Week 9 DVOA Ratings

Week 9 DVOA Ratings
Week 9 DVOA Ratings
Photo: USA Today Sports Images

by Aaron Schatz

New England is still on top of the Football Outsiders DVOA ratings after nine weeks of play, with Indianapolis fourth despite an 8-0 record. Anyone want to settle this on the field?

This week's results scramble the top ten a bit, with Dallas making the biggest move and climbing four spots to number five. You'll also notice that Pittsburgh's win over Denver leads to those teams switching spots.

This is the first week where WEIGHTED DVOA starts to consider Week 1 below 95 percent strength, so we start to see some small differences between the full-season ratings and the weighted ratings. Giving more strength to recent games moves both New Orleans and Indianapolis ahead of Philadelphia, and moves Houston up from 16th to 14th.

My apologies for the lack of extended commentary this week, as I'm off to work on some material for ESPN's midseason package. Individual and team stats pages are updated, as well as playoff odds. FO Premium should be updated later this afternoon. We'll announce the winner of Loser League Part I in tomorrow's Scramble for the Ball; don't forget to register your team for Loser League Part II before Saturday.

* * * * *

These are the Football Outsiders team efficiency ratings through nine 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. Because it is early in the season, opponent adjustments are currently at 90 percent strength. WEIGHTED DVOA slightly discounts the results of the first five weeks to get a better idea of how strong teams are 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>


TEAM TOTAL
DVOA
LAST
WEEK
WEIGHTED
DVOA
RANK W-L OFFENSE
DVOA
OFF.
RANK
DEFENSE
DVOA
DEF.
RANK
S.T.
DVOA
S.T.
RANK
1 NE 41.0% 1 43.0% 1 6-2 33.3% 1 -6.3% 8 1.5% 11
2 PHI 35.5% 2 32.9% 4 5-3 9.9% 15 -18.6% 1 7.0% 3
3 NO 35.2% 3 34.0% 2 8-0 28.6% 2 -11.3% 4 -4.8% 26
4 IND 32.1% 4 33.6% 3 8-0 23.6% 5 -6.9% 6 1.7% 10
5 DAL 27.7% 9 28.9% 5 6-2 27.1% 3 4.5% 20 5.1% 5
6 BAL 27.2% 5 27.5% 6 4-4 21.5% 6 -5.7% 10 -0.1% 17
7 MIN 25.9% 7 26.9% 7 7-1 16.3% 9 1.0% 17 10.5% 2
8 PIT 24.4% 10 25.7% 8 6-2 24.6% 4 -5.9% 9 -6.0% 30
9 GB 18.2% 6 17.5% 9 4-4 16.4% 8 -10.6% 5 -8.8% 32
10 DEN 18.1% 8 17.3% 10 6-2 9.2% 17 -14.3% 3 -5.3% 28
11 ARI 13.9% 12 14.6% 11 5-3 11.6% 12 -1.6% 14 0.7% 14
12 MIA 12.9% 11 14.5% 12 3-5 13.3% 11 5.3% 21 4.8% 6
13 CIN 12.2% 15 12.2% 13 6-2 19.0% 7 2.0% 18 -4.8% 27
14 NYG 9.2% 13 9.0% 15 5-4 8.9% 18 -2.5% 11 -2.2% 24
15 ATL 8.9% 14 7.7% 16 5-3 10.9% 14 2.0% 19 0.0% 16
16 HOU 7.1% 17 9.0% 14 5-4 13.8% 10 9.9% 24 3.1% 7
TEAM TOTAL
DVOA
LAST
WEEK
WEIGHTED
DVOA
RANK W-L OFFENSE
DVOA
OFF.
RANK
DEFENSE
DVOA
DEF.
RANK
S.T.
DVOA
S.T.
RANK
17 NYJ 6.1% 16 4.3% 17 4-4 -8.4% 21 -14.7% 2 -0.2% 20
18 SD -1.1% 19 0.2% 18 5-3 9.5% 16 9.5% 23 -1.1% 23
19 SF -2.4% 18 -2.5% 19 3-5 -8.4% 22 -6.7% 7 -0.7% 22
20 JAC -2.7% 20 -4.4% 20 4-4 11.4% 13 16.0% 28 1.9% 9
21 SEA -6.9% 21 -7.8% 21 3-5 -7.0% 20 0.5% 16 0.6% 15
22 WAS -12.9% 22 -13.8% 23 2-6 -14.8% 26 -2.0% 12 -0.1% 18
23 CAR -15.2% 25 -11.9% 22 3-5 -10.6% 23 -2.0% 13 -6.6% 31
24 CHI -16.5% 23 -16.6% 24 4-4 -13.2% 25 8.9% 22 5.6% 4
25 BUF -16.9% 24 -18.0% 26 3-5 -18.7% 27 -1.0% 15 0.7% 12
26 TEN -17.1% 26 -17.9% 25 2-6 -0.6% 19 13.0% 26 -3.5% 25
27 KC -33.9% 27 -34.5% 27 1-7 -21.2% 29 14.9% 27 2.1% 8
28 TB -34.7% 30 -35.7% 28 1-7 -12.7% 24 22.7% 31 0.7% 13
29 CLE -41.0% 28 -42.1% 29 1-7 -33.9% 31 17.7% 29 10.7% 1
30 STL -44.1% 29 -43.1% 30 1-7 -19.8% 28 24.2% 32 -0.2% 19
31 OAK -50.8% 31 -51.9% 31 2-6 -37.4% 32 12.8% 25 -0.6% 21
32 DET -54.0% 32 -53.6% 32 1-7 -29.1% 30 19.4% 30 -5.5% 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 haad 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).


TEAM TOTAL
DVOA
W-L NON-ADJ
TOT VOA
ESTIM.
WINS
RANK PAST
SCHED
RANK FUTURE
SCHED
RANK VAR. RANK
1 NE 41.0% 6-2 38.3% 6.9 2 0.5% 17 8.4% 6 18.9% 25
2 PHI 35.5% 5-3 43.8% 6.2 8 -8.4% 30 4.4% 11 26.5% 31
3 NO 35.2% 8-0 38.8% 7.6 1 -1.5% 19 -9.1% 28 6.2% 4
4 IND 32.1% 8-0 43.2% 6.7 4 -4.4% 24 9.0% 5 12.5% 16
5 DAL 27.7% 6-2 29.6% 6.6 5 -2.1% 20 2.9% 15 9.8% 10
6 BAL 27.2% 4-4 28.1% 6.3 7 3.7% 10 -9.0% 27 11.0% 12
7 MIN 25.9% 7-1 31.1% 6.7 3 -5.9% 27 -10.5% 29 4.1% 1
8 PIT 24.4% 6-2 33.4% 6.5 6 -8.2% 29 -4.0% 22 11.7% 14
9 GB 18.2% 4-4 30.4% 5.2 14 -14.0% 31 1.9% 17 17.6% 24
10 DEN 18.1% 6-2 17.1% 5.9 10 4.4% 9 -8.0% 25 13.9% 18
11 ARI 13.9% 5-3 15.3% 6.1 9 0.5% 16 -17.8% 32 22.6% 30
12 MIA 12.9% 3-5 4.6% 5.3 12 12.4% 1 -2.0% 20 11.4% 13
13 CIN 12.2% 6-2 11.9% 5.4 11 7.2% 6 -17.8% 31 20.2% 26
14 NYG 9.2% 5-4 8.9% 4.6 18 -2.4% 21 12.6% 3 20.4% 27
15 ATL 8.9% 5-3 11.0% 5.2 15 7.8% 5 -2.2% 21 10.4% 11
16 HOU 7.1% 5-4 10.9% 5.2 13 -2.9% 23 2.2% 16 8.4% 8
TEAM TOTAL
DVOA
W-L NON-ADJ
TOT VOA
ESTIM.
WINS
RANK PAST
SCHED
RANK FUTURE
SCHED
RANK VAR. RANK
17 NYJ 6.1% 4-4 13.4% 4.7 16 2.7% 12 3.5% 13 12.4% 15
18 SD -1.1% 5-3 0.7% 4.5 19 -4.9% 25 -1.6% 19 4.2% 2
19 SF -2.4% 3-5 -11.1% 3.9 20 2.2% 14 -8.1% 26 8.5% 9
20 JAC -2.7% 4-4 0.7% 4.6 17 -7.3% 28 5.6% 8 22.1% 29
21 SEA -6.9% 3-5 -4.5% 3.7 21 -5.1% 26 -4.8% 23 15.2% 20
22 WAS -12.9% 2-6 -3.3% 3.4 25 -14.3% 32 14.5% 2 6.5% 5
23 CAR -15.2% 3-5 -25.8% 3.3 26 6.3% 7 14.9% 1 17.2% 23
24 CHI -16.5% 4-4 -12.1% 3.5 24 -2.7% 22 4.6% 10 14.3% 19
25 BUF -16.9% 3-5 -17.5% 3.5 23 1.3% 15 6.8% 7 15.4% 21
26 TEN -17.1% 2-6 -29.8% 3.6 22 11.4% 2 -0.4% 18 52.0% 32
27 KC -33.9% 1-7 -28.8% 2.1 27 3.6% 11 -5.3% 24 6.9% 6
28 TB -34.7% 1-7 -35.7% 1.9 28 9.6% 4 12.2% 4 13.0% 17
29 CLE -41.0% 1-7 -51.8% 1.8 30 10.3% 3 -11.3% 30 21.5% 28
30 STL -44.1% 1-7 -43.4% 1.9 29 -0.3% 18 3.9% 12 4.9% 3
31 OAK -50.8% 2-6 -53.7% 1.1 32 4.4% 8 3.1% 14 16.7% 22
32 DET -54.0% 1-7 -56.6% 1.4 31 2.6% 13 5.4% 9 8.2% 7

Comments

225 comments, Last at 14 Nov 2009, 8:38am

#1 by Rich Arpin (not verified) // Nov 10, 2009 - 2:11pm

I'm surprised that the pats offence has a better dvoa than New Orleans. Not that I'm complaining. I wonder how the defensive dvoa of indy will look after this week? Do the pats go spread against indy even with inexperienced depth recievers?

