A probe bet is a turn or river lead by the out-of-position player after the previous bettor or preflop raiser checked back the preceding street. That check creates a new decision; it does not guarantee a weak opposing range. Compare the hands that checked, the new board card and your available bet sizes before choosing a probe.
This guide explains the range questions behind that decision. For a direct test, our Blank-Turn Bluffs: Start With the Checked Range examines 100bb button-versus-big-blind pots after rainbow flops check through. The test required at least 95% of eligible range weight to stay within 0.10bb of the best available action EV. Automatically betting 1.8bb with every eligible no-pair, no-direct-draw hand failed that screen on all eight selected turns. That is a selected-model result, not a frequency estimate for live play.
Disclosure: we build GTO Gecko. The strategy discussion was corrected on September 12, 2026 to distinguish a check-back from an automatic invitation to bluff.
Probe bet, donk bet, float bet: which is which
These three get mixed up constantly, and the distinction is entirely about who had the betting lead and when.
| Bet type | Street | Position | What just happened |
|---|---|---|---|
| Donk bet | Flop (or any street after calling a bet) | Out of position | You lead into the previous aggressor before they get to act |
| Probe bet | Turn or river only | Out of position | The preflop raiser checked back the previous street |
| Float bet | Turn or river | In position | You called a flop bet with a weak hand, then bet when they checked |
A probe can't happen on the flop, because nobody has checked back yet. That's the whole test. If you're first to act on the turn and your opponent passed on the flop, you're in a probe spot. If you lead into the previous aggressor after calling their bet, you're donk betting. The earlier action determines which decision you are studying.
Probe and donk frequencies depend on the configuration and available actions. Neither label supplies a betting frequency by itself.
What a check-back tells you
Strong hands often benefit from a continuation bet, but some check. Medium hands can also prefer checking because they have showdown value and little to gain from a bet. You need the composition of the checking range, not just the fact that a check occurred.
In GTO Wizard's 100bb lowjack-versus-big-blind analysis, a checked range on A-K-8 two-tone remains strong even though it contains fewer very strong hands than the betting range. The big blind does not gain a broad small-bet strategy merely because the raiser checked. A somewhat capped range can still be difficult to attack.
The opposing case matters too. GTO Wizard's T-7-2 rainbow example shows how some turn cards support much more probing. These examples use different boards and ranges. A result on one flop is not a rule for every missed c-bet.
What to look for in a turn card
Rank, pairing, suit and connectedness help describe a turn. Treat them as questions about the reached ranges, rather than a score that automatically selects a bet.
| Feature | Question to investigate |
|---|---|
| Rank | Which hands in each player's actual range improve? A low card need not weaken the raiser's retained top pairs. |
| Pairing | Who holds the paired rank, and what happens to the other made hands? A board pair does not automatically remove fold equity. |
| Suit | Does the card add or complete a flush draw, and which exact combinations hold it? |
| Connectedness | Which straights and direct draws become possible, and are those hands present after the earlier actions? |
For example, a 5 on T-7-2 can give 64 an open-ended draw; it does not complete a straight. Whether it changes the suit texture depends on the suits. An ace on the same flop can improve checked ace-high hands, but that does not mean every big-blind hand must check.
Counting how often a rank or texture appears is a separate calculation from valuing a probe. A selected flop catalog does not measure how often a betting opportunity occurs in dealt hands, and card counts alone do not establish a range or EV advantage.
How much to bet
Probe size depends on which part of the opponent's range your bet targets and which hands support that size in your own range. The turn card is one input, not a sizing instruction.
| Range question | Sizing implication to test |
|---|---|
| Can a wide betting range pressure many weak hands? | A smaller size may serve value, protection and bluffs; compare it with checking. |
| Do you have a strong upper end but little overall equity advantage? | A narrower, more polarized betting range may prefer a larger size. |
| Did the turn improve the opponent's checked hands? | Check which hands can still value bet and which bluffs can support them; neither size is automatic. |
A quarter-pot bluff that wins only when the opponent folds breaks even at a 20% fold rate in a simple model with no rake or further betting: 0.25 / (1 + 0.25). That calculation does not show how often an opponent folds, and it does not compare a real turn bet with the value of checking.
The T-7-2 analysis cited above includes a sparse, large probing range on an ace turn. That illustrates why "high card means never bet" is too strong. Our guide to bet sizing and polarized versus merged ranges explain the broader distinction; use the exact scenario to select a size.
