Final 2011 DVOA Ratings

Final 2011 DVOA Ratings
Final 2011 DVOA Ratings
Photo: USA Today Sports Images

by Aaron Schatz

We went through a large part of the year worrying about how our ratings looked when an undefeated Green Bay team wasn't at number one. Turns out that we were worrying for no reason. Green Bay didn't end up undefeated, and they eventually ascended to the top spot in DVOA nonetheless. In fact, the Packers end the year with a reasonable lead, more than five percentage points ahead of No. 2 Pittsburgh. Still, as we've noted all season, the Packers are not a historically great team. They just happen to be the best team this year. The Packers' final DVOA of 28.3% is the second-lowest ever for a team that finished at No. 1, behind only the 1993 San Francisco 49ers (25.6%).

The Packers also don't finish the year as the hottest team in the league. That would be the New Orleans Saints, who moved into the top spot in weighted DVOA this week. New Orleans has single-game DVOA ratings above 35% in five of their last six games, and ratings above 60% in three of those last six. Pittsburgh is second in weighted DVOA, followed by Green Bay and New England. The really hot teams, the teams with a big difference between total DVOA and weighted DVOA, were Seattle and Philadelphia, and they didn't make the playoffs.

One of the interesting things about 2011 is how similar it was to 2010. Although some different teams made the playoffs, and ratings were more condensed than last year, most of the top teams were the same. Eight of the top ten teams in 2010 are once again in the top ten in 2011. That includes a couple of teams that DVOA seems to particularly like, the Jets and Eagles. In addition, 11 of the top 13 teams are the same in both years. The teams that dropped out of the top 13 were San Diego and Tampa Bay. No team went from the top ten in 2010 to the bottom ten in 2011, and only San Francisco (just barely) went from the bottom 10 last year to the top 10 this year.

From 2001-2010, the year-to-year correlation of total DVOA was .526. For 2010-2011, it was .622. This actually isn't the highest year-to-year correlation between two recent years. The correlation between 2008 and 2009 was ridiculous, .744. However, part of the reason the correlation was higher for 2008-2009 is that ratings were higher and lower at the extremes in those seasons (especially at the low extreme). If we look at the correlation between a team's rank 1-32 in one year and its rank 1-32 the next year, the correlation in 2010-2011 is the highest ever at .673. Here's a look at the year-to-year correlation of DVOA ratings (not rankings) since 2000:

Years DVOA
08-09 0.744
01-02 0.634
10-11 0.622
02-03 0.591
06-07 0.525
04-05 0.522
05-06 0.501
07-08 0.497
00-01 0.485
09-10 0.418
03-04 0.405

This year's higher level of parity also played out in the individual games each week, as we didn't have a lot of spectacular individual game performances this year. This year there were only three games with DVOA over 100%. The highest was Houston at 119.1% when they dismantled Tennessee 41-7 in Week 7. Baltimore got 113.0% for its turnoverfest over Pittsburgh in Week 1. And the Eagles get 100.4% for their 45-19 win over the Jets in Week 15. There was only one game under -100% this year, as Washington gets -107.7% for its 23-0 loss to Buffalo in Week 8. By comparison, 2010 had seven games over 100% and nine games under 100%.

I know that a lot of people will look at this year's tables and think something is wrong with opponent adjustments. All the top teams played easy schedules this year, and the top six teams all have schedules in the bottom seven. However, this is a one-year fluke. Look back at 2010 and you'll notice that last year four of the top six teams actually played top ten schedules.

Like last year, I'm going to go through and show where the teams and players of 2011 fit historically -- at least as far as the regular season goes. Note that all these historical rankings come with a bit of an asterisk. One of the first projects I'm planning for the offseason is an update of DVOA that will normalize the ratings so every season averages out at 0%. This will adjust for the big uptick in offensive numbers in recent years, which is likely to move more seasons from 1992-2003 into the various top ten rankings. We may run updated "best ever" lists this summer once we've done the new normalized DVOA and added 1991 to our database. (That's another one of my February projects.)


Although the Packers aren't a historically great team, they are a historically great offense. So are the Saints and the Patriots. Here's where they finish up on the all-time DVOA lists:

NE 2010 46.1% x NE 2007 75.4%
NE 2007 45.2% x GB 2011 73.2%
GB 2011 39.2% x NE 2010 72.5%
KC 2002 38.0% x IND 2004 69.1%
NO 2011 37.8% x SD 2009 63.7%
NE 2011 36.8% x NE 2011 60.7%
IND 2004 33.2% x NE 2009 57.0%
KC 2004 32.9% x IND 2006 56.9%
KC 2003 32.2% x NO 2011 55.8%
DEN 1998 30.8% x SD 2008 54.9%

The Carolina Panthers set a new record for the largest year-to-year improvement in offensive DVOA, and they set it by leaps and bounds. The Panthers were dead last in 2010 at -31.9%. This year, they rank fourth at 23.0%. That's an increase of 54.9%. There is only one other team that has ever improved its offense by 40% DVOA in one season: the 1998-1999 Raiders, who went from -28.1% DVOA to 15.2% DVOA.

The Panthers also set a new record with the best running game in DVOA history, thanks to adding Cam Newton to their running back tandem of DeAngelo Williams and Jonathan Stewart. The Panthers led the league with 36.0% rushing DVOA. The previous record was 34.4% by the 2000 St. Louis Rams, followed by the 2002 Kansas City Chiefs at 32.0%.

Defenses were neither historically great nor historically terrible this year. Baltimore leads the league at -13.7%, only the second team to lead the league with defensive DVOA above 18%. (The other: the 2007 Titans at -13.3%.) Carolina edges out Tampa Bay and New England to rank as this year's worst defense, but it isn't close to the all-time worst. At one point it looked like Jacksonville and Houston were challenging for the title of "best year-to-year defensive improvement," but both teams declined a little bit on defense in the final weeks.

Chicago ends up with this year's best special teams at 7.3%. Finally, Atlanta ends up with a variance of 4.6%, which makes the Falcons the most consistent week-to-week team in DVOA history. Amazingly, the second most consistent week-to-week team in DVOA history was last year's Falcons at 4.9%.


We'll be running a season-review version of Quick Reads on Thursday, so for now let's just go through players who set records rather than listing all the best and worst players at every position.

Just like the pass offense ratings for teams were historically great, so too were the passing DVOA and DYAR numbers for the top three quarterbacks. The difference in FO stats is the same difference that's going to be discussed when it comes to the MVP race: Which is more important for a quarterback, quality or quantity? Not that Drew Brees' season wasn't filled with quality, but he threw far more passes than Aaron Rodgers. Brees finishes fifth all-time in pass attempts (he has four of the top nine seasons) and his DVOA and DYAR are based on 138 more pass plays than Rodgers' ratings. So Brees ends up with 2,544 passing DYAR, the second highest total we've ever measured, but he finishes second to Rodgers in DVOA. Here's how the three quarterbacks end up on the all-time lists. I've limited the DVOA list here to quarterbacks with at least 400 passes so we're not listing guys who had a smaller sample but still enough passes to make the FO yearly rankings, such as Boomer Esiason (53.6% DVOA on 197 passes in 1997) and Wade Wilson (55.3% DVOA on 169 passes in 1992).

Top 11 Passing DVOA, 1992-2011
(min. 400 passes)
  Top 10 Passing DYAR, 1992-2011
Year Player Team DVOA   Year Player Team DYAR
2004 P.Manning IND 60.6% x 2007 T.Brady NE 2,788
2007 T.Brady NE 56.9% x 2011 D.Brees NO 2,544
2010 T.Brady NE 53.3% x 2004 P.Manning IND 2,493
2011 A.Rodgers GB 52.6% x 2006 P.Manning IND 2,308
2006 P.Manning IND 51.0% x 2011 A.Rodgers GB 2,268
2009 P.Rivers SD 45.9% x 2011 T.Brady NE 2,235
2011 D.Brees NO 44.5% x 2009 T.Brady NE 2,170
2009 T.Brady NE 44.2% x 2010 T.Brady NE 2,137
1998 R.Cunningham MIN 42.9% x 2009 P.Manning IND 1,936
2011 T.Brady NE 41.0% x 2004 D.Culpepper MIN 1,929
2009 D.Brees NO 41.0%    

Blaine Gabbert, with -825 passing DYAR, finishes with the fifth worst season we've ever measured in that stat.

Pierre Thomas led the league with 32.6% rushing DVOA, which is the third highest rating ever for a running back with at least 100 carries, behind Marshall Faulk (35.0% in 2000) and Jamaal Charles (33.9% in 2010). LeSean McCoy led the league in rushing DYAR.

