Variance is the statistical measure of how far your actual results can deviate from your expected (true) winrate. In poker, even strong winning players experience prolonged losing streaks because of the inherent randomness in card distribution. Understanding variance is essential for bankroll management and mental game.
Standard deviation in No-Limit Hold'em typically ranges from 60 to 90 BB/100 for cash games. Tighter players tend to have lower standard deviations (around 60-70), while aggressive and loose players can see 80-100+. PLO (Pot-Limit Omaha) has even higher variance, often 120-150 BB/100.
How many hands do you need? The general rule of thumb is that you need at least 30,000-50,000 hands before your results start to give a reasonable indication of your true winrate. Even at 100,000 hands, the confidence intervals remain wide. Some professionals argue you need 200,000+ hands for a statistically meaningful sample at higher stakes.
Confidence intervals show the range where your results are likely to fall. A 95% confidence interval means that 95% of the time, your actual results will be within this range. The remaining 5% of the time, results will be even more extreme. This is why bankroll management matters: you need enough buyins to survive the downswings that variance guarantees will happen.
Note: This calculator assumes results follow a normal distribution, which is a reasonable approximation for large sample sizes. Actual poker results may have slightly fatter tails due to the nature of the game.
