Football player market value, multiple linear regression, interaction effects
Abstract
This article attempts to identify those factors contributing to football player market value. To analyze and furthermore determine significant factors or independent variables with around 500 samples of football player data from 2017 to 2020, Multiple Linear Regression is the method to choose. to analyze the significant factors. Based on an assumption, 14 variables that were chosen do correlate with market value. This paper also considers the interaction effects between minutes played and the times a player in a starting lineup, and uses Forward Stepwise Regression to solve the covariance problem caused by adding interaction terms. In order to test the effectiveness of this operation, the research compares the VIF value and significance of those variables. It turns out that the number of minutes played, assists efficiency (attackers), shot on target efficiency (attackers), height(defenders) and passes frequency have a significant linear relationship with prices, while times of dribbles past, goals efficiency, times of aerial won, touches, even height (attackers) fail the significance test. Overall, the volatility of market value of attackers and defenders can be considered by the extent to which these factors affect them.