388438,
        "unilateral_improvement_p1": 0.000558442653133,
        "nash_conv": 0.001369509041571,
        "exploitability": 0.000684754520786
      },
      "current": {
        "on_policy_value_p0": 0.086577148622056,
        "best_response_value_p0": 0.086577148622056,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.41991048195539,
        "nash_conv": 0.41991048195539,
        "exploitability": 0.209955240977695
      }
    },
    {
      "iteration": 98900,
      "average": {
        "on_policy_value_p0": -0.055555268029001,
        "best_response_value_p0": -0.054615666755369,
        "best_response_value_p1": 0.056037459619136,
        "unilateral_improvement_p0": 0.000939601273632,
        "unilateral_improvement_p1": 0.000482191590135,
        "nash_conv": 0.001421792863767,
        "exploitability": 0.000710896431884
      },
      "current": {
        "on_policy_value_p0": 0.058066669538924,
        "best_response_value_p0": 0.058066669538924,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0,
        "unilateral_improvement_p1": 0.391400002872257,
        "nash_conv": 0.391400002872257,
        "exploitability": 0.195700001436129
      }
    },
    {
      "iteration": 99000,
      "average": {
        "on_policy_value_p0": -0.055554672859417,
        "best_response_value_p0": -0.054520031912758,
        "best_response_value_p1": 0.055980856124572,
        "unilateral_improvement_p0": 0.001034640946659,
        "unilateral_improvement_p1": 0.000426183265155,
        "nash_conv": 0.001460824211814,
        "exploitability": 0.000730412105907
      },
      "current": {
        "on_policy_value_p0": 0.017696834757511,
        "best_response_value_p0": 0.131272997151645,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.113576162394133,
        "unilateral_improvement_p1": 0.351030168090844,
        "nash_conv": 0.464606330484978,
        "exploitability": 0.232303165242489
      }
    },
    {
      "iteration": 99100,
      "average": {
        "on_policy_value_p0": -0.055553995309207,
        "best_response_value_p0": -0.054462397124102,
        "best_response_value_p1": 0.055924366865113,
        "unilateral_improvement_p0": 0.001091598185105,
        "unilateral_improvement_p1": 0.000370371555906,
        "nash_conv": 0.001461969741011,
        "exploitability": 0.000730984870505
      },
      "current": {
        "on_policy_value_p0": 0,
        "best_response_value_p0": 0.166666666666667,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.166666666666667,
        "unilateral_improvement_p1": 0.333333333333333,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 99200,
      "average": {
        "on_policy_value_p0": -0.05555320596468,
        "best_response_value_p0": -0.054415420833476,
        "best_response_value_p1": 0.05586799149529,
        "unilateral_improvement_p0": 0.001137785131204,
        "unilateral_improvement_p1": 0.000314785530609,
        "nash_conv": 0.001452570661814,
        "exploitability": 0.000726285330907
      },
      "current": {
        "on_policy_value_p0": -0.059500462919832,
        "best_response_value_p0": 0.136916435206751,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.196416898126582,
        "unilateral_improvement_p1": 0.273832870413502,
        "nash_conv": 0.470249768540084,
        "exploitability": 0.235124884270042
      }
    },
    {
      "iteration": 99300,
      "average": {
        "on_policy_value_p0": -0.055552344442842,
        "best_response_value_p0": -0.054488958999363,
        "best_response_value_p1": 0.055870906912829,
        "unilateral_improvement_p0": 0.001063385443479,
        "unilateral_improvement_p1": 0.000318562469987,
        "nash_conv": 0.001381947913466,
        "exploitability": 0.000690973956733
      },
      "current": {
        "on_policy_value_p0": -0.18604228491769,
        "best_response_value_p0": 0.186042284917689,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.372084569835379,
        "unilateral_improvement_p1": 0.147291048415644,
