Profiting from the English Premier League: Predictive Elicitation, the Kelly Criterion, and Black Swans

Andrés Ramírez Hassan and Mateo Graciano Londoño

We present a complete betting strategy for football (soccer), and develop a free graphical user interface to apply it. We use betting odds to elicit the hyperparameters of our prior distributions that are mixed with historical records to estimate the probability of each outcome. Forecast probabilities are used to define the optimal bet size based on the Kelly Criterion and we establish a stopping rule for betting, based on atypical positive returns (referred to as “black swans”). We apply our method to the English Premier League, obtaining profits equal to 33.54%, 22.12%, and 49.01% for seasons 2013-14, 2014-15, and 2015-16, respectively.