Thursday, March 26, 2009

The Wonderlic is ridicu-lic

I took the ACT at about 7 a.m. on a Saturday morning in the middle of spring during my junior year in high school. It was way too early, but I took the practice test and felt relatively prepared, and I ended up surpassing my goal of a 30 (I scored a 32).

Years later, I look back and wonder why I cared. While it might have opened a few doors into Ivy League or other elite schools, my grades were already good and, at the end of the day, that test had no impact on my career choice and it had no ability to predict my success in any particular field.

The NFL's version of the ACT/SAT is the Wonderlic. There's always a player or two with a remarkably good or bad score, and that leads scouts and analysts to wonder whether that player is ranked appropriately.

This year, the score that jumped out was Matt Stafford's impressive 38 (the average is about a 21). Stafford was already in the running to go No. 1 overall to Detroit, and his test score only solidified that. On the flip side, Percy Harvin scored a 12 and is now suddenly being questioned for his ability to learn routes, and Rey Maualuga's 15 apparently has scouts concerned that he may not be able to run a defense as a middle linebacker.

Question: Why is Maualuga's test score linked to his ability to run a defense? For that matter, why does the Wonderlic affect a player's draft status at all?

Answer: I have no idea. Obviously, you'd prefer to have a smart player with good problem-solving abilities to a player who's a complete idiot. But as long as that player knows exactly what he's doing on the field -- and based on Maualuga's college performance, that'd be difficult to dispute -- does it matter?

While there's no official database (not that I can find, anyway), there's plenty of anecdotal evidence out there that should tell us something. Here are some noteworthy scores from QBs of the relatively recent past:

Drew Henson -- 42
Alex Smith -- 40
Eli Manning -- 39
Tom Brady -- 33
Joey Harrington -- 32
John Elway -- 29
Peyton Manning -- 28
Ryan Leaf -- 27
Brett Favre -- 22
Dan Marino -- 15
Jim Kelly -- 15
Steve McNair -- 15
Donovan McNabb -- 14

You can draw your own conclusions from that data, but a paper published by economic analyst Mac Mirabile went even further, debunking the idea that QB test scores revealed anything about past performance or future success:
The models reveal no statistically significant relationship between intelligence and college passing performance. Likewise, there is no evidence of higher compensation.
What about non-quarterbacks? If you're a numbers geek, you'll enjoy this paper -- which covers all positions -- published by a team of researchers:
The first hypothesis stated that GMA would be positively related to NFL performance. Results indicated that the WPT was unrelated to any of the NFL performance criteria.
And let's finish things off with some more results, just for kicks:

Michael Turner -- 35
Steven Jackson -- 28
DeAngelo Hall -- 23
Ronnie Brown -- 23
Demarcus Ware -- 20
Larry Fitzgerald -- 18
Roy Williams -- 17
Sean Taylor -- 10
Frank Gore -- 6

There's a wide range there, and that's pretty much the point. There's no correlation. I'm not saying that Stafford won't become a very good NFL QB -- I think he will, actually -- but if he does, it won't have anything to do with his test-taking ability, and the same holds true for both Harvin and Maualuga.

No comments: