Poker Bankroll & Session Log
Log your sessions and see the range your win rate actually supports, plus the bankroll your measured variance needs. Nothing leaves the browser.
Log a session
Everything stays in this browser — nothing is uploaded anywhere. Enter the result as profit or loss in whatever currency you play in; the big blind is what makes mixed stakes comparable.
How big a bankroll
—
Sessions
What a run of results can tell you
- Your tracker shows one number. The honest answer is a range, and over anything less than a very large sample that range usually includes zero — which is why so many players are certain of a win rate the data does not support.
- The range narrows with the square root of the hands played, not with the hands played. Quadrupling your sample only halves the uncertainty, and that is the whole reason proving a small edge takes hundreds of thousands of hands rather than tens of thousands.
- Risk of ruin has a closed form: exp(-2 x win rate x bankroll / variance). It assumes you never move up in stakes, never move down, and play forever, so treat it as the pessimistic bound it is rather than a forecast.
- A losing player goes broke with probability one, however large the bankroll. No amount of money makes a negative edge survivable, and the formula returns exactly that rather than a reassuring small number.
- The standard deviation matters as much as the win rate and gets a fraction of the attention. It enters the bankroll formula squared, so a loose game with twice the variance needs four times the bankroll for the same safety at the same win rate.
- None of this knows whether you were tilted, tired, playing a softer game, or learning as you went. It is the arithmetic of a sample, and it assumes every hand came from the same player in the same conditions — which is the assumption most likely to be wrong.
How to use
- Log each session with its hands, result and big blind.
- Read the RANGE your win rate supports, not the single number.
- Check the bankroll ladder against the risk you are willing to carry.
Frequently asked questions
Am I actually a winning player?
Your tracker shows one number; the honest answer is a range, and over anything short of a very large sample that range usually includes zero. This works out the range from your own results and says plainly whether it clears break-even yet — which for a modest edge takes well over a hundred thousand hands.
Why does it take so many hands to prove a win rate?
Because the uncertainty shrinks with the square root of the sample, not the sample. Quadrupling your hands only halves the margin, so going from a rough idea to a confident one costs four times the volume each time. That square-root law is the single reason poker samples have to be so enormous.
How is the standard deviation worked out?
By weighting each session by its own length before combining them. A five-hundred-hand session and a fifty-hand one are not comparable observations, so taking the plain standard deviation of session results gives a badly wrong answer — on test data it came out more than three times too high. Each session’s deviation is standardised by the square root of its own length first.
How big should my bankroll be?
That depends on the risk of going broke you are willing to carry, and there is a closed form for it: exp(-2 x win rate x bankroll / variance). The ladder shows what each level of safety costs in buy-ins, computed from your measured rate and variance rather than from a rule of thumb.
Does a big enough bankroll make me safe?
Only if you have a real edge. A losing player goes broke with probability one however large the bankroll, and this returns exactly that rather than a reassuring small number. The formula also assumes you never move up or down in stakes and play forever, so treat it as a pessimistic bound.
Why does variance matter so much?
Because it enters the bankroll formula squared. A loose game with twice the standard deviation needs four times the bankroll for the same safety at the same win rate — which is why game selection changes your bankroll requirement far more than a point of win rate does.
Is my data uploaded anywhere?
No. Sessions are stored in your own browser and never sent anywhere. You can export them to a spreadsheet at any time, and clearing them removes them from this device — there is no copy elsewhere to delete.
🔒 This tool runs entirely in your browser. Nothing you enter is uploaded, logged, or stored.