SHAP Values in Football Analytics – Explaining Player Valuations

What Are SHAP Values?

SHAP (SHapley Additive exPlanations) values help explain how each feature contributes to a model’s prediction. Unlike feature importances, which show overall impact, SHAP values can provide insight at an individual player level.

Key Findings from the RandomForest Model

🔹 Defensive Strength Matters – avg_attemptsConcededIbox_Per90_per_game had a wide range of SHAP values. High numbers (more conceded attempts) negatively impacted a player’s market value.

🔹 Team Success Drives Value – Players from teams with higher Team_points consistently had positive SHAP values, reinforcing the correlation between team success and player worth.

🔹 Forward Passing & Build-Up Play – avg_accurateFwdZonePass_value_per_game showed a strong positive impact, indicating that midfielders who successfully transition the ball forward see higher market valuations.

Why SHAP Matters

SHAP values help clubs go beyond just knowing which features matter—they show how those features affect each player’s valuation. This means more data-driven decisions in scouting and transfer negotiations.

🧐 Up Next: Feature Importances in the XGBoost Model – What Drives Market Value?