Writers of Pro Football Prospectus 2008

Most Recent FO Features


» The Deep Ball Project

Guest columnist John Kinsley breaks down the tape of every deep pass in the NFL in 2017 and comes away with a shocking conclusion: even without Andrew Luck, the Colts had the best long-ball quarterback in the league.

24 Nov 2009

Week 11 DVOA Ratings

by Aaron Schatz

One of the things that makes DVOA distinctly different from other "power ratings" across the Web is that, once you get past the midpoint of the season, it doesn't often have strong reactions to single games. That's because it takes every play into account rather than concentrating on just the last couple weeks. Even "weighted DVOA" counts the last eight weeks at 95 to 100 percent strength and doesn't seriously begin to discount plays until you get to games that were played two months ago. So when the best teams go out there and take care of business, the ratings are going to remain the same. This week's top four are the same as last week's top four -- New England, Indianapolis, New Orleans, and Philadelphia -- and while there's a little shuffling below, the same teams that were ranked 1-10 last week are ranked 1-10 this week.

That's not to say there aren't some changes. For example, this week San Diego passes Denver for the first time in 2009. The Denver collapse has been quite remarkable, especially if you look at the FO playoff odds. Just four weeks ago, we listed Denver with a 98.8 percent chance of making the playoffs, including a 96 percent chance of winning the division and a 71 percent chance of getting a first-round bye. This week, the Broncos' chances of making the playoffs are listed at 57.2 percent, with only a 20 percent chance of winning their division. San Diego's win over Denver raised the chances of the Chargers making the playoffs by 38 percent, and cut the chances of the Broncos making the playoffs by 24 percent.

This also marks the first week the chances of a team going 16-0 went over 10 percent. The Colts are about twice as likely as the Saints to pull it off, although the Saints' chances will get a lot better if they can beat New England this week. No particular Indianapolis opponent has a strong chance of a win, but still, the odds are good that the Colts will stumble in one of the six games. It happens, and not always when you expect it. Just ask the 1998 Broncos.

I don't have much time to get into commentary this week, because the short Thanksgiving week means I have to get the ESPN content done earlier. I'll make it up to everyone by doing a mailbag in next week's DVOA commentary. Usually I don't read all the comments closely -- too many to get through -- but this week I'll check out the comments, take a few questions you may have about why the ratings say specific things about specific teams, and do up some answers next Tuesday. 

* * * * *

These are the Football Outsiders team efficiency ratings through 11 weeks of 2009, 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 DVOA 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. WEIGHTED DVOA is adjusted so that earlier games in the season become gradually less important. It better reflects how well the team is playing right now. 

As always, positive numbers represent more points so DEFENSE is better when it is NEGATIVE.

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 NE 39.7% 1 42.0% 1 7-3 32.5% 1 -5.9% 9 1.2% 12
2 IND 35.9% 2 35.6% 2 10-0 26.9% 2 -8.5% 5 0.5% 16
3 NO 32.5% 3 30.4% 3 10-0 26.0% 3 -7.9% 7 -1.4% 23
4 PHI 30.4% 4 28.6% 5 6-4 14.2% 12 -10.9% 2 5.3% 4
5 MIN 27.6% 6 29.1% 4 9-1 17.1% 7 -1.6% 15 8.9% 1
6 BAL 25.4% 5 23.4% 7 5-5 16.6% 10 -8.8% 4 0.0% 19
7 GB 24.7% 9 24.1% 6 6-4 19.5% 4 -12.1% 1 -6.8% 31
8 ARI 20.6% 10 23.0% 8 7-3 16.9% 8 -3.0% 12 0.7% 14
9 DAL 18.8% 8 19.0% 10 7-3 18.7% 5 2.9% 18 3.0% 8
10 PIT 18.5% 7 20.7% 9 6-4 16.5% 11 -9.0% 3 -7.0% 32
11 NYG 12.4% 14 11.0% 14 6-4 12.7% 14 -2.7% 13 -3.0% 27
12 MIA 11.7% 13 13.6% 11 5-5 12.5% 16 4.5% 19 3.7% 5
13 SD 9.9% 16 12.9% 12 7-3 18.6% 6 7.7% 22 -0.9% 21
14 HOU 9.3% 15 12.2% 13 5-5 16.6% 9 10.4% 25 3.1% 7
15 CIN 8.0% 11 7.4% 15 7-3 12.4% 17 1.8% 17 -2.6% 26
16 DEN 6.8% 12 4.5% 16 6-4 5.6% 18 -4.5% 10 -3.3% 28
17 ATL 4.2% 17 2.9% 17 5-5 13.8% 13 10.0% 24 0.4% 17
18 JAC -1.5% 18 -3.0% 20 6-4 12.5% 15 13.9% 28 -0.1% 20
19 NYJ -1.6% 19 -5.0% 21 4-6 -11.3% 23 -7.5% 8 2.1% 10
20 SF -2.0% 20 -2.2% 18 4-6 -10.5% 22 -7.9% 6 0.6% 15
21 CAR -7.4% 22 -2.7% 19 4-6 -4.0% 20 -1.9% 14 -5.3% 30
22 WAS -9.4% 21 -8.7% 22 3-7 -11.4% 24 -3.0% 11 -1.0% 22
23 TEN -12.8% 24 -13.1% 23 4-6 3.3% 19 13.7% 27 -2.4% 25
24 CHI -17.1% 25 -17.8% 24 4-6 -17.2% 27 5.5% 20 5.6% 3
25 SEA -17.9% 23 -19.3% 25 3-7 -9.3% 21 9.5% 23 1.0% 13
26 BUF -20.9% 26 -23.7% 26 3-7 -22.6% 30 -0.2% 16 1.5% 11
27 KC -26.7% 27 -26.7% 27 3-7 -19.0% 28 10.9% 26 3.3% 6
28 TB -33.8% 28 -33.2% 28 1-9 -15.7% 26 20.5% 29 2.4% 9
29 OAK -38.4% 31 -38.2% 30 3-7 -32.9% 32 5.9% 21 0.4% 18
30 STL -38.5% 29 -37.2% 29 1-9 -15.0% 25 21.9% 31 -1.5% 24
31 CLE -41.6% 30 -41.0% 31 1-9 -28.7% 31 21.3% 30 8.4% 2
32 DET -52.3% 32 -53.6% 32 2-8 -22.4% 29 25.0% 32 -4.9% 29
  • ESTIMATED WINS uses a statistic known as "Forest Index" that emphasizes consistency as well as DVOA in the most important specific situations: red zone defense, first quarter offense, and performance in the second half when the score is close. It then projects a number of wins adjusted to a league-average schedule and a league-average rate of recovering fumbles. Teams that have had their bye week are projected as if they had played one game per week.
  • PAST SCHEDULE lists average DVOA of opponents played this season, ranked from hardest schedule (#1, most positive) to easiest schedule (#32, most negative). It is not adjusted for which games are home or road.
  • FUTURE SCHEDULE lists average DVOA of opponents still left to play this season, ranked from hardest schedule (#1, most positive) to easiest schedule (#32, most negative). It is not adjusted for which games are home or road.
  • VARIANCE measures the statistical variance of the team's weekly DVOA performance. Teams are ranked from most consistent (#1, lowest variance) to least consistent (#32, highest variance).
1 NE 39.7% 7-3 37.1% 7.5 4 1.2% 16 4.7% 11 16.5% 21
2 IND 35.9% 10-0 39.0% 8.0 2 3.1% 11 -4.1% 22 10.1% 15
3 NO 32.5% 10-0 41.1% 8.1 1 -8.7% 30 2.4% 16 6.1% 4
4 PHI 30.4% 6-4 36.7% 6.8 7 -5.4% 26 6.2% 9 23.2% 29
5 MIN 27.6% 9-1 35.1% 7.5 3 -10.1% 32 -0.1% 18 5.5% 1
6 BAL 25.4% 5-5 26.5% 7.0 6 2.4% 15 -9.2% 27 7.6% 6
7 GB 24.7% 6-4 33.9% 6.4 10 -9.4% 31 -4.5% 23 13.1% 18
8 ARI 20.6% 7-3 23.5% 7.4 5 -4.1% 25 -10.7% 29 16.8% 23
9 DAL 18.8% 7-3 19.3% 6.6 9 -1.5% 23 7.5% 7 9.5% 12
10 PIT 18.5% 6-4 27.6% 6.7 8 -8.2% 29 1.5% 17 10.0% 13
11 NYG 12.4% 6-4 10.6% 5.4 17 0.7% 19 13.4% 2 18.3% 25
12 MIA 11.7% 5-5 7.0% 5.8 14 5.2% 7 6.4% 8 8.1% 7
13 SD 9.9% 7-3 9.9% 5.9 12 0.8% 18 -12.7% 31 8.9% 11
14 HOU 9.3% 5-5 11.9% 6.0 11 -2.3% 24 5.9% 10 8.2% 9
15 CIN 8.0% 7-3 10.8% 5.8 13 2.9% 13 -16.9% 32 28.2% 31
16 DEN 6.8% 6-4 6.0% 5.7 15 3.7% 9 -2.6% 19 17.8% 24
17 ATL 4.2% 5-5 3.8% 5.4 16 6.5% 3 -5.4% 24 10.0% 14
18 JAC -1.5% 6-4 0.9% 5.1 18 -5.9% 27 10.6% 6 21.4% 28
19 NYJ -1.6% 4-6 1.7% 4.7 19 6.5% 4 -2.8% 20 13.4% 19
20 SF -2.0% 4-6 -8.3% 4.6 20 3.3% 10 -11.8% 30 8.4% 10
21 CAR -7.4% 4-6 -15.6% 4.3 21 5.3% 5 15.3% 1 19.4% 27
22 WAS -9.4% 3-7 -2.0% 4.2 22 -7.8% 28 13.1% 3 8.1% 8
23 TEN -12.8% 4-6 -22.1% 4.2 23 7.7% 2 4.4% 12 41.1% 32
24 CHI -17.1% 4-6 -14.9% 3.7 24 -0.7% 22 2.9% 14 11.3% 16
25 SEA -17.9% 3-7 -20.0% 2.9 27 1.1% 17 -10.6% 28 18.8% 26
26 BUF -20.9% 3-7 -17.0% 3.6 25 -0.5% 21 12.6% 4 14.2% 20
27 KC -26.7% 3-7 -23.3% 2.9 26 2.5% 14 -6.2% 26 6.1% 3
28 TB -33.8% 1-9 -37.9% 2.4 28 12.0% 1 2.8% 15 12.5% 17
29 OAK -38.4% 3-7 -41.9% 2.3 30 2.9% 12 3.7% 13 23.4% 30
30 STL -38.5% 1-9 -41.7% 2.4 29 5.3% 6 -4.0% 21 5.5% 2
31 CLE -41.6% 1-9 -47.6% 2.0 31 4.2% 8 -6.0% 25 16.8% 22
32 DET -52.3% 2-8 -50.3% 1.6 32 0.6% 20 11.9% 5 6.8% 5

