Estadística multivariante aplicada al análisis y predicción de partidos de fútbol en las principales ligas europeas
- 1 Universidad Complutense de Madrid, España
ISSN: 2174-0410
Year of publication: 2021
Volume: 11
Issue: 2
Type: Article
More publications in: Pensamiento Matemático
Abstract
The purpose of this study is to analyse main game-related statistics differences between the main European leagues and which factors are more determinant when predicting a match score, by means of multivariate statistical techniques, including principal component analysis and logistic regression. The first two principal components explain around the 70 % of variance, and over a 70 % of accuracy is obtained when predicting away-team wins, with these two principal components as predictive variables. This study also shows that in English Premier League, games are less equilibrated.