DVOA Analysis
Football Outsiders' revolutionary metrics that break down every single play of the NFL season

Week 2 DVOA Ratings

Week 2 DVOA Ratings
Photo: USA Today Sports Images

by Aaron Schatz

So far, 2013 has not been a very good season for the Football Outsiders preseason forecast. Actually, 2013 so far hasn't been a very good season for most anybody's preseason forecast... except at the very top. Many members of the NFL commentariat predicted a Super Bowl appearance for Denver, Seattle, or even both. Denver and Seattle started the season with the two highest mean win projections from Football Outsiders. Denver and Seattle have been the two best teams in the NFL through the first two weeks.

After that, things look a little bit goofy. Right now, Kansas City is third in VOA. The Chiefs were last year's 30th-ranked defense. This year, they rank No. 1. Obviously, we're not including opponent adjustments after just two weeks, and eventually there's going to be a big adjustment for every team that plays Jacksonville. Still, it certainly looks like the Chiefs have done a lot to plug the holes between all of those Pro Bowlers from a year ago. (It helps that one of those holes, Dontari Poe, certainly seems to have taken his own leap to a Pro Bowl level.)

Miami at 2-0 is also an early surprise, and the Dolphins have been very balanced. They rank in the top ten in all three phases of the game.

On the other hand, the six lowest teams in the league include three teams we had projected to go to the postseason -- Washington, San Francisco, and the defending champion Ravens -- plus the New York Giants, who everyone expected to be in playoff contention. Again, opponent adjustments will really change things here. The 49ers have played two of the best teams in the league. The Giants got clobbered by Denver, clearly the top team in the league right now. But did they have to get clobbered by quite so much? Yikes.

What's interesting here is that the strange results of 2013 so far have generally seemed like strange results because of how we projected teams coming into 2013 and what we know about teams historically. One thing we haven't seen is teams that are particularly deceiving at 2-0 or 0-2. Last year, for example, the Cardinals were the third-worst 2-0 team in DVOA history. This year, every 2-0 team has a positive VOA and every 0-2 team has a negative VOA.

That fact probably surprises you if you've been watching the New England Patriots. Given how bad the Patriots have looked in their first two games, it wouldn't be ridiculous to expect them to rank among the "worst 2-0 teams ever." You may be shocked to see them ranked 11th in VOA through Week 2. The reason they are so high? The units you don't see. We've all seen the Patriots receivers play like a bad high school team. However, you may not have noticed that the rest of the time, the Patriots were playing shockingly good defense and their usual great special teams. So the Patriots' ratings look completely wacko: 27th in offense, fourth in defense. This will change going forward, of course. Opponent adjustments will ding the defense because shutting down Geno Smith isn't that difficult. But at the same time, opponent adjustments may boost the early performance of the New England offense because they played the Jets. Yes, we saw the Jets secondary didn't play very well last Thursday night, so perhaps the Jets defense won't return to the greatness I was predicting before the season began. But that team has a great front line, and whatever else may be wrong with him, Rex Ryan still designs an excellent defensive scheme.

(The other change going forward: We'll have to see what the injury to Matthew Slater does to New England's special teams. Over the past few years, Chicago and New England have been the two positive exceptions to the rule that special teams performance is rarely consistent from year to year.)

One other team that's interesting to look at is Washington. There's a good debate to be had about how much to write-off Washington's ability to score points in the second half of games where they are getting blown out in the first half. The policy at Football Outsiders has always been to compare like with like. We want to adjust the offensive rating of a team playing from way behind by comparing them to the average team playing from way behind. And what we've found is that the difference between how well an offense plays when way behind and how well an offense plays in a close game is MUCH smaller than most fans would expect. Therefore, we end up giving credit to an offense like Washington's that comes back from a big deficit, even if they don't come back enough to win the game (in the case of Week 1) or make it close (in the case of Week 2). The DVOA system is based on the idea that this strong play is still evidence that the Washington offense is a good offense; they just happen to have a crazy first half/second half split through two weeks. And so, Washington below is listed as sixth in offense and dead last in defense.

