Does mid-season change of coach improve team performance? Evidence from the history of the NBA

Abstract

This research has advanced in the understanding of the effect of hiring new coaches on performance of NBA teams, when change is achieved in the middle of the season. Changing a coach is an important managerial decision which does not guarantee improvement of results in the short term, i.e. in the same season. Using an easily understandable procedure based on comparison of proportions in a finite population approach framework, we show that only about 15% of new coaches outperformed in a significantly way their predecessors. In order to maximize the probability of success, highly experienced coaches, with a long career as former NBA players should be signed. In addition, change should be made before season advances. These three factors slightly contribute to increase the probability of success, being the most important variable the winning percentage of the team at the moment of change. Therefore, worse teams are more probable to be successful than better teams when a change is made. Finally, implications and limitations are discussed.