Yeah, we're back at the playcalling again. This time, we're going to take a little bit of a different look at it (and, unlike so many of this series last year, this actually won't have pretty pictures, just walls of text) - this time, we'll look at the passing premium, which involves a little bit of game theory. Don't worry, we'll have pretty pictures soon enough.
What's Game Theory, and Why Does It Matter?
Game theory, long story short, is what happens when an offensive and defense coordinator decide to get together and call a bunch of plays against each other. The outcome - and success - of either guy depends in part on the actions of the other, and of course of the 22 guys on the field. (Football is more of a zero sum game if we want to get down to brass tacks, but that's neither here nor there.) Basically, both the offensive and defensive coordinator have a set of moves (plays) they can use, and they will both use their moves in such a way as to optimize their chances of success.
For our purposes, we're just going to consider the offensive side of the ball, and we'll put plays into two buckets: run and pass. We know there are a bunch of variants between run plays and pass plays - power does not equal stretch, four vertical does not equal all-curl, who let all the screens in here?, and so on - but for our purposes, those distinctions don't make that big a difference. If we're an offensive coordinator, we want to do one thing - get touchdowns. To do so, we need to get yardage (again, we'll just generalize heavily and say that TDs are a function of yardage), and to get yardage we need to maximize our gains from both running and passing plays. We can figure that out broadly by looking at average yards per rush and average yards per pass attempt. If they're about equal, presto! We're done, right?
Not so fast.
Oh, the Negatives
Yeah, we can't just look at average yards per rush and average yards per pass attempt; it's not quite that simple. Think about the bad things that can happen with a running play; the play can get blasted for negative yardage, no gain, or a fumble. Then again, it's basically the same for a pass play - except we're adding interceptions and sacks, too. The negative gain and no gain plays are essentially covered in the average yards per rush and yards per carry metrics, but what to do about the interceptions, fumbles, and sacks?
We can solve the fumble and interception problem by assigning a negative yardage value to those plays; the actual number isn't a big deal, but it's there. (Honestly, you can probably get rid of fumbles entirely; for simplicity's sake, I'll do that - it affects both run and pass numbers anyway.) Similarly, sacks get a negative value. However, passing TDs (and passing TDs only) can get a slight boost. Advanced NFL Stats has the details on how they came up with this.
Anyway, after doing all this math, you end up with two numbers: an average yards per rush and an average yards per pass attempt.
Oftentimes, the yards per pass attempt number will be a fair bit greater than the yards per rush attempt. This should imply some time of playcalling imbalance, right? After all, we know the theoretical optimum playcalling mixture would be such a mixture that those two averages would equal, right? It's not quite that simple. Remember that a passing play has more inherent risk than a running play (Woody Hayes is pointing and nodding at this), and so it should naturally have some kind of premium attached. The exact amount of this premium varies; Advanced NFL Stats found about a 0.5 yppa advantage, while Smart Football ran a few case studies a few years ago and found that in the college game, it's close to 1 yppa average advantage or a bit more. (He also followed up with an excellent set of comments talking about risk in more detail than I have here.)
However, it's not the case studies that interest us. It's this section that we care about:
The idea is if you are a very good passing team you pass most of the time, then you run when it is favorable and see positive results without having had to practice it too much. Same goes vice-versa--we all know how dangerous play-action passes are from heavy run teams, especially say a veer option team.
That first sentence sounds just a wee bit like what we're expecting, right?
Bringing It To Chaney
Our eventual goal is to be an elite passing team - or at least that's what the appearances would look like so far. As a result, we should expect defenses to be more concerned about the pass. This in turn opens up Tauren Poole to be a relative beast. It may have little to do with Poole improving, though - it may be because Tyler Bray and Justin Hunter have developed the kind of connection that romantic comedies are about. Of course, the converse may be true - Poole could have a great year, which free up Bray and Hunter to do their thing. Regardless, yardage per play needs to be maximized in such a way to properly account for risk (and if Bray doesn't get a handle on his interceptions, there will be a lot of risk.)
With that being said, it falls to Chaney to devise a gameplan to reach this theoretical maximum, and it's a useful tracking tool on our part to see if we're getting there. To some extent, if the passing premium is excessive (think greater than 2 yards) we're probably not passing enough (WOO!), and if the passing premium is minor (think less than, oh, 0.3 yards or so) then we're probably not running enough (woo!). That sounds counter-intuitive, but it's not. If we're averaging appreciably more yards per pass attempt, it makes sense that we should do that more until the defense adjusts (thus bringing the yppa down), right? Similarly, the converse is true as well. As long as we're in some kind of reasonable range, we should be somewhat okay.
The concept of a passing premium is a tool - it's a measurement of how effectively we're calling plays. It makes a lot of sense that we'd use it as such, no?