Predicting NBA Player Performance Post ACL Injury: A Data Driven Approach to Early Return Risk Management

NBA injury recovery is unpredictable at time requiring intuition (gut feeling) when establishing baseline for predicting a players ability to return to preinjury form. Using ML methods this project aims to predict player performances to their preinjury baseline. The scope used when projecting a players performance is the 1st 20 games after an ACL injury and metrics to be used to aid with ramp up (minutes). My model used KNN for a baseline and Random Forest for advanced analysis to balance accuracy and interpretability. These insights could aid NBA organizations, agents and medical staff with decisions on contracts, rehab, and player time effectively reducing financial as well as competitive risks.