UCLA

This thesis constructs a box-score–based framework to measure NBA player efficiency from 2010 to 2025 using three core metrics: Efficiency per Game (Eff PG), Points per Game (PTS PG), and unadjusted Player Efficiency Rating (uPER). After rigorous data cleaning, I analyze aging trends, positional effects, and usage patterns, finding that shooting efficiency and free-throw rate are key drivers of performance. K-means clustering reveals four player archetypes, and logistic regression shows that box-score metrics predict MVP winners with high accuracy. While recognizing limitations, this study affirms the enduring value of public box-score data for understanding player performance and league trends.