The Valuation of Football Players with Data Mining Technique (Case Study: Esteghlal Club)

Players are the most important and valuable assets of sports clubs; their contracts cover most of the clubs' budgets. The present study aimed to investigate the role of those factors related to players’ valuation and predict the amount of their contract. The research method was applied-survey and quantitative; the research sample were selected by census sampling method including 41 players of the Esteghlal Club football team. The data from the research were based on the text mining method of the players' performance data for two seasons. When applying data mining method, neural network algorithms, decision tree and average chi-square clustering algorithm were used for data categorization, data analysis and price prediction. Also, the obtained model was tested by predicting the price again with the data using algorithms made in different models and applying graphs and numerical analysis and the predicted value with the actual value in the Clementine software. According to the results, dive had the highest impact factor and total time played during a season had the lowest impact factor for players’ valuation in the neural network algorithm. Age was the factor with the highest effect on players’ price, and players’ position was had the lowest effect in the decision tree algorithm. Physical activity was also the first factor affecting the price. The difference between predicted values in algorithmic methods and the actual data is probably due to the lack of a scientific approach to determine the value of players' contracts. Decision tree algorithm is recommended when predicting players' prices with the club fixed budget and the neural network is the most appropriate method when the budget is varied.