As a starting point: 20–40 buy-ins for live cash, 100+ for online cash, and 50–100 for tournaments. But the right number is personal — it falls out of your win rate and your standard deviation, which you can run on the risk-of-ruin calculator.
Bankroll management is the single most important non-poker skill in poker. A winning player with too thin a roll still goes broke — not because they’re bad, but because variance got to them before their edge could. Good bankroll management is simply making sure you outlast the swings.
Why bankroll is a variance problem, not a wealth problem
The size of roll you need has little to do with how much money you want and everything to do with two numbers: your edge (win rate) and how wild your swings are (standard deviation). A bigger edge pulls the required roll down; bigger swings push it up. That trade-off is the entire content of the risk-of-ruin formula — the chance of losing a bankroll B is e^(−2·WR·B/SD²).
How many buy-ins per format
| Format | Common guidance | Why |
|---|---|---|
| Live cash | 20–40 buy-ins | Slower pace, softer games, fewer hands — swings accrue slowly. |
| Online cash | 100+ buy-ins | Tougher fields and far more hands per hour compress your edge and speed up variance. |
| Tournaments (MTT) | 50–100 buy-ins | Top-heavy payouts mean most buy-ins lose; the swings are brutal even for winners. |
These are pre-chewed answers for typical win rates and standard deviations. Treat them as the floor for a comfortable player, not a law — a thin edge or a swingy game wants more, every time.
The risk-of-ruin calculator turns your win rate, standard deviation, and bankroll into a concrete chance of going broke — and tells you how many buy-ins you’d need to push that risk under 5% or 1%.
Moving up — and the rule everyone forgets
Move up when your roll clears the buy-in threshold for the next stake with room to spare, not the moment you can technically afford one buy-in. And the rule that quietly saves bankrolls: be willing to move back down. The risk-of-ruin maths assumes you never drop stakes; in reality, dropping down when your roll shrinks cuts your true risk of ruin far below the never-adjust figure. Stubbornness at a stake you’ve dropped out of is how good players bust.
The honesty trap in bankroll planning
Every bankroll calculation assumes you know your win rate. You almost certainly don’t — early samples flatter you, and a roll sized off an inflated rate is thinner than it looks. A player who “knows” they win at 8 bb/100 off 10,000 hands, but truly wins at 2, has badly under-rolled without realising it.
So before you trust a bankroll plan, pressure-test the edge it’s built on. Check how wide your win-rate range really is on the confidence calculator, and read how many hands it takes before that rate means anything. An honest input is the whole game.
A workable routine
- Pick a buy-in threshold per stake and write it down.
- Track your real win rate and standard deviation as your sample grows — and treat the early numbers as provisional.
- Re-run your risk of ruin when those inputs change, not just when your balance does.
- Move up with a cushion; move down without ego.
Build your bankroll plan on honest numbers.
StackWise tracks your real win rate and swings as your sample grows — and labels the confidence on every figure, so you size your roll on fact, not hope. Free to start, no account.