Calvin Johnson had a huge 122 DYAR week in Week 17, which rocketed him past Wes Welker and into this year's top spot for wide receiver DYAR. In fact, Johnson's huge final week puts him into historic company, because it also rocketed him past Jerry Rice, Marvin Harrison, and Randy Moss. Johnson's 586 receiving DYAR are the second highest total we've ever measured, behind only Michael Irvin's 636 DYAR in 1995. Jordy Nelson finishes second for the season with 530 DYAR and ends up with 54.1% receiving DVOA. That's not technically the record, but it is a record for players with more than 60 pass targets. Our rankings for wide receivers start at 50 pass targets, and there was one wide receiver who had a higher DVOA on exactly 50 targets: Dennis Northcutt with 61.0% DVOA in 2002. However, no wide receiver had ever put up 50% receiving DVOA on more than 60 pass targets. In fact, this year we get three of the five highest wide receiver DVOAs with a minimum of 60 passes, as Malcom Floyd and Laurent Robinson also had great seasons with smaller sample sizes. (Floyd was injured for a few weeks; Robinson was the third option in Dallas.) Here is where everyone ends up in the record books:

Top 10 WR DYAR, 1992-2011   Top 10 WR DVOA, 1992-2011
(min. 60 passes)
Year Player Team DYAR   Year Player Team DVOA Passes
1995 M.Irvin DAL 636 x 2011 J.Nelson GB 54.1% 96
2011 C.Johnson DET 586 x 2011 M.Floyd SD 51.9% 70
2007 R.Moss NE 569 x 1993 J.Taylor SF 49.7% 74
1995 J.Rice SF 550 x 2010 M.Wallace PIT 48.8% 98
1994 J.Rice SF 545 x 2011 L.Robinson DAL 43.1% 80
2001 M.Harrison IND 537 x 2002 J.Porter OAK 41.5% 70
2011 J.Nelson GB 530 x 2009 R.Meachem NO 39.2% 64
2006 M.Harrison IND 510 x 1998 E.Moulds BUF 38.9% 116
2005 S.Smith CAR 497 x 2009 V.Jackson SD 38.7% 109
2004 R.Wayne IND 496 x 2004 R.Wayne IND 38.0% 115

Finally, tight ends. Rob Gronkowski absolutely destroyed the previous record for receiving DYAR by a tight end with 505 DYAR. The previous record was 371 DYAR by Antonio Gates last year, when he played only ten games. The difference between Gronkowski's total and the previous record is equal to the difference between the previous record and the season now ranked 21st all-time, which is Tony Gonzalez's 237 DYAR this season. Between Gronkowski and Gonzalez are two other tight ends who make the all-time top ten in receiving DYAR. Jimmy Graham finishes with 331 DYAR, which is fourth all-time. Antonio Gates finishes with 287 receiving DYAR, which is ninth all-time. Gates has four of the top nine tight end seasons, and five of the top 11.

Gronk doesn't get onto the all-time tight end DVOA top ten because the minimum number of passes to be ranked as a tight end is 25, and there are a number of tight ends who had 25-40 passes and DVOA ratings in the 50s and 60s (including Gronkowski himself last year). However, no tight end with at least 80 passes has ever come close to Gronk's 52.0% DVOA. The next-highest would be Gates with 38.4% DVOA in 2009.

* * * * *

All team and individual stats pages should be updated in the next few minutes after this article posts. FO Premium will be updated with final 2011 ratings later today, and there are already matchup views available so you can check out how the wild card games match up. We'll get all the 2011 stats onto the player pages sometime in the next few weeks, as well as updated similarity scores based on 2009-2011 rather than 2008-2010. Loser League results will be announced in Scramble for the Ball on Wednesday, and look for a brand new playoff fantasy game on FO coming this evening.

* * * * *

These are the Football Outsiders team efficiency ratings through the end of the 2011 regular season, 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.)

Please note that ratings may change in the future as we get a chance to incorporate stat changes from throughout the season.

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. LAST WEEK represents rank after Week 16, while LAST YEAR represents rank in 2010.

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 GB 28.3% 1 3 15-1 26.1% 3 39.2% 1 12.6% 24 1.6% 8
2 PIT 23.7% 2 2 12-4 29.9% 2 16.3% 6 -5.8% 7 1.6% 9
3 NO 23.3% 5 10 13-3 30.1% 1 37.8% 2 15.0% 28 0.6% 13
4 NE 22.5% 3 1 13-3 23.7% 4 36.8% 3 17.9% 30 3.7% 5
5 HOU 19.4% 4 13 10-6 17.2% 7 13.3% 9 -5.2% 8 0.9% 12
6 SF 19.0% 6 23 13-3 16.1% 8 1.6% 18 -10.3% 3 7.0% 2
7 BAL 17.2% 7 4 12-4 14.8% 9 8.2% 13 -13.7% 1 -4.7% 30
8 ATL 15.5% 9 8 10-6 19.5% 6 10.3% 12 -6.2% 6 -1.0% 22
9 NYJ 14.2% 8 6 8-8 9.9% 11 -2.7% 21 -11.9% 2 5.0% 4
10 PHI 14.1% 10 5 8-8 22.9% 5 14.3% 8 0.3% 12 0.1% 18
11 DET 11.6% 11 18 10-6 14.5% 10 11.8% 10 -4.0% 9 -4.3% 29
12 NYG 9.0% 13 9 9-7 4.8% 16 15.6% 7 6.9% 20 0.3% 16
13 TEN 7.2% 12 11 9-7 5.0% 15 6.0% 15 4.2% 15 5.5% 3
14 DAL 4.4% 14 24 8-8 3.9% 17 10.8% 11 4.8% 16 -1.6% 25
15 CHI 1.8% 15 16 8-8 -0.1% 20 -15.7% 30 -10.2% 4 7.3% 1
16 SD 0.6% 20 7 8-8 6.3% 13 17.7% 5 15.9% 29 -1.2% 23
17 CIN 0.5% 18 19 9-7 0.3% 18 3.9% 17 5.4% 17 2.0% 7
18 MIA -0.4% 19 14 6-10 5.6% 14 -2.2% 20 0.3% 11 2.2% 6
19 SEA -0.7% 16 30 7-9 8.7% 12 -4.1% 22 -3.1% 10 0.2% 17
20 CAR -2.4% 17 31 6-10 0.1% 19 23.0% 4 20.1% 32 -5.4% 32
21 WAS -5.9% 21 27 5-11 -6.6% 21 -2.1% 19 2.8% 14 -1.0% 21
22 OAK -6.8% 22 21 8-8 -10.8% 25 7.6% 14 13.7% 26 -0.7% 20
23 BUF -8.2% 23 28 6-10 -25.3% 28 5.5% 16 12.3% 23 -1.3% 24
24 DEN -11.0% 24 26 8-8 -8.8% 23 -5.1% 23 6.4% 19 0.5% 15
25 CLE -12.9% 25 20 4-12 -10.0% 24 -6.3% 25 8.0% 21 1.4% 10
26 KC -15.7% 26 17 7-9 -8.1% 22 -14.2% 29 0.9% 13 -0.7% 19
27 JAC -16.6% 27 22 5-11 -13.5% 26 -21.4% 31 -6.6% 5 -1.8% 26
28 ARI -18.9% 28 32 8-8 -15.4% 27 -13.7% 28 6.4% 18 1.1% 11
29 MIN -21.2% 30 25 3-13 -33.0% 30 -5.2% 24 12.7% 25 -3.3% 27
30 TB -25.0% 29 12 4-12 -38.4% 32 -6.7% 26 18.8% 31 0.6% 14
31 IND -32.0% 31 15 2-14 -30.9% 29 -12.5% 27 14.3% 27 -5.2% 31
32 STL -35.5% 32 29 2-14 -34.1% 31 -22.8% 32 8.7% 22 -4.0% 28
  • 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.
  • WEIGHTED DVOA is adjusted so that earlier games in the season become gradually less important. It better reflects how the team was playing at the end of the season.
  • 2011 SCHEDULE lists average DVOA of opponents played this season, ranked from hardest schedule (#1, most positive) to easiest schedule (#32, most negative).
  • PYTHAGOREAN WINS represent a projection of the team's expected wins based solely on points scored and allowed. Please note that for 2011, Pythagorean wins uses the new "Pythagenport" method where the projection takes into account the "offensive environment" that each team played in, rather than just using the same exponent to project for each team.
  • 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).
RANK 2011
1 GB 28.3% 15-1 28.3% 12.6 1 -2.7% 28 12.2 3 6.7% 4
2 PIT 23.7% 12-4 24.2% 11.4 4 -3.0% 29 11.1 6 13.9% 16
3 NO 23.3% 13-3 24.7% 11.6 3 -4.2% 31 12.4 1 14.2% 18
4 NE 22.5% 13-3 24.6% 11.8 2 -1.4% 26 11.9 4 9.5% 7
5 HOU 19.4% 10-6 25.1% 10.2 8 -3.6% 30 10.9 7 15.7% 20
6 SF 19.0% 13-3 22.6% 10.3 7 -4.2% 32 12.3 2 6.2% 2
7 BAL 17.2% 12-4 16.9% 10.9 5 -0.1% 21 11.2 5 23.3% 31
8 ATL 15.5% 10-6 14.1% 10.5 6 0.2% 17 9.4 10 4.6% 1
9 NYJ 14.2% 8-8 12.7% 8.8 12 1.1% 13 8.4 15 19.8% 27
10 PHI 14.1% 8-8 13.7% 9.2 10 1.3% 11 9.8 9 17.4% 24
11 DET 11.6% 10-6 14.1% 9.4 9 0.3% 16 10.1 8 9.5% 6
12 NYG 9.0% 9-7 4.8% 9.1 11 4.3% 3 7.8 19 15.8% 21
13 TEN 7.2% 9-7 11.5% 8.6 15 -2.4% 27 8.2 17 16.6% 22
14 DAL 4.4% 8-8 6.6% 8.7 14 0.1% 19 8.6 13 12.7% 13
15 CHI 1.8% 8-8 2.1% 7.9 18 1.8% 9 8.3 16 14.9% 19
16 SD 0.6% 8-8 4.5% 7.6 19 -0.7% 23 8.7 11 20.8% 28
RANK 2011
17 CIN 0.5% 9-7 4.5% 8.7 13 0.7% 14 8.6 12 7.0% 5
18 MIA -0.4% 6-10 -2.8% 8.2 16 3.3% 6 8.5 14 10.4% 11
19 SEA -0.7% 7-9 0.1% 8.2 17 -0.8% 24 8.2 18 16.9% 23
20 CAR -2.4% 6-10 -3.1% 7.6 20 -0.9% 25 7.4 20 19.5% 25
21 WAS -5.9% 5-11 -13.4% 7.1 24 0.3% 15 5.7 25 11.3% 12
22 OAK -6.8% 8-8 -3.2% 7.5 21 0.2% 18 6.1 23 19.7% 26
23 BUF -8.2% 6-10 -8.2% 7.2 23 3.8% 4 6.4 22 29.3% 32
24 DEN -11.0% 8-8 -9.1% 7.3 22 1.2% 12 5.8 24 9.7% 8
25 CLE -12.9% 4-12 -9.1% 6.8 25 -0.1% 20 5.0 28 6.3% 3
26 KC -15.7% 7-9 -17.3% 6.5 26 1.7% 10 4.1 29 21.5% 29
27 JAC -16.6% 5-11 -14.6% 5.9 27 2.8% 7 5.3 26 13.7% 15
28 ARI -18.9% 8-8 -14.6% 5.9 28 -0.5% 22 6.9 21 9.9% 9
29 MIN -21.2% 3-13 -21.5% 5.3 30 3.3% 5 5.3 27 14.0% 17
30 TB -25.0% 4-12 -31.5% 5.4 29 7.9% 1 3.2 30 22.8% 30
31 IND -32.0% 2-14 -35.5% 3.9 31 2.7% 8 3.2 31 13.4% 14
32 STL -35.5% 2-14 -42.3% 3.7 32 7.5% 2 2.3 32 9.9% 10

