Do In-Game Runs Increase the Probability of Winning in Professional Basketball? A Data-Driven Approach

Abstract

In professional basketball, an in-game run can be defined as a set of consecutive points that one team scores in a relatively short time frame. This study’s goal was to determine whether an in-game run increases a team’s probability of winning a game above and beyond a baseline value, which is calculated as the probability of winning observed for the same score advantage without a run. Data collected from 8,370 U.S. professional basketball games show that in-game runs rarely provide an additional increase in probability of winning above the baseline. In certain cases, in-game runs are associated with lower probabilities of winning despite their score advantage. Our analysis enables fans, sports enthusiasts, bettors, and TV networks to determine how in-game runs affect the probability of winning in real time.