An analysis is done on soccer games played by three top European soccer leagues: England, Spain, and Italy during the first 33 rounds of soccer during the 2011-2012 year. Each league has 20 teams playing two games with each other. Two regression models are developed in an effort to predict the point spread of a game between two teams (Team A and Team B) based on the following variables: sum of differences in the number of cards received by Team A and their opposing teams for the last k rounds, sum of differences in the number of cards received by Team B and their opposing teams for the last k rounds, sum of differences in the number of goals received by Team A and their opposing teams for the last k rounds, and sum of differences in the number of goals received by Team B and their opposing teams for the last k rounds, with the value of k always being even. The models developed were used to predict winners of games for the last five rounds of the 2011-2012 season. The models correctly predicted the winner of a game at 73% to 80% of the time. Of particular interest in this research is whether the sum of the differences in the number of cards received by each team and their opponents in the last k rounds of soccer has a significant effect on which team will win the soccer match.
© Copyright 2014 International Journal of Sports Science. Scientific & Academic Publishing. All rights reserved.