Hand review habit: a diary without fanaticism

Hand review fails when it tries to replay every pot. You will burn out before you fix a leak. The sustainable model is a diary, not an autopsy: flag three decisions per session, review them in 15 minutes, log one sentence each, and move on. Compounding comes from consistency, not volume of notes.
What deserves a flag
- Big pot where you were unsure at decision time — not where you already know the answer.
- Repeat spot — second 3-bet/fold mistake in a week beats reviewing a random cooler.
- Emotional spike — any hand where you typed a chat message or stood up from the chair.
The ledger columns that actually matter
I use four columns: date, spot tag (SRP BTN vs BB, 3BP IP, etc.), decision (fold turn), verdict (chart ok / exploit ok / spew). Optional fifth: repeat? (Y/N). That is enough to see patterns in 30 days without building a database nobody reads. Fancy HUDs are optional; repeat tags are mandatory.
Review protocol (timer on)
Export or screenshot
Pull the three flagged hands from client history — no rabbit-hunting neighbors' pots.
Solver or chart one spot
Only the worst mistake gets sim time. The others get 30 seconds each.
Write one fix
«Next session: no river bluff vs station tag.» Not a paragraph — a constraint.
Close the file
Review ends when the timer rings. Open-ended study is tomorrow's job.
I remember when I reviewed forty hands and played worse next day
I remember a «study Sunday» — 41 hands, color-coded ranges, three solver trees — and Monday I played 6 hours of scared, chart-bound poker and lost $380 at $2/$5. The diary was perfect; the player was exhausted. I cut to 3-15-1 and my spew tag rate dropped from 18% of flagged hands to 9% over six weeks — not because I got smarter faster, but because I kept showing up.

After your next NUTS POKER session, flag three hands before closing the client, review with a 15-minute timer, and log one constraint for tomorrow. The players who compound are not the ones who review most — they are the ones who review again next week.




