Examples are study prompts, not universal instructions. Card games contain incomplete information, and the best review is usually the one that makes the next similar decision more deliberate.
The shape of a good log
A decision log is not a diary entry about whether I felt good or bad. It is a compact reconstruction of one moment. I write the situation, the options I saw, the evidence I used, the evidence I skipped, and the next question. This structure keeps me from rewriting the hand around the result.
I also write the action before the explanation. “I checked the turn” is a fact about my choice. “I checked because the card was poor for my range” is an explanation that may be right, incomplete, or invented after the fact. Keeping those lines separate lets me test the explanation instead of treating it as history.
The last line is a cue, not a verdict. A cue might be “name the player still to act” or “count the value combinations before choosing a river response.” I want something observable that I can repeat. Broad promises such as “never make that mistake again” do not help me much.
Example: a preflop hesitation
Situation: I was in a middle position seat with a playable but awkward holding. An earlier player had entered, and two players remained behind me.
Choice: I entered the pot without first checking who could act after me.
Evidence noticed: My hand had reasonable coordination. I also noticed that the earlier player had been active.
Evidence skipped: I did not name the seats behind me, their recent tendencies, or how often I would face a raise.
Study question: When I describe a hand as “playable,” what position and action conditions must also be true?
Next cue: Before entering, say “position, action behind, plan if raised” in that order.
This log is useful because it does not depend on what happened after the decision. If the flop had been favorable, the process gap would still exist. I can review several hands from the same seat and see whether the missing position check repeats. The point is not to ban a category of hands; it is to make the conditions visible before I act.
Example: a turn plan changes
Situation: I continued to the turn after a flop action and the board gained a card that changed the available draws and relative strengths.
Choice: I continued with the same plan I had formed on the flop.
Evidence noticed: My original reason for continuing still felt familiar, and the opponent used a sizing I had seen before.
Evidence skipped: I did not ask which hands improved, which draws arrived, or whether the new card changed the value of position.
Study question: What turn cards should force me to rebuild the hand rather than carry forward my flop story?
Next cue: On every turn, name one improvement, one new draw, and one change in leverage.
I like this example because it catches narrative inertia. A plan is useful, but it is not a contract with the future. The board can change the relationship between ranges, and the opponent can choose a different action. My review therefore compares the original plan with the new information instead of asking whether the original plan was foolish.
For follow-up study, I collect a small set of turn cards from similar hands. I sort them into “little changes,” “major changes,” and “unclear.” The unclear group is often the best study group because it shows where I need a better question, not just a stronger opinion.
Example: a river decision
Situation: I reached the river after a line with limited information about the other player’s exact holding.
Choice: I called quickly because the line felt familiar.
Evidence noticed: I had a hand that beat some missed possibilities, and the final action was not the largest sizing I had seen.
Evidence skipped: I did not list the value hands that would choose this line, or compare them with the bluffs I believed were realistic.
Study question: Which observations make a river call more credible, and which ones are only emotional comfort?
Next cue: Before a close call, write one value hand and one missed-draw hand before deciding.
This is not a demand to perform a perfect calculation at the table. It is a review of whether I gave myself enough structure. After the session, I can expand the list and check assumptions. During play, the small cue simply interrupts the urge to treat familiarity as proof.
Example: a position leak
Situation: I noticed several difficult decisions from the same seat, even though the hands and boards were different.
Pattern: I was entering with a plan that worked better when acting later, then facing decisions before I had enough information.
Evidence: The repeated feature was the seat and action order, not one particular hand.
Skipped check: I reviewed individual outcomes but had not grouped the hands by position.
Study question: Which decisions become simpler when I act later, and what information am I missing when I act early?
Next cue: Tag the seat on every review so patterns can be grouped after the session.
Grouping by position can reveal a leak that a single-hand review hides. I do not need a dramatic conclusion. If one seat produces repeated “unclear” notes, I can study that seat for one week and compare the quality of my pre-action plans. The measure is process clarity, not a promise of a particular result.
When the log is incomplete
Sometimes I can identify the decision but not reconstruct every option. In that case I leave the missing evidence visible. I write “not recorded” beside the absent detail, keep the question narrow, and avoid using the log to support a broad rule. An incomplete log can still show a process gap, but it cannot settle a range or frequency question by itself.
I also distinguish “not observed” from “did not happen.” If I did not record a timing change, that does not prove there was none. This small wording habit prevents a clean-looking log from becoming stronger evidence than the original record allows.
How I review the logs
Once or twice a week, I read the newest logs without looking at the final outcomes first. I mark repeated missing questions: position, action behind, board change, sizing meaning, or opponent history. Then I choose one category for the next few sessions. Limiting the focus keeps the exercise practical.
I archive logs that contain a clear question and a follow-up cue. I return uncertain logs to a separate queue. I merge duplicates only when they truly describe the same decision pattern; otherwise I keep both because repeated confusion is information. A log-free session is also acceptable when there were no hands worth recording.
My final review sentence is simple: “The next time I see this shape, I will pause for ___.” If I cannot fill that blank with an observable action, the note is not finished yet. That small standard has made my hand reviews more consistent and much less dependent on hindsight.