GM Overconfidence and Inefficiency
Read a very interesting research paper recently by two professors, Cade Massey of Duke and Richard Thaler of the University of Chicago. The subject is "The Loser's Curse: Overconfidence vs. Market Efficiency in the NFL Draft."
Massey and Thaler apply heuristics to the player selection process. These are psychological factors or biases that effect our judgment. Biases include overconfidence, the winner's curse, false consensus, non-regressive predictions.... The conclusion is that NFL teams overvalue high first-round picks both in terms of what is required to acquire them via trade and in terms of what they pay higher-selected players. As a result, high first round picks are the least valuable in the NFL draft when you measure the surplus value that picks can be expected to produce (again, this is relative to cost, which is very, very high for players picked highly in the first round).
The authors conclude that teams have not come to grips with the implications of the salary cap. They say that there's no economic justification for the value that all the GMs place on moving up X-number of draft slots based on the formula devised by Jimmy Johnson. Teams don't realize that expensive players, even if they are great, "impose opportunity costs elsewhere on the roster." They rip the Giants for investing what they did in dollars and in draft picks for Eli Manning, saying that the Giants were typically overconfident in their ability to recognize greatness. They say that every team that trades up in the first round is making a mistake for that reason and also because they fall into the common trap of believing that everyone is thinking like they are. In other words, if GMs did nothing, there is a very good chance that the player they targeted would have fallen to them at their original pick. Even if that were not to happen, the authors research shows, the likelihood of GMs getting player appreciably worse later in the first round is barely greater than 50-50 and any increased odds over that coin flip are not worth the extra dollars that the higher-drafted player will surely cost.
I think the study is fascinating and will mine it for relevance to the fantasy GM later in the week.
Massey and Thaler apply heuristics to the player selection process. These are psychological factors or biases that effect our judgment. Biases include overconfidence, the winner's curse, false consensus, non-regressive predictions.... The conclusion is that NFL teams overvalue high first-round picks both in terms of what is required to acquire them via trade and in terms of what they pay higher-selected players. As a result, high first round picks are the least valuable in the NFL draft when you measure the surplus value that picks can be expected to produce (again, this is relative to cost, which is very, very high for players picked highly in the first round).
The authors conclude that teams have not come to grips with the implications of the salary cap. They say that there's no economic justification for the value that all the GMs place on moving up X-number of draft slots based on the formula devised by Jimmy Johnson. Teams don't realize that expensive players, even if they are great, "impose opportunity costs elsewhere on the roster." They rip the Giants for investing what they did in dollars and in draft picks for Eli Manning, saying that the Giants were typically overconfident in their ability to recognize greatness. They say that every team that trades up in the first round is making a mistake for that reason and also because they fall into the common trap of believing that everyone is thinking like they are. In other words, if GMs did nothing, there is a very good chance that the player they targeted would have fallen to them at their original pick. Even if that were not to happen, the authors research shows, the likelihood of GMs getting player appreciably worse later in the first round is barely greater than 50-50 and any increased odds over that coin flip are not worth the extra dollars that the higher-drafted player will surely cost.
I think the study is fascinating and will mine it for relevance to the fantasy GM later in the week.
Entry #112105377291452249 posted by Catherine Salfino at 11:27 PM
<< Home