Gaussian Process Priors for Dynamic Paired Comparison Modelling
This is an incremental improvement for sports analytics and ranking systems, offering better predictive accuracy in specific scenarios.
The authors tackled the problem of dynamic paired comparison modeling for sports prediction by introducing a Gaussian Process prior for time dynamics instead of Markovian assumptions, and the model outperformed Elo and Glicko on log loss in ATP tennis matches, especially with surface covariates.
Dynamic paired comparison models, such as Elo and Glicko, are frequently used for sports prediction and ranking players or teams. We present an alternative dynamic paired comparison model which uses a Gaussian Process (GP) as a prior for the time dynamics rather than the Markovian dynamics usually assumed. In addition, we show that the GP model can easily incorporate covariates. We derive an efficient approximate Bayesian inference procedure based on the Laplace Approximation and sparse linear algebra. We select hyperparameters by maximising their marginal likelihood using Bayesian Optimisation, comparing the results against random search. Finally, we fit and evaluate the model on the 2018 season of ATP tennis matches, where it performs competitively, outperforming Elo and Glicko on log loss, particularly when surface covariates are included.