Can machine learning algorithms predict football players´ market values? A data-driven approach

Football players transfer fees have been increasing in an unprecedented manner in the last decade. As so, it is crucial for football clubs to correctly assess the value of its players. To tackle this problem, this thesis proposes a data-driven solution. After assembling a dataset with data regarding football players’ characteristics, on-field performance indicators and market values, Machine Learning algorithms were used to construct a prediction model. The final proposed modelis a Random Forest regression, whichregistered a coefficient of determination (R2)of 0.88in the test set, displaying a promising outcome for future research