Analyzing Relationship Between Distance Covered in a Soccer Game and Defensive Performance by Regression Models

Abstract

Regression models are widely used in sports research, and there are mainly for two purposes: prediction and effect estimation. The article focuses on providing better strategies for soccer teams’ managers based on analyzing data, which is related to the team’s defense particularly. The methodology designed to find distance covered by a team as an indicator of its defensive performance. By using multiple regression models, the best fits for specific teams are created. Then based on the curve derived, the relationship between the distance covered and the players’ defensive performance can be found. It is concluded that for some teams the two variables do not relate to each other. By contrast, for the teams do, though the specific relationship for each team differs slightly (they fit in different models), the defensive performance increases in intensity as distance covered increases. This detailed result may assist managers in changing tactics, injury prevention, player selection and training management.