Bayesian functional mixed effects model for sports data

Abstract

The use of statistical methods in sport analytics is common practice nowadays. In this work, we propose a hierarchical Bayesian model for describing and predicting the evolution of performance over time for shot put athletes. We address seasonality and heterogeneity in results by means of a linear mixed effects model with heteroskedastic errors. The model provides an accurate description of the performance trajectories and allows for prediction of athletes’ performance in future seasons. We apply our method to an extensive real world data set on performance data of professional shot put athletes recorded at elite competitions.

Silvia Montagna

Assistant Professor in Statistics