(Note: Although this post is titled "Final DVOA Ratings," unofficial postseason ratings will continue each Monday through the playoffs.)


104 comments, Last at 08 Jan 2012, 6:01pm

#1 by bigtencrazy (not verified) // Jan 02, 2012 - 3:31pm

Thanks for all the hard work.

Points: 0

#2 by Purds // Jan 02, 2012 - 3:45pm

Oh, the irony. Don't want to start a P. Manning/Brady discussion, but as a Colts fan, I was always of the impression (from 2001-2005/6) that Brady wasn't even close to Manning in talent level when the debate was hot, from 2002-2005 or so. Yet, Brady had won championships, and everyone said he was the reason for NE's success (or used that tactic in saying he was one of the greats). Now, if you look at Brady's great DVOA and DYAR seasons of the past 4/5 years, you can tell Brady is clearly at Manning's, and yet he hasn't won any championships.

My point? Brady is now a MUCH better QB than he was in 2001-2004, and I can now accept him as Manning's equal.

(As a side note, it's pretty clear the the 2009-11 Patriots have become the 2003-5 Colts, from the passing offense to the terrible but opportunistic defense, to the playoff disappointments. We'll have to see if NE can break out this post season.)

Points: 0

#3 by shawnqa800720 (not verified) // Jan 02, 2012 - 4:03pm

Thanks for admitting you respect Mr. Brady. It's fans like you that make it easier for me, as a NE fan, to admit how much I respect Peyton. In my opinion, both suffer from their greatness - too much emphasis on one individual in a team sport spells disaster. I think you are dead on in your analysis that the 2009-11 Patriots have become the 2003-5 Colts.

For my money, there was no better thrower in the NFL than Dan Marino, but both Peyton and Tom blow him away as complete quarterbacks.

Go Pats, and I truly hope Peyton comes back so we can have so more historic duels! Even the crushing losses were historically memorable - from both sides.

Points: 0

#59 by FrontRunningPhinsFan // Jan 03, 2012 - 10:33am

After reading that this summer DVOA will be normalized to 0% for each year, I'll be excited (even though its years away I guess) to see how Marino's '84 season grades out.

The guy was just un-freaking-stoppable.

I'm sad to see his yardage record broken this year and was when his TD record was too. I'll always argue its the rule changes!

(please draft a 1st rd QB this year...)

Points: 0

#66 by Vincent Verhei // Jan 03, 2012 - 1:22pm

Um, do not hold your breath. Aaron hopes to add 1991 to the database this offseason. At a rate of one old season added every offseason, that means we'll get to 1984 about 2019.

In the meantime, by one era-adjusted metric, Marino 84 is the second-best season of all time behind Manning 04. Rodgers 11, by the way, is third.

Points: 0

#99 by archibaldcrane (not verified) // Jan 05, 2012 - 9:34pm

What's funny is that Brees' 2011 is...40th on that list - below his 2009.

Points: 0

#10 by RickD // Jan 02, 2012 - 5:07pm

That sounds about right. The media were too quick to place Brady at Peyton's level, but he's there now, and has been since 2007.

Points: 0

#91 by Mr Shush // Jan 05, 2012 - 8:18am

I think what changed in 2007 was the supporting cast, not Brady. He's been an elite player since at least 2005, maybe even 2004. I think 2006 may have been his most impressive season, given what he had to work with. Similarly, Manning's best play wasn't in 2004, great as he was then, with in-their-primes Harrison, Wayne, Clark, Saturday and Glenn. It was the more recent seasons, in which he managed to produce an outstanding offense with declining receiving talent and a shambles of an offensive line (seemingly now somewhat fixed by a solid but too-late 2011 offseason).

Points: 0

#94 by nat // Jan 05, 2012 - 10:44am

Supporting cast matters a lot. In a different thread, I saw a Colts fan lamenting how Garcon and Collie can't beat double coverage. I then realized that during Manning's career he always had at least one receiver who could routinely challenge and beat double coverage, and often had two or three.

Brady didn't have the luxury of even one until 2007. Even today, you don't think of Welker as the "beat the double" kind of receiver. He's more of a "destroy the zone" guy. Gronkowski might now be considered a "beat the double" guy, but not in the usual sense. Branch is nice and all, but he's no threat if/when double covered and never has been.

Points: 0

#17 by commissionerleaf // Jan 02, 2012 - 5:44pm

Unfortuately for the Patriots, they are more like the oh seven, oh eight Colts than the oh four version. Prolific passing offense coupled with terrible defense and no running game to speak of, held together by an aging but world class quarterback. If the offensive line doesn't come back from injury the transformation will be complete.

Points: 0

#22 by MJK // Jan 02, 2012 - 5:58pm

I don't agree with this statement. First, running game. The Patriots had the 4th ranked rushing offense this year (+17%) and the second ranked rushing offense last year (+27%). In '07-08, Indy had the 5th ranked rushing offense (~+11%) and the 27th ranked rushing offense (-6.5%). So I wouldn't say they have no running game to speak of. It's true that they don't have a great FANTASY running back, which many people seem to equate with not having a running game, but that's because they platoon it out among three or four guys.

QB age is roughly analogous, but at this point Brady has played one full season less as a starter than Manning had at the end of the '08 season. (2001 - 2011, excluding 2008 for Brady, 1998-2008 for Manning).

The teams are similar in defense, but not as having "terrible" defense. Both teams had terrible defense measured by yards, and fairly respectable defenses measured by points allowed (the 2007 Colts actually had the best scoring defense in the league by points allowed). This isn't surprising, because the Patriots this year were actually running predominantly a 4-3 with a cover-2 zone behind them...exactly the defense the Colts ran through their heyday.

The final difference is that the Colts started to fall apart when their offensive line did. The Pats offensive line fell about back about Week 4 (and similarly it fell apart last year, and the year before) due to injury, and they still have weathered the storm... The Pats have got to have one of the best O-line coaches in the league...

Points: 0

#23 by dmstorm22 // Jan 02, 2012 - 5:59pm

Both the '07 and '08 Colts had good defenses (#2 overall DVOA in 2007 and #10 in 2008). The 2004 Colts team was the all-offense, no-defense version.

Points: 0

#25 by Stats are for losers (not verified) // Jan 02, 2012 - 6:20pm

Weird how the '07 and '10 Patriots have a better Passing Offense (by DVOA) than the '04 Colts, while '04 Manning has a better QB DVOA than '07 or '10 Brady.

Points: 0

#30 by PackersRS (not verified) // Jan 02, 2012 - 7:24pm

It's not strange at all, when you consider Darren Sproles', Jimmy Graham's, Wes Welker's and Rob Gronkowski's DVOAs.

Points: 0

#34 by Aaron Schatz // Jan 02, 2012 - 7:58pm

Two reasons:

1) QB passing DVOA only compares passes to other passes. Team passing DVOA is based on comparing all plays to all plays, including penalties and runs. So a team might have a higher passing DVOA based on how often they pass compared to how often an average team passes.

