Improving upon NBA point-differential rankings

TL;DR

This paper proposes new capping and weighting functions, optimized via gradient descent, to enhance NBA team rankings based on point-differential, aiming to improve prediction accuracy of future success.

Contribution

It introduces novel capping and weighting functions and employs gradient descent to optimize their application to game scores for better rankings.

Findings

Improved ranking accuracy over traditional point-differential methods

Effective weighting functions discovered through optimization

Enhanced predictive power for NBA team success

Abstract

For some time, point-differential has been thought to be a better predictor for future NBA success than pure win-loss record. Most ranking and team performance predictions rely largely on point-differential, often with some normalizations built-in. In this work, various capping and weighting functions are proposed to further improve indicator performance. A gradient descent algorithm is also employed to discover the optimized weighting/capping function applied to individual game scores throughout the season.