Writers of Pro Football Prospectus 2008

12 Dec 2012

SI: Blessed Are the Geeks...

For those who did not see, there was a big article in the latest issue of Sports Illustrated about football analytics, mostly focused on myself and Brian Burke of Advanced NFL Stats, as well as Tony Khan of the Jaguars. It's now online. A lot of neat things have happened to me during my 10-year journey from random guy starting a website to accepted NFL "expert," but being the focus of an article in SI is pretty up there on the list of "things I dreamed about when I was a kid." The funny thing about the title, of course, is that I think Burke and I are pretty far away from stereotypical "geeks." He's ex-military, and I'm an ex-radio jock.

Posted by: Aaron Schatz on 12 Dec 2012

45 comments, Last at 14 Dec 2012, 3:07pm by Nick Bradley


by nuk :: Wed, 12/12/2012 - 4:40pm

Outside of "Big Bang Theory", does the stereotypical geek even exist? I'm pretty damn geeky, but I was a varsity athlete and have a pretty (and geeky) wife. Do those disqualify me?

by tuluse :: Wed, 12/12/2012 - 4:48pm

BBT characters are not stereotypical geeks, they're bad writers' idea of what a stereotypical geek would be.

Just walk into a 300 or higher level computer science class in any university and you'll see plenty of people who conform to a lot of the stereotypes.

by nuk :: Wed, 12/12/2012 - 4:55pm

I spent five years in classes like that. My point is that you can't just pick out one thing like "radio DJ" and say you're not a stereotypical geek. Get to know any of those computer geeks and you'll find something that goes against the stereotype.

by tuluse :: Wed, 12/12/2012 - 5:11pm

Oh sure, no one conforms 100% to any stereotype. People are people.

by justanothersteve :: Wed, 12/12/2012 - 5:27pm

Yup. I work for the DoD in a giant building with no windows. Even though several are ex-military and have deployed to Iraq, Afghanistan, and several other less hot spots around the world. There is still a lot of overlap between that group and geeks (however anyone would want to define it). Heck, I've worked with guys who make Sheldon look well-adjusted.

by Nick Bradley (not verified) :: Fri, 12/14/2012 - 3:07pm

Yeah, I work in a giant DOD building with no windows too, and you're 100% correct.

by rageon :: Wed, 12/12/2012 - 6:29pm

I tend to think qualification as a "geek" is more inclusionary than exclusionary.

I admit to being a geek. I don't think I'm excluded by having gone to law school, playing in a band, working as a wedding DJ, or marrying well over my head. Once you've competed in a computer programming competition, you're in for life.

(BTW - nothing says "hey ladies" like a 2000 DigiKey Computing Competition t-shirt.)

by nat :: Wed, 12/12/2012 - 9:05pm

Isn't "Radio DJ" just A.V. Club writ large, anyway?

by erniecohen :: Thu, 12/13/2012 - 2:56pm

The main characters in BBT are not just geeks but nerds. And far nerdier than anyone I met in 4 years at Caltech.

by tuluse :: Thu, 12/13/2012 - 3:16pm

They're not anything, the writers constantly get things wrong that "real" nerds wouldn't. Simple stuff too, like saying they're going to play Mario 64 multiplayer.

by JonFrum :: Thu, 12/13/2012 - 4:49pm

Only a nerd would know that, and only a nerd would care.

by jimbohead :: Wed, 12/12/2012 - 4:50pm

I can think of one guy in my chemical engineering program, but only one (as long as we're not considering high school, which i think appropriate). And that was a decent-sized program. I also feel it necessary to point out that this guy was average at best.

by Clemson Matt (not verified) :: Thu, 12/13/2012 - 3:15pm

I don't think Chemical Engineering counts. I'm a ChemE, and we were the cool engineers on campus.

The electrical engineers were the geeks.

And yes, stereotypical geeks exist. See DragonCon attendees. And my wife's nephew. He's a computer engineer, not a theoretical physicist, but he's alot like Sheldon. We even assume if he ever reproduces it will be by mitosis.

by drobviousso :: Wed, 12/12/2012 - 6:33pm

Yes, the stereotypical geek exists.

by andrew :: Wed, 12/12/2012 - 7:02pm

Remember the Venn Diagram that explains all this.

According to this, the three pillars are: Intelligence, Obsession and Social Ineptitude.

Those with both intelligence and social ineptitude are dweebs.
Those with both obsession and social ineptitude are dorks.
and those with both intelligence and obsession are geeks.

Someone with all three traits is a Nerd, something which you avoid with your lack of social ineptitude. I can't speak for Aaron's social skills, but I think it fair to say football outsiders is a product of both intelligence and obsession.