2) Jim Sorgi played most of Week 17 in 2004.

Points: 0

#4 by TomC // Jan 02, 2012 - 4:06pm

If Jay Cutler had been injured one week earlier (or earlier in the SD/CHI game), or if the Bears' schedule against the AFC West had been flipped, the Chargers would be in the playoffs, and the Steelers' first-round game would be *very* different. Instead, the Chargers got the peaking, Cutler-led Bears and lost, but not before turning them into the pathetic, Hanie-led Bears so that their AFCW rivals could all get an unexpected win. Not sure how to blame that one on Norv, but I'm guessing someone can figure out a way.

Points: 0

#5 by bigtencrazy (not verified) // Jan 02, 2012 - 4:09pm


If Cutler doesn't get hurt I am convinced it would be a rematch of last year's NFC championship game.

Points: 0

#26 by Duke // Jan 02, 2012 - 6:22pm

If Cutler AND Forte AND Urlacher (for the playoffs) had not gotten hurt, I would agree. But Cutler alone is probably not enough to offset the loss of Forte and Urlacher.

Points: 0

#62 by Steve in WI // Jan 03, 2012 - 11:54am

True. Of course, who knows if Forte and Urlacher would have gotten hurt if Cutler had been healthy? (I would like to think that Urlacher at least would not have been playing in the 4th quarter of the Vikings game if the Bears had clinched a playoff spot, which I'm assuming they would have had Cutler stayed healthy). And I think RB is probably the single position the Bears have the most depth at, so I wouldn't say it would have been impossible for Bell to have a good game in the playoffs.

One of the points on Chicago talk radio this morning is that if Cutler hadn't gotten hurt, Jerry Angelo almost certainly would not have been fired - I don't see how this team wouldn't have won 11 games and made the playoffs, and even assuming another loss (in the NFC championship game or not sooner) his job would have been safe. I'm certainly not happy Cutler was hurt, but long-term it may have been a good thing for the team, assuming he comes back at 100% next year.

Points: 0

#29 by TomC // Jan 02, 2012 - 7:20pm

Either of you could be right, but I was just pointing out how funny it is that the Cutler injury (and, to a lesser extent the Forte injury) affected the AFC playoff picture almost as much as it did the NFC playoff picture.

Points: 0

#56 by Eggwasp (not verified) // Jan 03, 2012 - 2:37am

Oh if we are playing pick the injury, then I'd keep Campbell, McFadden, J Ford, M Huff and D Moore fit so who knows what the Raiders do. Maybe even tackle someone to force a SD punt!

Points: 0

#6 by nath // Jan 02, 2012 - 4:22pm

I'm baffled how PIT finishes 2nd in weighted DVOA when their last six games look like this:

@ KC W 13-9
CIN W 35-7
CLE W 14-3
@ SF L 3-20
STL W 27-0
@ CLE W 13-9

The CIN win is the only one that seems legitimately strong. Playing a couple of mediocre-to-bad teams close and beating up another really bad team doesn't say "#2" to me.

Points: 0

#9 by Purds // Jan 02, 2012 - 5:01pm

Did any of the top 4 have an amazing finish stretch of schedule?

GB: Det, NYG, Oak, KC, Chi, Det
NO: NYG, Det, Ten, Min, Atl, Car
NE: Phi, Indy, Wash, Den, Mia, Buf

All of the top teams played some pretty bad teams down the stretch, or at least mediocre teams.

Points: 0

#14 by Eddo // Jan 02, 2012 - 5:13pm

I think the point was that the Steelers didn't look particularly good against those teams. At least not "#2 in weighted DVOA" good.

I didn't see enough of those games to say whether that's right or wrong, though.

Points: 0

#84 by Tehnico (not verified) // Jan 04, 2012 - 11:49am

Looks like NO and GB both had half their games against playoff teams. NE played one playoff team.

Points: 0

#11 by RickD // Jan 02, 2012 - 5:08pm

They had two blowout wins, one of which was against a playoff team. DVOA tends to reward blowouts, esp. against decent teams.

Points: 0

#98 by Scott C // Jan 05, 2012 - 7:42pm

Unless you are San Diego destroying Baltimore. Then it does nothing. Apparently rightly so as Detroit showed a week later.

Points: 0

#13 by drobviousso // Jan 02, 2012 - 5:12pm

It's a case of the scores not representing the play on the field. KC and the first CLE game were not as close as the score suggests. The SF game felt demoralizing, but the Steelers where in it a lot longer than the final score indicates.

As a fan, I can't put my finger on why, but the KC-CIN-CLE felt like the best football they played all year.

Points: 0

#15 by 'nonymous (not verified) // Jan 02, 2012 - 5:13pm

They are a 13-3 team, and they went 5-1 over those games, giving up an average of 8 points; that's pretty good, regardless of the opponents.

But there are a lot more statistics than points. My impression is that the Steelers dominated statistically (in yards, first downs, TOP) much more than the scoreboard indicates; turnovers are a big reason-- they've won a lot of games this year while losing the turnover battle. (For example, when I last checked yesterday's PIT-CLE stats, right before CLE's final drive, the Steelers had double the yards and TOP of the Browns, even though the score was 13-9.) I see no reason to doubt the ranking; but they have looked a bit weaker to me the last 3.5 games, with Ben injured.

Points: 0

#47 by RickD // Jan 02, 2012 - 10:03pm

Well, they're a 12-4 team at least.

Points: 0

#65 by troycapitated … // Jan 03, 2012 - 12:59pm

Well, I would probably have to pick the New England game as the best one of the season for the Steelers.

The games after the injury they have certainly looked weaker because the offense has failed to put much on the scoreboard, despite moving the ball fairly well in all the games. In their 4 losses and 6 close wins, the Steelers had 25 turnovers- Ben was responsible for 17 of those- while only coming up with 11 for a -14 differential in those 10 games.

Both Cleveland games might have turned into comfortable victories but for the 5 Steeler turnovers (1 from Ben) against 3 Browns turnovers.

Points: 0

#24 by cisforcookie (not verified) // Jan 02, 2012 - 6:18pm

I think this might be a case of something like the anti-marshawn lynch effect. The steelers, when I have watched them, haven't _looked_ dominating. They _look_ like they are limping along, old, slow, injured half the time, no offensive line to speak of. They look very vulnerable, and that's a great story to tell about them. Yet they're 3-4 against playoff teams, with 4 of those games (in which the steelers went 2-2) being decided by 1 score.

It doesn't help their image that the two biggest games of their season, opening night and that monday night game against san fran, were also their biggest embarrassments of the season. The rest of the country doesn't pay much attention when the steelers beat up on weaker opposition. Yet the two highest scoring games by opponents against the steelers this year were the 35 and the 23 put up by Baltimore. By comparison, the ravens have allowed more than 20 points 5 times, including 3 of their 4 losses. (the other being that turd sandwich in jacksonville) The pats have allowed over 20 11 times. This steeler defense is a shadow of what it once was, but that shadow still has teeth.

Points: 0

#35 by Aaron Schatz // Jan 02, 2012 - 7:59pm

Weighted DVOA looks over a longer period of time than most people think of as "recent." The last eight weeks are at mostly full strength. It's nine weeks and above where games start to lose strength.

Points: 0

#7 by Hot_Biscuit_Slim // Jan 02, 2012 - 4:25pm

The 2011 FO Almanac hit on 8 of the 12 playoff teams (based on estimated wins):

*New England
*New Orleans
*Green Bay
*San Francisco

It overestimated four teams that didn't end up making the playoffs:

*NY Jets
*San Diego

And it missed four sleepers:

*NY Giants

Eight out of 12 is not as good as 2010's 10-for-12, but still pretty damn good.

Points: 0

#12 by Eddo // Jan 02, 2012 - 5:11pm

The four overestimates aren't terrible, either. The Jets collapsed at the end, but still had an OK year. The Chargers are pretty widely viewed to be the best AFC West team as of right now. Ditto the Eagles and the NFC East. And the Bears were playing great football, sitting at 7-3, when Cutler went out for the year and the Caleb Hanie Slapdash Variety Show started.

Points: 0

#57 by Eggwasp (not verified) // Jan 03, 2012 - 3:08am

The Chargers are the only AFCW team that didnt replace its starting QB... Aside from replacing that Rivers-imposter that had kidnapped the real one in the first half of the season.

Points: 0

#63 by Steve in WI // Jan 03, 2012 - 11:56am

With a healthy Jay Cutler, I would say it's almost a certainty that the Bears would have made the playoffs. The Kansas City and Denver games would have been wins, and I would argue that they would have won the Oakland game too. 10 wins would have gotten them in, and I just don't see how the Bears don't win 2 from the 5-game losing streak if they'd had Cutler.

Points: 0

#70 by Thomas_beardown // Jan 03, 2012 - 2:38pm

It's also likely they win the Seahawks game. The only game I would have favored their opponent in would be the Packers game.

Points: 0

#88 by Steve in WI // Jan 04, 2012 - 3:36pm

I didn't see that one (mercifully) and I know Hanie threw a couple pick-6s, but wasn't the defense a little shaky overall that day?

Regarding GB, who knows what the score would've been with Cutler and if GB had had to put up more points, but I don't think it would have been impossible for the Bears to score more than 35. I agree that I would have favored the Packers regardless.

