Multivariate exploratory comparative analysis of LaLiga Teams: principal component analysis
Date
2019-03-19Keyword(s)
Performance analysisElite football
Multivariate analysis
Principal component analysis
LaLiga
Abstract
The use of principal component analysis (PCA) provides information about the main
characteristics of teams, based on a set of indicators, instead of displaying individualized information
for each of these indicators. In this work we have considered reducing an extensive data matrix to
improve interpretation, using PCA. Subsequently, with new components and with multiple linear
regression, we have carried out a comparative analysis between the best and bottom teams of LaLiga.
The sample consisted of the matches corresponding to the 2015/16, 2016/17 and 2017/18 seasons.
The results showed that the best teams were characterized and differentiated from bottom teams in
the realization of a greater number of successful passes and in the execution of a greater number of
dynamic offensive transitions. The bottom teams were characterized by executing more defensive
than offensive actions, showing fewer number of goals and a greater ball possession time in the final
third of the field. Goals, ball possession time in the final third of the field, number of effective shots
and crosses are the main discriminating performance factors of football. This information allows us
to increase knowledge about the key performance indicators (KPI) in football.