Kelly Criterion Explained — Optimal Bet Sizing

By POWERHOUSEUpdated Apr 20255 min read

The Kelly Criterion is a mathematical formula that calculates the optimal percentage of your bankroll to wager based on your edge and the odds offered. Developed by John L. Kelly Jr. at Bell Labs in 1956, it has become the gold standard for bet sizing among professional gamblers and investors. When used correctly, Kelly maximizes long-term bankroll growth while minimizing the risk of ruin.

The Kelly Formula Explained

The Kelly formula is: f* = (bp - q) / b, where f* is the fraction of your bankroll to bet, b is the decimal odds minus 1, p is your estimated probability of winning, and q is 1 minus p. For a bet at +150 (decimal 2.50, so b = 1.50) where you estimate a 45% chance of winning (p = 0.45, q = 0.55), Kelly says: f* = (1.50 * 0.45 - 0.55) / 1.50 = (0.675 - 0.55) / 1.50 = 0.083 or 8.3% of your bankroll.

If Kelly returns a negative number, the bet has negative expected value and you should not place it at all. A Kelly value of zero means the bet is exactly fair (no edge). Only positive Kelly values indicate a profitable wager.

Full Kelly maximizes the geometric growth rate of your bankroll over the long run. However, it also produces significant short-term volatility. A bad stretch can draw your bankroll down 50% or more even with correct probability estimates.

Fractional Kelly: Managing Volatility

Most professionals use fractional Kelly (typically quarter-Kelly or half-Kelly) to reduce variance while retaining most of the growth benefit. Quarter-Kelly sizes each bet at 25% of the full Kelly recommendation. This sacrifices some long-term growth but dramatically smooths the equity curve.

The reason for fractional Kelly is that probability estimates are never perfectly accurate. Even a small error in your estimated probability can turn an optimal Kelly bet into an overbet. Using a fraction provides a safety buffer against estimation errors.

As a practical guideline, if your Kelly calculation suggests betting 8% of your bankroll, quarter-Kelly would size the bet at 2% — which aligns with standard bankroll management advice. This is not a coincidence; the 1-2% unit sizing rule is essentially a simplified version of fractional Kelly.

When to Use and When to Avoid Kelly

Use Kelly when you have a reliable edge estimate based on a large sample of tracked bets. If you have 500+ bets at a verified 54% win rate at -110, you can reasonably apply Kelly to optimize your sizing. Without a proven track record, Kelly is a dangerous tool because garbage inputs produce garbage outputs.

Avoid Kelly for same-game parlays, longshot futures, and any bet where your probability estimate is highly uncertain. Kelly works best for high-volume, relatively standardized bets (like NFL spreads or NBA totals) where your historical performance gives you a reliable edge estimate.

Key Takeaway

The Kelly Criterion calculates optimal bet size based on your edge and odds. Use fractional Kelly (quarter or half) to reduce volatility. Only apply Kelly when you have a proven, data-backed edge estimate.

Frequently Asked Questions

Is the Kelly Criterion guaranteed to make money?

No. Kelly optimizes bet sizing assuming your probability estimates are correct. If your estimates are wrong, Kelly will still lead to losses. It also requires a true positive edge — no formula can make a losing strategy profitable.

What is the difference between full Kelly and half Kelly?

Full Kelly bets the mathematically optimal amount for maximum growth. Half Kelly bets 50% of that amount, which achieves about 75% of the growth rate while cutting variance roughly in half. Most professionals prefer quarter to half Kelly for a smoother experience.

Can I use the Kelly Criterion for parlays?

Technically yes, but it requires accurate probability estimates for the combined parlay outcome, which is much harder than estimating individual game probabilities. Most experts recommend flat-betting parlays at a small fixed percentage rather than applying Kelly.