Cream Puffs

Abstract

This article provides a novel answer to the question of why elite college football programs schedule so-called “cream puff” games against vastly inferior out-of-conference opponents. Using data on college football games from 2002 to 2010, I find that a team’s chances of winning are 5.3–15.6% greater in the game following their victory over a cream puff. In my preferred estimation, this “cream puff effect” is roughly half as large as the estimated home field advantage. I also show that the U.S. Today/Sagarin rating system, which I use to control for team abilities, penalizes teams for playing vastly inferior opponents. I devise two empirical strategies that deal with this potential problem and show that the cream puff effect is not simply an artifact of the rating system. These results contribute to the literature on dynamic contests by showing that not only does the timing of one’s efforts within a contest matter but so does the schedule of one’s opponents.