Predicting and Modelling Football Matches with the R Package footBayes

During the past few years, prediction of the outcome of football matches has attracted a lot of interest. The footBayes package offers a powerful and intuitive framework for modelling and predicting football matches using both frequentist and Bayesian approaches. footBayes supports multiple goal-based models, including Poisson models and their extensions that account for time-varying attack and defence parameters. In addition, the package facilitates the incorporation of historical external information on team strengths, using ranking points (relative strengths) as additional covariates. To this aim, a key feature is the implementation of both dynamic and static Bayesian Bradley-Terry-Davidson (BTD) models, which use past match data to estimate historical team strengths. The package also provides tools for graphically analysing a wide range of informative summaries and assessing model performance.