Points: 0

#32 by mm (not verified) // Nov 10, 2009 - 3:19pm

Well, the Saints offense can score a lot. Problem is, they've recently been turning it over a lot as well. The national media and the staff here at FO (read audibles) have been implying the Saints run defense has been the big problem these last 3 weeks, but it's actually been early turnovers, which is obviously (and justifiably) holding back the offensive DVOA for the Saints. (Yes, the Saint's defense ranks 25th in rushing DVOA, but 2.5% is hardly an awful number, particularly when their passing DVOA is -22.9%.)

Also, when you see the Saints' points scored, remember that has been aided with lots of takeaways by the defense, and a ridiculous number of defensive scores.

Points: 0

#61 by Still Alive (not verified) // Nov 10, 2009 - 4:30pm

Excellent points.

Points: 0

#102 by DomM (not verified) // Nov 10, 2009 - 5:56pm

The Pats have quite frequently used a spread formation with two tight ends this season. If Belichick doesn't trust his 3rd and 4th receivers (although I'm not sure he doesn't, Sam Aiken has been playing well recently) then he'll probably go with more of that. Of course if he feels a power running game gives them a better chance of winning he's quite capable of ignoring the spread altogether.

Points: 0

#131 by rms (not verified) // Nov 10, 2009 - 9:25pm

Could it help that the Bucs and Titans defense have been upgraded from abyssmal to abominable since their drubbing at the hands of the Pats?

Points: 0

#2 by sfckoski // Nov 10, 2009 - 2:13pm

Nit pick: SF's record is listed as 4-4. Unfortunately, they are 3-5.

Points: 0

#8 by The Guy You Do… (not verified) // Nov 10, 2009 - 2:24pm

Aaron obviously fears showing the 49ers with a losing record--there was a mistake on them last week showing them 4-3 instead of 3-4, too.

I can't say I blame him--Singletary is one of the scariest human beings on the planet.

Points: 0

#10 by Aaron Schatz // Nov 10, 2009 - 2:25pm

It was left over from last week's error, and is now fixed...

Points: 0

#11 by The Guy You Do… (not verified) // Nov 10, 2009 - 2:27pm

Watch out for Singletary.

Especially because you never know if he'll have pants.

Points: 0

#53 by EZE (not verified) // Nov 10, 2009 - 4:23pm

Can't blame Aaron for wishing the 49ers to be a better team than they are, they were the darlings of the FO Prospectus (BTW how's that Rams projection working out??). Until the team he despises most in the league wraps up the NFC West with 3 games to go, he'll be in denial. For the most part, I love this site, but the individual biases of the individual contributors need to be eliminated.

Points: 0

#59 by jmaron // Nov 10, 2009 - 4:28pm

"....but the individual biases of the individual contributors need to be eliminated."

Why? I'm not sure the biases are all that obvious but even if they are why on earth wouldn't you want them to express some opinion? That sounds incredibly boring to me.

Points: 0

#79 by EZE (not verified) // Nov 10, 2009 - 4:58pm

For a site that purports themselves to be a haven of statistical analysis, which means they rely less on fandom opinion to support opinions than factual analysis (which is quite refreshing nowadays, and the reason why many fans come to this site), I seem to be seeing more and more "facts" masquerading as viewpoints to support individuals' favorite teams.

I don't visit here for opinion (or chest thumping for validation of contributors projections), but there is virtually no analysis of how pre-season projections do or do not match reality.

Points: 0

#83 by Alvaden (not verified) // Nov 10, 2009 - 5:08pm

The biases are pretty obvious to any NFC West fan that heard Aarons anti-Cardinal rants last year, irregardless of what the stats say.

I wouldn't suggest that individual biases need to be eliminated. Just acknowledged.

Is it fate that to log in one of the words required was crowned, as in crown their ass!? This is getting a little spooky! :)

Points: 0

#108 by rk (not verified) // Nov 10, 2009 - 6:40pm

I don't know if what you said is accurate, but I will assume it isn't because of my extreme bias against people who use the word "irregardless."

Points: 0

#87 by Spielman // Nov 10, 2009 - 5:13pm

That is absolutely not true.

He'll be in denial long after the division is clinched.

Points: 0

#88 by Eddo // Nov 10, 2009 - 5:16pm

The 49ers were the darlings of FOA 2009? If anything, I'd say the Rams (NFC surprise team), Bengals (AFC surprise team), Bears (best NFC record), and Chargers (best AFC and overall record) fit that description much better than the 49ers.

Also, are you implying that the Cardinals are the team Aaron despises the most? What has given you that impression? He's a Patriots fan, so logically, shouldn't he despise the Jets, Dolphins, Colts, or even Steelers more?

Points: 0

#100 by ammek // Nov 10, 2009 - 5:43pm

FOA predicted 5.7 wins for the Niners. Perhaps you should read the book before getting on your high horse.

Points: 0

#112 by dbostedo // Nov 10, 2009 - 7:02pm

One more thing : do you really think that accidentally posting the wrong record for a team (that has since been corrected) is proof of any sort of bias?

Points: 0

#204 by Omaholic (not verified) // Nov 11, 2009 - 6:33pm

They're biased because they made an error on their W-L record? That proves it! Burn them all!

Points: 0

#3 by Will Allen (not verified) // Nov 10, 2009 - 2:15pm

The numbers are what they are, but my eyeballs, in the four games I've watched, don't tell me that the Packers have a top five defense.

The Cowboys have surprised me by how much better they have pass blocked.

Who would wager a lot on the Ravens beating the Bengals right now?

Football is hard to analyze.

Points: 0

#7 by Rich Arpin (not verified) // Nov 10, 2009 - 2:24pm

I don't care what dvoa says, when tampa lays 38 on you, it's not a good d. What happened to the aggresive corners and attacking d? Dom Capers. I glad Belichick let him go.

Points: 0

#15 by Arkaein // Nov 10, 2009 - 2:36pm

The D actually have very little to do with most of the points given up by the Packers to Tampa Bay:

1st TD: setup by INT return to within GB 15 yard line
2nd TD: blocked FG returned for TD
3rd TD: setup by 80+ yard kickoff return to within GB 15 yard line
4th TD: legitimate long drive, with TD scored on 4th down
5th TD: INT return

The defense only gave up 24 of the 38 points, and 14 of those points were on very short drives. Still not a good performance by the Packers against an inferior team with a rookie QB making his first start, but the special teams deserves the most blame (in addition to the long return and blocked punt, they also had a muffed kickoff returned to only the GB 4, and were outpunted all day), and the turnovers by the offense much of the remaining blame.

For more stats, TB was 3-12 on 3rd down, 1-1 on 4th down, averaged 3.2 yards per rush, and 6.2 yards per pass attempt, so really the defense performed okay given the field position they were given.

Points: 0

#18 by jmaron // Nov 10, 2009 - 2:40pm

you beat me to the post re: how TB scored.

I have watched two GB games against Minnesota and GB's defence was about on par with what I saw from Balt but not as good as SF or Pitt.

I'm curious as to your thoughts on GB's defence. My sense is it's a pretty good unit.

Points: 0

#28 by Will Allen (not verified) // Nov 10, 2009 - 2:56pm

They have appeared very average when I have watched them. No notable pass rushing results, which, in a passing league, is a notable defect.

O.K., I take that back. They rushed the passer well against the parking cones the Bears employ for an offensive line.

Points: 0

#52 by Arkaein // Nov 10, 2009 - 4:21pm

I think they're fairly solid, and should continue to improve as the rookies get better and the team gets more familiar with the 3-4.