Points: 0

#89 by Thomas_beardown // Jan 04, 2012 - 3:54pm

The defense let a few big plays happen in the 2nd half, but I doubt Cutler would have had trouble responding.

Points: 0

#16 by Dales // Jan 02, 2012 - 5:41pm

DVOA overestimating Philadelphia is a given.

Points: 0

#28 by Pat (filler) (not verified) // Jan 02, 2012 - 6:39pm

Pythagorean wins overestimated the Eagles, then, too. The Eagles had the highest point differential in the NFC East, by a large margin (+68 vs. the Cowboys +22) - heck, the Eagles scored more and allowed fewer points than any other team in the NFC East.

I think it's pretty clear if the season were 18 games long, the Eagles would've been the favorite to win the division. The only reason they underachieved is because the season's too short.

Points: 0

#33 by Kurt // Jan 02, 2012 - 7:51pm

Yes, it's a shame for the Eagles the season starts in September and not December. If my life ever depends on someone winning a meaningless game, or scoring another touchdown when it's 27-10, I know who I'm betting on.

Points: 0

#40 by Pat (filler) (not verified) // Jan 02, 2012 - 8:45pm

I think it's more the fact that the preseason was short. The offense was pretty consistent all year - they ended with essentially the same weighted DVOA as DVOA. The defense, however, went from awful to very good - basically average over the year, but the last games, one of the top defenses in the league.

I don't think it takes a huge leap of faith to think that the short offseason hurt the Eagles defense a ton - moreso than the top teams, who experienced very little turnover.

winning a meaningless game

Meaningless game? The only fully meaningless game the Eagles played this year was the last one. For the weeks before that, they were all must win games, and the Eagles won every one (and probably would've still beaten Dallas had the Jets won).

Points: 0

#61 by chemical burn // Jan 03, 2012 - 10:52am

Holy cow, Pat - we agree! The lockout hurt the Eagles just about more than any team in the league. The defensive improvement can't be overstated - they were solidly 32nd in DVOA after several games, just the worst by a significant margin. That they finished 12th is a pretty spectacular turn-around. Also, that dude is going to be sitting on his money for the bet on a meaningless game - the last time the Eagles played in a game with zero playoff implications before this last one was in 2005. (that is, a game where they weren't going to the playoffs and they team they were playing had no chance of going to the playoffs.)

Anyhoo, I have no idea how I want the off-season to proceed other than for Andy Reid and Howard Mudd to come back... I might even be able to convince myself that Juan Castillo has earned a second shot. (no... I just can't do it.)

Pat - how do you feel about the situation at LB? Do you think Rolle, Chaney, Matthews is viable? Or that having Jordan do so much spot duty means Matthews/Rolle aren't really viable options? Also, Coleman/Allen is working out pretty good at safety, right?

But finally, I think that this team played its heart out on Sunday is one of the best things that can be said in favor of Andy Reid - this team played their hardest even when it didn't matter and all they were left with this season was a giant pile of disappointment and tons of derision from the fans and media. Look at the Bucs for comparison...

Points: 0

#72 by Pat (filler) (not verified) // Jan 03, 2012 - 4:00pm

Holy cow, Pat - we agree!

And if you would've just seen reason before the season, your expectations would've been right in line with where they ended up! Granted, as you might've guessed if you had asked before the season, I wished I had been wrong, but my instincts were just flat out screaming that this was going to be a complete and total disaster until they managed to find a few linebackers/safeties that don't suck.

Seriously, it was really depressing losing all hope two weeks into the season seeing all my fears come true. I really, really wish you had been right.

Anyhoo, I have no idea how I want the off-season to proceed other than for Andy Reid and Howard Mudd to come back... I might even be able to convince myself that Juan Castillo has earned a second shot. (no... I just can't do it.)

Really? I'm fine with Juan sticking around another year. Replacing Juan with Random Guy (that includes Spags) is most likely to produce an average defense. Juan had them playing at top-10 level defense by the end of the year. I'll take that gamble.

Pat - how do you feel about the situation at LB?

How do I always feel about the situation at LB? Do I like Rolle/Chaney/Matthews? Um. No. They might deserve a shot next year, but guess what? That's yet another year (every... single... one under Reid) that they've had a new set of starting linebackers. In other words - I've felt like this before, and by the middle of next year, I've wanted to take the linebackers out back and put them out of my misery (see, I'm funny).

I like Rolle and Matthews... so far. Chaney never impressed me during the year. He's really, really slow at recognizing things (if he does at all), so he ends up wildly out of position a lot. But again, see above. Linebackers for the Eagles need to impress me for more than a half-year in order to make me accept that they aren't Yet Another Eagles Linebacker.

Also, Coleman/Allen is working out pretty good at safety, right?

Assuming they both stay healthy. I like Allen more than Coleman - Allen's easily got more range - but the knee just worries me.

Points: 0

#73 by chemical burn // Jan 03, 2012 - 4:34pm

You like Allen more than Coleman? I know you haven't liked Coleman as a rule, but Allen takes such bad angles on plays and is such a weak tackler - with him 20 yards gains frequently end up turning into 30 yarders or home runs.

But let me ask you - you aren't even a little excited for Rolle? I haven't seen an Eagles linebackers be so disruptive in the back-field since Trotter. He's rarely out of position and his instincts are just phenomenal. His main problem is that he's small and gets pushed off of plays, so he shares the same downside as Matthews, but he had a handful of "wow - great play!" moments this year. I'm not sure he didn't lead the team in them.

Anyway, we were both right in a funny way - you insisted that the Eagles didn't have the talent on the roster or depth to be good and I disagreed. They just needed time to gel and get their shit together. I thought they would do it before week 10. I also thought as with all reid teams, they didn't need great or even above average LB"s to be a good d and by the end of the year, that was proven to be true...

How do you feel about the impending Samuel and Jackson departures? Like I said, the only people in the organization I want 100% to come back are Mudd and Reid and literally almost every single other player or coach I'm ambivalent about. Well, not Trent Cole & Jason Peters. But if literally anyone else got traded or cut for good reasons/value, I wouldn't scream about it...

Points: 0

#76 by bubqr // Jan 03, 2012 - 5:14pm

Guys, If I can chime in on the LB/S situation:

Chaney: Athletic, but questionable instincts, takes bad angles and can have issues tackling. I was baffled to run that he thinks he had a great year. Still think he can be a good starter.

Matthews: REally don't like him... Not physical (always getting dragged on tackles), and quite slow for a small guy (watching Witten gaining yards on him after a dumpoff in the flat was depressing.

Rolle: Active, good hitter, good blitzer, seem to read the plays correctly, but still misses a high number of tackles. I really like him because I really liked him in college and reminds me of London Fletcher, who I absolutely love (target in FA?), but ideally you'd have him competing for a spot next year, not more.

Allen: Had some very, very bad games (Pats...), can still take bad angles and miss tackles, but can at times look quite good in coverage. I honestly don't know what to expect from him, but he should be starting next year (and getting challenged by the other 2)

Coleman: I like the guy, even with his inconsistency (games to games and tackling during the games). At least our 3rd safety next year.

Wish Colt Anderson will be back healthy, what a good STer he was this year...

I think adding a top LB within the next 3 picks (I see LB/WR/DT), and maybe even 2 is the major priority coming into the draft. Tommy Lawlor, whose opinion I value a lot is high on Luke Kuechly for the Eagles and from what I've seen I tend to trust his opinion. A good, solid young MLB would go a long way in stabilizing this defense

Points: 0

#87 by Kevin from Philly // Jan 04, 2012 - 2:59pm

There is zero chance the Eagles go after London Fletcher. 30+ vets with high salaries are typically not on their wish list. Vontaze Burfict in the draft, please. This D needs a lot of talent and a little crazy.

Points: 0

#90 by chemical burn // Jan 04, 2012 - 3:58pm

I actually agree with the "a little crazy" part. I think Andy Reid teams have a tendency to be a little subdued and while being even-keeled pays off for the offense in the long run, the defense sometimes feels soft and non-threatening. It's one of the reasons I like Coleman - in the Pats blowout, he seemed like he was going to go punch in a car window after the game, while everybody else was just sort of shrugging like "oh, well... he is Tom Brady, after all - what are you gonna do?"

Points: 0

#82 by Pat (filler) (not verified) // Jan 04, 2012 - 10:10am

What I wish is that I could shove Allen and Coleman together and make some transmogro-safety out of them. Allen's a better athlete who needs to work on a few techniques. Coleman's got very good technique, but is a limited athlete.

See previous comment on Rolle. I really do like him, but I liked Bradley, and I liked Trotter 2.0. Injuries killed both of them. I also liked Gaither, but his problem is that he was too small, which makes me worry about Rolle. He's the one linebacker of the three I'm most intrigued about, but he's also the one I think I'm going to be most disappointed with. I think Matthews is probably the most likely to be a longtime player with the Eagles.

Also, I don't disagree that they don't need great LBs to be a good defense. I completely agree there. The problem is that they need great LBs to be a great defense, and Philly hasn't had a quarterback consistent enough to be dominant with just a good defense.

re: Samuel/Jackson - I'm fine with Jackson leaving, I think they'll be fine without him. I also think they won't let him go. I really wish they'd mend fences with Samuel, but he's the one I think will leave. I have never liked the Eagles screwing around with cornerbacks. Cornerbacks are like quarterbacks - if you get even an average one, you keep him. Period.