        "nash_conv": 0.519375618251023,
        "exploitability": 0.259687809125511
      }
    },
    {
      "iteration": 99400,
      "average": {
        "on_policy_value_p0": -0.055551783427274,
        "best_response_value_p0": -0.054690395809834,
        "best_response_value_p1": 0.056150044162749,
        "unilateral_improvement_p0": 0.00086138761744,
        "unilateral_improvement_p1": 0.000598260735475,
        "nash_conv": 0.001459648352915,
        "exploitability": 0.000729824176457
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 99500,
      "average": {
        "on_policy_value_p0": -0.055551799824264,
        "best_response_value_p0": -0.054970438963124,
        "best_response_value_p1": 0.05642862033277,
        "unilateral_improvement_p0": 0.00058136086114,
        "unilateral_improvement_p1": 0.000876820508507,
        "nash_conv": 0.001458181369646,
        "exploitability": 0.000729090684823
      },
      "current": {
        "on_policy_value_p0": -0.333333333333333,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.333333333333333,
        "unilateral_improvement_p0": 0.666666666666667,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.666666666666667,
        "exploitability": 0.333333333333333
      }
    },
    {
      "iteration": 99600,
      "average": {
        "on_policy_value_p0": -0.055552386863156,
        "best_response_value_p0": -0.055249919780764,
        "best_response_value_p1": 0.056671762677292,
        "unilateral_improvement_p0": 0.000302467082392,
        "unilateral_improvement_p1": 0.001119375814136,
        "nash_conv": 0.001421842896528,
        "exploitability": 0.000710921448264
      },
      "current": {
        "on_policy_value_p0": -0.212477166142674,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.212477166142674,
        "unilateral_improvement_p0": 0.545810499476007,
        "unilateral_improvement_p1": 0,
        "nash_conv": 0.545810499476007,
        "exploitability": 0.272905249738004
      }
    },
    {
      "iteration": 99700,
      "average": {
        "on_policy_value_p0": -0.055553354269197,
        "best_response_value_p0": -0.055296417826219,
        "best_response_value_p1": 0.056737688804307,
        "unilateral_improvement_p0": 0.000256936442978,
        "unilateral_improvement_p1": 0.00118433453511,
        "nash_conv": 0.001441270978088,
        "exploitability": 0.000720635489044
      },
      "current": {
        "on_policy_value_p0": 0.007188510730893,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.146143026280382,
        "unilateral_improvement_p0": 0.32614482260244,
        "unilateral_improvement_p1": 0.153331537011275,
        "nash_conv": 0.479476359613715,
        "exploitability": 0.239738179806857
      }
    },
    {
      "iteration": 99800,
      "average": {
        "on_policy_value_p0": -0.055554199365853,
        "best_response_value_p0": -0.055171516356882,
        "best_response_value_p1": 0.056606171388475,
        "unilateral_improvement_p0": 0.000382683008971,
        "unilateral_improvement_p1": 0.001051972022622,
        "nash_conv": 0.001434655031593,
        "exploitability": 0.000717327515796
      },
      "current": {
        "on_policy_value_p0": 0.105718083507715,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.166666666666667,
        "unilateral_improvement_p0": 0.227615249825619,
        "unilateral_improvement_p1": 0.272384750174381,
        "nash_conv": 0.5,
        "exploitability": 0.25
      }
    },
    {
      "iteration": 99900,
      "average": {
        "on_policy_value_p0": -0.055554689622968,
        "best_response_value_p0": -0.054949456113616,
        "best_response_value_p1": 0.056400415423065,
        "unilateral_improvement_p0": 0.000605233509352,
        "unilateral_improvement_p1": 0.000845725800097,
        "nash_conv": 0.001450959309449,
        "exploitability": 0.000725479654724
      },
      "current": {
        "on_policy_value_p0": 0.198317822940768,