The run defense has been a pleasant surprise, mostly shutting down or at least limiting opponents. The only bad game was against Cincinnati, and that was due to poor tackling, which otherwise has been pretty good. Very assignment sure with good gap control.

Coverage has been pretty decent, not giving up many long gains but maybe giving up a bit more underneath. Al Harris can run down the sideline against go routes as good as anyone in the business, but he's not as comfortable against underneath routes in playing off zone coverage. LBs appear to still do pretty well against RB passes except when bishop is in their as his aggressive play has left a few guys open. Average against tight ends. Woodson is still great, but he hasn't had the chance to blitz like it looked like he would in the preseason, where he really looked like a difference maker.

Pass rush has been the letdown. Clay Matthews has looked good, Raji has been dissappointing so far but I think that's due initially to his holdout and more significantly to his ankle injury. I think he'll have a growing impact as he gets on the field more. They have to do something with Kampman. I wasn't worried in the preseason because I'd read an article saying how Kampman had historically gotten nearly all of his sacks in nickle and dime defenses, and that he would play as a down lineman in those packages this season. For whatever reason that hasn't seemed true until the past few weeks.

It's kind of bizarre. The defense was looking great in both preseason and the first game against the Bears. A lot of that can be attributed to lack of game planning and Chicago's awful o-line, but it's like all of the creativity and aggression was taken away from the defense, and I'm not sure why. Maybe it's because the run defense is playing well and the coaches don't want to risk using overload blitzes that sacrifice gap control.

I think the DBs are good enough to hold up against the occasional failed blitz, but it doesn't seem like the coaches are willing to risk it. They either need to turn Kampman loose on the pass rush or start taking chances with some less predictable fronts if the pass rush is to improve.

Points: 0

#16 by C (not verified) // Nov 10, 2009 - 2:36pm

Tampa had 14 points off of a blocked punt TD and a pick 6. That has nothing to do with the Packers defense. Giving up 24 on the road doesn't sound nearly as bad as giving up 38.

Points: 0

#17 by jmaron // Nov 10, 2009 - 2:38pm

while they didn't look like much defensively in the two games I watched them against the Vikings; against TB they didn't give up 38 points. There was a ST TD (blocked punt) interception return TD and TD drives of 8 and 17 yards.

It wasn't exactly the defence that lost that game in TB.

Points: 0

#96 by ammek // Nov 10, 2009 - 5:30pm

The Packers rank #4 in takeaways/drive. Opponents have completed 54% of their passes and are averaging 3.5 yards per rush. Adjusting for the Lions, Rams, Bucs, etc makes these stats less impressive than they seem, but it's still pretty good.

Your eyes might be deceiving you because the Packers' D isn't very good in the red zone. Or because the 32nd-ranked special teams keep on leaving it to defend a short field (30th in LoS according to drive stats). Or because they're inconsistent (15th in DSR).

For instance, in the second half vs Tampa, the Bucs had six drives. Four of them were three-and-out. Another began on the Packers' 17 and resulted in a TD. The other was a long TD drive sustained by a disputable flag which negated an interception, and by a fourth-down conversion. Sure, it was the Bucs, but the D has been the best unit on the Packers by far.

Personally, I was shocked to see the special teams ranked as high as #32.

Points: 0

#215 by Whatev (not verified) // Nov 12, 2009 - 7:28pm

You could stick some college teams in there, I guess.

Points: 0

#4 by Fan in Exile // Nov 10, 2009 - 2:16pm

I am really happy that the Broncos have come through the hard part of their schedule with a 6-2 record. I can't wait for the rest of the year. We still get to play Oakland once and KC twice. Good things are coming.

Points: 0

#6 by Gubdude // Nov 10, 2009 - 2:20pm

And the Redskins this week.

Points: 0

#5 by Anonymous Coward (not verified) // Nov 10, 2009 - 2:20pm

NE, NO, IND, BAL, PIT, GB are all in the top 10 in both offense and defense. Seems like a lot of teams with that kind of balance for this point of the year, does anyone know if that's an abnormal number?

Of course GB, PIT, and NO are being drug down by awful special teams play so far.

Points: 0

#9 by for realz (not verified) // Nov 10, 2009 - 2:25pm

Is there a typo for the Jets - the DVOA is much lower than the VOA despite the fact they faced a pretty tough schedule.

Points: 0

#80 by Tundrapaddy (not verified) // Nov 10, 2009 - 4:58pm

1. The difference is ~7 percent, which isn't abnormally large. Indy, for example, has an 11 percent difference between DVOA and non-weighted VOA. Others have similar 7-ish percent differences.

2. The Jets schedule wasn't really that tough thus far. They've played 3 bad teams (based on DVOA), 2 great teams and 2 mediocre teams (one of them twice).

Points: 0

#12 by Anonymous1 (not verified) // Nov 10, 2009 - 2:32pm

still not clear how/whether these rankings account for injuries. i'm thinking in the case of the pats who destroyed the titans on a week where they had 2 or 3 starters out in the secondary. that thumping makes the pats' game against the titans look much better than say, the colts'

Points: 0

#22 by countertorque // Nov 10, 2009 - 2:47pm

The rankings don't account for injuries. The teams get graded based on the people who were on the field making the plays. There's no data to track who was on the field for each play.

Points: 0

#150 by theshadowj // Nov 10, 2009 - 11:23pm

I don't think that's what he's talking about. Tennessee started some low-round rookies at corner against the Patriots because of injuries, and that makes the Pats great passing day somewhat less impressive.

Points: 0

#164 by Jerry // Nov 11, 2009 - 3:52am

countertorque's point stands. The classic example was last year, when the Cowboys with Tony Romo were very different than the Cowboys with Brad Johnson, but both QBs' plays counted for Dallas. Adjusting for personnel is a reasonable thing to do, but it's necessarily left to the reader.

Points: 0

#194 by theshadowj // Nov 11, 2009 - 1:44pm

I see what your saying, and they do some "adjusting for personnel" after the season with AGL.

However, one of the major distinctions of DVOA is that it adjusts for strength of opponent. Using your example of the Cowboys with Tony Romo and Brad Johnson last year, DVOA sees no difference in the two teams when it adjusts the teams that played them. Remember when the Rams blew out the Cowboys when Brad Johnson started? The Rams had a ridiculously high DVOA for that game, and it was partly because DVOA saw the Cowboys as a good team and gave the Rams a good adjustment, which really didn't make much sense. If DVOA is used to predict future performance, it seems that incorporating significant injuries of past opponents into it would be useful.

Granted, any adjustment for injuries they could make probably still wouldn't have affected that game that much. Normally the drop-off from starter to backup isn't so gigantic. That's partly the Cowboys fault, though, for not getting a better backup.

Also, I don't know why it would be up to the reader to do this. They don't leave it up to the reader to do the other opponent adjustments, so why this one?

Points: 0

#205 by Jerry // Nov 11, 2009 - 7:03pm

One big problem is that, with the exception of quarterbacks, participation data isn't available, and charters can't pull it off the broadcasts. Another is that if FO could slice and dice with personnel, some opponent adjustments would be based on very small sample sizes.

It would be nice if all this personnel information could be effectively added to DVOA, but "the best is the enemy of the better." I'm willing to accept what we have now for what it is despite its limitations.

Points: 0

#208 by nat // Nov 11, 2009 - 9:58pm

Search for "McNabb Injury Version". FO did try to adjust DVOA for an injury to a key player in 2006. They gave it up after two weeks. Teams all have injuries, or players who play hurt, or off-field distractions, etc. It turned out to be better and more entertaining to do DVOA straight, and leave it to the beholder to consider the specifics.

Points: 0

#13 by Led // Nov 10, 2009 - 2:34pm

It's a league of haves and have nots as far as offense goes. NYG is 18th at 8.9 while Seattle is 20th at -7.0. That's a lot of good to great teams and a lot of bad to terrible teams with only Tennessee is in the middle. Factor in the wider spread between best and worst on the offensive side of the ball and we get a sense where exactly parity died. Is it because there's not enough starter qualty QBs and LTs to go around, or is it just generally harder to manage the cap with offensive personnel so talent disparity in the front office yields a greater talent disparity in offense on the field? Or is it just a statistical fluke and will straighten out next year?

Points: 0

#19 by C (not verified) // Nov 10, 2009 - 2:42pm

1. Brady
2. Mcnabb
3. Brees
4. Manning
5. Romo
6. Flacco
7. Favre
8. Rothlisburger
9. Rodgers
10. Orton
11. Warner
12. Henne/Wildcat
13. Palmer
14. Manning
15. Ryan
16. Schaub

With a coule argued exceptions, what do these teams NOT have in common with the bottom 16 teams besides San Diego and maybe Seattle?

Points: 0

#29 by Rich Conley (not verified) // Nov 10, 2009 - 3:00pm

I'm not sure DVOA isn't just all getting pushed up because of how bad the bad teams are.

