In a previous article we broke down a study that showed us how statistical discrimination in the NFL Draft occurs based on the success of a players collegiate team, but today we will look into a few trickier kind of discrimination. First, is the NFL being IQaphobic by taking a players intelligence into account when considering hiring them? And on a more serious note, is there a chance that black and white NFL players are evaluated by different standards in the NFL Draft?

With all of the effort teams spend in trying to identify and draft productive college players, they
likely have a few shortcuts for scouting hard-to-see-with-the-naked-eye-intangibles like
leadership and cognitive ability.

For some positions that are more cerebral, like quarterback, NFL scouts use the Wonderlic test
as a general measure of a player's mental acuity and capacity to think quickly. It is believed that
a high Wonderlic score means a player will transition better into the NFL.

For those "unhip" the Wonderlic is a 50 question mental acuity test that players take at the NFL combine, usually at some point before working out in tight clothes for a bunch of guys who are there to see how their "bodies move" and after being forced to stand in line for 12 hours to get physicals done. Does the test really predict NFL Success? Who knows. Ryan Fitzpatrick scored a 48 and Vince Young scored a 4, for what its worth.

In Madden 2006 I would frequently take their version of the Wonderlic and do very well for my fictional franchise quarterback. The real Wonderlic is 50 questions in 12 minutes, and beyond seeing if you can do small logic games and math problems, the Wonderlic tests your fortitude. How well can you mentally performed when you are fatigued and under pressure?

The data set from this research shows evidence that high wonderlic scores are directly
correlated with higher draft position for quarterbacks, but it also delves into a deeper issue and
tests for statistical discrimination in regards to whether or not Wonderlic scores impact the draft
stock of white and black players differently

The methodology behind the data is that for every 1 point increase in Wonderlic score, draft
position would be compared between black and white players to see if there is a difference in
increased marginal jump when also regressed with 40 yard dash time, pre-draft grade, and
college team ranking (Can’t escape the college team ranking.)

This test found conclusive speed discrimination in that for every .1 second increase in 40 yard
dash time, a player rose 3.0 draft spots when everything else was held the same. However,
when testing the original statistical discrimination hypothesis regarding black players and white
players draft stock being affected differently by the wonderlic, the numbers found an (admittedly
weak) correlation in that a 10-point increase in a white players wonderlic resulted in a 1.47
increase in draft position, a 10 point increase in a black players wonderlic increased their draft
position only by .23.

The results of this show that while there is not a major statistical correlation, there is still some
evidence to suggest that NFL teams find the Wonderlic to be more reliable when evaluating
white players than when evaluating black players.

This would present a major problem in the real world, if for example law schools weighed white
students LSAT scores as more significant than black applicants or vice versa. Statistical
discrimination on the basis of race was at one point a major issue in American labor during the
20th century, but stars like Jackie Robinson paved the way for a new-look America where
statistical discrimination based on race is becoming a smaller issue every year.


Curious to see how you'd score? Click the link below to try a free sample Wonderlic and comment your score.

Wonderlic




The study containing the statistics in the article referenced above:

Andrew Gill, Victor Brajer “Wonderlic, Race, and the NFL Draft” Journal of Sports Economics,
First Published December 13, 2011. Web 8/17/2017.
(http://journals.sagepub.com.proxy-um.researchport.umd.edu/doi/full/10.1177/15270025114295
75)