Points: 0

#85 by chemical burn // Jan 04, 2012 - 12:47pm

The most interesting (depressing?) thing about the Samuel situation is that - and Eli Manning came out and said as much - is that teams were afraid of him and were targeting Asomugha instead of testing him. I think he's really, really an impact player, but he really fits in poorly with the defense they've built. The right side of the field went Babin, Fouku/Jordan/Matthews, Samuel so teams quickly learned that they could run right at those terrible tacklers and easily pick up between four and 46 yards a pop. Samuel was absolutely not protected in the run game by the players in front of him and it was really a problem, although the players in front of him were the REAL problem, so I can't fault him so much. I was surprised how good DRC was in run support considering his cousin is such a notorious "Ole!" tackler. DRC/Asomugha/Hanson makes more sense as a trio of corners than Asomugha/Samuel/DRC, weirdly. I'd hate to see Samuel go because he's one of their very best players, but they seem to be building the team all wrong to keep him.

And I basically agree with bubqr's assessment of everything, although Matthews really improved and Chaney was a B-/C+ backer, not a C-, which I think is an important distinction at the end of the day...

Points: 0

#20 by The Hypno-Toad // Jan 02, 2012 - 5:55pm

Sleeper seems like the wrong word for Denver. Something more along the lines of aberration, perhaps. I don't understand anything anymore... If I ever actually did.

Points: 0

#38 by Hot_Biscuit_Slim // Jan 02, 2012 - 8:28pm

The eyeball test would seem to make it appear that the four "out" teams are generally better than the four "in" teams (with the possible exception of the Giants). But all the "out" teams are widely acknowledged to come up woefully disappointing this season. I think that kinda proves that DVOA is still a pretty decent barometer.

And you're right, "Sleeper" might not be the word, but it was the only shorthand I could think of.

Points: 0

#42 by Aaron Brooks G… // Jan 02, 2012 - 9:09pm

Detroit went 2-1 against those "out" teams, including 1-1 against the 7-3 Cutler-version Bears, not their 1-5 Hanie-version namesakes.

Points: 0

#69 by The Hypno-Toad // Jan 03, 2012 - 2:23pm

That wasn't meant to be a critique of your word choice. More of a comment on how goddamned weird and unpredictable the Broncos' season was.

Points: 0

#21 by dmstorm22 // Jan 02, 2012 - 5:56pm

It's a hell of a lot better than 2009, where the Almanac hit 5 of the 12 (NE, IND, SD, PHI and MIN), projected JAX (finished 7-9), TEN (8-8), PIT (9-7), NYG (8-8), CHI (7-9), CAR (8-8) and SEA (5-11) to make the playoffs, and under=projected NYJ (projected at 6.2), BAL (8.8), CIN (6.9), DAL (8.0), GB (7.4), ARZ (5.6) and eventual Champions NO (7.8).

That, I'm guessing, was probably the low water-mark for FO preseason accuracy.

Points: 0

#8 by Jonadan // Jan 02, 2012 - 4:34pm

Nice to see the final rankings, & that Detroit stays in the top 12.

"When you absolutely don't know what to do any more, then it's time to panic." - Johann van der Wiel

Points: 0

#77 by LionInAZ (not verified) // Jan 03, 2012 - 8:11pm

I'm wondering, though, if it isn't time to start a new meme:

"DVOA hates the Lions"

Points: 0

#86 by chemical burn // Jan 04, 2012 - 12:50pm

Well for a team with awful special teams, an awful run game and a middling defense, I don't think 11 is bad at all. I think Lions fans should be able to see that they have a team that is excellent in many areas and still really struggles in many others...

(p.s. they happen to excel in the passing game, which is the best way to cover up flaws - if they made it to the superbowl, I don't think anyone would be shocked, even DVOA...)

Points: 0

#18 by MJK // Jan 02, 2012 - 5:44pm

Perhaps silly question. Since folks are always quick to point out that when we say "QB X had a passing DVOA of Y" we really mean "The team that QB X plays for had a passing DVOA of Y when QB X was passing the ball". In other words, we're not saying QB X is better than QB Z, but rather that QB X, defended by his team's O-line and throwing to his WR's, executing the plays his OC calls, is better than... and so forth.

So...why are NE, GB, and NO's pasisng DVOA numbers (61%, 73%, and 56%, respectively) dramatically different from Brady's, Rodgers', and Brees'? (41%, 53%, and 45%)?

Perhaps it's because other players sometimes pass for these teams? That can't be it, because the Patriots only ran one passing play this year without Tom Brady (the final pass yesterday, that Hoyer threw to give Gronkowski the TE receiving record). I don't think Brees failed to play any passing snaps. And Rodgers sat out Week 17, but I can't imagine Flynn's performance was so much BETTER than Rodgers to raise the Packers' overall passing DVOA 20% relative to where it was for Rodgers.

Is it a baseline issue? I.e. a team compares to the league average for TEAM passing offense, while an individual compares to the league average for individuals? This would imply that the league average for team passing DVOA is much higher than the average for individual passing DVOA...but I don't see how that's possible unless it's some weird manifestation of Simpson's paradox...

And in any case, a simple baseline difference wouldn't account for the ordering of the top 3 changing whether you looked at team or individual.

So please explain...why are individual QB passing DVOA's so different from their team's passing DVOA?

Points: 0

#31 by JForbes // Jan 02, 2012 - 7:25pm

I'd start with YAC, short passes turning in to huge gains - essentially, how much to credit the QB and how much to credit the receiver?

The QB hardly deserves a very high DVOA for a quick slant where the WR sheds a couple tacklers and gains 50+ yards. The QB deserves credit for an 8 yard pass or whatever it may be.

But the WR's ability to move the ball after the catch certainly makes their team's passing attack more effective.

Points: 0

#32 by MJK // Jan 02, 2012 - 7:44pm

But I don't think DVOA knows about those differences. All it knows is that the QB was facing 3rd and 10 on his 20, and now is facing 1st and 10 on the opponent's 30. I don't think DVOA actually divides success points up between the QB and the WR...doesn't it award both with the full amount (or lack thereof) of success points? Obviously, the two face different baselines...but I would expect the QB baseline to be very close to the TEAM baseline...

Maybe I'm fundamentally misunderstanding how team and QB passing DVOA are calculated...but it seems like if one QB plays just about every passing snap for a team, the two DVOA's should be almost identical...

Points: 0

#37 by Red (not verified) // Jan 02, 2012 - 8:14pm

Quarterback DVOA and team passing DVOA use a different baseline.

QB's are compared against other QB's only; in other words, the baseline is passing plays only, so the baseline is higher, and DVOA is lower.

Coversely, team passing DVOA is compared against ALL plays, since the team chooses whether to pass or run. That's why team passing DVOA is higher than QB DVOA - the baseline is lower when you include running plays.

Points: 0

#45 by JForbes // Jan 02, 2012 - 9:46pm

I thought it did but apparently it does not. Red's right.

Points: 0

#36 by Aaron Schatz // Jan 02, 2012 - 8:01pm

This question was asked earlier in the thread, and I answer it in comment 34 above.

Points: 0

#52 by MJK // Jan 02, 2012 - 10:59pm

Thanks Aaron. That explains part of it. So a team's passing DVOA is that teams defense-adjusted value over average total performance (where average total performance includes ALL plays), but only looking at plays where the team chose to pass. Whereas a QB's passing DVOA is the team's defense-adjusted value over average passing performance when that QB is throwing a pass.

What I still don't understand is how the relative rankings between Brees and Brady can be different than the relative rankings between New England and New Orleans (i.e. Brees has a better passing DVOA as a QB, but NE has a better passing DVOA as a team). The baseline difference between teams (which looks at all plays) and QB (which looks at passing plays only) should affect both teams equally, shouln't? As should the opponent adjustments? I.e. if one team has the second best team passing DVOA, its QB should have the second best QB passing DVOA, assuming he's the only guy throwing passes for that team?

Brian Hoyer had 1 attempt (compared to 611 for Brady), while Chase Daniels had 5 attempts (compared to 657 for Brees). So I can't imagine other people passing had anything to do with the difference, as was the case you mentioned above with Jim Sorgi playing most of Week 17 for the Colts.

Points: 0

#54 by Vincent Verhei // Jan 03, 2012 - 12:08am

What I still don't understand is how the relative rankings between Brees and Brady can be different than the relative rankings between New England and New Orleans (i.e. Brees has a better passing DVOA as a QB, but NE has a better passing DVOA as a team). The baseline difference between teams (which looks at all plays) and QB (which looks at passing plays only) should affect both teams equally, shouln't? As should the opponent adjustments? I.e. if one team has the second best team passing DVOA, its QB should have the second best QB passing DVOA, assuming he's the only guy throwing passes for that team?

Don't assume the baselines are the same. I'm going to make up a bunch of numbers to give an example. Suppose one team passes a lot on third-and-short, and the other almost always runs on third-and-short. Suppose both teams pick up a first down on half of their third-down throws, but 80 percent of their third-down runs. The two QBs will have equal DVOAs, because they performed the same in similar scenarios. However, the team that passes more often will have a lower passing DVOA, because we're comparing their passes to runs that would be more likely to pick up a first down.