I think the 0-20% teams aren't as good as a typical 20% dvoa team is.

Points: 0

#14 by C (not verified) // Nov 10, 2009 - 2:35pm

I still have trouble with Philly being #2 over Indy and New Orleans.

Carolina being lower than Washington...

Points: 0

#23 by Jimmy // Nov 10, 2009 - 2:47pm

Yeah Campbell is not good but neither is Delhomme at this point in his career.

Points: 0

#33 by Chief too // Nov 10, 2009 - 3:21pm

I've decided for myself that you need to look at DVOA and Variance in order to properly rank the teams. The Eagles have fewer estimated wins than the Saints by DVOA, and a model taking variance into account would probably give the Saints more wins head-to-head, even with the a higher total DVOA.

Points: 0

#20 by Temo // Nov 10, 2009 - 2:43pm

I was going to ask how the Dallas defense still grades out so low after a few good weeks of work, then I realized that they still haven't faced a top-12 offense this season.

Points: 0

#34 by Key19 // Nov 10, 2009 - 3:27pm

Still, I feel that they've been more impressive than 20th in the league. I guess we'll see this week when they face the supposed 5th best offense in the league.

Dallas could end up with back to back significant jumps in DVOA if they drub the Pack this week.

Points: 0

#72 by Temo // Nov 10, 2009 - 4:48pm

Then again, if you look at the difference between the 20th ranked defense and the best defenses, it's much lower than the spread between the offenses.

Points: 0

#84 by Tundrapaddy (not verified) // Nov 10, 2009 - 5:09pm

We're asking the same question of two different teams, then. I wonder the same about the Vikes. It appears to me that they are doing something I've familiar with as an Oregon Duck alum; they're playing to their competition's level, which unfortunately keeps opponents in games where good teams - Super Bowl caliber teams - would have put them away.

For example, if you put the Vikes against the Cowboys, neutral field - or even in Dallas - I'd wager on the Vikes winning. And I'm not usually a homer of raiderjoe proportions (excuse me - 'prooptoins'), as you might have noticed. But the 'Boys are showing a slight DVOA advantage, which I think is due to the Vikes playing to (or Leslie Frazier coaching to) the opponent's level.

Points: 0

#97 by Temo // Nov 10, 2009 - 5:31pm

Actually, I also do not feel that the Cowboys are the Vikes' level right now. We may have a better offense (I'd put the Cowboys' offense up against anyone when Romo's not melting down), but I do feel the Vikes are a better defense than their rating.

It was more the #20 ranking of the Cowboys' defense that got me. Although I do kind of understand it now.

Edit: That said, Brett Favre's never won in Dallas and I'd like to keep it that way. At the very least we are able to now say that Favre's never won in Texas Stadium and never will. Woo.

Points: 0

#119 by Key19 // Nov 10, 2009 - 7:29pm

I also think the Vikings would win (as of today). But that said, I don't think there's a team in the league that the Cowboys just flatout can't beat.

Points: 0

#21 by Gubdude // Nov 10, 2009 - 2:44pm

Dallas is 5th in special teams.

I'm attributing this to DeCamillis arriving and Buehler kicking off.

Does anyone think Crayton being demoted led to those two punt returns for TD's? Fresher legs?

Points: 0

#36 by Key19 // Nov 10, 2009 - 3:30pm

DeCamillis has been incredible for us, no doubt. He is also a big reason why we landed Gerald Sensabaugh supposedly, which has been great as well.

Crayton demotion is overblown. Other than stiffarming the punter to the ground on the 2nd TD return, he wasn't touched on either one. That tells me that blocking has been the difference, not his own performance. He pretty much just ran straight forward and there was no one available to tackle him. Credit to Joe D and the improvement of the players throwing the blocks.

Points: 0

#40 by Arson55 // Nov 10, 2009 - 4:00pm

DeCamillis and Buehler are my heroes. It is so freaking weird to see a good Cowboys special teams. I don't think they've ever had good special teams play since I started watching football.

Points: 0

#56 by Temo // Nov 10, 2009 - 4:26pm

The last time we had a top-10 special teams was 1998, when they ranked first: http://footballoutsiders.com/stats/teamst1998

That was back when Deion and Kevin Mathis were returning kicks and punts, Toby Gowin was booming kickoffs, and Richie Cunningham was doing FGs.

Toby Gowin and co were an absurd 22.3 points above average on kickoffs that year. Second most was the Jets with 10.6 points! Gowin actually didn't kick too deep (still top 10), but the Cowboys only allowed 18.5 yards per return, least in the league.

Actually, all of those Cowboys teams pre-Dave Campo had really good special teams units.

Points: 0

#57 by Arson55 // Nov 10, 2009 - 4:26pm

I started watching football in Campo's second year which would be...2001? So yeah, I guess I have never seen a good Cowboys special teams until this year.

Points: 0

#62 by Temo // Nov 10, 2009 - 4:33pm

Between you and Key19, I'm starting to feel really old despite still being in my early 20s.

Then again, it would make sense that the only Cowboys fans other than myself to visit this site are those that have not been weaned on the arrogance-building domination of the 90s 'boys.

Points: 0

#64 by Arson55 // Nov 10, 2009 - 4:37pm

Don't feel old; I'm also in my early twentys. I just did not get into football until high school.

Points: 0

#68 by The Powers That Be // Nov 10, 2009 - 4:44pm

Well, some of us were weaned on the dominance of the '70s Cowboys, if that makes you feel any better.

Points: 0

#74 by Temo // Nov 10, 2009 - 4:52pm

Not quite the same; everything I've heard/read of the pre-Jerry Jones Landry Cowboys was that they were more low-key than the current incarnation.

I've heard that the Steelers of that era were much worse.

Points: 0

#125 by Lance // Nov 10, 2009 - 8:16pm

Yeah, my earliest real Cowboys memory is "The Catch"-- not quite the 70's but close. And it's why I put the 49ers higher than the Redskins, Eagles, or Giants as most hated Cowboys rival...

Points: 0

#163 by Staubach12 // Nov 11, 2009 - 2:31am

Well, I'm 29, so I grew up watching the 90s team. My dad has always been a huge football fan, so I have Cowboy memories going back to the Landry years. However, I didn't REALLY start following the team until the Jimmy Johnson Era.

Points: 0

#121 by Key19 // Nov 10, 2009 - 7:32pm

See, I even missed the Campo years. I started my Cowboys fandom in 2004.

Points: 0

#206 by Rocco // Nov 11, 2009 - 8:43pm

I wouldn't say you "missed" anything there.

Points: 0

#209 by Arson55 // Nov 11, 2009 - 10:37pm

True, i suffered through two of those freaking Campo years. Nothing to miss. Of course, I started following the team during the Campo years, so kinda my fault.

By the way, I roll my eyes at bandwagon Cowboys fans.

Points: 0

#24 by starzero // Nov 10, 2009 - 2:48pm

after watching indy the past two games, i'm surprised they're as high as 4. maybe i'm just used to easier wins than we've had. maybe i expect more from these guys than they're delivering. to me, they don't look like an 8-0 team. the patriots game may settle this.

Points: 0

#124 by Ben // Nov 10, 2009 - 8:15pm

I agree. The offense is out of sync, and the defense is getting hit by injuries. I think next week is going to expose the Colts shortcomings.

Points: 0

#25 by jmaron // Nov 10, 2009 - 2:49pm

I think the Vikings offence is better than 9th. They've played some of the top defences (GB (5th), SF (7th), Pitt (9th), Balt (10th) and had very little trouble moving the ball in those games. They didn't score many points against Pittsburgh but that had a lot to do with two red zone turnovers.

What they didn't do is run up huge numbers against the lousy teams (Det, Clev in particular). I think the first few games were almost like pre season games (owing to Favre's late arrival at camp).

I'm very interested to see if they put up some big numbers in the coming weeks against Det, Sea and Chic.

I'm also hoping Reynaud their punt returner is healthy again. I think he has top level return talent (right up there with Harvin). Jaymar Johnson has been fine as the punt returner, but Reynaud is another game breaking type of player.

Points: 0

#37 by TGT2 (not verified) // Nov 10, 2009 - 3:40pm

An offense that moves the ball well most of the time, but commits turnovers? 9th sounds about right.

Remember that DVOA is tuned to correlate with future performance. We know from the Guts and Stomps research, that destroying bad teams is more correlated with future success than squeaking by good ones. Your evidence that the Vikings are better than their ranking is actually evidence they are ranked properly.

Points: 0

#39 by jmaron // Nov 10, 2009 - 3:56pm

I think they are ranked properly based on their actual output but I was making an observation that they improved considerably as the season progressed and that they are probably a better offence than their ranking.

The offence for the first three games was quite pedestrian - they weren't turning it over, but they were very conservative. The offence changed as of the 4th game against GB.

