Meet John Urschel, NFL Mathlete and Lover of Graph Laplacians
You previously examined the distribution of academic majors amongst players at major college football programs – Auburn and Stanford, specifically – as they related to upperclassmen at the same schools. Did your findings surprise you?
I can tell you how it went: not the way I thought it would! This happens a lot, less in math but more in science; you do an experiment and you expect a result and sometimes you don’t get that result. But science or math, whatever you get, it’s what you publish. I expected the result I got out of Auburn [nearly two-thirds of all players majored in Liberal Arts], but the reason I chose Stanford is that I thought I could contrast the two. I could say, “Hey some college football programs really just funnel students into the easy majors, but some really have a distribution similar to the general university.” I thought Stanford was going to be that for me, and frankly it was not. It turns out that neither had a distribution that represents the general university at all.
I wrote that article and I received all of this hate mail and it’s all from Stanford. And it’s the strangest type of hate mail that I have ever received in my life, because it’s not people telling me that I suck or cursing at me. The emails start along the lines of “Dear Mr. Urschel, did you know blah-blah-blah-blah about Stanford?” and I love Stanford, it’s a great university, I contemplated going there for my undergrad and my master’s and I might go there for my Ph.D. but I just found it very, very interesting because it’s the strangest type of hate mail I’ve ever received. I didn’t receive a single piece of hate mail from Auburn.
You scored a 43 out of 50 on the Wonderlic last year, the highest of anyone in your draft class. Were you still disappointed?
I was obviously very disappointed in myself about the 43! And I’m not going to blame it on the fact that this was the first year that they changed it to a new system, but what I will tell you is man I was prepared for the old system and I was primed to, if not get a 50, get close to it. That was a goal of mine.
Is the Wonderlic a good indicator of intelligence and future football success?
I think the old Wonderlic was not as good [as the new test] because some of the questions really depended on your upbringing and your familiarity with certain phrases or sayings, and I think it had some socioeconomic implications. And I think the current Wonderlic is much, much better and does not have any hint of that. Although I wish they would have done it the year after me! I think the Wonderlic, like any intelligence exam, is a very good measure around the average and the median, but once you start getting to the extremes – the extreme highs and extreme lows – I think it loses its accuracy very quickly. And that goes for any type of exam.
What do your teammates think about your interest in math?
They all embrace it. They’ll joke with me sometimes, they like to have a good time with me about it, but they very much embrace the math and some of them think it’s pretty cool. The one thing they will get on me about is my lifestyle. The source of a lot of these conversations is my car. I drive a used Nissan Versa hatchback and they’ll say, “Hey you should probably get a nicer car. We know you got this nice signing bonus, we know what your paychecks are, why are you driving this tiny car?” And even beyond the fact of “this thing is dirt cheap” is “it’s not safe.” They keep telling me it’s not safe, that a big truck is just gonna eat my little thing up.
Will you bow to the pressure at some point?
No, I love my car. This little guy is my baby. I’m gonna be driving this thing for like the next five years.
There are seemingly stats for every position in the NFL; do you think there should be similar numbers for offensive linemen?
All these stats, especially QBR, they’re just numbers. I think what you have to look more closely at is the performance itself. I think there’s never going to be a way to judge [an offensive lineman’s performance] fully. People need to realize that, not just in football but in anything, when you quantify these things that you know, it’s not fundamental. There’s a lot of moving parts, you’re only going to be able to capture so much of it. And that’s the way it is in applied math. What you do when you get these numbers of these players, these performances, you’re essentially doing a dimensionality reduction, and of course you’re going to lose things. I think for offensive-line play, you might be able to quantify better once player tracking really takes off. I think the field of football analytics as a whole should really take off once player tracking really becomes standard in the NFL.
When you wrote about the retirement of Chris Borland, you said you envied his ability to walk away from the game because you rely on football to get your aggression out. What exactly is that aggression and where do you think it comes from?
I hate to say this. I think it’s hardwired in my DNA, man. It’s just one of those things. People can talk and say that it’s natural for us to be docile all the time, but we’re descendants of cavemen who went out and hunted and killed their prey. I feel like this aggression was an inherent trait for survival back in the day, and I think this is something that a lot of us carry – some to a degree more than others. I don’t think there’s anything particularly wrong with that as long as you give it the correct outlets and you control it and you keep it under control when you encounter situations that might invoke that part of you. That’s something I haven’t ever had a problem with, especially because I’ve been blessed to be able to play this game and let out my aggression and just run around and just kind of get it out.
When will you stop?
I’m gonna play until no team will have me anymore.
And then what? A career in coaching or a career in math?
I think I’ll go into math and get my Ph.D. in it, preferably at somewhere like a Princeton, MIT, Stanford, Cal Tech. Something like that.