How to review a poker hand without judging the result

Record enough detail to recreate the decision

Write down the format, number of players, positions, effective stack, blinds or antes, exact cards and every action with its size. Effective stack means the most the relevant players can put at risk against each other, not necessarily your full stack. Include rake or tournament payouts when they affect the question.

Choose one decision point. A note such as “I lost with a flush draw” hides both the price and the action. “Button versus big blind; turn bet 20 into 60; opponent all-in; no other players” creates a calculation you can check.

Hide the outcome until the analysis is written. Information learned at showdown can help a later range update, but it was not available when you made the original decision.

Work one decision all the way through

Consider an illustrative heads-up turn. You hold A♥ Q♥ on 8♥ 4♥ 2♣ K♠. There are 60 chips in the pot before your opponent moves all-in for 20. No one else remains in the hand. Calling makes a 100-chip final pot, so the break-even equity is 20%.

Begin with a deliberately simple model: nine hearts win, all other river cards lose, and opposing cards are unknown. The draw completes on 9 of 46 unseen cards, about 19.57%. Under those assumptions, the call’s EV is (9 ÷ 46) × 80 − (37 ÷ 46) × 20 = about −0.43 chips.

That is a model result, not a final verdict on the real hand. An ace or queen may sometimes win against the actual range. The opponent may already hold a hand you beat. Their possible heart holdings affect the distribution of remaining cards. To answer the full question, calculate equity against a specified range rather than treating nine outs as complete information.

The value of the first calculation is that it shows the margin is close. It tells you where to spend study effort: checking the opponent’s range and the validity of the outs. It does not justify a confident call or fold from arithmetic alone.

Compare only the options actually available

In this example the opponent is all-in, so the meaningful options are call or fold. A raise cannot make an all-in player fold. If stacks remain instead, the decision changes: future bets, implied winnings and losses after improving become relevant.

Ask what checking or folding is worth relative to the current decision. Money already contributed is in the pot and is not recovered by making another poor call. Avoid reasoning that you have “put too much in to fold” without examining the new price.

When reviewing a betting decision, compare at least one alternative size or a check, using plausible opponent responses. Do not award every imagined line its most favourable outcome.

Finish with a reusable review note

Keep the conclusion short enough to revisit. State the model, what it suggests, and the main uncertainty. For the worked example: “The flush-only model is slightly below the price. Recalculate against a plausible turn-shoving range before treating the call as good or bad.”

Choose one next action that resolves the uncertainty: reconstruct the range, check card-removal effects, or compare the same hand against a different bet size. Replaying the loss without a question does not create a study plan.

Look for repeated decision types across several reviews. If many notes concern turn draws, focus the next session on one-card probabilities and all-in equity. The aim is a clearer process, not a promise that the next river will cooperate.