Sports Betting Strategy & Math
The math and discipline that underpin sustainable sports betting — expected value, Kelly criterion sizing, closing-line value, no-vig probability, bankroll management, and the basics of building a model.
The Complete Guide
The math and discipline that underpin sustainable sports betting — expected value, Kelly criterion sizing, closing-line value, no-vig probability, bankroll management, and the basics of building a model.
More Guides in This Topic
About Sports Betting Strategy & Math
The math that decides long-run results
Sports betting outcomes are dominated by three things in the long run: expected value, stake sizing, and execution price. Strategy that ignores any of the three loses to strategy that respects all three, regardless of how strong the underlying picks look on paper.
The math behind each is fixed and well-defined. Expected value is the probability-weighted average payoff of a bet. Stake sizing — most often via the Kelly criterion — is the share of bankroll that maximises growth without risking ruin. Execution price is the implied probability of the odds taken, measured against the no-vig fair value of the outcome.
Why models matter, and why they're not enough on their own
A model produces a probability estimate. A bet is profitable when the market's implied probability is lower than the model's estimate, which is positive expected value. The whole framework rests on the model being calibrated — a small calibration error compounds over thousands of bets into the dominant source of loss.
Closing line value (CLV) is the standard discipline check. If a bettor's average bet price beats the eventual closing line, the model is sharper than the market on average. If it doesn't, the model isn't ready for live deployment regardless of recent win rate. CLV is causal; recent results are noisy.
How exchange pricing changes the math
Sportsbook prices include a 4–6% margin per market. Exchanges typically don't. A 2% edge that's marginal on a sportsbook becomes profitable on an exchange purely from the price improvement, before any model adjustment. The implication: a model that breaks even against sportsbooks usually clears positive expected value on an exchange.
Kelly sizing also changes shape under no-vig pricing. The same model edge supports a larger Kelly fraction because the implied edge after costs is bigger. This isn't a free lunch — variance still scales with stake — but it's the structural reason serious bettors gravitate toward exchanges.