Betting calculatorWager planning
Kelly Criterion Sports Betting Calculator
Estimate a bankroll fraction from your odds and your own win probability.
- Free
- No sign-up
- Runs in your browser
Size a Kelly stake
Calculation result
Example inputsEnter the example values to calculate a result.
How it works
Kelly staking for sports bets
The Kelly Criterion estimates the bankroll fraction that maximizes long-run logarithmic growth when the odds and win probability are known. It does not identify good bets by itself; it converts an assumed edge into a stake size. If your estimated probability is below break-even, full Kelly returns a zero stake.
Because probability estimates are uncertain, many bettors use half Kelly or quarter Kelly. Fractional Kelly reduces the recommended stake and the size of bankroll swings, while also reducing modeled growth.
Kelly formula and inputs
For net decimal payout b and win probability p, the full Kelly fraction is (b × p − q) ÷ b, where q = 1 − p. Enter the bankroll, offered American odds, estimated win probability, and Kelly fraction. The calculator reports the suggested dollar stake and the full-Kelly percentage.
A small change in estimated win probability can cause a large change in stake size. Use a conservative probability estimate and recalculate when the price moves.
Bankroll risk and practical limits
Kelly assumes a repeatable edge, accurate probabilities, and a bankroll that can tolerate losses. Those assumptions are difficult to satisfy in real betting. Fractional Kelly and a maximum bet limit can reduce exposure, but neither guarantees profit or prevents a losing run.
Frequently asked questions
What is the Kelly Criterion formula?
For a binary wager, the full Kelly fraction is (b × p − q) ÷ b, where b is net payout per unit, p is win probability, and q is loss probability.
Should I use full Kelly for sports betting?
Full Kelly is sensitive to probability errors and can produce volatile stakes. Fractional Kelly uses less bankroll per bet, though the appropriate risk level is personal.
Why does Kelly recommend no bet?
The estimated win probability does not exceed the break-even threshold implied by the price, so the calculated growth-maximizing stake is zero.