A data-driven integer programming model for soccer clubs’ decision making on player transfers

This paper presents a mathematical model that provides soccer clubs with optimal player transfer recommendations. Using publicly available data on soccer clubs and players, we create a data-driven optimization model that aids soccer clubs in making informed decisions on how to optimally allocate their multi-million dollar transfer budgets. First, some performance attributes, market value, and salary of 7377 real players from across the world are forecasted using a simple moving average method. Next, these forecasted values are input into an integer programming model that solves for the optimal transfer decisions for any club with the objective of maximizing the club’s utility. This utility is a function of the forecasted values, and increases in the market values of the club’s players. Constraints are imposed to reflect the rules and regulations of club soccer. The model focuses on the Premier League (PL), the first-division soccer league in England and Wales. Numerical results are provided for twelve selected clubs in the PL, for the transfer periods during 2016–2017 and 2017–2018. The results show that transfers recommended by the model could help the clubs achieve better market values. The model can be easily customized to fit any club from any country. Our study provides significant contributions to the literature on optimal budget allocation and team selection in sports, by addressing some important ideas that are not considered in the literature.