Posted by: Aaron Schatz on 24 Nov 2009

121 comments, Last at 30 Nov 2009, 10:23am by After this week's games


by Keith (1) (not verified) :: Tue, 11/24/2009 - 3:52pm

The Packers are still the top-ranked defense (but not for much longer).

But why, oh why!, is their Special Teams 31st? Maybe my eyes are deceiving me, or are they just giving up average returns that are longer than most other average returns?

Still, on paper, those guys are great and looking to head down the charts (unless they do not miss Kampman and Harris, which is not likely).


by hansen9j :: Tue, 11/24/2009 - 5:18pm

We give up huge kickoff returns seemingly every game; the blocked punt against Tampa didn't help either. I'm just delighted that we're no longer 32nd; thank you Jamaal Charles!

How context-sensitive are opponent adjustments? Like, if a team has a poor special teams day against Cleveland, is the impact decreased because Cleveland is great at ST, or increased because Cleveland is overall awful?

by Arkaein :: Tue, 11/24/2009 - 5:19pm

As for special teams, they give up about one long kick return per game (off the top of my head Harvin, TB, SF, and I think Cribbs have all had at least one pretty big KO return, and I'm sure there are a few others I can't remember directly), their own kick and punt returns are below average, their punts are inconsistent, and their FG kicking is no better than average (good below 50 yards, bad beyond 50 yards).

The only real positive they have is kickoff distance.

by dank067 :: Tue, 11/24/2009 - 5:31pm

Packers' special teams are pretty awful, especially their kickoff and punt coverage, poor special teams play has been a huge part of all of their losses this year. Bengals had like a 10 yd td drive fairly early on set up by a punt return and had another good PR late, Harvin scorched the Pack in both games vs. Minnesota, especially game 2, and Tampa Bay had the blocked punt to keep them in the game in the first half and then the huge return just when it looked like the Pack were about to finally put them away. And is it a surprise that the catalyst of the 49ers near comeback was a huge kick return? We might be talking about the Packers in a totally different context this year if they could just cover kicks, but that's part of football

by BJR (not verified) :: Fri, 11/27/2009 - 9:41am

So, another solid outing for the Green Bay special teams yesterday, contributing a fumble from the opening kick off, a missed field goal and consistently giving up good field position on punt and kick returns (they were a very good Mason Crosby tackle away from conceding a TD at one point). I think I buy into GB being a top unit on both sides of the ball hamstrung by abysmal special teams.

by Vincent Verhei :: Tue, 11/24/2009 - 5:45pm

Check out the special teams page. They're about average in kicking for points, and below average in everything else. Most notably, they're the worst punting team in the league -- -14.7, and the gap is HUGE. The next worst team is Denver at -8.2. Denver is closer to number 15 San Francisco (-1.9) than they are to Green Bay.

Exactly two-thirds of Green Bay's punts are returned -- only Oakland is worse. Green Bay allows 11.4 yards per return -- only Seattle, San Diego, and Carolina are worse. Put it together, and Green Bay allows 7.6 return yards per punt, the worst figure in the league.

Or I can skip the details and point out they average 36.5 net yards per punt, worst in the league.

by MurphyZero :: Wed, 11/25/2009 - 12:34am

Well if GB's so bad on punting/punt coverage, maybe they should take a page out of Belichick's playbook and go for it on fourth. Every time. Well maybe not 4th and obscene, but it may well be worth it, especially against teams with good punt returners.