[ad placeholder 3]

Obviously, a lot of fans and even analysts will feel differently about this. And so, look at our quarterback stats page and the difference between DVOA and ESPN's Total QBR for Robert Griffin III. (Yes, I've got it listed as DVOA in some places, VOA in others... it's the same thing this early in the season, don't get caught up in how many letters are in the name.) Right now, Griffin ranks 11th in the league in YAR (total value) and 14th in VOA (value per play). Yet ESPN's Total QBR ranks him a dismal 31st out of 33 quarterbacks. That's a huge difference, based almost entirely on the question of what we learn or don't learn from Griffin's strong offensive play in the second half of the last two games. In the first half of games, Griffin right now has 1.9 QBR. In the second half, he has 75.9 QBR. Washington was never above 10 percent win expectancy after halftime in either game, so basically, RG3 starts out as Blaine Gabbert and then turns into Aaron Rodgers once his defense has enjoyed 30 minutes of total nuclear meltdown.

In many ways, the difference between VOA and QBR is caught up in how much you want your rating to be predictive vs. descriptive of the past. Both VOA and QBR are neither purely predictive nor purely descriptive. They both are sort of somewhere in between, but VOA is closer to predictive and QBR is closer to descriptive. This is not a controversial statement, and the folks at ESPN Stats and Information would certainly agree. They added the clutch weighting to QBR in order to have it better line up with a list of which quarterbacks were helping their teams win close games, even if those quarterbacks weren't necessarily actually playing better than other quarterbacks -- who may play better in the first half, or may share a locker room with a terrible defense. It's two ways to analyze the same data. Neither is more right. It's the difference between them that tells the story, or at least part of the story. In this case, it's the story of what is and is not going well for RG3 this season. (This is why I've always been an advocate of more stats, rather than trying to create one rating that synthesizes a player's entire performance into a single number.) 

As to the readers who will of course slag Total QBR because they are suspicious of anything created by the Worldwide Leader: Have you considered reading Basketball on Paper?

* * * * *

During the 2013 season, we'll be partnering with EA Sports to bring special Football Outsiders-branded items to Madden 25 Ultimate Team. Each week, we'll be picking out a handful of players who starred in that week's games. Some of them will be well-known players who stood out in DVOA and DYAR. Others will be under-the-radar players who only stood out with advanced stats. We'll announce the players each Tuesday in the DVOA commentary article, and the players will be available in Madden Ultimate Team packs the following weekend.

The Football Outsiders stars for Week 2 are:

  • WR Steve Johnson, BUF: 8 catches on 10 passes, 111 yards, game-winning touchdown, led all WR in DYAR (Defense-adjusted Yards above Replacement).
  • LG James Carpenter, SEA: Came off the bench to abuse the San Francisco defensive line as the Seahawks ran for 172 yards.
  • DT Dontari Poe, KC: Two sacks and overall domination of Cowboys rookie Travis Frederick.
  • DT Kawann Short, CAR: Three quarterback hits, two assists, and a tackle.
  • CB Richard Sherman, SEA: Only allowed one completion in man coverage according to game charting, which was a "failed completion" that didn't convert third-and-long; also had an interception. (He did allow a second completion, but that was in a zone.)

If you came to our article because you love Madden Ultimate Team and wanted to see who would be the special Football Outsiders players this weekend, and you want to know more about Football Outsiders advanced stats and analysis, please feel free to send us your questions (Contact Us) or ask the regular readers by contributing to the discussion thread below.

One thing I neglected to mention when we first announced this deal: special Football Outsiders players are available not only on XBox 360 and Playstation 3 but also in the mobile (iOS/Android) version of Madden 25 Ultimate Team. It's Football Outsiders on your phone, and Football Outsiders in your pants!

* * * * *

All stats pages are now updated with 2013, including playoff odds, snap counts, and for the first time this season, offensive line and defensive line. The FO Premium splits database will also be updated with 2013 sometime this evening... and, as a special bonus, it will be updated with 1989 and 1990 as well. Also, you may have noticed we have a new green drop-down stats menu at the top of the site. It condenses the stat listing to make it easier to find the right page. Mouse over the listings for either Drive Stats or College Stats to get a more expanded list of those two categories.

