A longitudinal analysis and data mining of the most representative external workload indicators of the whole elite Mexican soccer clubs

Abstract

ABSTRACT The present study aimed to select the most critical external workload variables that explain sports performance in the Mexican male professional soccer League (Liga BBVA-Bancomer) and analyse the effect of match period and match outcome in workload demands. 480 professional soccer players that belonged to 16 Liga BBVA-Bancomerteams were registered during the 2018–2019 opening and closing tournament by inertial measurement units. The statistical analysis was composed of: (1) principal component analysis (PCA) to select the most critical performance indicators, and (2) mixed analysis of variance to explore match outcome and match period effects. Through PCA analysis, eight variables were selected and distributed in three components that represented the 59.47% (opening tournament) and 58.85% (closing tournament) of total variance with three different PC´s as: high-intensity actions (RelDist, ExplDist, Speed18-21 km/h and Speed21-24 km/h), speed changes (Acc and Dec) and actions at maximum capacity (AccMax and SpeedMax). Differences in tournaments and match periods were found in almost all variables, finding the highest external workload in draw matches when match outcome was considered a factor. PCA analysis seems to be a helpful method to discriminate the most representative variables of performance.