Offensive DVOA by game:

Cle: 12.2
Det: -4.7
SF: -8.7
GB: 44.0
StL: 33.9
Bal: 29.1
Pit: -13.8
GB: 51.8

Points: 0

#50 by TGT2 (not verified) // Nov 10, 2009 - 4:17pm

Since their weighted offensive DVOA is greater than their offensive DVOA, I agree with your premise. :) I also have a habit of extrapolating continued change, but it digresses equally as much as it holds up. We'll have to see what happens moving forward. I haven't seen enough of their plays (or trust my eyes enough) to speak to it subjectively.

Objectively, the sample sizes are so small, that it's hard to tell. I wouldn't mind some DVOA confidence intervals (95% chance they're between 5% and 25% DVOA, 67% chance they're between 13% and 18%).

Points: 0

#63 by jmaron // Nov 10, 2009 - 4:37pm

one can get caught by extrapolating change - particularly when change is in the direction one is hoping for.

From a purely observational view. The pass protection has improved dramatically from the first three games. As well, or perhaps as a result of the pass blocking improvement, the Vikings have attacked down the field far more often and far more effectively. They have also had a much higher pass/run ratio when games were still in question.

Points: 0

#75 by peterplaysbass (not verified) // Nov 10, 2009 - 4:53pm

I think a large part of that is due to Sullivan and Loadholt learning on the job. They've both triple their experience (or so - injuries not withstanding) since after the first two weeks, and it shows.

Points: 0

#86 by Will Allen (not verified) // Nov 10, 2009 - 5:13pm

I also think the Zombie King started making better assignment calls at the line of scrimmage after a few weeks.

Points: 0

#90 by Tundrapaddy (not verified) // Nov 10, 2009 - 5:18pm

Well called. It's always best to look at the line first. Now if Childress would get Loadholt active on downfield run blocking in the 2nd half, that would really help them put games away.

The things that concerned me were 1)Birk's departure, and 2) knowing that Rick Spielman was still affecting the draft. Both of those concerns are somewhat alleviated.

Points: 0

#26 by Bobman // Nov 10, 2009 - 2:52pm

This is awesome--From a Colts fan perspective, Pats and Colts are FINALLY essentially even on D and ST, and surely the Colts have a sizeable advantag-- what's that? You're shitting me.

Okay, well, if only there were a way to settle this on the field.... like the BCS does it. Maybe they can do it soon, and on national TV so I can watch from Seattle.

Oh if only there were a way....

Points: 0

#43 by jebmak // Nov 10, 2009 - 4:05pm

Like the BSC?

So NE should play Philly?

(Yes, I realize that you are being sarcastic too.)

Points: 0

#109 by Fan in Exile // Nov 10, 2009 - 6:41pm

These are two fantastic posts

Points: 0

#156 by Bobman // Nov 11, 2009 - 12:19am

I agree about Jebmac's--I was being sarcastic but you took it to the next level. Excellent.

Although undefeated NCAA teams from major conferences, even if they've only played a lineup of HS squads and stiffs, tend to get the nod even if there's a potentially superior 11-1 team out there. Not sure if I'd consider the NFCS a "major conference" and Philly has losses, so maybe NE/Indy for the national title game? Hey, that means that after next week, the rest of the season is moot.

Points: 0

#27 by drobviousso // Nov 10, 2009 - 2:55pm

Playing Defense: It's easier with Troy Polamalu.

Points: 0

#30 by SteveVolk (not verified) // Nov 10, 2009 - 3:07pm

I have questions. Going forward, how can we factor in, for example, the seasonal collapse that visits itself upon Tony Romo each December? What is the quirk in the DVOA formula that allows Baltimore to be ranked so high, yet mired in the middle at 4-4? I'm not a hater. I subscribed because I was a big fan of the whole Bill James revolution in baseball. But I do wonder if football submits itself to this kind of statistical analysis.
I think some teams just prove to be greater than the sum of their parts, or at least greater than the sums of their accumulated statistics. Just eyeballing New England's D, for instance, they look more 14th to me than eighth. But we shall see. They're catching the Colts at the right time this year: just after the annual Bob Sanders is out for the season announcement (and tell me, how do we factor that into DVOA?).
—Cheers, Steve V.

Points: 0

#42 by TGT2 (not verified) // Nov 10, 2009 - 4:04pm

James had 80 years of detailed statistics to work with. FO so far has 14 years, and only a few years of charting data. It's an ongoing process. The small sample size each year is also a problem.

Saying your eyeballs don't match the DVOA ranking is silly for a number of reasons. First, eyeballs lie (See Leaf, Ryan and Russel, JaMarcus). Second, your brain lies (You don't remember all the minor plays, just the ones that stand out). Third, comparing by ordinal ranking is stupid. Go by actual % numbers. Fourth, if Defenses are worse than usual this year, than a defense that looks like it should be middle of the pack, can really be middle-high (that's all moving from 14th to 8th is, not even changing the quartile a team is in).

In Baltimore's case, my eyes tell me in their 8 games, they have 3 decisive wins, 1 decisive loss, and 4 games that were essentially tossups in the last minute. They went 1-3 in those games. If they'd gone 3-1 or 4-0 (75% chance at a field goal, dropped pass in the hands, 1 unpredictive penalty), they'd be 6-2 or 7-1 and you wouldn't have a problem with them. Your issue here is that for the question of team strength, you're checking the answers you get from DVOA against the 'back of the book', but the back of the book doesn't print how good teams are, just their win-loss record. Unfortunately, that's a very flawed representation of actual team strength, especially in a sample size of 8.

The "quirk" in the DVOA formula is that it isn't another way of telling you which team won. If you want that, just read the standings.

Points: 0

#49 by RickD // Nov 10, 2009 - 4:17pm

Our minds are tricking us w.r.t. Baltimore because they really looked weak on Sunday in Cincinnati. At no point did I think they had a shot of winning that game.

And the most recent observation gets overblown in our minds.

Points: 0

#129 by Oldcat (not verified) // Nov 10, 2009 - 9:00pm

The first meeting was almost the same. Subtract the Ed Reed pick 6 and the final score is 17-7. Their 1 touchdown in that game was on a play only slightly less fluky than the Denver game winner, being that 2 Bengals tackling Ray Rice collided and he ran to the score. This was on a 2nd down and 24, so their drive wasn't a model of efficiency.

I could accept that the offense could look bad on the average because of Carson's turnover (a pick 6 to Reed, 1 pick 6 and another close to that in the Green Bay game), and the special teams because of the longsnapper. I'm wondering more on how stopping GB, Baltimore x 2, and Denver's offense for most of the game leads to no improvement on that side of it.

Points: 0

#58 by Will Allen (not verified) // Nov 10, 2009 - 4:27pm

Funny, my eyeballs never lied to me about Ryan Leaf or Jamarcus Russell. It is silly to say that intuition, or eyeball perception, is a silly way of looking at DVOA rankings. DVOA is not perfect, especially in regards to understanding what is likely to happen in the future. It is no knock on DVOA to state that perceptive eyeballs were not nearly as surprised at the Giants' victories against the Packers and Patriots in the playoffs a few years ago, compared to a strict DVOA analysis.

Yes, the problems with eyeball perception that you describe are very real, which is why a quantitative approach is so valuable, but that doesn't mean that there are not inherent problems with any quantitative approach which can be mitigated with a perceptive set of eyeballs. Which is why I will say again that football analysis is hard.

Points: 0

#70 by Scott P. (not verified) // Nov 10, 2009 - 4:46pm

"It is no knock on DVOA to state that perceptive eyeballs were not nearly as surprised at the Giants' victories against the Packers and Patriots in the playoffs a few years ago, compared to a strict DVOA analysis."

Well, eyeballs said that the Patriots defeated the Giants in New York pretty handily (the final score was deceptively close), which is why there was surprise they lost in the Super Bowl.

Points: 0

#98 by Will Allen (not verified) // Nov 10, 2009 - 5:34pm

Eyeballs that don't understand the difference between a week 17 game without playoff implications, and a playoff game, are really missing something, and I believe my eyeballs told me that the Giants led that game 28-16 in the fourth quarter.

Points: 0

#117 by dbostedo // Nov 10, 2009 - 7:23pm

Will - I believe that a lot of people look at teams, get an idea in their head and run with it, even though they really aren't all that sure about it. Then if it pans out, it seems like it was obvious to them all along. I believe this happened a lot with the Giants SB win over the Pats.

Just a guess on my part, but I think when you start to try to actually make weekly picks (straight up or against the spread) you start to realize how truly uncertain you are about particular teams/games. For instance, say you could have gotten a big group of people together before the Giants/Pats super bowl who said they were CERTAIN the Giants would win. I'm betting that if you had offered them a lot of money if they correctly picked the game, most would have been a lot less certain of the outcome, and a lot of people would have taken the Patriots.