Likewise, the Steelers are SO bad on kickoffs this year, onside kick it. Twice a game or more. Especially if Polamalu is in the game. With him, the defense is more likely to handle the shorter field if it fails. Once the threat of the onside kick is high enough, other team having the hands team on the field, in position for every kick should (hopefully) improve coverage when you choose not to onside kick it.

by Andrew Potter :: Wed, 11/25/2009 - 3:53am

It sounds like what you're saying is "kick it deep if the other team's hands team is on the field, onside kick it if their return team is". That would be an immensely entertaining kickoff strategy.

by aa (not verified) :: Tue, 11/24/2009 - 6:03pm

If you look at the defensive ratings, the packers are already 31st against #1 wide receivers...so maybe losing Harris isn't going to be the end of the world. How much worse can it get? Aside from Calvin Johnson and the Arizona combo platter none of their remaining opponents have anyone that's too scary.

by S :: Tue, 11/24/2009 - 6:09pm

Did the Packers still use Harris exclusively against #1's? I know they did back in the day, but I thought that changed when Woodson came on board. Any game charters know this?

by Arkaein :: Wed, 11/25/2009 - 12:46pm

The packers tend to line up Harris on the right side of the defense, so he's not always covering the #1. Also GB is playing more zone this year, and in several games Woodson has been playing a fair amount against the TE instead of WRs (Winslow, Witten, Vernon Davis before Harris was injured).

I think these matchups have somewhat distorted the effectiveness of GB against top WRs, while boosting their rating against TEs, as well as all other types of receiver (no worse than 8th against WR2, TE, other WR, or RB).

by Dej (not verified) :: Tue, 11/24/2009 - 7:50pm


Your eyes are deceiving you. The special teams may not be giving up too many TDs, but they have let opposing teams start drives in the red zone on an almost weekly basis. It's so painful to watch. In addition, our two bright spots, Will Blackmon and Jordy Nelson haven't played much this year (and Blackmon is done).


A couple of Packer related topics that I would love to see addressed in your mailbag:
(1) The special teams of the Packers and Pittsburgh are atrocious. Is there any hope for one of the bottom two ST units to still make the playoffs by historical standards? In fact the Packers are in the bottom two of STs, Penalties, and Adjusted Sack Rate. Quite the trifecta. Are any of those strong predictors? (2) Pretty much from the moment that the Packers cut Tracy White (ST guru) in the middle of last season, the units have gone down the tubes. Is there any data that you could offer on the decline without him, and also any historical parallels for a non-returner on STs? (3)Losing Al Harris and Aaron Kampman will be tough on many levels, but the one place it could help would be in zone defense. They were our worst LB and DB in zone coverage (IMO). Is there any insight that can be drawn from our game against Dallas (no Kampman) about how it might affect the GB defense without them? Also, we are bad on WR1s, does that track how often Al Harris is lined up on the WR1?

Happy Thanksgiving, and thanks in advance for any reply.


by justanothersteve :: Wed, 11/25/2009 - 12:40am

On Dec 20, GB plays in Pittsburgh. What will happen when special teams take the field? Could we see some sort of dimensional rift in the space-time continuum? What's the over/under on returns for TDs? I'll be on the other side of the planet that week, so I'll unfortunately miss the game. Should be interesting.

by Keith (1) (not verified) :: Tue, 11/24/2009 - 8:52pm

Holy Packers following, Batman!

I guess I just wipe out all the long returns. The detailed stuff that I cannot really see with my eyes (thanks, Vince!) was great. Thanks ;]

Also, CAPTCHA -- garfield grandmother (terrible joke buried in there).

by Concerned Citizen (not verified) :: Tue, 11/24/2009 - 4:09pm

Hey Aaron,

Next time you're on XL 950, could you mention to the guy on there that Jacksonville is not that bad? Pretty much all year, he's been catagorizing them as just a step up from the Detroit/Cleveland/StL crowd. I don't even like the Jaguars, so I don't know why this bothers me so much, but it's just really obvious from watching the games that Jacksonville is miles ahead of those teams. DVOA backs that up as well, as does plain old W-L. They're not awesome or anything, but there are at least 10 teams that are indisputably worse than Jacksonville.

by djanyreason :: Tue, 11/24/2009 - 4:10pm

Aaaaaaand.... Pittsburgh passes the 2000 Bills for worst kickoff coverage unit in the history of DVOA.

by FireOmarTomlin :: Tue, 11/24/2009 - 5:05pm

FIRE BRUCE LIGASHESKY !!!!!!!!!!!! (sp?)

by dedkrikit (not verified) :: Tue, 11/24/2009 - 4:24pm

"That's because it takes every play into account rather than concentrating on just the last couple weeks."
What about the surging Titans? One could easily argue that they are a MUCH different team than they were during (most) of their losing streak. The defensive backs are healthy now and Vince Young changes their offense quite a bit (especially by opening holes for CJ).

I don't believe a rating of their whole season is valid when comparing to their opponents - especially when it affects another teams offensive ratings. A poorer performance against the post-Bye Titans is more understandable than against the injury-depleted and Collins-led team.

by Still Alive (not verified) :: Tue, 11/24/2009 - 4:29pm

And your point is? I don't get the impression anyone would disagree with what you said.

DVOA does not take such things into account, and attempts to have it do so do not improve it. Just simply adjust the figures with your mind if you feel that strongly about it. This is supposed to be one good but imperfect tool among many, not the TRUTH.

by Theseus (not verified) :: Tue, 11/24/2009 - 5:31pm

Regardless of DVOA's abilities on its own, I'm curious as to whether there's any way that Aaron and/or the rest of the team can pull out anything interesting about the Titans' performance vis a vis the rest of the league or previous teams in the DVOA era to start under similar circumstances. If there's anything they can do to illuminate where the team is now and where they're going, it would be really useful: the talking heads are talking like they have a real, if slim, chance of making the playoffs, and even if that's possible, are they even playing well enough now to justify that kind of talk? Is Vince Young really playing better than he was during Vince Young 1.0? That seems like the kind of thing a site like FO is designed to discuss intelligently.

Maybe DVOA isn't perfect, but it's not completely inflexible, either, and the team seems pretty talented at taking on special projects using the stats.

by Still Alive (not verified) :: Tue, 11/24/2009 - 6:27pm

Yeah I would love to see that as well. A whole article about TEN would be interesting. Maybe they are waiting to see if TEN can squeeze out one more win?

Sometimes I get a bit oversensitive to all the "This screwdriver does not pound nails like a hammer." criticisms of DVOA.

by Tom Gower :: Tue, 11/24/2009 - 8:34pm

This is on my radar to do when I get/make the time. In the interim, if you're an ESPN Insider, check out the piece that went up a couple days ago on the Titans' chances of making the playoffs. In terms non-FO writings, I also did a recap and liveblog for last night's game, both of which touch on what the Titans did well and areas that have improved.

by Will Allen (not verified) :: Tue, 11/24/2009 - 7:12pm

Vince Young, I think, is a much more formidable qb, much more so than other NFL qbs, when he is paired with a running back who is a threat to go the distance on each touch. It makes defending Young's running much more difficult, and having two outstanding runners on the field, one the qb, makes it a lot easier for the Titans receivers to get large seperation.

by MC2 :: Tue, 11/24/2009 - 9:16pm

Another reason why Young has seemed so much better than Collins is that he has been asked to do so much less than Collins was. It's a very fortunate coincidence that Young took over just as the Titans defense was getting healthy, allowing him to be much more of a "game manager" than Collins, who often had to spend the entire second half airing it out in catch-up mode. That's not really the strength of either QB.

ETA: Young is averaging 23 attempts per start, compared to 34 for Collins.

by dedkrikit (not verified) :: Wed, 11/25/2009 - 11:18pm

My point is not that the Titans are ranked too low or any of that nonsense. It is more of making the statistics more accurate. A complete game against the post-Bye Titans is tougher than the checked-out-in-a-snowstorm group.

Perhaps an example will help illuminate things...
Say the Titans continue to win and their defense continues to get pressure and turnovers... Brady's DVOA will soar until early season games lose value.

