Quick answer
Enter the odds and your own probability estimate. The Kelly criterion gives the fraction of a bankroll that maximises long-run growth: f* = (b·p − q) / b with b = decimal odds − 1 [1]. At 2.00 with p = 55% it is 10%; half Kelly is 5%.
Calculator
How it works, with an example
Kelly sizes a position by edge relative to odds. It only recommends a stake when your probability is above the break-even probability (1/odds), i.e. when Expected value is positive. Example: odds 2.50 (b = 1.5), p = 45% → f* = (1.5 × 0.45 − 0.55) / 1.5 = 8.33%; quarter Kelly = 2.08%. Because full Kelly is very sensitive to errors in p, many people use a fraction such as half or quarter Kelly. The result is only as good as your probability estimate.
Formula
f* = (b · p − q) / b
b = decimal odds − 1, p = your probability, q = 1 − p
f* ≤ 0 → no position; fractional Kelly = k × f*Feed it live odds
Kelly needs a probability estimate; a common starting point is the no-vig consensus of sharp books. Odds-API.io Partner link's value-bets endpoint computes fair probabilities from a sharp panel and returns the EV of each price, which you can feed into this formula [2].
Partner link — we may earn a commission. See our affiliate disclosure.
Frequently asked questions
What if Kelly is negative?
A zero or negative fraction means the price offers no edge at your probability, so Kelly says not to take the position.
Why use half or quarter Kelly?
Full Kelly maximises long-run growth only if your probability is exactly right. Probability estimates are noisy, so a fraction reduces swings and the cost of overestimating your edge.
Does Kelly work for prediction markets?
Yes. Convert the share price to decimal odds (a 0.40 share is 2.50 before fees) and enter your probability.
About this guide
Pred-Data publishes independent, source-linked explainers about prediction markets and the AI tools used around them. Every factual claim on this page links to a primary or reputable secondary source, checked on the date shown above. We do not run a trading venue, take bets, or give personal advice.
Author: Ken Lin · Corrections: if you spot an error, we review it and date-stamp the fix.
Affiliate disclosure
Some links on this site (marked “Partner link”) may be affiliate links: if you sign up through them we may earn a commission at no extra cost to you. Partner links use rel="sponsored nofollow" and never change our assessment. Platform links in our comparison tables are plain, non-affiliate links to official sites.
Sources
- J. L. Kelly Jr., "A New Interpretation of Information Rate", Bell System Technical Journal 35(4), 1956 — Bell Labs (PDF copy hosted by Princeton), accessed 2 Nov 2026
- Value Bets guide (EV formula, response format, limitations) — Odds-API.io docs, accessed 2 Nov 2026