Variance

Also known as: swing, swings

The statistical spread of results around your true win rate — the reason short-term outcomes lie about your edge.

Variance is the dispersion of your results around expectation. In poker it's quantified by standard deviation, usually expressed in bb/100 (big blinds per 100 hands) for cash, or in buy-ins / ROI points for tournaments.

The key fact players underrate: variance scales with the square root of sample size, while your edge scales linearly. Over \(N\) hands your expected profit is \(\text{win rate}\times N\), but the standard deviation of that profit is \(\sigma\sqrt{N}\). So the ratio of skill to luck grows only as \(\sqrt{N}\) — which is why a 5 bb/100 winner can lose over tens of thousands of hands and a break-even player can run hot for a month.

Typical figures:

• 6-max NLHE cash: \(\sigma \approx 80\text{–}100\) bb/100.
• Full-ring cash: lower, ~60–75 bb/100.
• MTTs: enormous — ROI standard deviation per tournament can be many multiples of the mean ROI, so even strong players go hundreds of cashless tournaments deep.

Variance is why your bankroll is sized in buy-ins and why a downswing tells you almost nothing in a small sample. You can't reduce variance much (game selection helps a little); you can only fund it.