Free tool · variance

Poker Variance Calculator: Win Rate, Swings & Downswings

How much can a winning player lose just from bad luck? Enter your win rate, your standard deviation and a number of hands. The calculator gives the range your results will land in, the chance you are still behind, the chance of a downswing of any size, and how far your measured win rate could be from your real one.

Poker variance calculator

bb/100 from your results

Model: results per 100 hands are independent with the win rate and SD you enter, so totals are close to normal over thousands of hands.

How big is variance? A 5 bb/100 winner

Win rate 5 bb/100, standard deviation 100 bb/100 (a typical No-Limit cash profile):

HandsExpected result95% of results95% range for measured win rate (bb/100)Chance of being behind
10,000+500 bb−1,460 bb to +2,460 bb−14.6 to 24.630.9%
50,000+2,500 bb−1,883 bb to +6,883 bb−3.8 to 13.813.2%
100,000+5,000 bb−1,198 bb to +11,198 bb−1.2 to 11.25.7%
250,000+12,500 bb+2,700 bb to +22,300 bb1.1 to 8.90.6%

After 10,000 hands a real 5 bb/100 winner is behind 31% of the time. Even after 100,000 hands, the 95% range of the measured win rate runs from −1.2 to 11.2 bb/100. It takes about 108,000 hands before this player is 95% likely to be ahead (about 433,000 at 2.5 bb/100 and 27,000 at 10 bb/100).

Expected = win rate × hands ÷ 100. Spread (1 SD) = SD × √(hands ÷ 100).

How likely is a downswing?

Chance of falling at least this far below where you start at some point during the next 100,000 hands (SD 100 bb/100):

Win rate1,000 bb (10 buy-ins)2,000 bb (20 buy-ins)3,000 bb (30 buy-ins)
2.5 bb/10054.8%28.4%13.9%
5 bb/10035.9%12.6%4.2%
10 bb/10013.5%1.8%0.2%

Formula (first passage of a random walk with drift): P = Φ((−D − μ) ÷ σ) + e−2 × win rate × D ÷ SD² × Φ((−D + μ) ÷ σ), where μ and σ are the expected result and spread over the hands. A swing from a peak (not from today) is more likely still.

A 20-buy-in hole is a real possibility even for a solid winner. Knowing that ahead of time is the best tilt protection there is: you will know the downswing is math, not proof that your game is broken. Set your bankroll with the bankroll & risk of ruin calculator, and keep a stop rule for the sessions inside the swing.

What variance means for aggressive players

FAQ

How do you calculate variance in poker?

Over N hands, your expected result is win rate × N ÷ 100 and the standard deviation of your result is SD × √(N ÷ 100), with both the win rate and the SD in bb/100. About 95% of the time, your result lands within 1.96 standard deviations of the expected result.

How many hands do I need to know my win rate?

More than most players think. With an SD of 100 bb/100, the 95% range on your measured win rate is ±196 ÷ √(hands ÷ 100) bb/100. After 100,000 hands that is still about ±6 bb/100, so a measured 5 bb/100 could be anything from a small loser to a big winner.

What is a normal downswing in poker?

For a winning No-Limit cash player, downswings of 10–20 buy-ins over a big sample are expected, not a sign of bad play. The calculator shows the chance of falling a given number of big blinds below where you started over the hands you choose.

How do I calculate my bb/100?

Divide your profit by the big blind to get it in big blinds, divide by the number of hands, then multiply by 100. Winning $1,200 at $1/$2 over 20,000 hands: 1,200 ÷ 2 = 600 bb; 600 ÷ 20,000 × 100 = 3 bb/100.

What standard deviation should I use?

Your own, from PokerTracker or Hold’em Manager. Many No-Limit cash players report roughly 80–120 bb/100; loose-aggressive players and short-handed games sit at the high end. If unsure, use the higher number.

Related free tools

Unleash the Madman by Jack Maniac, book cover

Variance is math. Tilt is a choice.

Unleash the Madman by Jack Maniac is a short No-Limit Hold’em playbook for aggressive players: when to fire, when to stop, and how to keep a downswing from turning into tilt. $9.99 on Kindle, or read it on Kindle+.

Get Unleash the Madman: $9.99 on Kindleor read it on Kindle+Read a free sample