PropPrizmFree Tools › Kelly Criterion Calculator
Stake Sizing

Kelly Criterion Calculator

Find the mathematically optimal stake for a bet using the Kelly criterion. Enter the odds, your win probability and bankroll — full Kelly maximises long-run growth; most bettors use a fraction to cut variance.

Recommended stake
% of bankroll
Full Kelly stake
Edge

What the Kelly criterion does

The Kelly criterion tells you what fraction of your bankroll to stake to maximise long-run growth. Bet more than Kelly and you risk ruin; bet less and you grow slower but smoother.

Kelly fraction = (decimal odds − 1) × p − (1 − p)) ÷ (decimal odds − 1), where p is your win probability.

Why almost nobody bets full Kelly

Full Kelly is optimal only if your probability estimate is exactly right — and it never is. Overestimate your edge and full Kelly overbets badly. Half or quarter Kelly keeps most of the growth while cutting the swings and protecting you from your own estimation error. That's why the default here is half Kelly.

Worked example

Bankroll $1,000, odds +120 (decimal 2.2), win probability 50%. Full Kelly = ((1.2 × 0.5) − 0.5) / 1.2 = 8.3%$83. Half Kelly → $41.67.

Stop doing this by hand

These calculators do one bet at a time. PropPrizm's EV+ board runs the de-vig and EV math across every book, on every prop, live — and surfaces only the ones where a book is soft versus the market consensus.

See the EV+ board →

Frequently asked questions

Should I use full or fractional Kelly?

Most disciplined bettors use half or quarter Kelly. Full Kelly assumes your win probability is exact; since it never is, fractional Kelly protects you from overbetting when you have overestimated your edge, at a small cost to growth rate.

What happens if Kelly returns a negative number?

A negative Kelly fraction means the bet is −EV — your win probability is below what the odds imply — so the correct stake is zero. Do not bet.

Does Kelly work across many simultaneous bets?

Basic Kelly assumes one bet at a time. With multiple correlated bets you should scale down further; simultaneous uncorrelated bets can each take a fraction of full Kelly.