I think it's easy to build it up in your mind with nothing on the line, and want to feel vindicated when it happens, like you "knew" it all along, when you were really wishing, or guessing, or picking randomly, or just plain being a fan and rooting.

Just a theory, but makes sense to me based on my experience...

Points: 0

#120 by Danish Denver-Fan // Nov 10, 2009 - 7:32pm

The "sitting-at-home-watching-who-wants-to-be-a-millionaire"-syndrom.

Points: 0

#133 by Will Allen (not verified) // Nov 10, 2009 - 9:43pm

Well, I never said I "knew" it all along, so I don't what you are referring to. I didn't wager anything on that game, so, no I wasn't putting my money where my skepticism lied, with regard to the overwhelming likelihood of a convincing Patriot victory. However, a previous Super Bowl, which DVOA has also declared a huge upset, Rams vs. Patriots, was one where I made a substantial wager on the underdog, simply because I thought the public was not giving sufficient weight to the difference in coaching quality between the two teams. DVOA didn't either, retroactively.

I'm not trying to rip quantitative analysis. I'm simply stating that it is not a perfect tool, and that judicious subjective analysis can augment quantitative methods. Like I've said a few times now, however, it is damned hard.

Points: 0

#183 by Rich Conley (not verified) // Nov 11, 2009 - 11:48am

So you 'felt' that 2 of the Patriots starting offensive lineman were going to be hurt in the first half? You knew that Brady was going to aggravate his ankle injury in the first half?

because those were probably the biggest factors in the outcome of that game.

Points: 0

#192 by Will Allen (not verified) // Nov 11, 2009 - 1:24pm

No, Rich, I felt that the Giants defense was playing well enough that it was not nearly as likely, as perceived by the public, that the Patriots would win easily. One did not need to predict injuries to come to that feeling.

Points: 0

#196 by dbostedo // Nov 11, 2009 - 2:27pm

I think we mainly agree Will - I was just giving an example that's been bantered around here recently including in the discussion leading to my post, not trying to say you were of the "I-knew-for-sure-that-the-Giants-would-win" ilk. Maybe I did that poorly.

In any case, it still speaks to what you're talking about with trusting your eyeballs or intuition. I'd guess most people's intuition isn't as certain as they think it is. Yes, many people were right when they looked at Ryan Leaf or (so far) JaMarcus Russell. However, a lot of people (not necessarily you) will comment that they knew all along that Leaf wasn't going to be any good. My point was that they may not have known all along - they were just blustering, and if they had had to put money on it, they would have realized they were a lot less certain.

I've certainly found that to be the case with me in trying to pick games - even straight up for very small amounts.

Points: 0

#99 by drobviousso // Nov 10, 2009 - 5:36pm

Third, comparing by ordinal ranking is stupid. Go by actual % numbers.
When reading these, it's probably helpful to mentally group the rankings, and compare the groups. If you do that, you can see that NE is maybe a bit ahead of the pack, PHI, NO, and IND are jockying for second, etc. This week is a pretty good example of a bunch of clusters, but it happens often.
EDIT: http://spreadsheets.google.com/pub?key=tEx59qiOzl4j7mIbc_k6zFQ&single=true&gid=0&output=html

Points: 0

#81 by peterplaysbass (not verified) // Nov 10, 2009 - 5:00pm

These are great questions, and I'll respond off the cuff for fun.

"Going forward, how can we factor in, for example, the seasonal collapse that visits itself upon Tony Romo each December? What is the quirk in the DVOA formula that allows Baltimore to be ranked so high, yet mired in the middle at 4-4?"

A model that factors in a player's annual collapse is likely to hurt the accuracy more than ignoring such unpredictable things. I think we're meant to look at DVOA as a starting point, and then dialogue (yes, Dallas rates higher than so and so, but since they play in December I think it could easily be a loss for the Cowboys.

Also, Baltimore lost three "heartbreakers" (as the media loves to say), but there's some truth there. Baltimore played very well against good teams and had some events that DVOA considers luck/fluke bounce the wrong way for them. Remember, last year's version of this model predicted a strong Baltimore team coming off of the 5-11 season in 2007. Few people thought they'd do well with that team lead by a rookie QB named Flacco, and then they went 13-3. (Something like that anyhow, I might be a little off on the W-L records and I'm too lazy to look it up right now)

This site makes some funny declarations, and sometimes they're way off, but sometimes they hit the nail on the head when everyone else missed, and retrospection makes it seem so obvious. "Why didn't I think of that?" All in all, the numbers are mroe logical than most other analysis I've found and make excellent starting points for dialogue/debate. The forums on this site are intelligent and (or?) entertaining. That's why I read.

Points: 0

#105 by willy72480 (not verified) // Nov 10, 2009 - 6:12pm

In keeping with Baltimore's high DVOA despite a 4-4 record, I was wondering the following: how do these ratings account for penalties?

My reason for asking is, I've watched every Ravens game and penalties have played a crucial role in every loss (mainly, costly penalties occurring on game-changing plays). When I checked the overall penalty stats for this year, I realized the Ravens are indeed the most-penalized team in the NFL, with 590 total penalty yards so far, that's including the 2 teams who still haven't had a bye yet and essentially have had an extra game to commit penalties (Texans, Giants). The Ravens have 81 more total penalty yards than the next team on the list (Packers) - that's a whole game's worth more penalty yards than the Packers, which is really amazing when you think about it.

If you go down and compute the differences in total penalty yards between consecutive teams on the list, you'll see that the Ravens difference of 81 total penalty yards between them and the Packers is the biggest difference in total penalty yards between consecutive teams on the entire list. The next highest difference between consecutive teams is only 28 yards.

Is this statistically significant? It would seem to me that the Ravens 590 Total Penalty Yards is a statistical outlier of the group that deviates significantly from the norm for the rest of the league. (Of course, it may be that my memory of statistics has weakened considerably since college.)

The Ravens average almost 10 yards/penalty, almost a full yard per penalty more than the rest of the league. They have a league-leading 7 Pass Interference penalties, which accounts for almost 100 penalty yards of their total (not actual, just an estimate).

I'm just wondering how DVOA accounts for penalties, especially crucial pass interference penalties that impact huge swings in field position and penalties on crucial plays like fourth down, incomplete pass on third down, etc. If anyone can give an abridged explanation without me having to go research this, I'd be greatly appreciative.

Link for penalty stats: http://www.nfl.com/stats/categorystats?season=2009&seasonType=REG&d-447263-o=2&conference=ALL&tabSeq=2&role=TM&statisticCategory=GAME_STATS&d-447263-p=1&d-447263-s=PENALTIES_YARDS_PENALIZED&d-447263-n=1

Points: 0

#31 by Cabbage // Nov 10, 2009 - 3:13pm

The Bears are clearly ranked too accurately because I am a fan boy and I came here to hear that Cutler has been goin' all Muhammed-Ali-Rope-A-Dope on the NFC Norris -- wearing the division down before solidifying his grip on the crown. Satisfying my residual Cub fan angst is way better than this. Garret Wolfe is teh skaterbak wiz!!11

Points: 0

#94 by Tundrapaddy (not verified) // Nov 10, 2009 - 5:26pm

You know, it's really not a complete DVOA comment thread without at least a few of these. Although the 'rated too accurately' part about made me lose it.

Points: 0

#104 by Arson55 // Nov 10, 2009 - 6:12pm

"Rated too accurately" was rather amusing.

Points: 0

#35 by Led // Nov 10, 2009 - 3:29pm

Baltimore's defensive DVOA is +18.8 from last year and the Jets' defensive DVOA is -17.2. Personnel changes don't appear to be hugely significant (but let me know if I've overlooked anything). Looks like Ryan and Pettine know something about defense.

Points: 0

#38 by Danish Denver-Fan // Nov 10, 2009 - 3:49pm

So estimated wins gives us, sort of, the teams' "true" wins, correct?

So if I were to claim that the Broncos are extremely lucky to be 6-2, I'd be wrong, right? I have the feeling the Broncos have had a lot of calls and bounces go their way, but estimated wins seems to put the Broncos right at 6-2 (5.9-2.1).

Points: 0

#46 by merlinofchaos // Nov 10, 2009 - 4:10pm

I'm pretty surprised that Denver didn't fall further. The last 2 losses included a LOT of bumbling by the Broncos offense. But still ranked #3 in defense. That's nice. That might lead us to a 1st round playoff exit. =)

Points: 0

#51 by Danish Denver-Fan // Nov 10, 2009 - 4:20pm

I 100% agree!

Per the playoff-odds page:
Chance of making the playoffs = 94.7%
Chance of appearing in the AFC championship game = 31.1%
That leaves a 94.7-31.1 = 63.6% of loosing in the first two rounds.

That actually seems low. Very low. Once weighted DVOA kicks in for real, that will drop.

Points: 0

#47 by jebmak // Nov 10, 2009 - 4:13pm

I think that Est Wins is currently out of nine games and that they would be 5.9-3.1.

