The Kelly Criterion Meets AI: Optimal Betting Strategies for Sports Prediction Models

The Kelly Criterion, developed by Bell Labs mathematician John Kelly in 1956, remains one of the most elegant solutions to a fundamental question: given an edge, how much should you bet?

In the age of AI-powered sports prediction, this question has evolved. Now we’re asking: how do we optimally size positions when our edge comes from machine learning models with probabilistic outputs and time-varying confidence levels?