Kelly Criterion Calculator
The growth-optimal stake for a bet, from decimal odds and your estimated win probability — with the fractional variants serious bettors actually use.
Full Kelly
10.0%
≈ 500
Half Kelly (common)
5.0%
≈ 250
Quarter Kelly (cautious)
2.5%
≈ 125
Expected value: 10.0% per unit staked · Break-even win rate at these odds: 50.0%
Kelly assumes your probability estimate is correct — it usually isn't, which is why serious bettors stake half Kelly or less. If your estimated edge comes from feel rather than data, treat the output as an upper bound.
The input is the hard part
Kelly's formula is trivial; the win probability you feed it is everything. If it comes from wishful thinking, Kelly will size your wishful thinking optimally. The honest sources of a probability estimate — the market, your own tracked record, or a model — are covered in our expected value guide, and the case for capping stakes regardless is in bankroll management.
One way to ground the estimate in reality: your own history. After a few hundred tracked bets, your actual hit rates by league, market and odds band are evidence. SmartBet Lab computes exactly that from a one-click sync of your betting history — free, no credit card.
Frequently asked questions
What is the Kelly criterion formula?
f* = (b·p − q) ÷ b, where b is decimal odds minus 1, p is your win probability, and q = 1 − p. At odds 2.00 with a 55% win probability: (1×0.55 − 0.45) ÷ 1 = 10% of bankroll.
Why do serious bettors use half or quarter Kelly?
Full Kelly assumes your win probability estimate is exactly right, and it produces very large swings. Fractional Kelly gives up a little growth for much smaller losses when your estimate is wrong — almost always a good trade.
What if the calculator says my edge is negative?
Kelly returning zero means the odds demand a higher win rate than your estimate — the bet is negative expected value and the profitable action is not betting. Passing is a skill.