Points: 0

#48 by Danish Denver-Fan // Nov 10, 2009 - 4:15pm

Good point - I knew something was up...

Points: 0

#91 by mm (not verified) // Nov 10, 2009 - 5:21pm

estimated wins gives us, sort of, the teams' "true" wins, correct?

Remember that it's a projection for a 'league average schedule', which makes it good for comparing teams that have played different schedules. If you think your team has played more good teams than 'average' you might think their "true" wins should be lower; vice versa if you think they've played more poor teams.

Points: 0

#41 by jmaron // Nov 10, 2009 - 4:02pm

the wonderful thing about the NFL is how important a game in week 9 can be or any week for that matter. Balt's playoff chances dropped 27.8% after a loss to Cinn. GB down 25%. Atlanta, Pittsburgh and Dall up around 20%.

Unlike Baseball, Basketball and Hockey - where one game in a long season usually isn't of much importance; most games in the NFL season are extremely meaningful.

Points: 0

#44 by TGT2 (not verified) // Nov 10, 2009 - 4:06pm

Careful with the word meaningful. Yes, each game is more meaningful in making the playoffs, but they aren't any more meaningful in team strength.

Points: 0

#55 by jmaron // Nov 10, 2009 - 4:25pm

I was simply commenting on the entertainment value derived from watching a game in the regular season.

I'm not sure I understand what you're getting at by saying "they aren't any more meaningful in team strength."

Points: 0

#92 by Rich Conley (not verified) // Nov 10, 2009 - 5:24pm

"I'm not sure I understand what you're getting at by saying "they aren't any more meaningful in team strength."

Beating a team by less than about 14 points doesn't really mean you're any better than that team.

Points: 0

#45 by chemical burn // Nov 10, 2009 - 4:07pm

So, I know Eagles fans have been notoriously unfair in the past... but at what point can we legitimately start blaming the offense for struggling down to a 5-3 record? Their under-performance is clearly the result of an insanely inconsistent passing game... (not that the 9th ranked running game is some unstoppable force...)

Points: 0

#54 by Ven (not verified) // Nov 10, 2009 - 4:23pm

These ratings are a joke. I'll have to stop visiting this site.
You guys must be from Philly.

And a 4-4 team near the top? LOL

Points: 0

#60 by chemical burn // Nov 10, 2009 - 4:29pm

Nooooo!!!! Please continue to visit the site and post comments! You've added so much to the dialog - you're not hurting FO, only us other readers!!!

Points: 0

#157 by Bobman // Nov 11, 2009 - 12:26am

Awesome response, but too many words in excess of one syllable to get the message across.

Points: 0

#65 by Shedds (not verified) // Nov 10, 2009 - 4:41pm

Found near the top of the article:

To save people some time, please use the following format for all complaints: is clearly ranked because . is way better than this.

Points: 0

#66 by Shedds (not verified) // Nov 10, 2009 - 4:41pm

Haha, that'll learn me. You know what I meant.

Points: 0

#69 by Dean Wormer (not verified) // Nov 10, 2009 - 4:45pm

Ray Ventrone? That you?

Points: 0

#71 by Anonymous1 (not verified) // Nov 10, 2009 - 4:47pm

You will be missed.

Points: 0

#67 by MattinDenver (not verified) // Nov 10, 2009 - 4:42pm

Ok so the whole Eagles being overranked is getting to be ridiculous. They have ONE quality win - against the Giants at home - they lose to New Orleans by 26. They lose to the RAIDERS - and they are #2 in the league? Are you kidding me????
I think the way the Eagles are being ranked is turning people off to FO in general and to the DVOA rankings. Last year it was odd, this year it's just outrageous.
Somebody, anybody, please, please tell me how this is remotely possible! It's just crazy.

Points: 0

#73 by Anonymous1 (not verified) // Nov 10, 2009 - 4:49pm

While the Eagles looked odd last year, they did do OK in the playoffs. Doesn't this justify somewhat their "odd" high ranking?

Points: 0

#82 by Kurt // Nov 10, 2009 - 5:06pm

Maybe. The Giants had a bit of playoff success the year before; it didn't prevent FO from spilling tens of thousands of words explaining that none of it meant anything, and that their low DVOA ranking was absolutely correct. And Arizona had some playoff success last year too; the FO reaction was more along the lines of "the NFL playoff system is broken" than "DVOA is something other than the pinnacle of human achievement."

Points: 0

#93 by mm (not verified) // Nov 10, 2009 - 5:25pm

Exactly.

It is certainly bizarre that people are still complaining about the Eagles' high ranking last year. DVOA is supposed to project future performance and their performance in the playoffs matched what DVOA predicted.

Now if people wanted to complain about Arizona's DVOA last year, that might make more sense.

Points: 0

#85 by peterplaysbass (not verified) // Nov 10, 2009 - 5:11pm

Read the threads from past weeks' DVOA articles. There are numerous examples of teams that goto/win the SB after losing to some cruddy punk-ass team during the regular season.

Logic dictated by thoughts like "but, but they lost to the RAIDERS!" is emotional drivel. The numbers provide context and clarity.

And I HATE the Eagles.

Points: 0

#116 by HostileGospel // Nov 10, 2009 - 7:21pm

Hi, you must be new here? This happens *every* year. As an Eagles fan, I can't explain it either, but please believe I'd rather DVOA ranked the Eagles last every year and they won a goddamned Super Bowl once in a while.

Also, why do people come to FO to complain that the rankings don't match up well enough with subjective rankings for their tastes? I will never understand this.

--
Overall, I'd be kind of embarrassed to critique something when I didn't know what the hell I was talking about, but then, oh yeah, my NAME is on what I write, isn't it?

-Les Bowen

Points: 0

#166 by Vincent Verhei // Nov 11, 2009 - 5:32am

Week 1: The Eagles blow out the Panthers by 28 points. It's still the worst Panthers loss of the year. Eagles' DVOA for the game is 91.3%,

Week 2: The Eagles lose big to the Saints, but most teams lose big to the Saints. Eagles' DVOA for the game is 1.2%.

Week 3: The Eagles blow out the Chiefs by 20 points. Only one other team (the Chargers) had a better margin against Kansas City. Eagles' DVOA for the game is 28.8%.

Week 4: Bye.

Week 5: The Eagles beat Tampa Bay by 19 points. Only the Giants and Patriots have done that. Eagles' DVOA for the game is 90.9%.

Week 6: The Eagles lose to Oakland. That sucks. Eagles' DVOA for the game is -25.7%.

Week 7: The Eagles beat Washington by only 10 points. Despite winning, Eagles' DVOA for the game is negative, -11.0%.

Week 8: The Eagles beat the holy hell out of the Giants by 23. Even the Saints only beat the Giants by 21. Eagles' DVOA for the game is a mighty 104.1%.

Week 9: The Eagles lose a very close game to the Cowboys. Despite losing, the Eagles' DVOA for the game is positive, 12.3%.

So there you go -- the Kansas City win basically cancels out the Oakland loss, and the Dallas loss basically cancels out the Washington win. In every other game they've had a positive DVOA, usually by a huge margin. That's why their DVOA for the season is so high.

Points: 0

#213 by zip.4chan.org/sp/ (not verified) // Nov 12, 2009 - 12:04pm

Why is trouncing cupcakes worth so much?

Points: 0

#214 by nat // Nov 12, 2009 - 1:47pm

Because failing to trounce them is worth so little.

Points: 0

#216 by Vincent Verhei // Nov 13, 2009 - 2:09am

Because trouncing cupcakes is a better indicator of future success than close wins over good teams. See Guts and Stomps in the glossary.

Points: 0

#217 by Spielman // Nov 13, 2009 - 10:17am

Not that you don't know this, but that's the key to the "DVOA just loves the Eagles" issue from the last couple years. The Eagles were 2-5 in games decided by 7 or less in 2007, 1-5-1 (and 0-1 in the playoffs) in 2008, and are 0-2 this year. DVOA sees that as lousy luck. Our eyeballs (God, if that word appears in this thread much more, it'll cease to have meaning.) tend to see it as having enormous meaning, not least because close games involving good teams are the ones we tend to watch the most.

Points: 0

#223 by Whatev (not verified) // Nov 14, 2009 - 8:27am

Well, this is part of what drives concerns about the Eagles. Subjectively, they seem to suffer from an inability to seal the deal. I was wondering if the Eagles actually had a bimodal distribution, but glancing at the chart from last year seemed to indicate that it's just really high variance.