But the defense he faced is fairly different than what the Cardinals will face next Sunday.

Obviously, a request to redo the whole system based on this one anomaly is insane. However, I (as a Titans and FO fan) am fairly interested in what these guys can do with a study on it.

by jmaron :: Tue, 11/24/2009 - 4:25pm

I haven't watched Dallas much this season but it seems odd that they are ranked 5th on offence in DVOA after scoring a grand total of 14 pts in two games.

Like I said I didn't watch the game against Washington, but how do you come out with a 6.6% DVOA when you only score 7 pts?

I get that DVOA is more predictive than a bunch of other stats, but you score 7 pts and the system kicks out better than average offence? Sorry that just doesn't wash.

by mm (not verified) :: Tue, 11/24/2009 - 4:33pm

Their weighted offense is 10th. If they continue to do poorly they will continue to fall.

by bravehoptoad :: Tue, 11/24/2009 - 4:46pm

Maybe because they've played 8 other games.

by jmaron :: Tue, 11/24/2009 - 5:15pm

oh some sarcasm - fair enough, but I think scoring 14 points in two games is a sign that their offence isn't anywhere close to the 5th best or even 10th best.

by RickD :: Tue, 11/24/2009 - 8:13pm

GB has a good defense and, believe it or not, so do the Redskins.

by bravehoptoad :: Wed, 11/25/2009 - 11:50am

So, you're not talking about DVOA at all. What you're asking is, how can a team play so well some games and so crappy other games?

Well, it happens. The Pats had a real stinker against Buffalo, New Orleans had trouble beating the Rams, a couple good teams have actually lost to the Raiders, etc.

There's an old principle you might have heard of. It starts, "Any given Sunday..." and ends with "...and that's why they play the games." We here know it partly as "variance."

by Pat (filler) (not verified) :: Wed, 11/25/2009 - 5:19pm

No, it's a sign that their starting field position sucked, the field goal kicker is in the bottom 5 in the league currently, and both the Washington and Green Bay game were short (10 real drives).

by Temo :: Tue, 11/24/2009 - 5:06pm

Like I said I didn't watch the game against Washington, but how do you come out with a 6.6% DVOA when you only score 7 pts

The story of the game, as written by many beat writers following the game, was that both teams mounted drives with successful plays, but neither could sustain long enough drives to actually get in the end zone and would get derailed along the way.

by jmaron :: Tue, 11/24/2009 - 5:27pm

Dallas - 305 yards of total offence - 5.0 yards per play.

St Louis Rams average game - 305 Yards of offence - 5 yards per play.

I did look at the drives and out of 9 real drives (didn't include end of half drives)

2 - 3 and outs,
4 - 20-39 yard drives,
3 - 40-60 yard drives

I see the point - but does that strike you as above average offence in the NFL? If my teams defence held teams to that kind of production every week I'd be thrilled.

by crack (not verified) :: Tue, 11/24/2009 - 5:36pm

But your team would be way behind on success points.

by Key19 :: Wed, 11/25/2009 - 1:53am

Well, let's see.

In Green Bay, Dallas drove early to I'd say probably the 20 yard line (drove maybe 60 yards let's assume?), then stalled because of penalties and missed a FG. Penalties are ignored by DVOA as far as I know, so that's a very good drive in DVOA's eyes.

Dallas then had a huge play from Roy Williams, only to have him fumble the ball. Lots of yards and in scoring position again, but no points. Not sure how much the fumble is penalized for in DVOA, but the play was huge and I think it still might outweight the fumble.

Dallas then drove again only to have Roy Williams have the ball strike him in the face after "losing it in the lights." That drive stalled I believe, but it got a good amount of yards. Good DVOA.

Dallas then drove to the Green Bay goal line, only to throw a pick on about the 2 yard line. Drive outweighs negativity of pick in DVOA I'm sure.

Dallas then drives and scores at the end of the game. I know garbage yards/TDs are not weighted as heavily, but it was a pretty good drive so I'm sure it counted for something.

All of these things result in 7 points.

Now, for Washington.

Dallas drives early, absolutely killing the Redskins with the run. The Redskins are rated highly in run defense, so DVOA is going off the charts. Marion Barber than fumbles on the 15 yard line or so and they lose the ball. Romo gets hurt on the tackle. Good DVOA, no points.

The Cowboys later drive down into FG range. Nick Folk misses the kick, Green Bay style. Solid drive for DVOA, but 0 points still.

Dallas continues to run the ball well throughout the game, even though the drives are hit and miss. These are largely the result of penalties, which are not accounted for in DVOA. The Redskins' high defensive DVOA and the Cowboys' success running the ball on them takes some the edge off of what would be a pretty bleak offensive day overall.

Dallas then drives down and scores near the end of the game (in a close game), which has more weight in DVOA than earlier drives. I'm sure it counts greatly. Finally, 7 points, and DVOA is smiling again.

So, a mediocre day DVOA-wise against the #1 defense in the NFL and a great rushing day against a top-10 rushing defense = two games that don't really take away much from Dallas' offensive DVOA.

It is safe to say, however, that they will significantly drop if they score 7 points against the Raiders on Thursday.

by Dales :: Wed, 11/25/2009 - 8:32am

Safe to say because they aren't the Eagles, that is.

by Pat (filler) (not verified) :: Wed, 11/25/2009 - 5:09pm

I haven't watched Dallas much this season but it seems odd that they are ranked 5th on offence in DVOA after scoring a grand total of 14 pts in two games.

Points scored is a function of offense, defense, and special teams. DVOA judges an offense by itself - points scored doesn't. Note the poor defense, the correspondingly bad starting field position for the offense (27th in the league), and the poor field goal kicking (again 27th in the league by DVOA).

by zlionsfan :: Tue, 11/24/2009 - 4:26pm

Don't look now, but the Lions are the best 2-8 team in the NFL.

by MilkmanDanimal :: Tue, 11/24/2009 - 4:32pm

Dream the dream, Lions fans. The Bucs are the best 1-9 team, and there's even competition for that. In your face, St. Louis!

by BucNasty :: Tue, 11/24/2009 - 5:18pm

We're goin' all the way to the Senior Bowl!

by Paulo Sanchotene, RS, Brazil (not verified) :: Wed, 11/25/2009 - 12:10pm

Say something nice's Walkthrough spirit!

by Nuk (not verified) :: Tue, 11/24/2009 - 4:28pm

The plummeting of Denver's playoff odds supports my belief that the system puts too much confidence in the accuracy of the DVOA ratings. IIRC, there used to be a random element that affected the ratings in the Monte Carlo simulation, and that randomness was removed this season.

Why was that done? You can't seriously think that these ratings are perfect, so the random element reflects the uncertainty in the measurements, and would give more realistic odds for teams making the playoffs.

by Still Alive (not verified) :: Tue, 11/24/2009 - 4:32pm

I think any model you use looking at past data is going to have a hard time predicting a 6-0 team to follow it up with an 0-4 run. If DEN wins even 1 of those games the %'s don't seem so odd.

by M :: Tue, 11/24/2009 - 5:08pm

I think there have been only two teams since the 16 game seasons started in 1978 where a 6-0 team failed to make the playoffs: 1978 Washington Redskins & 2003 Minnesota Vikings. Denver is doing their best to make it a third.

My guess (without looking) is that that is about 2 of (at most) 50 teams, or roughly 4% to 5%.

