Total distance covered during football (soccer) matches predict performance in the FIFA World Cup Brazil 2014

Abstract

Purpose: The purpose of this study was to examine predictor variables of performance during the FIFA World Cup Brazil 2014. Methods: Individual data from 611 players from 32 teams participating in the tournament were retrieved from the FIFA’s website (http://www.fifa.com/worldcup/statistics/index.html). A performance prediction model using total points in the tournament as the dependent variable and the following predictor variables was studied: 1) goal attempts, 2) passes completed, 3) recovered balls, 4) tackles, 5) goalkeeper saves, 6) maximal running speed, and 7) total distance covered. Pearson product-moment intercorrelation coefficients were computed for the variables before multiple regression analyses were obtained. Basic assumptions for multiple regression analysis were studied. Results: Assumptions of normality, homoscedasticity and linearity were studied by visual scatter plots of standardized predicted scores and residual scores unstandardized scores. There were no influential cases (Cook’s distance, C <1.0) or autocorrelation (Durbin-Watson, D <0.43). Passes completed (r = 0.33, p = 0.031) and total distance covered (r = 0.41, p = 0.009) during games were related to total points in the tournament. The stepwise multiple linear regression analysis indicated that only the total distance covered was a significant predictor of performance in the tournament (β = 0.413; p = 0.019). The final equation for this model is Y = 23,986 + 0.0037 (X) ± 4.77 (R-squared = 0.17). Conclusions: Performance during the World Cup Brazil 2014 was partially explained (17%) by the total distance covered by players during matches.