The National Basketball Association, also known as the NBA, is a professional basketball league composed of 30 professional teams in North America and one of the four major professional sports leagues in the United States. This research delves into predicting winning percentages in NBA games using a data-driven approach. It utilizes three years worth of comprehensive game data from 30 NBA teams (2014-2016), including actual outcomes and expert predictions. The study aims to acquire valuable teamwork and data research skills while contributing to sports analytics. Data preparation involves cleaning and selecting recent seasons for model training. The prediction model utilizes an iterative algorithm to estimate team strength coefficients based on game results. Though the model shows promise, its accuracy falls slightly short compared to bookies predictions. The research provides valuable insights for predicting NBA game outcomes and suggests avenues for improving the models accuracy through additional variables
No takes yet. Share an insight, caveat, or question.
Li et al. (2024) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: