Poker Bankroll Calculator
A bankroll is the buffer that lets a winning player survive variance long enough for the win rate to show up. Enter your win rate and standard deviation — the calculator returns the bankroll, in big blinds, buy-ins and dollars, that holds your risk of going broke to the level you choose. Free, no signup.
Model: classic risk-of-ruin formula, assumes your win rate is real and stationary. Typical std dev: ~80 bb/100 full-ring/6-max cash, ~100+ for loose-aggressive styles.
This site also has a free practice mode — play hands against AI opponents. No signup. Try the practice table →
▸Embed this tool on your site or forum — free
Paste this into any blog post, forum thread, or coaching site. It stays free, always up to date, and needs no account. Please keep the attribution link.
<iframe src="https://holdempro.app/tools/bankroll?embed=1" width="100%" height="620" style="border:0;border-radius:12px;background:#0a0a0b" title="Poker bankroll calculator by HoldemPro" loading="lazy"></iframe> <p style="font-size:12px;margin:4px 0 0">Powered by <a href="https://holdempro.app/tools/bankroll">HoldemPro</a></p>
Buy-ins needed by win rate (reference table)
At 5% risk of ruin and a typical 80bb/100 standard deviation, with a buy-in of 100 big blinds. The pattern to internalize: the required bankroll scales with 1 ÷ win rate.
| Win rate | Buy-ins needed (5% RoR) |
|---|---|
| 1 bb/100 | ≈96 |
| 2 bb/100 | ≈48 |
| 3 bb/100 | ≈32 |
| 5 bb/100 | ≈19 |
| 7 bb/100 | ≈14 |
| 10 bb/100 | ≈10 |
How does the risk-of-ruin formula work?
The model treats your results as a random walk with drift: risk of ruin = e^(−2 · wr · BR / σ²), where wr is your win rate and σ your standard deviation per 100 hands, and BR your bankroll in big blinds. Solved for bankroll: BR = (σ² / 2wr) · ln(1/RoR). Two honest caveats: it assumes your win rate estimate is correct (most players' samples are too small to know), and it assumes you never move down in stakes — in practice, dropping stakes during downswings makes real-world ruin much less likely than the model says. And note the lever the formula hands you: required bankroll scales with 1 ÷ win rate, so improving your win rate is the cheapest way to need less bankroll. If you want to know what's holding yours down, the free 2-minute leak quiz shows your most expensive mistakes.
Bankroll FAQ
How many buy-ins do you need for cash games?
It depends on your win rate, not just a universal number. At a solid 5bb/100 win rate with typical 80bb/100 variance, a 5% risk of ruin needs roughly 19 buy-ins; a thin 2bb/100 win rate needs nearly 50. The often-quoted '20–30 buy-ins' rule is really a statement about being a modest winner — weaker win rates need far deeper reserves.
What is risk of ruin?
The probability that normal variance busts your entire bankroll before your edge can express itself, assuming you keep playing the same stakes. The classic model: risk of ruin = e^(−2 × win rate × bankroll ÷ variance). It falls exponentially as the bankroll grows — which is why doubling your cushion buys far more safety than doubling your caution.
What standard deviation should I use?
Most full-ring and 6-max no-limit cash players run 70–90bb/100; loose-aggressive styles and deep-stacked games run 100+. Tracking software reports your actual number. When in doubt use 80 — and remember variance enters the formula squared, so underestimating it flatters the answer a lot.
Does bankroll management help a losing player?
No — that's the uncomfortable part of the math. With a zero or negative win rate the formula returns infinity: ruin is eventually certain at any bankroll size. Bankroll management protects an edge; it cannot create one. If the honest inputs say you're not beating your game, the fix is finding the leaks, not adding buy-ins.
Should tournament players use the same numbers?
Tournaments have far higher variance — a typical MTT grinder's downswings run 5–10× a cash player's in buy-ins. Common guidance is 100+ average buy-ins for small-field MTTs and several hundred for large fields. This calculator's model fits cash games; for MTTs treat its output as a hard floor, not a target.