I was about to try to map this into an Ultima avatar virtue scale but I realize that would probably land me in the middle.

by tuluse :: Wed, 12/12/2012 - 7:49pm

What are the virtues of geekdom?

by Aaron Brooks Go... :: Thu, 12/13/2012 - 12:22pm

To crush your enemies, see them driven before you, and to hear the lamentation of their women.

by JonFrum :: Thu, 12/13/2012 - 4:51pm

Because after all - they get women.

by Anon. (not verified) :: Wed, 12/12/2012 - 9:59pm

Well played.

by Jerry :: Thu, 12/13/2012 - 4:42am

My experience has been that to the general population, all of us go in the nerd bucket. Within the community, there's a lot more differentiation (in a couple of senses, he punned geekily), and most of us don't think of ourselves as total nerds.

by Kevin from Philly :: Thu, 12/13/2012 - 12:02pm

Wrong - it's not what WE think, it's what hot chicks think. Unfortunately, that puts me way, way down below "geek".

by Aaron Brooks Go... :: Thu, 12/13/2012 - 12:23pm

Do you bite the head of live animals?

If not, you're not a geek.

by bravehoptoad :: Thu, 12/13/2012 - 1:07pm

No more than absolutely necessary.

by Will Allen :: Wed, 12/12/2012 - 6:36pm

The injury reduction stuff is wonderful to contemplate, and I hope it continues to pan out.

by Jerry :: Thu, 12/13/2012 - 4:45am

When the Jaguars' owner's son is in charge of analytics, it seems likely that football men who'd ordinarily be dismissive will have to embrace new ideas if they want to keep their jobs.

by ChrisS (not verified) :: Thu, 12/13/2012 - 11:09am

I had the opposite reaction. If its the owners son the results will be treated as the product of nepotism and disregarded.

by Kevin from Philly :: Thu, 12/13/2012 - 12:05pm

Owner's son or not, should anything that the Jags are doing not be disregarded/avoided?

by Andrew Potter :: Thu, 12/13/2012 - 12:10pm

Unless the team starts winning. I expect it ultimately to all depend on that. If the Jaguars take an unorthodox approach to game strategy and in so doing make the playoffs when they otherwise wouldn't, that's by far the most likely - and perhaps the only - way to convince NFL people to shed their preconceptions on this. Even then, I'd expect to see it initially dismissed as a fluke and unsustainable until and unless it's proven to win them multiple games over multiple seasons - and, as I mentioned, to get the franchise into the playoffs. Winning four games instead of two won't make a big enough impression.

Of course, however progressive their game strategy is, the biggest reason the Jaguars aren't in the playoff hunt is a very traditional one. Blaine Gabbert is not currently, and may never be, good enough to consistently play quarterback in the NFL. The cleverest game strategy in the world won't fix that.

by BJR :: Fri, 12/14/2012 - 8:01am

For a bad team like the Jags, trying to maximise their chance of winning games through optimal game-strategy may well lead to more blowout losses, especially over a small sample. For example, when they play the Patriots next week, the Jags stand virtually no chance of winning the game if they stick to conventional strategy. However, if they adopt high-variance strategies (onside kicking, no punting and generally making aggressive play calls) they could improve their chance of winning the game by, say 5%, but would also substantially increase their chance of losing by 30+. Would their coach have the courage to do that, or be happier to accept a 14 point loss?

by erniecohen :: Thu, 12/13/2012 - 11:04am

- I am appalled to see SI fawn over what amount to minor tweaks to what was essentially done in "The Hidden Game of Football" without even mentioning that groundbreaking work. (DVOA doesn't fall into this category, but things like WP and 4th down strategy analysis do.) And I would have expected one or all of the article subjects to point it out to them. It's like a replay of the media frenzy over Romer's article.

- I don't understand how non-military GPS, which has an uncertainty on the order of 3m (and that only under favorable conditions), can possibly give useful information about the accuracy of route-running or the closing speed of defensive backs.

by Rikki (not verified) :: Thu, 12/13/2012 - 11:40am

Well, all gamblers where it is legal, put some money on Army, Navy and Air Force...

by Aaron Brooks Go... :: Thu, 12/13/2012 - 12:25pm

Virgil Carter was publishing about 4th down strategy and the value of field position a full decade before the Hidden Game of Football.

by erniecohen :: Thu, 12/13/2012 - 2:46pm

While Carter did indeed analyze the value of field position, and did recommend running on 4th and short at the opponent's goal line, he did not really talk much about the value of going for 4th downs in other situations, or about win probability. But you're right, he should also have been mentioned.

