Comparing probabilistic predictive models applied to football. Issue 5 (4th May 2019)

• Record Type: Journal Article
• Authors: Diniz, Marcio Alves; Izbicki, Rafael; Lopes, Danilo; Salasar, Luis Ernesto
• Is Part Of: Journal of the Operational Research Society. Volume 70:Issue 5(2019)
• Page Start: 770
• Page End: 782
• Publication Date: 2019-05-04
• Subjects: Bayesian inference -- predictive inference -- probabilistic prediction -- scoring rules -- soccer prediction; Operations research -- Periodicals; 658.4034
• DOI: 10.1080/01605682.2018.1457485
• Languages: English
• ISSNs: 0160-5682

Abstract

Abstract : We propose two Bayesian multinomial-Dirichlet models to predict the final outcome of football (soccer) matches and compare them to three well-known models regarding their predictive power. All the models predicted the full-time results of 1710 matches of the first division of the Brazilian football championship and the comparison used three proper scoring rules, the proportion of errors and a calibration assessment. We also provide a goodness of fit measure. Our results show that multinomial-Dirichlet models are not only competitive with standard approaches, but they are also well calibrated and present reasonable goodness of fit.