The Price of a Pick: Examining the Efficiency of the NFL Draft Pick Market by Measuring the Elasticity of the Relationship Between Draft Pick Value and Player Performance

In the NFL Draft, teams take turns (known as “picks”) selecting (or “drafting”) college football players with the intention of drafting players that will perform well in the NFL and thus cause them to win more games. Since the pool of available college football talent is limited, players who are projected to have greater NFL performance are selected with earlier picks, causing the value of a draft pick to decrease as its number increases. With my research, I seek to determine a.) the relationship between the value of a particular pick in the draft (DPV) and the performance of the player selected with that pick (PP) and b.) whether or not this relationship differs between quarterbacks and non-quarterbacks. First, I devise appropriate measures for DPV and PP. Second, I regress the latter on the former using a log-log regression, including a dummy QB variable and an interaction term to determine if this relationship differs between quarterbacks and non-quarterbacks. I also include a control for team quality. Third, I constrain the model to include only drafted players. I find PP to be relatively inelastic to changes in DPV for all players, but I do not find a different relationship for quarterbacks and non-quarterbacks, nor do I find team quality to impact PP