While I think that Denver's playoff odds after their last win implied less than a 2% failure rate, I think that may be more indicative of the likelihood that San Diego was really a crappy team - if San Diego had been more highly rated (4 & 2 with a blowout win or two), it's possible that the projected failure rate would have been closer to 4% or 5%, which is more in line with history.

by mm (not verified) :: Tue, 11/24/2009 - 4:41pm

Playoff odds for some teams: 100%, 100%, 99.9%, 97.4%, 95.1%, 93.9%, 90.8%.

Not shocking, judging from the records, but still surprising to see them that high with so many games left.

by DrewTS (not verified) :: Tue, 11/24/2009 - 4:58pm

The only one I find surprisingly high is Cincinnati. Maybe it's just ~20 years of history talking, but I have to think there's a greater than 6.1% chance they miss the playoffs. But when you step back and realize that they have a 1 game lead in their division and all the tiebreakers in hand, and Pittsburgh and Baltimore don't really seem to want to step up and take it... maybe it's not so crazy. That loss to the Raiders was a crusher though. You have to win the freebies.

As for the 100%s, I can't see anything other than a season-ending Manning injury keeping the Colts out of the playoffs. And even then, they likely still get in at 10-6 (and are the team everyone hopes to face). Same with the Saints.

by turbohappy (not verified) :: Tue, 11/24/2009 - 5:24pm

I think the Colts have a chance to beat the Bills playing ALL 2nd-stringers, I'm pretty confident they'd beat them with 1st-stringers except Sorgi under center. So I would say the lowest win total they can possibly have is 11, which will almost certainly make the playoffs.

by S :: Tue, 11/24/2009 - 6:01pm

The Colts can go 1-5 down the stretch, and as long as Jacksonville goes 4-2 or worse and Houston goes 5-1 or worse, they'll win the division.

Similar deal with New Orleans; if they win one more game in their next 6 and the Falcons lose one more game in their next six, they win the NFC South.

Essentially, the playoff odds are showing that Indy and New Orleans are probably going to win their division.

(Also, the fact that Jacksonville is listed as having a 0.1% shot of winning their division probably means that Indy's odds for the division are probably closer to 99.95% than 100%)

by Brendan Scolari :: Tue, 11/24/2009 - 6:39pm

Really? You really think the Colts would go 0-6 the rest of the way if they lost Manning? They're a really talented team, I'd think they could pick up at least a few wins.

by Bobman :: Tue, 11/24/2009 - 6:52pm

Agreed--that slate of opponents is not exactly daunting. NYJ have a tough D but their O seems to be imploding. Plus Buf's struggling. And without Manning they could well win one of the division games. Just let Addai throw it (but not Reggie Wayne). Den is in free-fall but still talented.

I'd guess that without Manning the rest of the way they'd hit 12 wins again, maybe 13.

by Mr Shush :: Wed, 11/25/2009 - 7:07am

I agree that Sorgi and Company would have a decent chance of winning 12 games, which might even be enough for #1 seeding, but I really wouldn't like their chances against any of the other AFC South teams if the Titans defense can continue to play at its current level.

by DrewTS (not verified) :: Wed, 11/25/2009 - 9:31am

It wasn't really my intent to say that the Colts would lose every game without Manning, but it may have come out that way. Regardless, I think you guys are a little higher on Sorgi than I am. Of the remaining teams on the Colts' schedule, the only one I would actually pick them to beat with Sorgi at QB would be Buffalo, and that's not accounting for the weather factor.

If Manning snapped his femur tomorrow, I can't imagine the ceiling realistically being any higher than 12-4. No way they win any of the division games, especially with the way Tennessee is playing lately.

Anyway, the point I was originally trying to make is that the Colts would quite possibly still make the playoffs even if they lost every game from here on out.

by jmaron :: Tue, 11/24/2009 - 4:41pm

I watched Philly for the first time this year. I thought Chicago looked like the better team that night with one exception - Cutler was so horribly inaccurate.

Philly has 3 huge DVOA games against Car, TB, NYG (96.2, 87.7, 106.6).

Other than those 3 games Philly is essentially an average team (1.7 DVOA average over those 7 games).

It's easy to see why many people from an observational standpoint wouldn't be very impressed with Philly. Because 70% of the time this year they haven't been particularly impressive.

I wonder if capping DVOA totals in any one game would make the system more predictive? Or does putting up DVOA's around 100 really mean much more than putting up DVOA's of 50 or so?

by mm (not verified) :: Tue, 11/24/2009 - 4:46pm

We have this discussion seemingly every other week. I'll let others debate it with you, but if you want a statistic that combines DVOA with consistency (the opposite of variance), check out the Estimated Wins Column, Eagles fall to 7th, Saints rise to 1st, Giants fall to 17th).

by jmaron :: Tue, 11/24/2009 - 4:55pm

It's not the Eagles argument I'm really interested in - rather would capping DVOA's at a certain level make the system more predictive of future performance. Just curious if FO has looked at that yet.

by zip.4chan.org/sp/ (not verified) :: Tue, 11/24/2009 - 5:11pm

I think DVOA thinks that stomping bad teams *is* predictive of the potential strength of a team. Isn't this the perennial GUTS vs STOMPS thing?

by jmaron :: Tue, 11/24/2009 - 5:44pm

The article I read on Guts and Stomps qualified stomps as more than 14 or 20 points or something like that. It didn't look at the level of stomp as a factor for predicting future success.

by t.d. :: Tue, 11/24/2009 - 5:13pm

I understand Philadelphia not being hurt by the Oakland loss, because they didn't really play all that terribly in that game. I don't get Philly grading at about 0% DVOA for the New Orleans game, where they were thoroughly pantsed. New Orleans scores out as having a great game that week, but Philly doesn't score as not playing well.

by Pat (filler) (not verified) :: Tue, 11/24/2009 - 6:12pm

It's because they were playing the Saints. Since Philly/NO have roughly similar DVOA, the difference in their DVOA for that specific game is basically the difference in VOA (how they played) for that game.

by Rick (not verified) :: Tue, 11/24/2009 - 11:19pm

I don't know. For most of the first half, Philly looked like they would wind up completely owning the Bears. A few mishaps and they are looking at a close game.
In the end, a relatively close game turned on 2 things:
1. Philly dominated the first quarter
2. Philly dominated the fourth quarter.

In between, Chicago dominated the scoreboard in the second, but not the plays. Philly held them to FGs. The third quarter was pretty much a wash.

Overall, Philly definitely looked better.

But better of what? 2 average to above average teams?

by Peder (not verified) :: Tue, 11/24/2009 - 4:47pm

Have you considered lessening the power of DAVE? I'm just wondering if the system corrects quickly enough from preseason predictions, which can be wildly off.

by DAVE. (not verified) :: Tue, 11/24/2009 - 5:01pm

I know I've made some very poor decisions recently, but I can give you my complete assurance that my work will be back to normal. I've still got the greatest enthusiasm and confidence in the mission. And I want to help you.

by The Guy You Don't Want to Hear (not verified) :: Tue, 11/24/2009 - 5:10pm


by Peder (not verified) :: Tue, 11/24/2009 - 6:12pm


by Still Alive (not verified) :: Tue, 11/24/2009 - 6:31pm


by zlionsfan :: Tue, 11/24/2009 - 7:30pm

If we had the ability, I'd give you +1 for that.

by DrewTS (not verified) :: Tue, 11/24/2009 - 5:04pm

Kind of a strange time to ask that, since DAVE is no longer in effect.

