Modelling clubs' financial investment in association football players

Abstract

Previous attempts at modelling players' value to association football clubs have to a large extent been occupied with average market behaviour, while specific clubs' situations have been afforded less attention. In this thesis, a new approach is documented. Separate models are devised for rating individual players' performances, forecasting result probabilities and simulating a league competition based on fixed squads of available players. Individual player performance standards affect the result forecasts, and so by making marginal changes to the squads used in simulation, the added value clubs experience from transfers can be estimated.

A selection of observed transfer cases is presented and estimated values of player contributions to simulated club performances are obtained. The players in question and their effect on different clubs' seasonal expectations are analysed, with the circumstances surrounding the observed transfers explained. Marginal revenue estimates of player values are compared to outputs from an updated linear market value function. By comparing the marginal revenue figures to both observed valuations and the different market value approach, a basis for evaluating the new model's capabilities is formed. Areas in which improvements or extensions to the valuation framework seem possible or appropriate are recognised, and highlighted for recommendations for further research.