Kelly Criterion Calculator

The Kelly formula is:

f = (bp - q) / b

Where:

- f = fraction of bankroll to bet

- b = net odds received (decimal odds − 1)

- p = probability of winning

- q = probability of losing (1 − p)

Example:

- Decimal odds: 2.50 (b = 1.50)

- Your estimated win probability: 50% (p = 0.50, q = 0.50)

- Kelly stake: (1.50 × 0.50 − 0.50) / 1.50 = 16.7% of bankroll

If the Kelly result is negative, the formula is telling you the bet has no edge — don't take it.

Full Kelly maximizes the rate of bankroll growth in theory, but in practice it creates brutal variance that most bettors can't stomach psychologically.

| Fraction | Stake | Use Case |

|---|---|---|

| Full Kelly (1.0) | Maximum recommended | Theoretical max growth, very high variance |

| Half Kelly (0.5) | Most common for professionals | Good growth rate, manageable drawdowns |

| Quarter Kelly (0.25) | Conservative | Lower drawdowns, suits smaller edges |

Recommendation: Use Half Kelly (0.5) as your default. It reduces your maximum drawdown by roughly 50% while sacrificing only a fraction of long-run growth rate. The reduced variance makes it much easier to stick to your staking plan.

The Kelly Criterion is only as good as your edge estimate. If your p (win probability) is wrong, the Kelly output is wrong.

Over-estimating your edge is dangerous. If you think you have 60% but actually have 50%, Kelly will recommend an aggressive stake that over time destroys your bankroll.

Best practice:

1. Use a sharp market (Pinnacle) as your baseline true probability via the No-Vig Calculator

2. Apply Kelly only when you have a genuine reason to believe your probability exceeds the sharp market's implied probability

3. Start with Half Kelly or Quarter Kelly until you've validated your edge over a large sample (500+ bets)

Tracking your bets with a tool like mybets.gg lets you measure whether your actual win rate is matching your estimated win probability over time.

The Kelly Criterion was originally developed by John L. Kelly Jr. at Bell Labs in 1956 for information theory, then adapted for gambling and investing.

In investing, Kelly is applied to long-run portfolio allocation. In sports betting, it applies to individual bet sizing within a bankroll.

The key difference: in betting, every result is binary (win or lose), so the formula applies cleanly. In investing, outcomes are continuous distributions, requiring modified versions (like the Log-Optimal Portfolio).

For sports bettors, the standard Kelly formula is the correct version to use.