Points: 0

#218 by MattinDenver (not verified) // Nov 13, 2009 - 1:23pm

Thanks Vince! It makes more sense with a game-by-game breakdown. I guess the Eagles are just bad at finishing games, bad in the red zone....... but very good in other areas? I realize dvoa looks at every single play, and not the final score. Which is why people like dvoa, which is why I like it. I have to admit I haven't seen a whole lot of the Eagles this year. The parts of games I watched they looked very average. But I can't say I watched all of their games, or even a majority of them.
Like someone said it probably has more to do with Andy Reid play-calling, Donavan in the red zone, etc., etc.
Note to other posters - The reason the "philly bias" frustrates people is because they like dvoa as a concept and this site in general. I am in that category. This is about the most objective and in-depth football stats site I've ever found. (I don't think the "philly bias", if it exists, is an actual glitch with dvoa, more of a reflection of the Eagles poor finishing of games, red zone, etc., etc.)
Note to poster above - Been reading FO since 2004, thanks for your input.

Points: 0

#76 by Jeff Fogle // Nov 10, 2009 - 4:53pm

Was wondering why Cincinnati would rank four spots below GB given:

*Cincy has 5.4 estimated wins vs. the 6th ranked schedule

*GB has 5.2 estimated wins vs. the 31st ranked schedule

Cincy won their heads up meeting at GB 31-24, with a 319-311 yardage edge (151-89 on the ground)

Any guidance? What's leapfrogging GB four spots ahead of a team they trail in estimated wins despite a much more favorable schedule?

Points: 0

#77 by S // Nov 10, 2009 - 4:56pm

Don't have premium or game by game DVOA, but my guess is that GB is getting helped by high VOA from clobbering the Browns, Lions and Rams. I think the schedule adjustments aren't yet at full strength either.

Points: 0

#95 by ammek // Nov 10, 2009 - 5:28pm

Read the blurb about Estimated Wins between the boxes. Then work it out for yourself.

Points: 0

#103 by Jeff Fogle // Nov 10, 2009 - 6:00pm

Thanks ammek, was just reading the chart and hadn't noticed estimated wins was pre-adjusted to a league average schedule rather than the schedule they had played. Not sure how to work the rest out for myself yet. The original question would now ask how Green Bay jumps so far ahead with a slightly less estimated win number vs. league average schedules, and a straight up loss head to head.

Points: 0

#106 by ammek // Nov 10, 2009 - 6:13pm

DVOA just flat-out thinks GB is a better team. Head-to-head fixtures aren't given any special consideration — to take the obvious example, the Raiders' win over Philly was obviously a fluke. Also, DVOA doesn't take account of home/road splits, so the fact that the Bengals won at Lambeau is basically irrelevant to the rankings (though not to the fan).

The big difference is on defense. DVOA observes, for example, that Cleveland put up 20 points on the Bengals and took them to a near-tie, while the Pack conceded only 3 in a laugher that, nonetheless, had me sobbing tears of boredom.

As Will Allen & co remark above, common sense says DVOA may be over-rating the Packer D and slightly under-rating Cincinnati. Time will tell.

Points: 0

#122 by Jeff Fogle // Nov 10, 2009 - 7:44pm

Thanks again ammek. Think Cincy fans would have a right to feel skeptical though. It's not like the Cincy win at Green Bay was a fluke that was out of line with other results. Cincy is 2-0 against #6 Baltimore (using DVOA rankings), 1-0 against #8 Pittsburgh, and they lost at the buzzer to #10 Denver. I understand three of those games were tight, not reflecting any sort of Cincy dominance. But, they were playing at the level of teams in the 6-10 range if the games were going down to the wire. Cincy at least "belongs" in that range in the sense of playing very competitive games against teams in that range. Green Bay is 0-2 against #6 Minnesota (margins of 12 and 7 points), then lost to Cincy (7 points). Those are the only three opponents ranked better than #24 Green Bay has played. Green Bay's had three chances to perform vs. that class, and hasn't done so yet.

Cincy is 6-2 vs. the 6th rated schedule
Green Bay is 4-4 vs. the 31st rated schedule

They faced each other. In Green Bay. And Cincinnati had a clean win...319-311 in total yardage, 64% to 46% on third downs, 151-89 in rushing, 21-7 in points scored on drives of 60 yards or more.

There really have to be some hidden kickers in there in my view to justify the Packers at #9 in the league (when they're -8.67 points per game in their only three efforts vs. the top half), while Cincy sits at #13 with a much more tested pedigree in terms of number of minutes vs. quality opposition.

Points: 0

#151 by theshadowj // Nov 10, 2009 - 11:45pm

I think the main difference is that the Bengals have a tendency to play poorly against bad teams. They barely squeaked by the Browns and got beaten soundly by the Texans (although that doesn't look as bad now as it did then). People write off those games as fluky or whatever, but DVOA sees all the games as the same.

Points: 0

#78 by S // Nov 10, 2009 - 4:58pm

Anyone else notice Tennessee's variance? I'm guessing it's because of the stink bomb against NE coupled with their most recent solid if unspectacular wins. What's the record for variance in one season?

Points: 0

#128 by wr (not verified) // Nov 10, 2009 - 8:22pm

I am actually suprised the Cardinals are only #30, given their away/home Jekyll/Hyde act this year.

Points: 0

#89 by JoeD (not verified) // Nov 10, 2009 - 5:17pm

In the preseason, DVOA projected a top 10 rating for Colts special teams. I was highly skeptical about this given their previous performances. This kind of 180 is fairly remarkable, as was the prediction.

Points: 0

#101 by BlueStarDude // Nov 10, 2009 - 5:44pm

Dallas is clearly rated too high because Jerry Jones is a cretin, Tony Romo can't win in December, and Roy Williams is STILL wearing Danny White's jersey number.

Points: 0

#107 by just4fun (not verified) // Nov 10, 2009 - 6:30pm

You'd think they'd have fixed the "philly bias" by now. It's a bit disturbing, that after getting killed by the saints, beat at home by the cowboys, and losing to the lowly raiders, they still have them rated so high. Remember, philly is ranked high everyyear. When one team is show to be better than they are by the statistical analysis, the statistical analysis becomes kind of pointless.............. because it's wrong.

Points: 0

#111 by Danish Denver-Fan // Nov 10, 2009 - 6:59pm

First of all: I don't agree that it is wrong.

Second: Even if there was something magical about the Eagles that always makes their DVOA higher, there wouldn't be much to do about that, until somebody comes up with a somewhat reasonable adjustment. If you are going to put a Philly-only adjustment in the formula, that would just be fitting your results with the thesis. On the other hand, if *you* had a theory about how the weather at the Linc., how their defense works or how Reid is calling the game on offense is affecting their DVOA, I'd be willing to listen - and I'm fairly sure Aaron would too.

Points: 0

#114 by just4fun (not verified) // Nov 10, 2009 - 7:11pm

It was a running joke last year, and the problem with the philly bias is that there are probably many other inaccuracies but no one really cares how the bottom fifteen teams stack up, so it's just assumed to be right. When, in fact, they could be wrong, just not addressed in the form of complaints.

The whole problem is that statistical analysis in football is a lot tricker than many other sports. In baseball, a 300 avg hitter is a 300 hitter on any team. In football a probowl caliber qb depends on the offensive line, the defense/special teams to put him in good field position/game situations. WR to catch to ball and run good routes. Any one of these things can change the entire team's success. The most glaring thing about the philly bias is that it assumes they are on par with new orleans. Philly may be a playoff team but they are no where near the level of the saints or vikings right now. they have offensive line issues, an injury prone rb. and none of those are accounted for.

Points: 0

#118 by Danish Denver-Fan // Nov 10, 2009 - 7:28pm

Yes it is accounted for!! The bad oline play is reflected all over the DVOA rating, and Westbrook is reflected in McCoy and Weaver getting more carries.
Plus DVOA is designed to predict future results - it is, for an example, unlikely that the NO defense will keep producing turnovers at a near-record pace.

If you agree that football is a lot harder to analyse statistically, then you got to take ratings with a grain of salt. Nowhere do FO claim to have Baseball-caliber precision stat-wise, they only claim to have the best football-metrics anywhere.

Points: 0

#113 by MilkmanDanimal // Nov 10, 2009 - 7:07pm

The Eagles have the #1 ranked defense by DVOA. NO killed them, but the most points the team has given up outside of that game was 20. Even in the stinkfest vs. Oakland, they only gave up 13 points. They utterly clubbed the Panthers, Bucs, Chiefs, and Giants. I mean . . . why does this seem to be a big deal? Yeah, they're 5-3. Two losses are by four points. The whole point of DVOA is to look at more than "5-3", right? I mean, if you're just going to talk by record, why come to FO at all?

Doesn't Philly have the second or third most wins this decade? Perhaps the stats say they're good because, well, they're good?

Points: 0

#115 by Danish Denver-Fan // Nov 10, 2009 - 7:21pm

By quick count:
Pats: 102
Colts: 101
Steelers :94
Philly: 92

Including 2000, but not counting 2009.

Points: 0

Save 10%
& Support Aaron
Support Football Outsiders' independent media and Aaron Schatz. Use promo code SCHATZ to save 10% on any FO+ membership and give half the cost of your membership to tip Aaron.