Points: 0

#60 by matt w (not verified) // Jan 03, 2012 - 10:50am

Thanks, that's a very clear explanation.

Points: 0

#64 by MJK // Jan 03, 2012 - 12:05pm

Second the thanks. This makes it clearer.

Would it be fair to say, then, that by comparing team passing DVOA to that team's QB passing DVOA, you can say something about the quality of the team's offensive playcalling? Is this just another way of saying that team passing DVOA takes run/pass play selection into account, whereas QB passing DVOA does not?

To use your example, the first team has a lower passing DVOA not because they pass worse, but because their OC (or QB, when he audibles) is too pass happy.

In real life, where Brady ended up being ranked behind Brees, but NE ended up being ranked ahead of NO...does this therefore imply that New England's run/pass play selection was better than New Orleans' over the season?

Do you think you could use selective data filtering like what you're doing here to actually start rating coaching staff play selection? (I'm curious because, subjectively, I've never liked Bill O'Brien's playcalling, despite the obvious successful results evident from the DVOA tables at the top of this article. It's relevant now because he is being interviewed for head coaching positions, college and pro. It would be nice to have some way of objectively ranking coordinators...).

Points: 0

#68 by Vincent Verhei // Jan 03, 2012 - 1:30pm

Would it be fair to say, then, that by comparing team passing DVOA to that team's QB passing DVOA, you can say something about the quality of the team's offensive playcalling? Is this just another way of saying that team passing DVOA takes run/pass play selection into account, whereas QB passing DVOA does not?

Maybe, and yes. A team that tends to pass when the numbers say they should run would have a lower passing DVOA than they would otherwise. I wouldn't rush into using that as a play-call evaluator, though. The baselines are built around league-average data, not team-specific data. If, for example, Detroit has a lousy short-yardage running game, then maybe they should pass on third-and-1. When you start breaking things down to that level of detail, things like injuries and specific personnel can skew things. Again, it might be an effective evaluator, but without looking at that information specifically, I wouldn't put a ton of faith in it right away.

Also, since passing is generally more effective than running, I suspect that a team that simply passed every play would have the best "play-calling score."

Points: 0

#19 by ammek // Jan 02, 2012 - 5:54pm

The flipside to the Packers' historically great pass offense is that they're probably the most one-dimensional "great" team of recent years. The defense has been much discussed, but it's also worth noting that Green Bay had a peculiarly low number of rushing attempts for a team that won so many games and (as we've repeatedly heard) played with a lead for so much of the season. In fact, since 1978 among teams that have made the playoffs and had a points differential of more than +100, only two teams have had fewer RB carries than Green Bay's 319. One of those was the 1991 Oilers; the other the 2004 Eagles, who called 315 RB rushes at the height of Andy Reid's pass-wackiness. But although they won 13 games, those Eagles were very different: they scored fewer than 400 points (compared with the Packers' 560), and when they did run they were pretty good, with a 4.4 yards per carry average — much better than the Packers' pedestrian 3.9 this year.

The other arguably dominant offense almost as heavily imbalanced was the 2000 Rams in Mike Martz's first year as head coach. But as Aaron's chart shows, when those Rams ran, they were outstanding.

Points: 0

#27 by Stats are for losers (not verified) // Jan 02, 2012 - 6:25pm

...can't be good for their fan-base's collective self-loathing.

Points: 0

#39 by mansteel (not verified) // Jan 02, 2012 - 8:30pm

The (inverse) correlation between a team's strength of schedule and its DVOA is absurdly strong. The top six teams in DVOA played six of the seven easiest schedules, and five of the bottom six teams played top-eight schedules. Is the "D" in DVOA not strong enough, or is this just fluky?

Relatedly, kudos to the Browns, who managed to finish 4-12 with a below-average schedule, and (/sarcasm) the Giants, who were the only team with a top-ten schedule to finish with a winning record.

Points: 0

#41 by Thomas_beardown // Jan 02, 2012 - 9:00pm

I think there is a problem of granularity when it comes to opponent adjustments.

Look at the Packers against the Bears for instance. They get credit for playing a team who's strength is averaged over the whole season, but in both the games they played the Bears were were weaker than the average (well maybe equal to the average in the first game). In the first game the offensive line was in disarray from the injury, and the safety position was a mess. These problems would get addressed 3 weeks later for the Vikings game. Then in the rematch, Forte, Cutler, Conte, Knox, and even more offensive linemen were hurt. They lucked out terms of when the played the Bears but they get as much credit as the Chargers or Bucs do when they had to play the Bears at full strength.

Points: 0

#43 by Aaron Brooks G… // Jan 02, 2012 - 9:20pm

Contrast the Lions, the only team to play the full-strength Bears twice, whose strength of schedule was watered down by the 6 injured games.

Points: 0

#46 by BaronFoobarstein // Jan 02, 2012 - 9:53pm

That's a real possibility. Also possible is that the fact that teams never play against themselves is a factor.

Points: 0

#49 by RickD // Jan 02, 2012 - 10:14pm

You ought to expect a team's DVOA to be inversely correlated to its DVOA.

Consider only the intra-division games.

If you have Team A > Team B > Team C > Team D in terms of DVOA, then you would expect the standings to reflect A > B > C > D, but strength of schedule to be D > C > B > A.

This isn't absurd, it's natural.

In addition to the intra-division games, teams in the same division have 8 common opponents, one intra-conference division and one inter-conference division. And then each is paired up with the team that finished in the same relative position in its division for the other two divisions. The last two games are the only ones that would counter the natural forces of the intra-division games to create an inverse relationship.

Admittedly this argument only addresses the inverse relationship between teams in the same division. To extend it to the league as a whole, you'd have to consider the issue division by division.

Points: 0

#44 by ScottyB (not verified) // Jan 02, 2012 - 9:46pm

Pardon my ignorance, but does FO have a stat for something like "disaster plays", that is, giving up REALLY big plays on defense, or interception/fumbles/punt muffs that lead to opponent scores, etc.? I know FO would call many of these plays non-predictive, and they probably are in most cases, but I'm wondering if these are tracked.

If so, I believe that the teams with high DVOA but meh records probably exhibit a very high number of these plays. For example, the Jets/Giants game- aside from 2 crippling turnovers and a 10 yard pass turning into a 99-yard Cruz, the Jets almost assuredly outplayed the Giants (same this past week vs the Fish, in which after each of Sanchez' 3 horrible picks, the D holds Miami without a first down, but still surrendered 9 points; also losing 6 ST fumbles and 6 pick-sixes this season). Dallas, Philly and SD also seemed to have more than their share of these disater plays this year.

Points: 0

#51 by Dales // Jan 02, 2012 - 10:55pm

the Jets almost assuredly outplayed the Giants

We didn't see the same game. I saw the Jets doing better until the Cruz touchdown, and then being solidly outplayed the rest of the way.

Points: 0

#58 by armchair journ… // Jan 03, 2012 - 6:30am

I saw the game Dales saw. What I heard, however, from the stellar announcing crew, was how the Jets were beating the pants off the Giants, no matter what it looked like. Perhaps you are remembering the narrative and not the game.


Points: 0

#83 by ScottyB (not verified) // Jan 04, 2012 - 11:32am

Dales & Armchair- the Jets definately folded after getting down in the 2nd half. But up to that point, they had a great TD drive, moved the ball on most other drives, and held the Giants to mostly 3 and outs. I see what you are saying, though.

However, my larger question remains- does anyone track these "disater plays"?

Points: 0

#48 by PatsFan // Jan 02, 2012 - 10:06pm

I'm vaguely amused that with all the press about NE's "historically bad" defense, GB (courtesy of Stafford) ended up edging out NE and finished #32 in yards given up and passing yards given up, as well as setting records in both.

That said, I agree with DVOA that GB's defense is better than NE's.

Points: 0

#50 by NYMike // Jan 02, 2012 - 10:50pm

The only consolation I have is that Matthews and Woodson did not play, and Detroit, unlike Buffalo, didn't quit. Not much to go on ...

Points: 0

#53 by Vincent Verhei // Jan 02, 2012 - 11:56pm

Green Bay led the league in interceptions with 31. New England was tied for second (with San Francisco), but it was a distant second, just 23.

Points: 0

#55 by MilkmanDanimal // Jan 03, 2012 - 12:38am

Indianapolis lost Peyton Manning, and not shockingly had a 16-position drop in DVOA for the year. Tampa suffered no particularly significant injuries (far fewer than last year, in fact), and dropped 18 spots.

Points: 0

#67 by JasonG (not verified) // Jan 03, 2012 - 1:26pm

Thanks for all the great content this year guys. I would be curious to see what the Bears DVOA (rushing O, passing O, total O, total D and ST) splits are for their 7-3 stretch and their 1-5 stretch. Thanks.

Points: 0

#71 by aaa (not verified) // Jan 03, 2012 - 2:59pm

What was Houston's final DVOA if you calculate the injury to Schaub in? Similar to what was done in the Playoff odds report the last few weeks.

