Lowly Ravens, Falcons Fly High in College Draft

Senior Fellow Kevin A. Hassett

To listen to the National Football League coaches after the college draft last week, you would think that every team acquired precisely the player it rated the highest, each one a sure Hall of Famer.

As disappointed fans have learned over the years, projecting professional sports performance based on college statistics and player characteristics is an inexact science. For every quarterbacking legend like Peyton Manning, there’s a Ryan Leaf, the player chosen right after Manning in the 1998 draft who had a short and famously awful career.

Yet the latest research suggests that the economics of the draft isn’t such an inexact science at all. While it’s impossible to say with any certainty which players will succeed or fail, the overall performance of specific teams can be measured with surprising accuracy.

For the last two years, I have been tracking the performance of an economic model of the draft developed by economists Richard Thaler of the University of Chicago and Cade Massey of Yale University. They are remarkably successful at linking past draft performance to future team success.

Thaler’s and Massey’s insight is that because NFL teams all have to operate under a salary cap, a team can only become better than its competitors by having a number of players who perform far better than their salaries would indicate.

Virtual Superstars

You can’t have a superstar at every position, because paying 22 superstar salaries would violate the salary cap. The problem then becomes finding a way to get superstar performance at bargain-basement prices. How can a team do that?

One way is to draft shrewdly. If you select a running back in the later rounds and pay him $500,000 a year, and he plays at a level normally associated with a $5 million-a-year veteran, then that player has given the team a “surplus” of $4.5 million.

The teams with the biggest surpluses should improve the most. Thaler and Massey examined the history of drafting and found an interesting pattern in the data. Where the teams pick is more important than who they pick.

Teams that chose players at the top of the draft tended not to get the kind of performance from their players that significantly exceeded their enormous salaries. Players chosen later in the draft, especially in the second round, could regularly be expected to outperform their salaries.

Outperformance means surplus value, and surplus value means wins.

Striking Results

To evaluate their model’s predictive ability, I collected data on the NFL draft for the last five years, estimated a Thaler and Massey score for every team, and then used regression analysis to examine whether teams that won the draft according to their model ended up having better seasons the next year. The results were striking.

Teams that did better in the draft tended to improve much more the following season than teams that did worse. The effect was statistically significant. Each year, the teams that did the best in the draft improved by about four wins relative to teams that did the worst.

While the model has worked well over the years, it isn’t perfect. Last year, it predicted the New England Patriots would see their win total decline; the Pats went 16-0 in the regular season. The team’s draft was terrible, but it was offset by the signing of Randy Moss.

The model also missed the collapse of the Atlanta Falcons, which, given the imprisonment of Michael Vick, clearly had little to do with the draft.

Jets Down, Bucs Up

It did score a number of hits as well, foreseeing correctly a down year for the New York Jets and an up year for the Tampa Bay Buccaneers. So who won this year?

The three big winners of the 2008 draft were the Kansas City Chiefs, Baltimore Ravens and the Falcons.

The Chiefs, coming off a disappointing 4-12 campaign in 2007, were looking to rebuild their roster through the draft. Because they traded 2007 NFL sacks leader Jared Allen to Minnesota for three picks, the Chiefs picked seven times in the first four rounds, including the fifth and 15th picks in the first round.

Baltimore had seven picks in the first four rounds, five coming in the third and fourth rounds. Atlanta had five picks in the first three rounds–including the third overall pick, which they used to select Boston College quarterback Matt Ryan.

Biggest Losers

The big losers were the Minnesota Vikings, Cleveland Browns and Jacksonville Jaguars.

Minnesota, which traded three picks to the Chiefs for Allen, only had a single pick in the first four rounds. Allen might be a great player, but his high veteran salary precludes him from providing much of a surplus.

The Jaguars and Browns each had only two picks in the first four rounds.

How will the draft record translate into wins? If the normal correlation holds next year, Kansas City will win three more games than it did this year, lifting its win total from four to seven. Baltimore and Atlanta will each win two more games. And the cost to Minnesota of having only one pick should be three fewer wins.

Interestingly, the model once again weighs the evidence in a way that’s quite different from the “experts.” ESPN’s Mel Kiper Jr. agreed that Kansas City had a great draft, but Baltimore and Atlanta did almost as well, according to the model. Yet each received only a “B” from Kiper, the same grade he gave to last-place Minnesota.

Kiper, it seems, isn’t much different from many NFL coaches who, Thaler and Massey have shown, regularly emphasize the wrong things on draft day.

Kevin A. Hassett is a senior fellow and the director of economic policy studies at AEI.