Rodney Fort and Robert Rosenman
We are interested in how resources are allocated in the presence of perceived streaks in professional sports. Other works have searched for streak-related inefficiencies in sports betting markets, 1 but we look directly at daily game outcomes in Major League Baseball (MLB) over the 1989 and 1990 seasons. We find that team winning and losing streaks are, by and large, random occurrences. However, there is evidence that some managers may be better than others at extending winning streaks and ending losing streaks (holding win percentage constant, as detailed below) or what we refer to as "streak management." In addition, we find that winning streaks increase attendance while losing streaks negatively impact the gate for baseball teams.
This leads us to wonder whether or not managers might have an incentive to portray themselves as streak managers and allocate talent inefficiently. That some managers, namely, successful streak managers, allocate talent differently than other managers can be consistent with efficient resource use, given our results. However, if all managers act as if they are successful streak managers, there is the potential for misallocation. Our only evidence that this may be occurring is from managerial efficiency analysis. We find that streaks are not managed according to the quality of a given manager's team; streak management strategies that are on net valuable to a winning team may be devastating to a losing team, but some managers practice them regardless of winning record.
We offer no explanation of how some managers may be able to manage streaks. But the scenario that is consistent with our findings is as follows. A given manager sets the pitching rotation and rest intervals for fielders, yielding daily lineups that maximize owner profits. Effective streak management involves a plan that results in wins that occur in streaks (or quickly ended losing streaks). Of course, the costs that must enter