Comparative Analysis of Machine Learning Models for Predicting Top Goal Scorer and Goalkeeper Performance in Football

The use of football analytics has transformed how players are evaluated, strategies are developed, and recruitment processes are planned. Machine learning (ML) has become a vital tool in analyzing player performance, especially in identifying leading goal scorers and highly effective goalkeepers. This research focuses on understanding how team statistics influence goals and saves, examining not only individual player metrics but also the relationship between team attacks, team goals, and the impact of each player's threatening attacks on their goal-scoring output. Additionally, metrics related to goals conceded by the team were analyzed to establish any correlation with the number of saves made by goalkeepers. The study aims to predict the top goal scorer and the best goalkeeper using seven ML models: Random Forest, XG Boost, Linear Regression, K-Nearest Neighbor (KNN), Decision Trees, Support Vector Regression (SVR), and Gradient Boosting. Data was gathered through web scraping and API integration from five major leagues: La Liga, Bundesliga, Premier League, Ligue 1, and Serie A covering the seasons from 2017/2018 to 2022/2023. The models' performance was assessed using statistical measures like Mean Squared Error (MSE), Mean Absolute Error (MAE), R-Squared (R<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup>), Root Mean Squared Error (RMSE), and Mean Absolute Percentage Error (MAPE). Findings indicate that XG Boost performed best for predicting goal scorers, while Linear Regression was most effective for forecasting goalkeeper saves. This research enhances decision-making for managers and analysts by providing data-driven insights that improve strategic planning and player evaluation.