* * * * *

These are the Football Outsiders team efficiency ratings through two weeks of 2013, measured by our proprietary Defense-adjusted Value Over Average (DVOA) system that breaks down every single play and compares a team's performance to the league average based on situation in order to determine value over average. (Explained further here.)

OFFENSE and DEFENSE VOA are adjusted to consider all fumbles, kept or lost, as equal value. SPECIAL TEAMS DVOA is adjusted for type of stadium (warm, cold, dome, Denver) and week of season.

There are no opponent adjustments in VOA until the fourth week of the season, which is why offense and defense are listed as VOA right now rather than DVOA. As always, positive numbers represent more points so DEFENSE is better when it is NEGATIVE.

DAVE is a formula which combines our preseason projection with current VOA to get a more accurate forecast of how a team will play the rest of the season. Right now, the preseason projection makes up 75 percent of DAVE.

To save people some time, please use the following format for all complaints:

<team> is clearly ranked <too high/too low> because <reason unrelated to DVOA>. <subjective ranking system> is way better than this. <unrelated team-supporting or -denigrating comment, preferably with poor spelling and/or chat-acceptable spelling>

1 DEN 72.1% 2 33.0% 1 2-0 47.1% 2 -13.6% 7 11.4% 1
2 SEA 47.9% 9 27.9% 2 2-0 5.8% 13 -32.6% 2 9.5% 2
3 KC 44.3% 1 -1.4% 17 2-0 7.9% 12 -37.5% 1 -1.1% 18
4 GB 30.8% 16 25.5% 3 1-1 63.7% 1 31.8% 31 -1.1% 17
5 MIA 30.0% 3 0.9% 15 2-0 11.7% 9 -12.1% 9 6.2% 6
6 PHI 23.1% 8 4.4% 11 1-1 37.8% 3 21.9% 29 7.1% 5
7 CHI 22.4% 5 5.0% 9 2-0 13.6% 7 -7.1% 12 1.8% 13
8 BUF 21.9% 7 -5.6% 21 1-1 9.4% 10 -11.1% 10 1.5% 14
9 IND 14.8% 6 -1.5% 18 1-1 26.3% 4 9.7% 24 -1.7% 21
10 CIN 11.8% 18 7.1% 7 1-1 8.9% 11 -0.9% 15 2.0% 12
11 NE 10.9% 21 21.2% 4 2-0 -17.8% 27 -24.6% 4 4.2% 8
12 NO 7.9% 15 7.9% 6 2-0 -3.3% 19 -12.3% 8 -1.2% 19
13 ATL 6.9% 20 5.0% 10 1-1 12.4% 8 11.3% 26 5.8% 7
14 STL 3.7% 10 -11.9% 26 1-1 5.0% 14 9.4% 23 8.2% 4
15 TEN 3.2% 4 -12.1% 27 1-1 -17.2% 26 -20.0% 6 0.4% 16
16 HOU 1.5% 11 5.0% 8 2-0 4.8% 15 -9.4% 11 -12.7% 31
17 DET -2.8% 13 -2.5% 19 1-1 4.7% 16 -5.1% 14 -12.6% 30
18 SD -4.5% 23 -6.5% 22 1-1 24.8% 5 29.8% 30 0.4% 15
19 TB -7.4% 17 1.5% 14 0-2 -34.0% 29 -23.4% 5 3.2% 11
20 CAR -8.6% 19 9.6% 5 0-2 1.5% 17 6.7% 21 -3.5% 22
21 DAL -8.7% 22 -7.1% 23 1-1 -14.9% 25 2.2% 17 8.4% 3
22 NYJ -9.3% 14 -7.5% 24 1-1 -39.4% 31 -26.2% 3 3.9% 9
23 OAK -13.4% 26 -15.7% 29 1-1 -0.8% 18 2.4% 19 -10.1% 29
24 ARI -16.6% 25 -12.4% 28 1-1 -9.7% 21 2.3% 18 -4.6% 24
25 PIT -19.1% 28 -4.7% 20 0-2 -21.3% 28 1.5% 16 3.7% 10
26 MIN -24.0% 27 -16.9% 31 0-2 -13.5% 23 9.1% 22 -1.4% 20
27 BAL -29.3% 31 2.1% 13 1-1 -8.5% 20 12.7% 27 -8.2% 27
28 SF -32.2% 12 3.3% 12 1-1 -10.4% 22 13.8% 28 -8.0% 26
29 CLE -37.8% 30 -16.9% 30 0-2 -37.0% 30 -6.8% 13 -7.5% 25
30 NYG -39.5% 24 -7.9% 25 0-2 -14.4% 24 9.9% 25 -15.2% 32
31 WAS -45.5% 29 0.4% 16 0-2 15.7% 6 51.3% 32 -9.8% 28
32 JAC -74.5% 32 -35.0% 32 0-2 -67.6% 32 2.8% 20 -4.0% 23