        "best_response_value_p0": 0.333333333333333,
        "best_response_value_p1": 0.22996897921487,
        "unilateral_improvement_p0": 0.135015510392565,
        "unilateral_improvement_p1": 0.428286802155639,
        "nash_conv": 0.563302312548204,
        "exploitability": 0.281651156274102
      }
    },
    {
      "iteration": 100000,
      "average": {
        "on_policy_value_p0": -0.05555470109818,
        "best_response_value_p0": -0.05476979504503,
        "best_response_value_p1": 0.056122234405666,
        "unilateral_improvement_p0": 0.00078490605315,
        "unilateral_improvement_p1": 0.000567533307486,
        "nash_conv": 0.001352439360636,
        "exploitability": 0.000676219680318
      },
      "current": {
        "on_policy_value_p0": 0.126127974981742,
        "best_response_value_p0": 0.217263089911864,
        "best_response_value_p1": 0.317729770139755,
        "unilateral_improvement_p0": 0.091135114930122,
        "unilateral_improvement_p1": 0.443857745121497,
        "nash_conv": 0.53499286005162,
        "exploitability": 0.26749643002581
      }
    }
  ],
  "final_strategies": [
    {
      "profile": "reach_weighted_average",
      "information_set": "P0|J|root",
      "acting_player": 0,
      "private_card": "J",
      "public_history": "root",
      "action_0": "check",
      "probability_0": 0.794668260640333,
      "action_1": "bet",
      "probability_1": 0.205331739359667
    },
    {
      "profile": "reach_weighted_average",
      "information_set": "P0|J|cb",
      "acting_player": 0,
      "private_card": "J",
      "public_history": "cb",
      "action_0": "fold",
      "probability_0": 0.999996854033156,
      "action_1": "call",
      "probability_1": 0.000003145966844
    },
    {
      "profile": "reach_weighted_average",
      "information_set": "P0|Q|root",
      "acting_player": 0,
      "private_card": "Q",
      "public_history": "root",
      "action_0": "check",
      "probability_0": 0.999965416666667,
      "action_1": "bet",
      "probability_1": 0.000034583333333
    },
    {
      "profile": "reach_weighted_average",
      "information_set": "P0|Q|cb",
      "acting_player": 0,
      "private_card": "Q",
      "public_history": "cb",
      "action_0": "fold",
      "probability_0": 0.458844072079089,
      "action_1": "call",
      "probability_1": 0.541155927920911
    },
    {
      "profile": "reach_weighted_average",
      "information_set": "P0|K|root",
      "acting_player": 0,
      "private_card": "K",
      "public_history": "root",
      "action_0": "check",
      "probability_0": 0.375635844982863,
      "action_1": "bet",
      "probability_1": 0.624364155017137
    },
    {
      "profile": "reach_weighted_average",
      "information_set": "P0|K|cb",
      "acting_player": 0,
      "private_card": "K",
      "public_history": "cb",
      "action_0": "fold",
      "probability_0": 0.00000665538189,
      "action_1": "call",
      "probability_1": 0.99999334461811
    },
    {
      "profile": "reach_weighted_average",
      "information_set": "P1|J|c",
      "acting_player": 1,
      "private_card": "J",
      "public_history": "c",
      "action_0": "check",
      "probability_0": 0.664330748771452,
      "action_1": "bet",
      "probability_1": 0.335669251228548
    },
    {
      "profile": "reach_weighted_average",
      "information_set": "P1|J|b",
      "acting_player": 1,
      "private_card": "J",
      "public_history": "b",
      "action_0": "fold",
      "probability_0": 0.999995,
      "action_1": "call",
      "probability_1": 0.000005
    },
    {
      "profile": "reach_weighted_average",
      "information_set": "P1|Q|c",
      "acting_player": 1,
      "private_card": "Q",
      "public_history": "c",
      "action_0": "check",
      "probability_0": 0.999962272727273,
      "action_1": "bet",
      "probability_1": 0.000037727272727
    },
    {
      "profile": "reach_weighted_average",
      "information_set": "P1|Q|b",
      "acting_player": 1,