What to probe with
Build the probing range the way you'd build any betting range: value that wants to be called, bluffs that want a fold, and a clear reason for each.
- Value hands. Identify worse hands that can continue against the proposed size. A weak top pair or second pair is not automatically a value bet.
- Draws. A turn can improve a backdoor into a direct draw. Compare the value of betting with the value of checking; having a draw does not settle the choice.
- Bluff candidates. Inspect both the hands your cards remove from the opponent's continues and the hands they remove from its folds. A blocker can help one effect while hurting another.
No-pair hands can sometimes bluff, while hands with showdown value can sometimes prefer checking. Use exact cards, the reached ranges and action values rather than excluding or betting an entire high-card class.
When to check instead
A check deserves consideration when your proposed bet has little value to collect, poor bluff prospects, or less EV than preserving your hand's other options. These are comparisons, not four automatic check rules.
A high card arrives. It may improve hands retained in the raiser's checking range. Investigate how much that changes the range and which value hands or bluffs still want to bet.
The turn pairs the board. Reassess both ranges. The paired rank, earlier action and exact holdings matter; the card alone does not establish that a probe loses value.
The opponent calls too widely. If you have credible evidence for that read at this size and line, reduce bluffs that depend on folds and reassess your value threshold. Do not infer the read from a player label alone.
You hold a weak hand on a blank. Compare checking with each offered bet. The low-turn study gives both costly and inexpensive examples within its no-pair, no-direct-draw group.
River probes
A river probe follows a turn check-back; the flop need not also have checked through. Reconstruct the ranges from the complete action sequence. A second check can change their composition, but it does not automatically remove every strong hand.
No cards remain after the river decision, so future draws cannot improve a called bluff. With no showdown value, such a bluff loses its wager when called, not the entire existing pot from this decision point. Compare checking, available bet sizes and possible raises. Neither a fixed lower betting frequency nor a universal blocker rule follows just from reaching the river.
Adjusting to who checked back
An exploit requires a specific departure from the reference strategy. An opponent who folds too much to a particular probe size presents a different problem from one who checks strong hands and calls too often. Earlier checks and later responses both matter.
Model incentives and player labels are not measured action frequencies. Without a measured or explicitly modeled response, do not infer extra fold equity or a capped range from the label alone.
Three probing mistakes to review
Probing every turn with weak hands. A check-back leaves you a decision. Test the exact eligible hands rather than assuming every low card supports the same bluff.
Betting big to punish the check. Compare what each size folds out and what continues. A larger bet can have a different target; a smaller bet is not automatically sufficient.
Deciding the river in advance. After a probe gets called, the calling range and river card create a new decision. Plan relevant runouts, but allow both further bets and checks when the evidence calls for them.
Drilling probe spots
For a focused study session, record the earlier actions and predict the probe size before revealing the strategy. Then compare the action values and explain one hand that bets and one that checks.
GTO Gecko's paid study plans let you compare precomputed actions, frequencies, EV estimates and ranges in available scenarios. Access depends on the format and tier covered by your plan; the current App Store listing describes these study tools. Open GTO Gecko and use a comparable available scenario rather than assuming a particular example is included.
Our solver study workflow explains how to record a question and compare the available answers.
Frequently asked questions
What is a probe bet in poker?
A probe bet is a turn or river lead after the previous bettor or preflop raiser checked back the preceding street. It gives the out-of-position player the opportunity to start betting on the next street.
What is the difference between a probe bet and a donk bet?
A donk bet leads into the previous aggressor, such as leading the flop after calling a preflop raise. A probe follows a check-back on the preceding street, so it occurs on the turn or river. Their frequencies depend on the modeled situation.
How often should you probe bet on the turn?
There is no general frequency from the check-back alone. Start with the reached ranges, then assess the turn and available sizes. Even selected low, unpaired turns can contain weak hands that lose materially by betting.
What size should a probe bet be?
Test the size against its target. Smaller probes can pressure a weak part of the range; larger probes can support a narrower, polarized range. Compare both with checking in the actual spot instead of assigning one size to every turn texture.
Should you probe bet on a paired turn?
A paired turn is not an automatic check. Which rank pairs and how each player reached the turn determine the relevant hands. Evaluate the resulting ranges before choosing a lead.
Can you probe bet on the river?
Yes. A river probe follows a turn check-back. Earlier streets can have contained bets. Reassess the river ranges and size menu rather than assuming another check guarantees weakness or dictates a lower frequency.