Football is no longer just a sport, but also a major business. Player transfers are important for the clubs both in a sporting and a business sense. In the market, many millions of dollars are often paid for football players. Up to now in professional football, market values have primarily been estimated by sports magazines, or fans on internet platforms such as Transfermarkt.de. Both expert and fan estimates have repeatedly been shown to be highly accurate, but the estimation process is usually extremely complex. Market values are therefore often only updated once or twice a year.
Information at the push of a button
In contrast, the algorithm developed by Dr. Alexander Simons and Dr. Markus Weinmann from the University of Liechtenstein together with Dr. Oliver Müller from the IT University of Copenhagen can calculate market values “at the push of a button”. The algorithm thus provides a much more objective and efficient method for evaluating players. Although this is at present only available in the form of a prototype, the research results illustrate the potential of data-based market value estimates in professional football – in particular for identifying talent in scouting or for calculating transfer fees during contract negotiations.
Which players are over or undervalued by the market??
The algorithm was developed on the basis of player data from the five major European leagues over a period of six seasons. It includes twenty-nine performance and popularity indicators ranging from age, position, goals and passes to player-related news headlines, and shows how these indicators affect a player's market value. “A goal increases the market value of a player by around 2.5 percent, and a pass by around 1.5 percent” says Markus Weinmann. Another advantage is that the algorithm can be used to identify over or undervalued players in the market.
Are professional footballers overpaid?
A comparison with the transfer fees actually paid shows that the new approach allows a similarly accurate but much more efficient and transparent market value estimation than previous methods. Although the algorithm generated more accurate estimates with transfer fees of up to 18 million euros, the estimates on Transfermarkt.de had a greater level of accuracy for transfer fees that were higher than this. On the one hand, this may mean that the algorithm will have to be enhanced with additional factors to rate top players. On the other hand, however, the evaluation results may indicate that transfer fees above a certain level cannot be justified in terms of performance or popularity.
Further information
The results will be published shortly in the European Journal of Operational Research.
Background image: Marcello Casal Jr/Agência Brasil; Source: Wikimedia; License: CC BY 3.0 BR
