Poker Variance Calculator
Cash-game results look calm compared with tournaments, but they still swing hard. Enter your estimated standard deviation per 100 hands (in dollars) and a hand count to see how wide plausible outcomes are before you mistake noise for a leak.
The tool assumes independent identically distributed hands—an approximation. Real sessions have tilt, table selection, and fatigue. Use σ from your database when possible; if you are guessing, start with values your stake and site reports typically show (often tens of dollars per 100 hands at low/mid stakes online).
Cash Game σ & 95% Band
σ grows with √(hands). Doubling volume multiplies dollar SD by √2.
Update hands and σ/100 to refresh the estimate.
Table of Contents
How to Use This Calculator
Enter hands in sample and estimated standard deviation per 100 hands (in dollars or big blinds—stay consistent). The tool scales σ to your sample length and shows a rough 95% band.
σ/100 from HUD
Import from HEM/PT or similar trackers when possible.
Hands
Larger samples shrink relative noise but never remove downswings.
Win rate note
Variance depends on play style and stakes—win rate shifts mean, not necessarily σ.
Compare to sports
See binary bet variance for a simpler discrete model.
σ over Samples
Total σ scales proportional to √(hands/100) × σ/100. That is why you need enormous samples to prove live win rate.
Examples
Example 1 — Defaults
5000 hands, σ/100 = $85—check total σ and 95% band output.
Example 2 — Online micros
Lower σ/100 than live—adjust inputs accordingly.
Example 3 — Short trip
500 hands keeps σ large—avoid overfitting results.
Frequently Asked Questions
Stay consistent—if σ/100 is in BB, interpret band in BB.
No—cash-game style σ. Tournaments need MTT variance tool.
Rule-of-thumb visualization—results not truly normal but close enough for planning.
Not applicable—poker variance inputs are σ and hand counts.
Yes—lineup changes, format changes, and tilt all shift realized dispersion.