161 comments, Last at 22 Sep 2013, 11:52am

85 Re: Week 2 DVOA Ratings

One quick comment I have is that the quick blog says that teams that do well on Offense in blowouts is about the same as offense in close games, but what I might argue or offer in a different way is that defenses have a different motivation. Are they really playing as hard when they're up 24. Is it that big of a deal to give up a TD when your offense is continuing to score and just keeping the distance?

Looking at the Washington adjustments that need to be made I'd say certainly that probably no adjustments need to be made to their defense DVOA but I'd say there needs to be some value adjusted to the offensive DVOA once down a certain amount (what that amount is I don't know).

Perhaps looking at Defensive DVOA when up by >16 points (2 scores when having to go for two) versus <14 points (Two normal scores) and using those differences in Defensive DVOA to adjust the offensive power of teams that excel when down by more than two scores.

96 Re: Week 2 DVOA Ratings

"...the quick blog says that teams that do well on Offense in blowouts is about the same as offense in close games."


That's not really what it says. It's actually "...the difference between how well an offense plays when way behind and how well an offense plays in a close game is MUCH smaller than most fans would expect."

They're only saying the difference isn't as big as they think most people would expect. It doesn't say there's no difference, or that they treat the offense equally in those cases.

And there IS some adjustment to the offensive team's DVOA when they are down a lot of points (based on the typical performance change in offenses for that scenario). So the change in defensive tactics is already accounted for there. (As is the change in defense itself, when accounted for being up by a lot of points.)

114 Re: Week 2 DVOA Ratings

To piggyback off what is being said here, people often forget that while defenses change strategy while they're up by a bunch of points, so does the offense that's working to get back. The offense have to throw the ball more, play more uptempo, and take risks down the field that they wouldn't take if they had the lead, in which they can play more conservatively. So things tend to balance out.

125 Re: Week 2 DVOA Ratings

Ouch. I just noticed that VOA has the Washington defense rating almost as badly as the Jacksonville offense. I'm sure opponent adjustments will see that rise a touch, but yeowsers.

131 Re: Week 2 DVOA Ratings

"Kickoff returns and punt returns are judged on return yardage only."

Watching the NE-NY game and then SF-Seattle game, it seems that this way of judging punt returns leaves out the yards gained from not allowing an unfielded ball to roll forwards. I would think that Tate's insane rushes into traffic to field punts probably saved a few yards, while the Jets had one the the worst non-fumbling punt return games I had ever seen (either no one was back at all, or Wilson would not even bother to field it, and when he did, he got tackled hilariously easily).

132 Re: Week 2 DVOA Ratings

Beyond the usual issue of sticking to the play-by-play, there's absolutely no way of even estimating how many yards a given punt would have rolled if not picked up.

143 Re: Week 2 DVOA Ratings

Jeez, SF looks terrible this year; Colts are sure to kill 'em this weekend!

Either that, or these results will look a bit different for each team this time next week.....

151 Week 2 DVOA Ratings/ Playoff Odds

I like that Jacksonville is so bad that you don't feel the need to include them on the payoff odds report. Well done FO, well done.

161 Re: Week 2 DVOA Ratings

Denver sitting up there with a monster DVOA is playing way over their heads, as evidenced by the best teams in the DVOA era being around 40.

Denver is playing on a unsustainable level and has only one way to go and that is down.

Expect a huge regression this week or next for Denver.

Same principle would apply to the Jags.

Have fun and enjoy the games.