Variance and Downswings: Understanding the Swings

Poker is a game of skill played with cards dealt at random. Over thousands of hands, skill wins. Over a session, a week or even a month, luck can dominate. Variance is the name for those short-term swings, and understanding just how big they are is one of the most important — and most underestimated — parts of becoming a serious player. It’s what separates players who keep improving from those who quit or tilt after a bad month.

Key takeaways

• Variance is the natural gap between your results and your expected results.
• Online 6-max cash games typically have a standard deviation of roughly 80–120bb per 100 hands.
• A solid winner can easily be losing after 10,000 hands — about 3 times in 10 in our example.
• Downswings of 20–30+ buy-ins are normal for winning players.
• Judge your play on decisions and large samples, not recent results.

Expected results vs actual results

In Expected Value you saw that queens against ace-king wins about 57% of the time and gains around +14bb on average. But in any single all-in, you win the whole pot or lose your whole stack. Add up thousands of those spots and your actual results wander around your expected results — sometimes far above, sometimes far below.

Measuring the swings

Players measure their results in big blinds per 100 hands (bb/100) — see Tracking Results. Swings are measured by standard deviation (SD), also in bb/100. For online No-Limit Hold’em 6-max cash games, an SD somewhere around 80–120bb/100 is typical; looser, more aggressive styles swing more.

The key fact: your expected winnings grow in proportion to the number of hands, but the swings only grow with the square root. So over time, skill wins — but it takes much longer than most people think.

A worked example

Imagine a solid winner: 5bb/100 win rate and an SD of 90bb/100.

After 10,000 hands (100 × 100 hands):

• Expected result: 5 × 100 = +500bb (5 buy-ins)
• Standard deviation: 90 × √100 = 900bb
• That’s a range of about −1,300bb to +2,300bb for 95% of players with this win rate.
• Chance of being behind after 10,000 hands: roughly 29%.

After 100,000 hands:

• Expected: +5,000bb
• SD: 90 × √1,000 ≈ 2,850bb
• Chance of still being behind: about 4%.

So a genuinely winning player has almost a one-in-three chance of showing a loss after 10,000 hands — which is several months of part-time play. That’s variance.

Try it yourself: open this exact example in the variance calculator and draw a few sets of sample runs. Then change the hands to 100,000 and watch the band tighten around the expected line.

Downswings are normal

A downswing is a stretch where results fall well below expectation. For winning cash-game players, downswings of 20–30 buy-ins happen, and longer ones aren’t rare over a career. In tournaments, going 100 or more events without a big score is normal even for strong players, because most of the profit comes from rare deep runs.

This is exactly why Bankroll Management asks for so many buy-ins.

To see how deep your own downswings could go, enter your win rate and standard deviation in the variance calculator — it shows your range of results and your risk of going broke for a given bankroll.

Reducing the noise

You can’t remove variance, but you can see through it:

• All-in EV: tracking software can show what you’d have won if every all-in paid out at your equity. It strips out some luck and tells you whether you’re running above or below expectation.
• Big samples: judge results over tens of thousands of hands, or hundreds of tournaments.
• Decision quality: reviewing hands tells you whether you played well, independent of the outcome. See Session Reviews.

Handling a downswing

1. Check your bankroll rules. If you’ve dropped below your level, move down — it’s a strategy, not a failure.
2. Review, don’t panic. Look at a sample of big losing hands. Were they mistakes, or coolers?
3. Reduce volume or take a break if tilt is creeping in — see Tilt Control.
4. Focus on inputs you control: study time, sleep, game selection, table focus.
5. Talk about it with other players. You’ll find everyone has been there. Our community is a good place to start.

Quick check

A player wins 5bb/100 with an SD of 90bb/100. After 10,000 hands, roughly how likely is it that they're showing a loss?

Almost 0%About 5%About 30%About 50%

Check answer

Expected +500bb with an SD of 900bb: a loss is about 0.56 standard deviations below expectation, which happens roughly 29% of the time.

Quick check

Why does skill eventually win in poker?

Because luck evens out every sessionBecause expected winnings grow with the number of hands, while swings grow only with its square rootBecause good players get better cardsBecause variance disappears after 1,000 hands

Check answer

Over many hands, your edge grows faster than the swings around it. But it takes a large sample to show.

Next up: how to measure your results properly — Tracking Results.