Points: 0

#74 by Anonymiss (not verified) // Jan 03, 2012 - 4:46pm

Does anyone know when the last time a team has finished 5-1 (or better) in Division games and NOT won the Division? Pretty crazy.

Points: 0

#75 by Eddo // Jan 03, 2012 - 5:10pm

The Raiders went 6-0 in the AFC West last year.

Points: 0

#79 by LionInAZ (not verified) // Jan 03, 2012 - 8:17pm

Beat me to it. That was one that stood out in my mind from last year.

Points: 0

#80 by dmstorm22 // Jan 03, 2012 - 8:28pm

2002: Giants (5-1)
2003: Seahawks, Cowboys, Broncos (5-1)
2004: Rams (5-1)
2005: Redksins, Vikings (5-1 - MIN also didn't make the playoffs)
2006: Packers, Panthers (5-1 - both didn't make the playoffs)
2009: 49ers (5-1)

I'm surprised it has happened this often.

Obviously, different scheduling formula, but the 1999 Titans went 9-1 in division and didn't win it (13-3 record, vs. Jax who was 14-2 - both losses to the Titans as well).

Points: 0

#81 by Anonymiss (not verified) // Jan 04, 2012 - 7:36am

Whoa, I'm surprised it's that often also. Thanks for the answers

Points: 0

#78 by milo // Jan 03, 2012 - 8:14pm

"FO Premium will be updated with final 2011 ratings later today"
Forgive me, english is not my native language. What does this mean?

Points: 0

#92 by Mr Shush // Jan 05, 2012 - 9:12am

"FO Premium" is the paid-for, subscription-only part of the site, with more detailed statistical breakdowns. These numbers were still reflecting only Weeks 1-16 at the time this was written, but were to be altered to include the events of Week 17 later in the day. This update has presumably now happened.

Points: 0

#95 by milo // Jan 05, 2012 - 11:35am

You are only two clicks away from not presuming something that is wrong.

Points: 0

#97 by Anonymous12345 (not verified) // Jan 05, 2012 - 5:20pm

Good God you're a dick.

Points: 0

#96 by Schmoker (not verified) // Jan 05, 2012 - 11:57am

This is my first year following DVOA, and I have very much enjoyed it. Your process seems eminently logical and fairly predictive in the short term. It helped me make a few very, very good decisions against the grain in my football pick 'em league. Alas, some very bad decisions (some DVOA related, most not) cost me the title by a slim margin. But I want to thank you for all the work you put in. I also have a question about what seems to me to be an anomaly.

Trying to sync up DVOA with the old fashioned "eye-test" works pretty well in almost every case. Perhaps that is because I am influence by the DVOA so much now that I've already mentally labeled teams to coincide with what their final DVOA turned out to be. But there is one team that totally mystifies me, and I would love to hear your thoughts on them. "Them" being the New York Jets.

This is a perfect test, by the way, as right next to the Jets is another team in the Eagles whose number of wins does not match their DVOA remotely. But I look at the Eagles and I see it. I see why the DVOA rates them so highly despite their many problems this year. I see the way talent and measurables did not translate into wins. I see how they are certainly worthy of such a high ranking in DVOA even though the hype around them most of the season---combined with some of their mystifying losses---would seem to suggest they weren't that good. Yet after paying careful attention all season long, I feel that I can understand it. I'm not certain they'd be that much more successful if given a do-over, but I could believe that under the right circumstances they could have gone 12-4 just as easily as they went 8-8.

By the way, not from Philly here, and not a fan at all of the team or the organization, so that has nothing to do with anything. I'm from Akron, and I could care spit about the Jets or the Eagles.

Anyway, when it comes to the Jets I see none of the things I see with the Eagles. When it comes to the Jets, I don't get DVOA's opinion of them. When it comes to the Jets, I think they could more easily have gone 4-12 than 12-4. And not only do I not think they are deserving of the #9 ranking, I think the weighted DVOA is even more egregious, as they appear to have ended the season amongst the bottom feeders based on their play on the field vs. the level of competition they faced. If they are currently the 11th best team in football, I'll eat Rex Ryan's tear stained Jets cap.

By the way, I'm perfectly willing to believe there may be cap eating in my future. I understand I could be way off the mark. But when I look at every other team in the DVOA rating system, I don't see reason for argument. Or at least not for serious argument. There might be a team or two I'd alter by a spot or two, but there is no team that I'd drop a dozen places, as I would do with the Jets.

So how am I, as I assume I am, so off base here? Why is it that the Jets appear to be so thoroughly over valued by DVOA, even though the similarly disappointing Eagles do not appear to be so? Why is DVOA so in love with a team whose performance on the field do not seem to remotely back up their ranking? Is their some stat in particular that the Jets excel in which skews their ranking?

Or am I just looking at the New York team through the wrong pair of glasses?

Points: 0

#100 by nat // Jan 06, 2012 - 10:07am

Two thoughts:

The Jets have a very good ST DVOA. As fans, we rarely consider special teams when we think about how teams stack up. That accounts for 3-4 slots in the DVOA rankings.

Before adjusting for opponent, VOA considers both what a team does and what the average result is for the situations it gets into. While you didn't see the Jets offense looking like a -2.7% offense, you did see them in a lot of situations where they get an 'easy' grading scale. In seven of their eight losses, the Jets had more than one set of downs while down by two scores with five minutes or less. That's 50 plays worth of easy baselines to inflate their 'success over average'. In those situations, they played quite well compared to teams bad enough to be losing late by multiple scores.

Points: 0

#101 by Schmoker (not verified) // Jan 07, 2012 - 10:57am

That certainly makes some sense, but it would seem to reveal a serious flaw in DVOA. Or maybe not so serious, since it appears only one team is seriously affected by it.

Still, is it possible to rework DVOA to eliminate the overvaluing of garbage stats earned against soft prevent defenses, if that is what the problem was? Or do I still just undervalue the Jets too much?

One thing I heard this week that was interesting came from Chad Millman's Behind the Bets podcast. He had on a gambler named Warren Sharp. This guy uses his own system to judge red zone performance that takes into account not just what a team does in a the red zone, but also how often they get there. He uses to two as counter weights for each other, and he does the same for defenses. Seemed pretty sensible. And using that system, he lowered the Jets red zone efficiency because even though they had a high red zone success rate, they actually got into the red zone fewer than almost any team in football.

Points: 0

#102 by nat // Jan 08, 2012 - 12:18am

It's only a serious flaw if one of two effects turn out to be both common and large:

(1) the late game baselines are overly biased away from an "average team" towards an "average bad team"
(2) the success points are skewed too far away from what a good coach and team should be trying to do late in the game

It's hard to know how large those effects are in general, and how close to a worse case the Jets were this season. While I'm pretty sure these effects are real and at least partially responsible for the strange Jets results, I have no way of knowing how large or common these effects are in general. I'm hoping Aaron and company look at this in the off-season. There should be ways to remove the selection bias in the baselines, if they turn out to be important.

My guess is that it is only a "serious" flaw in unusual situations. But every now and then, a team is - by luck or style of play - an outlier as far as the baselines go.

About that gambler's system.... Isn't that just a fancy way to figure out who scored the most points from in the red zone? Success rate * opportunities = total successes?

Points: 0

#103 by Schmoker (not verified) // Jan 08, 2012 - 5:34pm

Good point about the gambler's system, but I guess I have heard so much about the Jets red zone efficiency this season that it has made me sick. For a such a poor offensive team, to make such a big deal out of such a small number of their overall plays always seemed really obtuse to me. I know that there is a difference between playing in the red zone and outside the red zone, as in the compressed field does change things. But it doesn't change things so much that one of the worst offensive teams in football should be lauded because during an exceedingly small sample of their overall number of plays they did well.

I guess I thought one of the components of DVOA was red zone efficiency. Maybe I am wrong about that. I have so much to still understand about DVOA and what goes into it. And like I said, aside from the Jets (which you rightly point out are an wacky outlier), it seems like DVOA is pretty much on the money for most of the rest of the teams.

On a personal note, DVOA did keep biting me in the ass with the Jets this season, which I think is leading me to a personal bias on this whole question. Pretty much every Jets loss this year I wanted to pick against them, but pretty much every time I let DVOA talk me into picking for them. Up until the final week of the season, I was letting that crazy Jets DVOA number mess up my weekly picks.

Of course, there were a couple of Jet wins in there that I would never have picked on my own. Still, on balance, the Jets were the one team where DVOA led me astray more often than it helped me this season.

Another thing I do wonder about is ST rankings. I'd like to see special teams rankings broken down into two groupings, with one being for kicking and one for coverage. Kicking is so dependent on two players that I wonder if it doesn't skew the overall ST ranking. Case in point being my hometown team, the Cleveland Browns. They are ranked 10th this season, and they were ranked 10th last season. But if you watched every Browns game the past two seasons then you would know that special teams were not close to being as good this year as they were last year. But the two kickers, Dawson and Maynard, could have been said to have had better seasons than last year. So I wonder, did their work skew things so that overall the ST ranking does not come close to showing us how poor the kick coverage was this year in relation to last year?

Probably there is an answer to this question here on the site and I have just not yet found it.

Points: 0

#104 by Thomas_beardown // Jan 08, 2012 - 6:01pm

You can look at the special teams page to see special teams broken down.

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.