NHL Players Rating Forecasting with Integrated Machine Learning Model

The following study explores the National Hockey League (NHL) with the objective of predicting player's overall rating using machine learning. Utilizing XGB Regressor, Decision Tree Regressor, Snap Decision Tree Regressor and LGBM Regressor, results reveal the direct strengths of each algorithm, with ensemble methods providing superior predictive accuracy and linear regression variants providing valuable interpretability also with regularization techniques increasing robustness. This study provides an insight into which algorithm to choose for precise player rating predictions. Analysis of each algorithm's performances based on various Performance Metrices like RMSE, R2, Explained Variance, MSE, MAE, MedAE, MSLE and RMSLE. The following paper also contributes to the burgeoning field of sports analytics by displaying the efficacy of machine learning in unravelling the complexities of player evaluation in professional Hockey.