| Fraction | Of bankroll | Stake | Growth |
|---|---|---|---|
| Full Kelly | 8.33% | $83.33 | +0.516% |
| Half Kelly | 4.17% | $41.67 | +0.388% |
| Third Kelly | 2.78% | $27.78 | +0.288% |
| Quarter Kelly | 2.08% | $20.83 | +0.227% |
Growth is how much the bankroll grows per bet, on average, compounded over many bets. Full Kelly maximises it; half Kelly gets about three quarters of the growth with half the swings.
What Kelly does
The Kelly criterion sizes a bet to grow your bankroll as fast as possible over the long run. Bet more when your edge is bigger, nothing when there's no edge.
It's only as good as your win probability. Overestimate it and Kelly overbets, which is why many bettors use half or a quarter.
Keep in mind
Never bet more than full Kelly: past it you take on more risk for less growth, and at around twice Kelly growth falls to zero.
A tiny edge gives a tiny stake. Check it's worth the time and any fees.
Using the kelly criterion calculator
Size a bet from your edge: the Kelly criterion gives the stake that grows a bankroll fastest over time, and fractional Kelly a safer share of it.
- 1Enter the odds you can get and your honest estimate of the chance it wins.
- 2Enter your bankroll and pick a Kelly fraction (full, half, third or quarter).
- 3Read the stake, your edge, the expected value and the break-even chance.
Questions bettors ask
What is the Kelly criterion formula?
Kelly = (b × p − q) ÷ b, where b is the profit per $1 (decimal odds − 1), p is your win chance and q = 1 − p. At +110 with a 55% chance: (1.10 × 0.55 − 0.45) ÷ 1.10 = 14.1% of your bankroll.
Why use half or quarter Kelly?
Full Kelly assumes your win probability is exactly right, and its swings are large. Overestimate your edge and it overbets; at around twice Kelly, growth falls to zero. Half Kelly keeps most of the growth with far smaller swings.
What if Kelly says zero?
A zero or negative result means the price gives you no edge at your estimate, so Kelly says don't bet. You need a win chance above the break-even, 1 ÷ the decimal odds.
How do I estimate my win probability?
Start from the market: the no-vig fair odds at a sharp book are a good baseline. Only bet when your own model or information says the chance is clearly higher than that.
