NHL Player Contract Calculation and Player Evaluation using Advanced Metric Analysis

This project aims to address the challenges of evaluating National Hockey League (NHL) skaters and forecasting future value in a decision environment where box-score statistics, reputation, and imperfect comparables drive negotiations. The project pipeline integrates multi-season NHL play-by-play, shift-level, and on-ice advanced statistical data with historical contract information to build a unified player-market dataset for skaters. Individual player impact is quantified via Regularized Adjusted Plus Minus (RAPM), which is translated into Goals, Wins, and Standing Points above Replacement (GAR/WAR/SPAR). Empirical age curves constructed from industry findings project each player’s future value, while a market-slope ratio links WAR and SPAR to cap hit percentage before converting into dollar values based on the seasonal cap ceiling. A hybrid framework combining K-nearest-neighbors and per-term ridge regression estimates cap-hit percentages for candidate contracts, and results are displayed on an interactive Dash-based Python dashboard. This system improves contract prediction accuracy relative to simple baselines and provides actionable and clearly interpretable insights for contract negotiations.