clock menu more-arrow no yes mobile

Filed under:

Countdown to Study Hall: College football and the Sloan Conference

New, 9 comments

My first book, Study Hall: College Football, Its Stats and Its Stories, will be released on (or around) next Monday.

Bill Connelly

But there is virtually no college football presence at Sloan. There is a research paper proposing a “new football efficiency metric,” but it is simply yards per point, which erstwhile college football magazine publisher Phil Steele has been using for a couple of decades.

In a few different, often unrelated, panels, someone will mention how random football is and how difficult it is to get a grasp for compared to other sports. “The ball’s not even round,” says professional gambler Bob Voulgaris in a Sloan panel. Voulgaris sticks mostly to NBA games because of both his enjoyment of the sport and the way he perceives the sport’s general reliability.

In The Success Equation (2012), author Michael J. Mauboussin went into a little more detail but basically reached the same conclusion. His analysis determined that while the contribution of luck to an NBA basketball game determines about 12 percent of the result, it was closer to 38 percent for an NFL football game.

Randomness aside, football is generally a more complex game as well. You’ve got 11 players on each side of the ball, and some will only touch the ball if something went drastically wrong, so evaluating individuals, with different roles and jobs on a given play, is somewhere between difficult and a total lost cause. Whereas in baseball you’ve got some true, well-defined outcomes – player pitches ball, player hits ball, player fields ball – football is quite a bit more complicated.

“You’ve got 22 guys running around at the same time, and it’s incredibly difficult to figure out how those things fit together with numbers,” says Rob Neyer. “The down, the distance, the should-we-go-for-it stuff is easy, but how do you figure out how the left guard is doing? It’s really hard. In basketball, you face the same sorts of issues, only there are half as many players. Baseball is easiest because you have this series of distinct actions that are relatively easy to put into a sequence.”

Indeed, some of the things that make football such an appealing sport to so many Americans can cloud the perceived usefulness, or direction, of advanced football stats. “There is a set number of outs in baseball,” points out The Hidden Game’s John Thorn. “The absence of a clock, the unpredictability of the outcome until the very last – these are virtues for baseball. But they are not for football, and that might be better for the modern fan.

“In football, stats can help you win in the offseason as much as anything. And they can help you understand the ramifications of certain strategies – going for two points here, going for it on fourth-and-2 there.”

From Chapter 4 of Study Hall, "You, Me, and Stats."

I've attended the Sloan Sports Analytics Conference each of the last two years, and I base part of Chapter 4 of Study Hall around my March trip to Boston. The Sloan Conference is an interesting beast. It is very much a "Let's celebrate ourselves" experience, but it is also a time for interesting ideas to percolate and for the stat movement, so to speak, to claim a couple of days in the spotlight. It's a pretty big deal at this point, with professional teams sending representatives by the truckload, just in case the Next Big Idea pops up. (All but one NBA team was in attendance this past March, which is staggering. And yes, the one team not there was the Lakers, because of course.)

While NFL has certainly begun to find a presence at Sloan, college football's is almost nil. As referenced above, the measure discussed during the only college football presentation at the conference can be rather quickly shot down. It is rather perplexing -- I mean, there was a Motorsports Analytics panel in 2013 -- but there's no denying that college football is complicated and weird, and honestly, it's hard to know where to start with the sport.

So today's discussion topic, the second-to-last one of the countdown (!), is this:

Let's say there is a College Football Analytics panel at next year's Sloan Conference. And let's assume that the makeup of the panel is pretty standard -- one or two nerds, one or two current/former coaches, and an ESPN personality. What would make for the most interesting hourlong discussion? The "going for it on fourth-and-2" stuff has indeed been discussed in random NFL-based panels in the past, so there wouldn't be much point for that. The college numbers are different but not different enough to matter.

The advanced stats chapter (and yes, there's only one) in Study Hall put a focus on how coaches could use such measures as part of the scouting and self-scouting process; is that a worthwhile angle? Maybe focus on talent identification and projection (from high school to college or from college to pro)? Stats as part of fan engagement and understanding (this is an ESPN-heavy conference, after all)? Have Mark May rant about Johnny Manziel bringing shame to the game for an hour?

(That last one was a joke. I hope.)

I'm reasonably confident that college football will have a seat at the Sloan table next year; the panel itself, however, could go in any number of different directions.