      "private_card": "Q",
      "public_history": "b",
      "action_0": "fold",
      "probability_0": 0.666588287827583,
      "action_1": "call",
      "probability_1": 0.333411712172417
    },
    {
      "profile": "reach_weighted_average",
      "information_set": "P1|K|c",
      "acting_player": 1,
      "private_card": "K",
      "public_history": "c",
      "action_0": "check",
      "probability_0": 0.000005,
      "action_1": "bet",
      "probability_1": 0.999995
    },
    {
      "profile": "reach_weighted_average",
      "information_set": "P1|K|b",
      "acting_player": 1,
      "private_card": "K",
      "public_history": "b",
      "action_0": "fold",
      "probability_0": 0.000005,
      "action_1": "call",
      "probability_1": 0.999995
    },
    {
      "profile": "current",
      "information_set": "P0|J|root",
      "acting_player": 0,
      "private_card": "J",
      "public_history": "root",
      "action_0": "check",
      "probability_0": 0.546810689580734,
      "action_1": "bet",
      "probability_1": 0.453189310419266
    },
    {
      "profile": "current",
      "information_set": "P0|J|cb",
      "acting_player": 0,
      "private_card": "J",
      "public_history": "cb",
      "action_0": "fold",
      "probability_0": 1,
      "action_1": "call",
      "probability_1": 0
    },
    {
      "profile": "current",
      "information_set": "P0|Q|root",
      "acting_player": 0,
      "private_card": "Q",
      "public_history": "root",
      "action_0": "check",
      "probability_0": 1,
      "action_1": "bet",
      "probability_1": 0
    },
    {
      "profile": "current",
      "information_set": "P0|Q|cb",
      "acting_player": 0,
      "private_card": "Q",
      "public_history": "cb",
      "action_0": "fold",
      "probability_0": 0,
      "action_1": "call",
      "probability_1": 1
    },
    {
      "profile": "current",
      "information_set": "P0|K|root",
      "acting_player": 0,
      "private_card": "K",
      "public_history": "root",
      "action_0": "check",
      "probability_0": 1,
      "action_1": "bet",
      "probability_1": 0
    },
    {
      "profile": "current",
      "information_set": "P0|K|cb",
      "acting_player": 0,
      "private_card": "K",
      "public_history": "cb",
      "action_0": "fold",
      "probability_0": 0,
      "action_1": "call",
      "probability_1": 1
    },
    {
      "profile": "current",
      "information_set": "P1|J|c",
      "acting_player": 1,
      "private_card": "J",
      "public_history": "c",
      "action_0": "check",
      "probability_0": 0.348210730264407,
      "action_1": "bet",
      "probability_1": 0.651789269735592
    },
    {
      "profile": "current",
      "information_set": "P1|J|b",
      "acting_player": 1,
      "private_card": "J",
      "public_history": "b",
      "action_0": "fold",
      "probability_0": 1,
      "action_1": "call",
      "probability_1": 0
    },
    {
      "profile": "current",
      "information_set": "P1|Q|c",
      "acting_player": 1,
      "private_card": "Q",
      "public_history": "c",
      "action_0": "check",
      "probability_0": 1,
      "action_1": "bet",
      "probability_1": 0
    },
    {
      "profile": "current",
      "information_set": "P1|Q|b",
      "acting_player": 1,
      "private_card": "Q",
      "public_history": "b",
      "action_0": "fold",
      "probability_0": 1,
      "action_1": "call",
      "probability_1": 0
    },
    {
      "profile": "current",
      "information_set": "P1|K|c",
      "acting_player": 1,
      "private_card": "K",
      "public_history": "c",
      "action_0": "check",
      "probability_0": 0,
      "action_1": "bet",
      "probability_1": 1
    },
    {
      "profile": "current",
      "information_set": "P1|K|b",
      "acting_player": 1,
      "private_card": "K",
      "public_history": "b",
      "action_0": "fold",
      "probability_0": 0,
      "action_1": "call",
      "probability_1": 1
    }
  ]
}
