Estimating Iranian Professional Football Players' Prices A Neural Networks Approach

The study aimed to estimate the price of Iranian professional football league players.

The research method was mixed exploratory designs, qualitative and quantitative methods. The research participants in the qualitative section included managers, club coaches, and experts familiar with the football players' market. Therefore, fourteen participants were selected by snowball until we reached saturation. The research tool in the qualitative section included in-depth interviews, so the reliability of which the re-test method was 81%. In the quantitative section, the statistical population had all the football players in the Iranian Football Professional League during 2016-2020, and random sampling was done. So, 863 players were selected to use their data for analysis. The quantitative methods were also collected from valid sites and the Iranian Football League Organization. The model was also designed through radial neural networks using software SPSS and R.

Qualitative section showed that the player's performance, personal characteristics and abilities, club characteristics, and bubble-creating factors are influential in determining the price of football players. In the quantitative section, a model with three hidden layers was designed, which had a nominal error rate in predicting the price of players.

Originality: 

Today, one of the main problems in the field of transfer in football leagues is the small number of appropriate criteria for pricing players. This study aims to estimate the price of professional football players using artificial neural networks.