Team Heterogeneity and Player Performance in the NFL

Lecture Center 13

Success in the National Football League revolves around teams putting their players into the best matchups possible. This concept of a matchup can be explored by looking at offensive positions vs. a team’s defense. Every defense has its strength and weakness so that some offensive players will play better against certain defenses than others. I propose that characteristics like a player’s speed, strength, endurance, and ball handling identify offensive players that will perform well. However, the importance of these characteristics varies with the type of defense. Some characteristics are more relevant against specific types of defenses. I use a linear regression model to examine the connection between characteristics of an offensive player and their effectiveness measured by NFL Fantasy Points. The basic idea is similar to the analysis of earnings as a function of education and human capital, or the estimation of hedonic pricing equations in industrial organization I use a team fixed-effect model to demonstrate that not all defenses are alike when it comes to points allowed after controlling for the offensive players. I believe defenses will allow a different amount of Fantasy Points to players with different characteristics and therefore, I will break the study down by position; wide receiver, quarterback, running back, and tight end. Every position requires a different skill, so it is necessary to split them up; the regression models will assess this hypothesis. Once the positional differences are identified, I will look for trends based on a set of characteristics that would return the highest number of fantasy points. A possible trend could be that running backs who are small and fast play better against some defenses. If that running back, who is small and fast, acquires more points against a defense than a running back who is larger, this would imply that playing a running back who is smaller would earn more points in subsequent games.