(Interestingly, the thing that he did analyze that I haven't seen elsewhere is the question of when to call timeouts at the end of a close game. Of course the payoffs are rather different now from what they were in 1971.)

by Aaron Brooks Go... :: Fri, 12/14/2012 - 11:32am

What's hilarious is that for as much as Andy Reid gets laughed at for his clock-management, what he does is shockingly similar to what Carter advocated.

by Aaron Schatz :: Thu, 12/13/2012 - 2:01pm

Actually DVOA is somewhat based on the Hidden Game of Football also. It adds in the element of comparing to a baseline and adjusting for opponent. I do think I mentioned Hidden Game to Albert in the interview I did, but you know how it is, if he put in everything I mentioned the article would be 20 pages long.

by CBPodge :: Thu, 12/13/2012 - 11:09am

Aaron, you do get referred to as a 10-year old maths prodigy at one point. I suspect that alone qualifies you as a geek!

Nice article though.

The point in there about the current generation of coaches being brought up through the 70s when conservatism was good and aggressiveness was bad does raise an interesting point - how long do you think it will take for teams to actually become more statistically orientated in practice? Is it a case of the coaches currently being brought up in this more statistical age getting promoted to levels where they are able to implement it?

I doubt the Billy Beane scenario will ever happen in football, simply because I can't imagine a football team being in the same position that the As were due to revenue sharing etc. They'll be looking to gain what advantage they can, but the A's situation seemed to be more along the lines of "well look, we don't have any choice but to try this - we can either do the same as everyone else and suck, or try something completely radical and probably suck" I can't really see an NFL team being in that situation.

by Jerry :: Thu, 12/13/2012 - 7:14pm

Michael Lewis always claims that Moneyball was about how Beane took advantage of market inefficiencies, and that sabermetrics were just the means to that end. A classic NFL example would be the Steelers being one of a few teams playing the 3-4 and being able to draft undersized defensive ends relatively late and then turn them into OLBs.

And any team that can come up with metrics that can help them evaluate personnel will have an advantage, however small, over teams that don't.

by Stats are for losers (not verified) :: Thu, 12/13/2012 - 2:01pm

I'm glad you linked to this, since that issue of SI went straight into the recycling when I saw the cover.

by BJR :: Fri, 12/14/2012 - 7:49am

'Straight in the recycling' doesn't have quite the same effect as 'straight in the trash'

by Bright Blue Shorts :: Thu, 12/13/2012 - 5:50pm

My understanding of this stats stuff is decidedly average ... but the thing I wonder about these probability tables is how well they apply to different teams.

I mean if the standard probabilities tell you to go for it on 4th down, is that still relevant even if you have the best offense or the worst offense in the league? Doesn't the probability change based on your offense, as well as the quality of your other units (i.e. defense/special teams) and the opponent's units?

by Ryan D. :: Thu, 12/13/2012 - 6:51pm

Absolutely. This is why it's crazy to watch Ron Rivera trot out the worst special teams on the planet to punt the ball away and let his poor exhausted defense try to defend a miniscule lead when he has Cam Newton/Jonathan Stewart/Mike Tolbert available to convert a 4th-and-short at midfield in the final minute of a game.

Punting to Atlanta (in their first game) and to Tampa Bay (in their second game) was a terrible decision both times because of the quality of his offense, especially in short-yardage power situations, and the decided lack of quality of his (prevent) defense. His defense gave up both games in the final minute, when his offense would have almost certainly kept him from losing at least one of the two games had he simply tried to convert.

by Aloysius Mephis... :: Thu, 12/13/2012 - 11:19pm

And yet, the heat Rivera took for that decision didn't amount to much. It’s astonishing the level of vitriol coaches get subjected to when they make an aggressive fourth down call and get a bad result, and aggressive fourth down calls that go badly are remembered and referred back to for years. But making a cowardly fourth down call and getting a bad result draws muted criticism at worst, and then is quickly forgotten. I guess it’s in the nature of human society that it's better to be wrong than to be unconventional.

by Bright Blue Shorts :: Fri, 12/14/2012 - 6:23am

And of course it follows on that for the journalists it's easier to criticise along with the crowd than to risk criticising against it. To do that might open themselves up to putting their own jobs/reputations at risk!

Thank-you for both confirming what I thought was the case about individual team differences, and not just that I don't understand the stats well.

by Aaron Brooks Go... :: Fri, 12/14/2012 - 11:36am

Atlanta only made it into scoring position because Matt Ryan threw into triple coverage and got away with it.

The odds didn't expect Matty Checkdown to turn into Favre.

by Ryan D. :: Fri, 12/14/2012 - 2:10pm

Well, it was the final minute of the game, and the Falcons were on the 1-yard line. They didn't have a lot of options other than "throw-deep-and-pray."

Also, no coverage that includes Haruki Nakumura can be called triple-coverage, unless there are three other defenders, plus him.