The idea of a passing premium isn't new, at least not in recent years. It's been studied fairly extensively in NFL circles, although less so in the college ranks. There are some reasons behind this - there aren't as many people doing that kind of analysis now for one, less so since CFBStats dropped their down/distance catalog (which still saddens me), and the diversity of offenses leads to likely divergent variants for two.
Okay, that was a lot of writing. Let me back up a second. I should start with a simple question: what the heck is the passing premium anyway? Well, we take a couple of basic, easily obtainable stats - average yards per running attempt (let's go with ypra), and average yards per passing attempt (and yppa). We then subtract ypra from yppa and see what the number is. That gives us an idea of what the passing premium for a certain offense is. In reality, the calculation is a little more detailed - I talked about it a bit last year, so go there for the dirt.
Great! That's useful information, in a vacuum. However, we're less concerned about what the premium is as much as it what it tells us about how an offense works, and more importantly, can it broadly work better than it is now? There is probably an optimal number for each offense, especially in college, and my suspicion is that it's somewhere around 1, maybe a little higher than that - Smart Football wrote on this a while ago and kind of assumed the same number, and for the NFL, it's a little closer to zero but not horribly so. For our purposes, the difference between 1 and 1.1 is pretty low, as we'll find out.
So, what we do we do once we know our passing premium? We can use it to broadly assess how our offense is operating - i.e., are we calling broadly the right number of passes and broadly the right number of runs? Obviously, context matters, and don't think I'm ignoring that, but this is a bit of a general stat as I'm presenting it; for a 3rd and 1, you want a play that consistently gets 2 yards and consistently doesn't get less than 1 yard. So, for now, let's think a bit broadly.
Matter of fact, let's think three years (or two years and four games) broadly. I compared Georgia's yrpa, yppa, and passing premium with Tennessee's yrpa, yppa, and passing premium over the 2010-12 range, and I threw in the number of total plays for good measure (all of these numbers are courtesy of cfbstats, in case you had any doubt):
|Georgia Run||Georgia Pass||Georgia YPRA||Georgia YPPA||Georgia Premium|
|Tennessee Run||Tennessee Pass||Tennessee YPRA||Tennessee YPPA||Tennessee Premium|
See, data! Now, what can we learn from it?
- Tennessee is actually pretty balanced on the whole. I know, I'm as shocked as you are about it, but those run/pass balances are pretty even since 2010.
- Georgia will run. A lot. I didn't run this for a bunch of teams, but at first glance, that looks like a lot of running. Like, to the tune to 57% in 2010 being the least amount of run.
- Georgia should probably pass a bit more than they do. Again, this is just looking at the numbers, but it also makes sense. (This is what I was talking about a few paragraphs ago about how a passing premium of 1 or 1.1 won't make much difference.) Again, we don't quite know the level of risk that Georgia OC Mike Bobo is willing to tolerate, but that ....is a lot of risk he isn't willing to take. Matter of fact, with Aaron Murray at the controls, I'd be tempted to pass quite a bit more than they do. As it stands, their offense is pretty dangerous - close to record-setting for Georgia - so the fact they're not capitalizing on that is probably for the best.
- Meanwhile, the same goes for Tennessee. Save the gnashing of teeth. We've got numbers now, and if the premium Dooley is comfortable with is in the 4 range, well, that should be adjusted. That means, broadly, Tennessee isn't taking advantage of the weapons they have in the passing game enough. Considering that we're lucky enough to see a litany of options in the passing game, this really doesn't make a whole lot of sense to me.