⚠️ Our AI model is still learning from match data. All predictions are experimental statistical estimates for information purposes only — not financial advice and not an invitation to bet. Outcomes are never guaranteed. 18+ · Gamble responsibly.
Colour scheme
Language
English ✓ Español ✓ Português ✓ Deutsch ✓ Français ✓
Home/Tools/Kelly criterion

Kelly criterion calculator

Enter the price, your probability estimate and your bankroll to see the Kelly stake — plus the fractional versions that survive contact with real-world estimate error.

Stake calculator

—Edge (EV per unit)
—Full Kelly stake
—Half Kelly
—Quarter Kelly (recommended)

—

How to use this without blowing up

Kelly answers a narrow question precisely: given a known edge, what stake maximises long-run growth? The catch is the word known. Your probability estimate carries error, and the formula is unforgiving about it — overstate a 5% edge by half and full Kelly has you betting roughly twice the optimal amount, which turns normal variance into ruinous drawdowns. That asymmetry is why the quarter-Kelly figure above is the one worth using.

Two sanity checks before you act on any number here. First, if the calculator returns zero, the bet is negative expected value at your own estimate — no stake sizing fixes that. Second, if it recommends a stake that makes you uncomfortable, your bankroll figure is probably not honest: a bankroll is money you can lose in full without consequence, not everything in the account.

We track our own model in flat 1-unit stakes rather than Kelly, deliberately. Flat staking makes a published record comparable and hard to flatter, and it does not silently amplify the model's own estimate errors. The staking guide goes deeper: bankroll management.

Kelly FAQ

What is the Kelly criterion?+

A staking formula that maximises the long-run growth rate of a bankroll. For a simple bet it is (bp − q) ÷ b, where b is the decimal odds minus 1, p your estimated win probability and q is 1 − p. It only recommends a stake when your probability beats the price.

Why use fractional Kelly instead of full Kelly?+

Full Kelly assumes your probability is exactly right. It never is. Because Kelly is aggressive near the optimum, a modest overestimate of your edge produces heavy over-betting and brutal drawdowns, so experienced bettors use a quarter or half of the recommendation.

What if Kelly tells me to bet zero?+

It means the price does not beat your probability estimate — the bet has negative expected value and the correct stake is nothing. A staking formula cannot rescue a bad price.

Is Kelly better than flat staking?+

Mathematically it grows a bankroll faster when your probability estimates are well calibrated. Flat staking is more forgiving when they are not, which is why we track our own model in flat 1-unit stakes: it keeps the published record honest and comparable.