Size a bet against your bankroll using the Kelly Criterion. Enter your win probability and the odds; see full, half, and quarter Kelly stakes side by side.
3
STAKE SIZES SHOWN
1/4
MOST COMMON KELLY FRACTION
TOOLS · NO SIGNUP
Kelly assumes your probability is accurate. It is a sizing formula, not a way to find an edge. Garbage probability in, garbage stake out.
%
$
Full Kelly
$160.00
Half Kelly
$80.00
Quarter Kelly
$40.00
Edge
+7.6pp (16.0% of bankroll at full Kelly)
Kelly only sizes correctly if your probability is actually right. A confident but wrong estimate produces a confidently wrong stake. Full Kelly also assumes you can stomach real drawdowns to grow bankroll fastest over the long run; most bettors who use Kelly bet a fraction of it (half or quarter, shown above) for exactly that reason. See bankroll management for the fuller version of this tradeoff.
The Kelly Criterion answers one question: given a probability estimate and a price, what fraction of your bankroll maximizes long-run growth? The formula, f* = (bp - q) / b, only needs three inputs: your probability, the payout odds, and your bankroll. It is deliberately aggressive by design.
Full Kelly produces real drawdowns even when your edge is genuine, because variance is real. That is why the calculator above always shows half and quarter Kelly alongside the full number. Most practical use of Kelly sizing is fractional. See bankroll management for the fuller tradeoff between growth rate and variance.
QUESTIONS
The Kelly Criterion is a formula for sizing a bet as a fraction of your bankroll to maximize long-run growth, given a probability estimate and the odds offered. The formula is f* = (bp - q) / b, where b is the net decimal odds, p is your win probability, and q is 1 minus p.
Full Kelly maximizes growth rate ONLY if your probability estimate is exactly right, and it produces large, uncomfortable swings even then. Most bettors size at half or quarter Kelly to trade some growth rate for a smoother bankroll, since real-world probability estimates always carry error.
Kelly sizing is only as good as the input. An overconfident probability produces an overconfident, too-large stake, which can be seriously damaging over a full Kelly bet. This is the single biggest risk in using Kelly sizing at all, and it's why fractional Kelly is the common practical choice.
A negative or zero Kelly fraction means your estimated probability is at or below the market's implied probability: there's no mathematical edge, and Kelly says bet nothing on that price.
NoPunt’s own Edge plays ship with a recommended unit size already computed against the model’s tier and a de-vigged fair price, using the same underlying formula. See the Edge for that in context.