Kelly Criterion Calculator for Sports Betting

Calculate the mathematically optimal bet size using the Kelly Criterion. Enter your odds, estimated win probability, and bankroll to see exact stake recommendations across four risk levels.

How to Use the Kelly Criterion Calculator

The calculator has two modes: Single Bet for individual wagers, and Parlay for multi-leg accumulators.

For single bets: Enter the decimal odds offered by your sportsbook, your estimated true win probability (as a percentage), and your total bankroll. The calculator outputs the optimal Kelly fraction plus half-Kelly and quarter-Kelly alternatives.

For parlays: Enter each leg’s decimal odds and your estimated win probability for that individual leg. The calculator computes the combined parlay odds and combined probability, then applies the Kelly formula to the combined bet.

What is the Kelly Criterion?

The Kelly Criterion is a mathematical formula developed by John L. Kelly Jr. at Bell Labs in 1956. Originally designed for information theory (optimizing signal transmission over noisy channels), it was quickly adopted by gamblers and investors because it answers a fundamental question: given an edge, how much should you bet to maximize long-term growth?

The formula is:

f* = (bp – q) / b

Where:

• f* = fraction of bankroll to wager
• b = decimal odds minus 1 (net profit per unit staked)
• p = probability of winning (your true estimate)
• q = probability of losing (1 – p)

If f* is negative, the Kelly Criterion tells you not to bet at all — the odds do not offer positive expected value given your probability estimate.

Full Kelly vs. Half Kelly vs. Quarter Kelly

Full Kelly maximizes the long-term growth rate of your bankroll in theory. In practice, most professional bettors and fund managers use fractional Kelly because:

• Full Kelly is volatile. It is mathematically optimal but produces large drawdowns. Simulations show that a full Kelly bettor will experience 50%+ drawdowns regularly, even with a genuine edge.
• Half Kelly reduces variance by 75% while only sacrificing about 25% of the theoretical growth rate. This is the sweet spot recommended by most professionals, including Edward Thorp (who used Kelly to beat blackjack and later ran a hedge fund).
• Quarter Kelly is ultra-conservative. It sacrifices more growth but makes the ride much smoother. Suitable for bettors with less confidence in their probability estimates.

The visual risk bar shows where your full Kelly stake falls on the spectrum from conservative (under 5% of bankroll) to reckless (over 25% of bankroll).

How to Estimate Win Probability

The Kelly Criterion is only as good as your probability estimate. Here are practical methods for estimating true probabilities:

Method 1: Remove the Margin from Sharp Books

Pinnacle is the sharpest sportsbook. If Pinnacle offers Team A at 1.90 and Team B at 2.00, the no-vig probabilities are approximately: Team A = 52.0%, Team B = 48.0%. Use these as your baseline and adjust based on your own research.

Method 2: Model-Based

Build or use a statistical model (Elo ratings, regression models, machine learning) that outputs win probabilities. Compare your model’s output to the bookmaker’s implied probability. If your model says 58% and the book implies 52%, you have a potential 6% edge.

Method 3: Closing Line Value (CLV)

Track where you bet vs. where the line closes. If you consistently bet at better odds than the closing line, your probability estimates are likely better than the market’s. CLV is the single best predictor of long-term profitability in sports betting.

Kelly Criterion for Parlays

Applying Kelly to parlays requires multiplying the individual probabilities together and using the combined decimal odds. This makes Kelly fractions for parlays much smaller because:

1. Combined probability drops exponentially with each leg (60% x 60% = 36% for a 2-leg parlay)
2. The edge must overcome the compounding uncertainty across all legs
3. If any single leg has negative EV, the entire parlay becomes -EV

In practice, Kelly often recommends very small stakes for parlays (under 1% of bankroll), which is sound mathematical advice — parlays are high-variance bets.

Common Mistakes with the Kelly Criterion

Overestimating your edge

The most dangerous mistake. If you think your true probability is 55% but it is actually 50%, full Kelly will lead to overbetting and eventual ruin. This is why fractional Kelly exists — it provides a safety margin against estimation error.

Ignoring correlated bets

Kelly assumes independent bets. If you have 5 NFL bets on the same Sunday that are correlated (e.g., you are overweight on favorites), your true risk is higher than Kelly calculates. Account for correlation by reducing your Kelly fraction further.

Using Kelly on -EV bets

If the calculator shows 0% Kelly fraction, the bet has negative expected value. No staking system can turn a -EV bet into a profitable one. Kelly is a sizing tool, not an edge-finding tool.

Frequently Asked Questions

What is a safe Kelly percentage?

Most professionals keep individual bet sizes between 1-5% of bankroll. If full Kelly suggests more than 10%, your probability estimate is either very confident or very wrong. Half Kelly in the 1-5% range is the standard among sharp bettors.

Should I use Kelly for every bet?

If you have reliable probability estimates, yes. Kelly is optimal for bankroll growth over time. However, if your estimates are rough guesses, flat staking (same amount every bet) may be safer because it does not amplify estimation errors.

Does Kelly work for sports betting?

Kelly was mathematically proven optimal for binary outcomes with known probabilities. Sports betting probabilities are estimated, not known. This is why fractional Kelly (half or quarter) is standard practice — it accounts for the uncertainty in your estimates while still growing your bankroll faster than flat staking.

What happens if I bet more than Kelly suggests?

Over-betting relative to Kelly reduces your long-term growth rate and increases risk of ruin. Betting double the Kelly fraction produces ZERO expected growth — it is mathematically equivalent to burning money, even with a genuine edge. This is sometimes called “over-Kelly” and is the most common way skilled bettors go broke.

How does Kelly apply to simultaneous bets?

If you are placing multiple bets at the same time, the total Kelly allocation across all bets should not exceed your bankroll. In practice, scale each individual Kelly recommendation proportionally so the sum stays within 100%. Many professionals cap total exposure at 20-30% of bankroll across all open positions.