Eight weeks ago, the Vikings were undefeated and ranked No. 1 in DVOA. Now they're 6-6 and probably done. What went wrong?
17 Nov 2011
by Aaron Schatz
So, regular readers know we've struggled all season trying to explain why are metrics are so uniquely positive about the 5-4 New York Jets. We have the Jets significantly higher than any other metric. Last week they were number one, even this week they are number three.
A lot of people also know that part of the problem is related to the Week 5 loss to the Patriots, where the DVOA system ended up scoring the Jets as the better team. (With current opponent adjustments, the Jets come out with 37.9% total DVOA in that game. The Patriots come out at 14.3% total DVOA.) What's particularly odd is that the system rates the Jets offense as above-average for the game, even though the Jets went three-and-out on seven different drives.
Some readers (for all I know, it may only be two of you, but whatever) are crying out, "Aaron, show us the numbers! Unless we see them, we can't believe!"
Therefore, I present to you a bit of a look into the black box that is DVOA. I can't show you how I figure each baseline, or what the specific equation is to figure play value for each play. However, this table gives you every Jets pass and run from that Week 5 game, plus the one penalty that gets counted in DVOA (Matt Mulligan's false start). It shows you the play value, the baseline, and the difference. To save room, I'm not showing opponent adjustment, so this is technically just VOA, not DVOA.
Go through it and you'll see that there really aren't any plays where the numbers stand out as obviously wrong. What sort of happened is that when we count each play equally, the successes the Jets had on three touchdown drives end up outweighing the failures they had in all those three-and-outs. The three touchdown drives make up 27 plays. The rest of the Jets offense makes up only 26 plays.
Lines in red are red zone plays where "value over baseline" is multiplied by 1.25.
There are three things we may be able to test in the offseason that could make DVOA more accurate for this game.
1) It is possible we are giving too much credit for "partial success" plays like a 4-yard gain on first-and-10.
2) It is possible that the bonus for red zone plays is too large.
3) It is possible that we need to include some sort of drive-measuring component that penalizes teams who go mix long drives with three-and-out drives and don't have many medium-length drives.
I hope this helps explain why the Jets come out so high for Week 5, and helps satisfy those of you who are asking questions.. And remember, the Jets' high DVOA this season isn't only caused by this single game. Take out Week 5 for both the Jets and the Patriots, and the Jets would still be third in DVOA. Reverse the Jets' offensive rating and the Patriots' defensive rating for every play during the Week 5 game -- thus making it count as a game worth around -34.7% for both units -- and the Jets would still be fourth in DVOA, falling behind only the Patriots.
Whatever is going on with the Jets this season -- whatever they are doing that the DVOA system likes better than other advanced metrics -- they're doing it in pretty much every game.
69 comments, Last at 18 Nov 2011, 3:05pm by Jeff Fogle