by tally :: Tue, 11/24/2009 - 5:27pm

He's asking about the period of time when preseason projections are factored in, which is DAVE. The fact that it comes up now is probably because teams right now look so different from their preseason projections that in retrospect, perhaps too much emphasis was placed on the preseason projections in the DAVE rating.

by Peder (not verified) :: Tue, 11/24/2009 - 6:09pm

I'm asking now because Aaron is taking questions. The current rankings don't look too wonky for me but earlier in the season they very much did.

by DrewTS (not verified) :: Tue, 11/24/2009 - 5:09pm

Here's a question for Aaron --

Are we going to see a column entitled "They Are Who We Thought They Were", retracting all the apologies for Denver's low preseason projection? Because I would read that.

by Still Alive (not verified) :: Tue, 11/24/2009 - 6:33pm


by Brendan Scolari :: Tue, 11/24/2009 - 6:45pm

Um, the projection for Denver still looks way off. They were projeted to be 31st or 32nd in DVOA (can't remember which) and they're currently 16th. They've also already got more wins than they were projected to win all season.

by DrewTS (not verified) :: Wed, 11/25/2009 - 9:40am

Give it time. The projection is looking a lot less off than it did a month ago. If they keep playing like they have been, they probably will be in the high 20s in weighted DVOA by the end of the season.

by Anonymousse (not verified) :: Tue, 11/24/2009 - 5:15pm

The top variance teams, except for the Eagles, are have Defensive DVOAs in the lower half. Does having a poor defensive team lend to more variance, historically?

I know it has been shown that Defense is variable (less predictable) year to year, but does a bad defense show variability week to week?

As I say that, I notice that the worst 2 teams by Defense are among the most consistent in the league, so perhaps this bad defense = more variable is merely noise?

by bsr (not verified) :: Tue, 11/24/2009 - 5:15pm

Its amazing to see Brady still as the #1 ranked QB despite coming back from that knee injury.

by Will Allen (not verified) :: Tue, 11/24/2009 - 5:48pm

Anybody remember whether it is somewhat typical for the top 50% of the rankings at this point in the season to have so few teams which have played a top 10 schedule, in terms of opponent strength?

by MC2 :: Tue, 11/24/2009 - 6:03pm

Well, there are 2 this year, compared to 3 last year, 6 in 2007, and 4 in 2006, so this year seems a little odd, but certainly within the realm of random variance.

by Will Allen (not verified) :: Tue, 11/24/2009 - 6:10pm

I know Aaron slaves over the data, to try to find the strongest correlations, but it just intuitively seems to me that if there are not years in which the top half has seven or so teams which have faced top ten schedules, then the opponent strength adjustment may be less than what it should be. Of course, my intution, among my other shortcomings, has thus far failed to put me on the Forbes 400, so big deal, right?

by Still Alive (not verified) :: Tue, 11/24/2009 - 6:34pm

I think the effect of not having to play yourself when you are good is rather large. The overall schedules are pretty balanced so if you are a top 5 team never having to play yourself is a huge plus.

by Will Allen (not verified) :: Tue, 11/24/2009 - 5:57pm

The reason for Vikings fans to be optimistic, if injury luck holds, is that there is reason to think their defensive backfield can make late season strides, especially as Winfield gets back. This is the part of their defense which has put them in the middle of the rankings. Combine that with the fact that a rookie like Harvin may still see significant improvement, and, well, if nobody gets in a good head shot on the Zombie King, I for one am looking forward to a Saints/Vikings conference championship. That would be a game filled with fascinating match-ups, players and coaches alike. Hope we see it.

by jmaron :: Tue, 11/24/2009 - 8:17pm

The Vikings are 1st in defensive DVOA against non shotgun formations. They are 20th defending the shotgun and they face the shotgun more than any other team.

The Vikings defence by Score Gap:

up by 8 or more: 14th
up by 0-7: 2nd
down by 0-7: 1st

Late and Close - all games 2nd half on - score within 8 - 3rd

One of two things is happening here:

1) The Vikes get conservative once they have a lead and play soft, or
2) Teams are stupid and don't just abandon the run and simply spread the field and throw on every down the way NE torched them on Monday night in 2006.

I'm honestly not sure which it is. What do you think?

by The Guy You Don't Want to Hear (not verified) :: Tue, 11/24/2009 - 10:18pm

I'm betting you'd rather hear from Will Allen (and I don't blame you), but since I've seen all but one Vikings game this year, I will throw my two cents in. I think the answer is somewhat both. The strategy of spreading the field and throwing all game would probably be less viable once Winfield gets back, but even with him it at least seemed like they really played softer once they got a lead. I guess that means I'm saying it's option one mostly and option 2 thrown in since Winfield went out.

by Jimmy :: Tue, 11/24/2009 - 10:37pm

You are right, not enough coaches are willing to be unconventional about game planning on offense. Belichick may be a long way from being my favourite human being but he is a very good coach. He took one look at those Vikings tackles and decided bothering to run the ball was essentially a waste of time (other than as a change up). If your QB is calm enough to keep making the reads then why waste 20 snaps moving the ball 45 yards when you could get at least 100 through the air?

by jmaron :: Wed, 11/25/2009 - 9:19am

I didn't used to be a Belichick fan but his obvious intelligence and willingness to challenge conventional wisdom has changed my mind.

by jmaron :: Wed, 11/25/2009 - 9:20am

I've always felt Winfield was a solid pass defender not a great one, but one of the best ever against the run.

by Will Allen (not verified) :: Tue, 11/24/2009 - 11:00pm

I think both factors are in play, along with possibly their safeties not having great ball skills. It really is crazy that those two guys don't have more interceptions, given a consistently good pass rush, and how often the Vikings have a had a substantial lead.

by jmaron :: Wed, 11/25/2009 - 9:25am

Ever since Frazier has been the def coordinator I've felt the safety play was crappy.

Since he's been here the ints have plummeted despite the pass rush getting better every year

Ints - ranking in league

2006 - 5th
2007 - 22nd
2008 - 25th
2009 - 24th

Sharper's ints dropped right along with the arrival of Frazier (to be fair they started to drop under Tomlin)

2005 - 9
2006 - 4
2007 - 4
2008 - 1

2009 with NO - 7

I strongly suspect this has to do with scheme and not being allowed to gamble. How many times do you see a receiver behind a safety on this team? Not very often - almost all the big plays are underneath throws with lots of YAC.

I'm not a big Frazier fan - I think a 15th rating with the talent on the defence is not great.

by Will Allen (not verified) :: Tue, 11/24/2009 - 7:03pm

Does the fact that the Vikings and Saints are ranked #1 and #4, respectively, in terms of lack of variance, and the Rams and Lions are ranked #2 and #5, in that category, mean that, when Joe Buck goes to hell, he will be, for eternity, be calling games which feature a team from the first pair against a team from the second? Will Peyton Manning perform the coin toss?

by Q (not verified) :: Tue, 11/24/2009 - 7:39pm

"Its amazing to see Brady still as the #1 ranked QB despite coming back from that knee injury"

Not as amazing when Brady is still throwing passes and tds in the 2nd Half when up 45-0 whereas other teams take out their qbs when up by a few tds or simply run the ball in to the line repeatedly when up by 21+ points (for example Favre coming out after Minn took a 28-3 lead vs Sea)

by RickD :: Tue, 11/24/2009 - 8:23pm

Uh, right. You are certainly talking about a statistically signifcant part of the Patriots' season. What would that be - the third quarter of the Bucs game and the third quarter of the Titans game?

Favre came out of that game with a minute left in the 3rd quarter, which is about when Brady comes out of a game when the Pats have a big lead - at the end of the 3rd quarter.

