Standard deviation measures how much your results bounce around their average. For online 6-max No-Limit it’s about 100 bb/100; full ring is lower (~70–90). The bigger it is, the wider your swings — and the more hands you need before a win rate is trustworthy.
If win rate is the signal, standard deviation is the noise. It’s a single number that captures how far individual results stray from your long-run average. Two players can share the same 5 bb/100 win rate and have wildly different experiences at the table — the one with higher standard deviation rides bigger upswings and uglier downswings to get there.
Typical values
- Online 6-max NLHE: ~100 bb/100. The common default if you don’t know your own.
- Full-ring NLHE: lower, often 70–90 bb/100 — more players, more folding, tighter spread.
- Loose, aggressive, or short-stacked games: higher — more big all-in pots means more variance.
- Tournaments: in a different league entirely. Measured in buy-ins, their standard deviation is huge, which is why MTT results take so long to settle.
Why it controls your confidence range
Standard deviation is the engine behind your win rate’s margin of error. The standard error of a win rate is SD ÷ √(hands ÷ 100) — so a higher SD directly widens the range your true rate could occupy, and means you need more hands to pin it down. Halve someone’s standard deviation and you’d roughly halve the hands they need for the same confidence.
In the win-rate confidence calculator, nudge the standard deviation up and down and watch the range stretch and shrink. That’s variance, made visible.
It also drives risk of ruin
The same swings standard deviation describes are what can wipe out an underfunded bankroll during a normal downswing. Higher variance games demand more buy-ins behind you — a major reason bankroll guidelines call for 20–40 buy-ins for live cash and far more online. It’s not that you’re a worse player on a downswing; it’s that the noise is doing exactly what the math says it will.
The honest point
You can’t out-skill variance, only out-sample it. Standard deviation is the reason a real win rate hides inside a wide range for tens of thousands of hands. Understanding it is what separates players who panic on a downswing from players who keep logging and let the sample grow.
Honest stats, swings and all.
StackWise reads your variance into every figure it shows — confidence ranges, sample-size labels, and a win rate it only reveals once the noise is under control. Free to start.
