For poker coaches and study groups

The hour starts with clean data

Every coach knows the first fifteen minutes of an unprepared hour. The student narrates a downswing from memory, you try to reconstruct three weeks of sessions from a story, and the actual poker starts late. A two-field result log cannot show whether the student is running bad or playing bad. That is the student who arrives saying they are card dead, with nothing you can check.

A Coach Pack replaces the story. One export from the student's account: net by stake and venue, the hands they marked at the table, the leaks tagged against those hands, and the session notes underneath. PDF for the call, CSV if you want the raw rows. You spend the hour on poker, not on assembling the picture, and the student stops paying coaching rates for admin.

A coaching week, end to end

Tuesday night your student is in a 2/5 game. A hand goes wrong in a 3-bet pot, so he marks it between orbits: ten seconds on his phone, no writing at the table, no trying to hold four streets in his head until midnight. When he racks up he does the cool-down in the car park: final count, tips, one line on the marked hand while it is fresh, and a tilt flag on the last hour because he knows you will ask.

Friday he plays again, same routine. Sunday evening he exports a Coach Pack and sends it over. Monday at your desk you prep in fifteen minutes instead of forty-five: net by stake, the marked hands with positions and action, the open leak tags with instance counts. You pick three hands, two that touch the turn overfold you tagged last month and one new spot he flagged himself.

Wednesday's call spends the whole hour on the poker. You tag one new leak during the review, it goes on his sheet with a drill attached, and next month's pack will tell you both whether the count is falling. No screenshots, no photographed notebook pages, no excuses about the data.

Inside a one-month Coach Pack

Here is what a typical pack looks like for a 2/5 student playing three or four sessions a week. Every line is something you would otherwise be extracting by interrogation.

Argue about leak counts, not win rates

A month of live volume cannot settle a win rate, and the pack is honest about that. Take the sample above: 2,200 hands is 22 blocks of 100. At a live standard deviation of 100 BB/100, the standard error on the observed win rate is 100 divided by the square root of 22, about 21 BB/100, so the 95% interval is roughly plus or minus 42 BB/100 around whatever the month shows. Any coach arguing from one month of BB/hr is arguing with noise.

Leak counts are different. Six turn overfolds against eleven last month is a real measurement of a real behaviour, and it is visible inside a single month. Because Overshove tags leaks to specific hands rather than to a general impression, you can watch the count fall, or not, and redirect the work accordingly. The coaching gets a scoreboard, the student gets evidence the hours are doing something, and neither of you has to wait 50,000 hands to know.