<![CDATA[The persistent discrepancy between actual transfer fees and the theoretical market values of football players highlights the need for a more objective and data-driven framework for player valuation. This study aims to classify the market value tiers of Indonesian Liga 1 players in the 2024/2025 season using an ensemble-based machine learning approach integrated with SHAP interpretability analysis. The dataset comprises 226 players with 27 attributes encompassing demographic, career, performance, physiological, and socio-economic dimensions. The research process involved secondary data collection, preprocessing, feature engineering, and percentile-based label construction, followed by model training using Random Forest, XGBoost, CatBoost, and a Stacking Ensemble. Experimental results show that the CatBoost model achieved the best performance, attaining an accuracy of 89%, a Macro-F1 score of 0.85, and an F1(High-Tier) of 0.78, demonstrating its robustness in handling heterogeneous and imbalanced data. SHAP analysis identified minutes played, age, and social media exposure as the most influential variables determining market value tiers. These findings demonstrate that combining ensemble learning with model interpretability can yield a transparent, adaptive, and practical framework for data-driven player valuation. The proposed approach provides actionable insights for football clubs and analysts in optimising player recruitment and developing fairer, evidence-based transfer strategies.]]>