Get over yourself.

by Kasmir :: Tue, 11/24/2009 - 9:32pm

Patriots offense is #1 in variance (most predictable), while their defense is #32 (least predictable)!

by Q (not verified) :: Tue, 11/24/2009 - 10:15pm

"Favre came out of that game with a minute left in the 3rd quarter, which is about when Brady comes out of a game when the Pats have a big lead - at the end of the 3rd quarter."

Historically Brady continues to pile up stats while holding a big lead much more than other qbs do which accordingly improves his stats. There is no reason for a starting qb to be even in the game much less throwing passes when up 35-0, regardless of whether it is the 2nd, 3rd, or 4th quarter (particularly against a team without a high powered offense.) It is also poor coaching for exposing a franchise player to injury in a game that is already over. It isn't just the game against Tenn, check back to 2007, the game vs Washington, etc.

**Manning also has compiled many more td passes when up by 4+ scores in a game than one would suspect

by theshadowj :: Tue, 11/24/2009 - 11:27pm

I thought we were talking about 2009, not 2007. Obviously you don't like the way the Patriots usually operate, but I think it would be hard to argue that it has had a significant affect on Brady's ranking this season.

by Andrew Potter :: Wed, 11/25/2009 - 4:10am

Historically Brady continues to pile up stats while holding a big lead much more than other qbs...

Manning also has compiled many more td passes when up by 4+ scores in a game than one would suspect

Wow, so top class players continue to play at a top class level even when they're winning. Who knew?

This season has already seen several large comebacks (comebacks of more than two touchdowns) by teams of all standards, including one against the Patriots. Why would you remove your top players from the game before you're absolutely certain the other team is dead and buried? Especially with the Patriots defense which, as another poster has pointed out, is #32 in variance - capable of playing at a high level, but also liable to implode at any time. If you're relying on Brian Hoyer and BenJarvus Green-Ellis to close out games for you, you're setting yourself up for a fall unless the other team has clearly thrown in the towel.

Or to put it another way, I'll continue to play hard against you as long as you continue to play hard against me.

Finally, I fail to see how the Washington game in 2007 affects the Patriots' offensive numbers this season. What exactly does that have to do with them being #1 in offensive DVOA?

by tonic889 (not verified) :: Tue, 11/24/2009 - 11:00pm

I'd love to see an article breaking up team's DVOA into two windows. For example, DVOA for Tennessee during 6 losses vs. subsequent wins. How much higher would the Titans DVOA be in the 4 wins? Are they playing much better, a little better, or not at all? Similar questions for teams like DEN, NYJ.

by mrh :: Tue, 11/24/2009 - 11:37pm

Since 1978, only 2 QBs have more than 1 TD pass with a lead of 35 points or more - Brady (3), Steve Young (2).

If I change the criteria to a lead of 28 pts or more, 9 QBs have thrown 3 or more TDS with that big a lead.
Brady - 8
Young - 5
Gannon - 4
Peyton Manning - 4
Tomczak (!) - 4
Moon - 4
DeBerg - 3
Erik Wilhelm - 3 (Bengals back-up; all 3 were in 1989)
Quinn Gray (!) - 3

Source: Player TD Finder on p-f-r

by Purds :: Wed, 11/25/2009 - 12:01am

Can I alter the QB conversation. I posted this in Quick Reads by mistake:

Speaking of quarterback play ..... What strikes me is that we know so little about defenses because, at least in a small sample of "good" teams (Colts/NE), they've played against such poor QB's, that it's hard to tell if the defenses are good, or the QB's just really bad. Look:

Colts opposing QB:

Jax - Garrad scored 12 points
Mia - Pennington - 23 pts
Ari - Warner - 10
Sea - Wallace - 17
Ten - Collins - 9
St.L - Bulger - 6
San F - Smith - 14
Hou - Schwab - 17
NE - Brady - 34
Balt - Flacco - 15
So, the numbers look good (15.7 points against average), but who is a good QB on that list? Brady, Warner, ... Schwab? (Only 3 QB's in top 10 DVOA, 6 are 20th or worst in DVOA!)

NE's the same:

Buf - Edwards - 24
NYJ - Sanchez - 16
Atl - Ryan - 10
Balt - Flacco - 21
Den - Orton - 20
Ten - Collins/Young - 0
TB - Johnson - 7
Mai - Henne - 17
Ind - Manning - 35
NJY - Sanchez - 14

Again, numbers overall are good (16.4 points against average), but so few good QB's: P. Manning...Flacco? (Only 1 opposing QB in top 10 DVOA, 5 are 20th or worse, counting Sanchez twice.)
I guess I am frustrated about knowing how good or bad the AFC defenses of Indy and NE are, with such awful QB competition. Not sure if the other AFC "powers" are facing such bad QB play, but it's hard to figure much out against such bad QB's.

In any case, a quick look shows that Baltimore had actually played a lot of good QB's (Rivers, Brady, Palmer x 2, Farve, Manning). Pitt, not as much (Palmer x 2, Farve). I'm worried (as a Colt fan) that the NFC is going to be better battle tested this year.

Now wait, NO has not gone against anyone good yet (unless you count Eli), Minny has faced only Rodgers x2 and Big Ben, and Philly has played Brees, Eli, Romo and Rivers.

So, maybe this applies to the NFC as well?

by Jerry :: Wed, 11/25/2009 - 3:32am

The Steelers did play Rivers and the Chargers. By finishing second, Baltimore got to face Manning and Brady, while Pittsburgh gets Collins and whoever's starting for Miami at the end of the year.

by utvikefan (not verified) :: Wed, 11/25/2009 - 1:57am

So, since your reading the posts, Aaron, my question is...

What is the difference between your computer and the BCS's computer, and do both like cupcakes? Would you improve your computer if you added human voting? I mean, could you like have all the media types vote for the Pats and Steelers and have that eliminate the Eagles bias? Or is Philly at home vs. Florida International or The Citadel the way to know that a team is REALLY good.

by sethburn :: Wed, 11/25/2009 - 2:46am

The absolute values for the offensive DVOA's are about 1.9 times the absolute values for the defensive DVOA's. In 2005 offense was about 1.5 times defense, and in 2006 it was 1.3. The top regular season DVOA in 2006 belonged to the Ravens, who were 14th in offense, 1st in defense, and 4th in special teams. Such a combination this year would be the Giants offense, the Packers defense, and the Eagles special teams. That team would rank 5th, right behind the Eagles, in unweighted DVOA.

I just checked 2004 and 2003. In 2004 the offensive absolute value was 1.3 times defense, and it was about 1.15 in 2003. I have a feeling we have entered an era where defenses simply cannot match up against great offenses, but I have no way of proving this.

by Mr Shush :: Wed, 11/25/2009 - 7:31am

I'm not convinced. In 2008 the top three teams in total DVOA were the top three defenses, and two of those top three defenses (PIT and BAL) were "better" than the best offense (SD) - and met in the AFC Championship game. The Steelers were 3-1 against elite passing attacks (IND, ARI, SD twice), allowing an average of 20.25 points per game in those games, and never more than 24. True, the Ravens gave up 31 in a blowout loss to the Colts, but their five offensive turnovers in that game were clearly a major factor. Great defenses are rarer than great offenses, but I think it's fairly clear they can both still exist and still be extremely valuable.

by sethburn :: Wed, 11/25/2009 - 2:01pm

Yeah, checked out 2008. Defense and offense were almost tied. Checked out 2007 as well and found offense was about 1.6 times defense. This year is probably just an anomaly due to all the completely awful offenses.

by ammek :: Wed, 11/25/2009 - 4:49am

A question about bad teams:

From 1994-97 no team finished with DVOA worse than -33%: not the expansion teams, not even the Rich Kotite Jets. So far this season, there are five such teams.

