Variance is the random spread of results around your true skill. It never goes away — over a bigger sample it just shrinks as a share of the total. Even a clear winner spends long stretches below break-even, so a losing month is usually variance, not a verdict.
Poker is a game of edges measured in big blinds and outcomes measured in luck. Variance is the gap between the two: the reason your results bounce around your real win rate instead of tracking it in a clean line. Understanding it is what separates players who survive downswings from players who panic, move up to “win it back,” and go broke.
Where variance comes from
Every hand is a small bet with a random payoff. String thousands together and the run-good and run-bad mostly — but never fully — cancel out. The statistic that captures how big those swings are is standard deviation, usually quoted in bb/100. Typical values:
- ~80–100 bb/100 for online 6-max No-Limit Hold’em.
- ~70 bb/100 for tighter full-ring games.
- Higher still for loose live games and for tournaments, where most of your buy-ins return nothing and a few return a great deal.
The counterintuitive part: variance doesn’t shrink
People assume a bigger sample “averages out” the swings. In absolute terms it does the opposite — your total result spreads wider the longer you play, because the standard deviation of a sum grows with the square root of the hands. What shrinks is variance relative to your expected win: skill slowly overtakes luck as the dominant term. That’s why a real edge needs volume to become visible.
The variance simulator runs thousands of samples for a win rate and standard deviation you choose, and shows the full fan of outcomes — including how deep the downswings get and how much of the ride sits below break-even, even for a winner.
Downswings are bigger than you think
A 2.5 bb/100 winner — a genuinely good player — can still run 20–30+ buy-ins below their peak over an ordinary sample, and can spend a third of their hands underwater along the way. None of that means they’re playing badly. It means they’re playing poker. This is precisely why bankroll management exists: the roll is there to outlast the swings.
Variance vs a leak: telling them apart
The honest answer is that you usually can’t, in the short term — and anyone who claims certainty off a few thousand hands is guessing. What you can do:
- Check the sample. Under ~50,000 hands, assume any swing is variance until proven otherwise. See how many hands it takes.
- Look at decisions, not results. Review whether your choices were sound, independent of how the hands turned out. Process is signal; one month’s graph is noise.
- Watch the confidence range, not the point. If your win-rate range on the confidence calculator still crosses zero, the data simply can’t tell you yet.
How to actually live with it
Keep a bankroll sized for the swings your game implies, judge yourself on decisions over a long horizon, and treat your tracker’s headline number with suspicion until the sample backs it. Variance is not your enemy — it’s the fog that hides edges, and the thing that keeps weaker players coming back. The job is to keep playing well long enough for it to clear.
Track the swing, not just the score.
StackWise shows your results with the confidence range around them, so a normal downswing reads as variance — not as a reason to tilt. Free to start, no account.
