Xtabet logo
    Glossary

    Kelly criterion in betting

    The Kelly criterion sets stake size in proportion to your edge and the odds, maximising long-term bankroll growth. Most practitioners use a fraction of it because full Kelly is uncomfortably volatile.

    Staking & bankroll

    What kelly criterion means

    The Kelly criterion sets stake size in proportion to your edge and the odds, maximising long-term bankroll growth. Most practitioners use a fraction of it because full Kelly is uncomfortably volatile.

    How it works

    The formula asks two questions: how big is your edge, and how much does the price pay? A large edge on a short price justifies a bigger share of bank than a small edge on a long one. Kelly is mathematically optimal only if your probability estimates are accurate — and since they never are, half or quarter Kelly is the practical standard.

    f = (bp - q) / b, where b = odds - 1, p = win probability, q = 1 - p.

    A worked example

    Odds 3.00, your probability 40%. b = 2, p = 0.4, q = 0.6, so f = (2 x 0.4 - 0.6) / 2 = 10% of bankroll. On ₦300,000 that is ₦30,000 — far more than most bettors would risk, which is exactly why quarter Kelly (₦7,500) is the sane version of the same instruction.

    Why it changes your returns

    Kelly links staking to edge instead of to feeling. Even the fractional version enforces the right relationship: bigger edge, bigger stake; no edge, no bet.

    Kelly criterion for Nigerian bettors

    If your edge estimates come from watching NPFL matches rather than from a validated model, use quarter Kelly at most. Overconfident inputs are the fastest route from an optimal formula to a ruined bankroll.

    Common mistakes

    • Feeding in an optimistic probability and letting the formula justify a huge stake.
    • Running full Kelly and abandoning it after the first 40% drawdown.
    • Applying Kelly per bet while ignoring correlation across simultaneous positions.

    Questions bettors ask

    What is fractional Kelly?+

    Staking a fixed share of the Kelly recommendation — typically a half or a quarter — to cut volatility at a small cost in growth rate.

    What if Kelly returns a negative number?+

    It means no edge. Do not bet.

    Is Kelly better than flat staking?+

    In theory yes, with accurate probabilities. With shaky estimates, flat staking is often safer.

    Put it to work

    Related terms

    Back to the full betting glossary, or read the key betting terms guide.