Is it Time to Abolish the NFL Combine?

If you’re into the NFL Draft, you might have heard of the data put together by University of Chicago economist Richard Thaler and Yale professor Cade Massey. The two say that high-end draft choices are overvalued. Their great piece of data? In their first five years on the field, the odds that a higher pick will outperform the guy selected before him is just 52 percent.

Translation? Trade down for more picks to increase your chances of getting a better player. The chance for the best player is in the second round, they say. Though of course, it still comes down to good scouting, or what I now believe is just blind luck. The Detroit Lions have had three more second round picks over the last 10 years than the New England Patriots.

With Jamarcus Russell, the No. 1 overall pick of 2007, adding to the bust list after being released from the Raiders with $37 million in his pocket, we summoned David Berri, sports economist and co-author of the new book “Stumbling on Wins” to crunch the numbers on quarterbacks specifically.

Teams obviously believe that a quarterback is a natural for a No. 1 pick. Since 1998, a quarterback has been taken first overall 77 percent of the time (10 out of 13). On the surface, the great vs. bust ratio suggests that choosing a signal caller first isn’t always the best call. Both Mannings, Michael Vick and Carson Palmer are balanced out with Tim Couch, Alex Smith, David Carr and now Russell.

Berri, in his book co-authored by Martin B. Schmidt, says that, when looking back, the characteristics that teams value in quarterbacks -- when considering whether or not to draft them -- are not accurate indicators of their future success. The two say that a player’s height, 40-yard dash time and Wonderlic score all fail to predict how the player will do in the league.

In order to calculate how quarterbacks do per draft position, Berri offers up the chart below. What he did was take all the statistically significant plays made by every quarterback drafted from 1980-2009. He then comes up with how many wins those significant stats produced (relative wins) and then broke it down into how many wins the quarterbacks produced per 100 plays. What does the data show over 30 years? Good news, Tim Tebow and Jimmy Clausen, quarterbacks picked 11-50 outperformed quarterbacks picked 1-10.

Performance of Quarterbacks Selected from 1980-2009

1-10 281 104084 442.7 0.425
11-50 325 102009 456.6 0.448
51-90 259 42660 146.1 0.343
91-150 294 54800 207.3 0.378
151-250 334 58835 229.3 0.390
Punch-line: Quarterbacks selected between picks 11-50 outperform picks 1-10 (and cost less)Source: Dave Berri/Stumbling On Wins

Berri then wanted to see if the order in which the quarterbacks were taken indicated future success. In the chart below, if we were thinking about the 2010 draft, Sam Bradford would be No. 1 and Tim Tebow would be No. 2 (since Tebow was the second quarterback taken). In the chart below, “pick” is the actual order of the draft. Meaning for 2010 purposes, Bradford is No. 1 and Tebow is No. 25, the spot he was actually taken by the Broncos. Berri looked at career performance from years 2-8 and found that performance and draft position isn’t necessarily correlated.

Correlation between Relative WP100 and wherea Quarterback was Selected in NFL Draft: 1980-2009

2 121 -0.02 -0.03
3 123 -0.02 -0.01
4 124 -0.06 -0.06
5 115 -0.03 -0.07
6 96 0.09 0.08
7 78 0.06 0.12
8 70 0.02 0.06
Punch-line: Where a Quarterback is selected is really not correlated with per-play performersSource: Dave Berri/Stumbling On Wins

Maybe it’s time to do away with the tapes and the combines and just admit that it’s a game of luck. Of course, that would leave Mel Kiper, Todd McShay and Mike Mayock looking for new jobs.

Questions? Comments?