This free poker bankroll variance calculator models the downswings every winning player runs into. Set your win rate and standard deviation to see the range of results, your risk of ruin, and how big a bankroll you really need.
Assumes results are normally distributed with a fixed win rate — a good model over tens of thousands of hands. Real edges drift, and short samples swing far more, so treat the bands as a guide, not a guarantee.
A winning player can lose for months and a losing player can win for weeks — understanding the size of that swing is what keeps you sane and solvent.
Poker pays out its edge in a fog of luck. A genuine 5 bb/100 winner — a strong result — can still be stuck after 25,000 hands purely from variance. That isn't a leak, a rigged site or a downswing you caused; it's the poker math of a game where any single hand can go either way. Understanding the size of that fog is what keeps you from quitting a winning game in a normal cold streak, or bragging about a win rate built on a lucky week.
The gap between your results and your true win rate shrinks with the square root of hands played. Play four times as many hands and the swing (in bb) only doubles — so relative to your growing expected profit, the noise steadily fades. That's why serious players judge themselves over hundreds of thousands of hands, not sessions, and why a big bankroll matters: it lets you survive the downswings long enough for your edge to show up in the count. Playing a tighter starting hands chart lowers your standard deviation, and running the pot odds calculator before big calls keeps that edge intact.
Standard deviation depends on how you play. Use it to set the simulator and to gauge how big a bankroll you really need.
| Style of play | Std deviation (bb/100) | What it means |
|---|---|---|
| Tight, straightforward | ~70–85 | Fewer big pots and marginal spots — smoother results, smaller swings. |
| Standard winning reg | ~90–110 | The typical range for solid 6-max cash play; ~100 is the usual rule of thumb. |
| Loose-aggressive / high 3-bet | ~120–150 | More big pots and coin-flips — a higher ceiling, but brutal downswings. |
| Multi-table tournaments | very high | Top-heavy payouts mean long stretches of losses broken by rare big scores — variance dwarfs cash games. |
Figures are rough bb/100 ranges for online 6-max cash play. Full-ring runs a little lower; heads-up and tournaments run much higher. Your own tracking software will give you an exact number.
Five steps to go from a fan of squiggly lines to a real read on your game and bankroll.
Use your real long-run bb/100 over a big sample, not last week's heater. If you don't have the hands yet, assume a modest edge — most players overrate their win rate.
Around 100 bb/100 is standard for 6-max cash; tighter play runs lower, loose-aggressive runs higher. This number sets how wide your results can swing.
The expected line is the least likely single outcome. Read the 95% band — that spread is the realistic set of results luck can hand you over the sample.
Check the risk-of-ruin figure. If it's uncomfortably high, you're under-rolled — drop stakes or add funds so a normal cold streak can't end your game.
Don't evaluate your game on a session or even a month. Only tens of thousands of hands separate real edge from noise — keep the sample large before drawing conclusions.
Say you're a strong 5 bb/100 winner with a standard deviation of 100 bb/100 — a genuinely good, beatable-game result. Run 25,000 hands and your expected profit is 1,250 bb. But the 95% range of outcomes stretches from about −1,850 bb all the way to +4,350 bb, and the model says there's roughly a 21% chance you're still in the red at the end. One in five times, a legitimately winning player looks like a loser over 25,000 hands. Stretch the same edge to 100,000 hands and the chance of being down falls to under 6% — the fog thins, but only because you put in the volume.
Verdict — a losing stretch inside this band is normal variance, not a broken game or a leakFix A night — or a month — is almost pure noise. Real win rate only emerges over tens of thousands of hands; anything shorter is variance talking.
Fix Too few buy-ins and a normal downswing busts you before your edge pays off. Check the risk-of-ruin number and keep it low.
Fix A losing stretch inside the 95% band is expected, not evidence of rigging. Look for actual leaks in your play before doubting the deal.
Fix A 20 bb/100 rate over 5,000 hands is a lucky week, not your edge. Regression to the mean is coming — don't move up on a heater.
No arithmetic needed here — just decide what variance can and can't be blamed for.
You've won at 5 bb/100 over a big sample. After 25,000 hands, roughly how likely is it you're actually down money?
With a 100 bb/100 standard deviation, the swing over 25,000 hands is large enough that a real 5 bb/100 winner is in the red about 21% of the time. That's the sample size talking, not your game.
You play four times as many hands. How does the size of your swing (in bb) change?
Standard deviation grows with the square root of hands. Four times the hands means the square root of four — double — the swing in bb, while your expected profit grows fourfold, so the noise shrinks relative to the signal.
What's the surest way to make your true win rate show up in your results?
Volume is the only cure for variance. The more hands you play, the smaller the gap between your results and your real edge — there's no shortcut around the sample size.
You scored 0 / 3. The more you drill this, the faster the maths becomes instinct.
A measure of how swingy your results are, usually quoted in bb/100. Around 100 is typical for 6-max cash: tight, straightforward play runs lower; loose-aggressive play with lots of big pots runs higher.
The probability that a normal downswing wipes out your bankroll before your edge pays off, assuming you keep playing the same stake. A positive win rate and a bigger bankroll both push it down; a break-even game makes ruin a certainty over time.
There's no hard line, but tens of thousands of hands are needed before your win rate is meaningful, and hundreds of thousands before it's reliable. Sessions and even months are dominated by luck.
Absolutely — and a winning player can have a losing one. Over small samples, variance easily overwhelms a modest edge in either direction. That's exactly why bankroll management and volume matter.