From 1994-98 only one team finished with double-digit negative DVOA for offense and double-digit positive (ie, bad) DVOA for defense (the 1994 Skins). So far this season, there are five such teams.

Is there anything in the data that would explain why the worst teams of the 2000s are more extremely and holistically horrible than their ancestors? Any visible changes in playcalling, turnovers, big plays? I remember an article from years back that found no link between the presence of Mike Vick as the Falcons' QB and the improved play of the Atlanta defense. But whatever happened to the 'half-bad' team?

by Mr Shush :: Wed, 11/25/2009 - 7:43am

The traditional "half bad" team is still with us - Buffalo are a classic example. They're just higher up the rankings than they used to be thanks to the blossoming of the "absolutely bloody awful" team below them. I think it has to be a roster management issue of some sort. Possibly the salary cap has risen so fast that it is no longer necessary to blow up good teams for cap reasons, while low revenue teams may not be able to afford to spend up to the cap; possibly players have increasingly realised that simply taking the biggest contract available is a suboptimal strategy, as playing for a well-run team will make them happier in the short run and richer in the long; possibly the lack of a cap on coaches and personnel men, who are more valuable thanks to the existence of same for players, has led to an advantage for the higher revenue teams on that front.

by ammek :: Wed, 11/25/2009 - 9:26am

All good and valid arguments.

I'm wondering what are the most significant situation-adjusted statistical manifestations of the polarization of NFL performance. For instance, using conventional stats, we can observe the following changes between 1994 and 2009 (projecting the current year to 448 games):

— Offenses are attempting the same number of passes, averaging one extra completion and five extra yards per game; yet passing DVOA is polarized as never before;
— Big pass plays are more frequent: an average team will produce 8.85 passes of 40+ yards in 2009, versus just 8.04 in 1994; to what extent is this non-situational?
— The distribution of big plays has narrowed: the team with most pass plays of 20+ yards in 1994 had 57 such, whereas the worst team had 30; in 2009, there is a clear elite of a dozen teams that have 34-40 such plays through ten games, while four teams have fewer than 20 and two have no more than 13; is that the difference?
— Touchdowns on kick and punt returns are down by about 15%, but pick-sixes have increased: they represent 10.8% of all interceptions in 2009, against only 9.5% in 1994; how does that affect the correlation of DVOA to wins?

by jebmak :: Wed, 11/25/2009 - 10:44am

Well, the Lions don't have Barry Sanders anymore. So, he can't save the owner from his own stupidity, that's one.

Maybe the quick change in the amount of cap allowed the smart teams to adjust while leaving dumb teams behind (as opposed to the (relatively) poor vs. rich scenario).

by Birdman (not verified) :: Wed, 11/25/2009 - 9:12am

Aaron, how is variance calculated? Is it based on the total DVOA for each game played, or averaging the variance of the three individual units. In other words, would a team that has a decent game with poor defense and great offense following another decent overall game with great defense and poor offense have low or high variance? To me, it would make sense to count that as higher variance, since the units are independent.

by R O (not verified) :: Wed, 11/25/2009 - 9:48am

I realize this may be a silly question but:

How can a Defense have a DEFENSE Adjusted Value Over Average? They don't play other defenses. This has never made sense to me.

by DrewTS (not verified) :: Wed, 11/25/2009 - 9:55am

It's just to keep the terminology standardized. It used to be called OVOA for defenses and DVOA for offenses, and I guess people found that confusing. I've always wondered why the compromise wasn't just AVOA -- Adjusted Value Over Average.

by DeltaWhiskey :: Wed, 11/25/2009 - 10:43am

They were afraid it would be confused w/ these people:


by R O (not verified) :: Wed, 11/25/2009 - 11:46am

Thought it was something like that. Thanks!

by Waverly :: Thu, 11/26/2009 - 11:44am

As opposed to http://www.dvoa.org?

by peterplaysbass (verified?) (not verified) :: Wed, 11/25/2009 - 11:53am

The Saints and Colts have 'TOT' = 100.0% on the playoffs odds page whereas Minnesota has 99.9%. Does this mean that the Colts and Saints have cinched? I don't think they have. Does it simply mean that FO's 10,000 simulations (or however many) didn't produce any scenarios in which those teams missed the playoffs? Or perhaps they produced fewer than 50 such scenarios, and the percentages are rounded up to the nearest 10th? (i.e. 99.95% or higher)

by DrewTS (not verified) :: Wed, 11/25/2009 - 1:00pm

I think it's that the 10k simulations didn't produce any scenarios where they missed the playoffs, or at least not enough that it didn't round back to 100%. (There's obviously some rounding error going on, because the Colts have a 100% chance of winning the division, and Jacksonville has a .1% chance.)

There's definitely nothing clinched yet though. By my count, there are 11 teams that still have a mathematical chance to pass the Colts. That includes Tennessee, but I might be wrong about how the tiebreakers would go in that case. A win at Houston this week would eliminate 5 of those and almost wrap it up.

by Temo :: Wed, 11/25/2009 - 1:45pm

If the Saints lost every single game and finished 10-6, they would need the Falcons and Philly or Giants (they play each other) to win out AND the Packers to all but one to not make the playoffs.

by JG (not verified) :: Wed, 11/25/2009 - 2:28pm

Question for next week. There's a lot of talk about the Patriots 1st Half/2nd Half productivity. Does their DVOA change significantly between the halves? If so, does the DVOA indicate why? How about by quarter? The points for, points against by quarter:

1st: 71-27
2nd: 125-52
3rd: 39-37
4th: 55-45

They've come from behind after being down at halftime once (Buffalo) and lost after being up at halftime three times (New York Jets, Denver and Indianapolis). They also were outscored in the second half by Baltimore and Baltimore was a drop away from first and goal with 30 seconds left and a timeout.

by jmaron :: Wed, 11/25/2009 - 3:26pm

NE by half DVOA

1st Half - 39.8 1st
2nd Half - 24.2 5th

There offence by quarter, 32.3, 45.4, 25.1, 23.2


1st Half - -19.2 3rd
2nd Half - 6.6 23rd

There defence by quarter

-25.4, -13.8, 14.4, .1

by Andrew Potter :: Wed, 11/25/2009 - 4:39pm

Those defensive numbers fit with my perception: Lack of depth meaning they struggle in the second half, particularly on the defensive line. They seem (from what I've seen, which isn't a massive amount) to struggle to counter offensive adjustments too, which seems wierdly atypical of the Belichick-coached Patriots.

by Aussie Bengal (not verified) :: Thu, 11/26/2009 - 7:27pm

The Cincinnati Bengals are clearly ranked TOO HIGH because they literally gave the game to the Raiders. The RAIDERS! @#@%*&(@!!! Drafting Ki-Jana Carter is way better than this. Ah well, at least we swept the Stealers. WHO DEY!

by After this week's games (not verified) :: Mon, 11/30/2009 - 10:23am

After Vince Young's incredible comeback last night, I find myself wondering how much the Patriots' huge win over the Kerry Collins-led Titans affects their DVOA. I guess specifically, I imagine that win is looking better and better as the Titans look better and better...but clearly that would be undeserved, since the Titans were simply not the team they are now when the Patriots shellacked them.