FanPost

The Most Important Down

Troy Wayrynen-USA TODAY Sports

This one begins with a trick question: what is the most important down? Like any good trick question, the answer is arcane. From the Latin arcanum, it can roughly be translated as an enclosing, though it connotes something hidden. What this means for the purpose of this writing is that the answer is concealed, because the question points to a common framing, and the assumption - which is fair because questions shouldn’t be so nefarious, ergo the trickery part - is that the answer will fit somewhere in this frame.

Except here it doesn’t.

The most important down in football is any down in which a team scores points. Obvious, I know, but obvious in its simplicity, and so simple that it often goes underappreciated. Because of all the effort we spend trying to decipher game film, listening to coaches and analysts talk about different personnel packages and schemes, it’s easy to lose sight of the fact that the ultimate goal in this dumb game is to score more damn points than the other team. Extending this notion statistically, we can rest assured that the only stat that perfectly correlates with victory is point differential: the team that scores more points than its opponent will definitely, 100% of the time win that game.

Now of course the actual game isn’t that simple: coaches can’t simply dial up scoring plays on every attempt and, even if they could, there’s no guarantee that players could execute those plays perfectly every time. This inherent complexity is why coaches try to break the game down into more meaningful particulates: quarters of play, down to drives, down to plays. And luckily - it actually has nothing to do with luck, and everything to do with the insight and willingness of the analytics crowd to take raw data and make it meaningful - we have a stat that corresponds to this notion: EPA, or Expected Points Added. This statistic is designed to indicate whether a play makes it more (positive EPA) or less (negative EPA) likely that a team will score points on its next down. Not only is the statistic enormously helpful for gauging whether a play could be construed as successful, with respect to the actual goal of the game, but it is also (which should be unsurprising) the one statistic that most closely correlates with victory.

With apologies to those for whom that’s too introductory, what I’m trying to convey is that this is the framework needed to understand the non-trick answer to the question of this writing: by EPA, the most important down is the third one. (This sets aside 4th downs, about which a great deal has already been written, particularly about how most coaches are far too conservative about going for it on 4th down, which is why we’re setting it aside for this writing.) What emerges is that some of the wildest swings in terms of EPA occur on that down. And even though the formula for calculating EPA is complex (and also proprietary?), the notion it conveys isn’t: giving an offense a new set of downs will almost always improve their chance of scoring. Of course, this is only generally true - pinning an opposing team deep on its side of the field can actually have a better chance of leading to points than a quick run on a short yardage third down that gains a first, especially if the offense is closer to its own endzone than its opponent’s endzone. EPA also coincides to a degree with some definitions of Success Rate - one of the more common approaches is to say that a successful play is one that results in positive EPA.

This has been an excessively long introduction to counter what may be the popular narrative coming from the Seahawks victory on Sunday: that the key victory was running the ball, or bullying the Cowboys, or establishing an identity. This isn’t true - the key to victory was 3rd down conversion rate, and especially the differential. On Week 3, the Seahawks converted almost 44% of third downs, compared to 23% for the Cowboys. Of the Seahawks’ seven conversions, three were made on the first TD drive, and two on the second TD drive. That rate is up from 38% on Week 2, and from a horrific 17% on Week 1. In fact, of the Seahawks 8 TDs in 2018 so far, three of them have coincided with third down conversions.

The conventional wisdom behind Pete-Carroll-ball is that football is, at its best, a war of attrition: you win the game by imposing your will upon your opponent, a Clausewitz-inspired breed of victory in which, at the end of the day, your opponent is so afraid of you that they just can’t cope anymore. This philosophy holds that small, and ultimately unsuccessful, runs in the first quarter, even if they do not yield points, turn into larger, and increasingly successful runs in the second and third quarters. These runs tire out opposing defenses, and, as they become more successful, they allow the offense to take greater risks, especially when facing 2nd and short situations, which opportunities lead to more explosive plays, and thus to more points. Then, as opposing offenses fall more and more behind, they take greater risks but because of necessity rather than as the beneficiary of exhaustion, which leads to more turnovers.

The problem with this wisdom is that, increasingly, it doesn’t appear to be true. The first five run plays of the game resulted in yardage counts of 9, 3 (1st down), 5, 5, and 5. If the Cowboys defense was tired, they were tired early. The first TD drive resulted in run plays with yardage counts of 4, 4, 3, and 0; the Cowboys D seemed to be less tired 17 minutes later, and all of the Seahawks conversions came on passes, which resulted in 6, 12, 16, and 16 yards. The second TD drive, which started 8 minutes later, resulted in run plays with yardage counts of 5, 4, 2 (1st down), 1, and 0; somehow, the Seahawks running seemed to reinvigorate the Cowboys D, until a 52 yard dime on poor coverage put them back to sleep. And, of course, the last three possessions resulted in 3 three-and-outs, in which run plays resulted in 0, 6, 0, 5, 1, 8, -1, 11, 1, 5, 4, and -1.

No, what seems to be more and more true is that football is a war of deception and execution. The Seahawks showed just as much commitment to the run in the beginning of Weeks 1 and 2 as they did last Sunday; the difference was that, in Week 3, they converted more. Towards the end of the game, when the Cowboys knew they were going to run the ball as much as possible in order to protect their lead, they were less successful than not, which goes to show simple and true it is that you shouldn't do what the other team expects you to do, unless you can do it way better than they can. (The Seahawks cannot, and did not.) In the first two weeks, Russell Wilson was not protected nearly enough in those games, which either directly contributed to an inability to convert on 3rd, or which led to sacks that made those 3rd downs so much harder to convert. Better, and less injured, defenses made it harder to scheme open a WR corps that, without ADB, appears to be mediocre at best. And falling behind more or less required the Seahawks to abandon the run in the first two weeks - the converse of this lies at the heart of what Carroll wants to do, so it only makes sense that it would work this way in the reverse as well. So let’s not get too carried away with our praises for victory a la Pete Carroll in the face of so much tangible evidence that his style, identity, philosophy, whatever you want to call it may no longer be the best way to actually win games.