Title: Plus-minus a couple of millions: A machine learning model for transfer fee analysis
Authors: Senthil Murugan Nagarajan - University of Luxembourg (Luxembourg) [presenting]
Arne Maes - BNP Paribas Fortis (Austria)
Dries Goossens - Ghent University (Belgium)
Lars Magnus Hvattum - Molde University College (Norway)
Christophe Ley - Ghent University (Belgium)
Abstract: One of the most popular sports is known to be Football, where team managers have major concerns for making important decisions about the player transfers, valuation-related issues, determination of market value, and transfer fees. Football clubs invest massive amounts of money in releasing or hiring players from clubs. However, it becomes a crucial task for club managers to estimate the value of a player in the transfer market. We propose various Machine Learning (ML) techniques, accompanied by feature selection methods, to build a model for predicting the players' transfer fees. More concretely, we used distinct regression techniques such as Lasso Regression, Ridge Regression, Random Forest Regression, Support Vector Regression, and Partial Least Square Regression, which we accompany with Interpretable ML techniques such as Variable Importance and Partial Dependence Plots in order to get insights into the most important predictors. The latter insight is particularly important for team managers.