# Poker hand notes: reproducible teaching examples

Created and sources accessed 2026-09-06. Two synthetic examples, four explicit branches. No personal/player data or proprietary solver values. The editorial statuses and manual workflow are proposed utility, not empirically validated memory or training interventions.

## Use the worksheet

Open worksheet.html in a browser, or copy note-template.txt into your own notes. No scripts, input forms, storage or upload functionality are included in this worksheet. It is not a PHH file or an app integration. Keep source and precision beside every detail. Follow room/platform rules for retained records and consider privacy before sharing.

## Verify the published package

Download all files from this directory into one directory. Python 3.9+ uses only its standard library:

Files: [examples.json](examples.json), [verify.py](verify.py), [note-template.txt](note-template.txt), [worksheet.html](worksheet.html), [note-ambiguity.svg](note-ambiguity.svg), this [README.md](README.md), and [SHA256SUMS](SHA256SUMS). Only examples.json and verify.py are required to run the numerical/card checks; retain all seven for the complete manual package.

    python verify.py
    python -O verify.py

Expected: two chip branches conserve 760; 92 single-river checks give 9 and 0 flush completions. The script reads examples.json beside itself and changes no files. No installation, credentials, network access or random seed is needed. The JSON is a literal input-and-result fixture; there is no generator to hide what was assumed.

The chip check independently derives contribution totals and remaining stacks, compares a literal ledger, checks affordable minimum full raises in the stipulated opening-bet situation, and conserves chips for the three contributors. Other dealt players folded without contributing. It does not validate arbitrary action sequences or make a ruling about spoken words.

The suit check constructs the standard 52-card deck and excludes two Hero cards and four board cards. For each specified turn it enumerates 46 compatible single rivers and counts whether any suit appears at least five times in Hero's seven available cards. Nine spades complete Hero's flush after 2s; no river completes Hero's flush after 2h. Other players' hidden cards are unspecified: counts prove completion possibilities in this partial information set, not probabilities, equity or action values. Hypotheses are not assigned probability weights. The two suits are illustrative alternatives, not an exhaustive list of possible suits for the remembered two.

## Source and asset ledger

- PHH v0.0.2 required fields, initial state, ordered actions and card notation: https://phh.readthedocs.io/en/stable/required.html (accessed 2026-09-06). Our prose template is NOT PHH syntax.
- PokerStars official Hold'em rules, best five of seven: https://www.pokerstars.com/poker/games/texas-holdem/ (accessed 2026-09-06 through web rendering; direct requests in this environment redirect geographically, a recorded limitation).
- GTO Gecko official US Apple listing: https://apps.apple.com/us/app/gto-gecko-poker-study/id6738748434 (accessed 2026-09-06, version 1.0.134). Bounded separate off-table study recommendation only; no template-import or exact-hand reconstruction claim.
- note-ambiguity.svg: original GTO Gecko editorial shapes/text, created 2026-09-06. No third-party image material. The article hero WebP is rendered from this source. Its totals are the chip_case branch interpretations in examples.json. Equivalent meaning and exact pot/call values are in the article and worksheet.

## Limits

Two teaching fixtures only; no optimal action, range/equity calculation, full hand replay, app binary test, learning efficacy, memory accuracy, rake model, live floor ruling, or universal validation. Real current-state facts can sometimes be established even when starting-state facts are missing. Derived values inherit uncertainty; balanced arithmetic cannot select the true history among multiple consistent histories.

SHA256SUMS lists the raw bytes of the other files. Verify the downloaded bytes before modifying them. The public verifier itself checks the worked examples, not the manifest; the publication evidence also records an independent clean-folder and extracted-delivery check.