Understanding Poker Downswings Mathematically
Downswings are math, not curses. Here's how variance, standard deviation, and risk of ruin work — with a table showing how long swings really last.
On this page · 6 sections
A downswing is not bad luck punishing you — it’s variance, the predictable statistical noise around your true win rate. Even a genuine winner will spend long stretches below expectation, because results bounce around the average before settling on it. Understanding the math turns a downswing from a crisis into an expected cost of playing. Here’s how variance, standard deviation, and risk of ruin actually work.
Win rate is a signal buried in noise
Your true win rate is a fixed edge. But every session adds a random swing on top of it. Over a small sample, the noise dwarfs the signal; over a huge sample, the noise averages out and the signal shows through.
That’s why a 5 bb/100 winner can be down over 30,000 hands and up huge over the next 30,000, with nothing about their skill having changed. The edge is real; it just needs volume to surface. For how to measure that edge, see win rate and ROI.
Standard deviation: the size of the swings
Standard deviation (SD) measures how wide the swings are. In online cash, a typical SD is around 80–100 bb/100 — meaning your results routinely bounce ~90 big blinds per 100 hands around your average. Compare that to a 5 bb/100 win rate and the scale of the problem is obvious: the noise is roughly 18 times the size of the edge over a short window.
Higher-variance formats have bigger SDs:
- Tight full-ring cash: lower SD (~70–80 bb/100).
- Loose 6-max / heads-up: higher SD (100+ bb/100).
- Tournaments: enormous — a few big finishes carry your entire profit, so short-run results are wildly swingy.
The higher the SD relative to your win rate, the bigger the bankroll you need to ride it out.
How long downswings really last
Here’s the part that surprises people. Using a 5 bb/100 win rate and 90 bb/100 SD, here’s roughly how far below expectation a genuine winner can run at different sample sizes (a rough one-standard-deviation downswing):
| Hands played | Expected profit | A normal bad run can be | Net result feels like |
|---|---|---|---|
| 10,000 | +500 bb | −900 bb swing | Down ~400 bb |
| 25,000 | +1,250 bb | −1,420 bb swing | Down ~170 bb |
| 50,000 | +2,500 bb | −2,010 bb swing | Up ~490 bb |
| 100,000 | +5,000 bb | −2,850 bb swing | Up ~2,150 bb |
The lesson: over 10,000–25,000 hands, a real winner can genuinely be in the red and it’s completely normal. Only as the sample grows past ~50,000 hands does the edge reliably outrun the noise. Downswings of 20+ buy-ins are not rare events — they’re baked into the math.
Risk of ruin: why bankroll size is a probability
Risk of ruin is the chance variance busts your entire bankroll before your edge pays off. It depends on three things: your win rate, your standard deviation, and your bankroll in buy-ins. The relationship is stark:
- A small bankroll relative to your SD → high risk of ruin, even for a winner.
- Growing the bankroll cuts risk of ruin exponentially, not linearly — each extra buy-in helps more than the last, up to a point.
- A losing player has a risk of ruin approaching 100% no matter how big the roll — no bankroll saves a negative edge.
This is the mathematical justification for the standard 20–30 buy-in cash guideline: it pushes risk of ruin down to a small percentage for a modest, real winner. The full sizing method is in how much bankroll you need.
What the math tells you to do
- Expect downswings; don’t panic at them. A 15–20 buy-in dip is within normal variance for a winner. Reacting as if you’ve suddenly gone broke is the real danger.
- Keep enough buy-ins that ruin is unlikely. The bankroll exists precisely so variance can’t end you before the average asserts itself.
- Move down during a downswing, not out. Dropping a stake lowers your risk of ruin instantly and keeps you playing.
- Separate variance from tilt. Losing to math is survivable; losing to bad decisions triggered by the downswing is not. That fight is a mental game one.
Put it together
Downswings are the price of an edge that only pays over the long run: variance and standard deviation make short-run results noisy, and risk of ruin is why a large enough bankroll is non-negotiable. Size the roll so ruin is improbable using how much bankroll you need, confirm your edge with an honest win rate, and see how it all fits in the bankroll management hub.
Frequently asked
How long can a poker downswing last?
Longer than most players expect. A winning cash player can run 20+ buy-ins below expectation over tens of thousands of hands, and tournament players routinely go dozens of games without a meaningful cash. Downswings are normal, not a sign you've stopped winning.
What is variance in poker?
Variance is the natural swing of results around your true win rate. Even with a fixed edge, luck makes actual results bounce above and below the average — the bigger the swings, the more bankroll you need.
What is risk of ruin in poker?
Risk of ruin is the probability that variance wipes out your entire bankroll before your edge pays off. It falls sharply as your bankroll grows relative to your win rate and standard deviation.
Am I a losing player if I'm on a downswing?
Not necessarily. A downswing is about results, not skill. The only way to know is a large sample and honest study of your play — variance alone can produce long losing stretches for genuine winners.