Ep 80: Martin Ingram on Predicting Match Outcomes, Bayesian Style

Jeff chats with Martin Ingram (@xenophar), a PhD student in statistics and author of a recent academic paper presenting a new approach to predicting tennis match outcomes. We talk about his model, what makes it different from other common approaches to match prediction such as Elo, and the simplifying assumptions that make it possible. Martin explains the benefits of a technique that allows to incorporate the effects of surface and even specific tournaments, while considering what data we might include in a more comprehensive model.