Final 2010 DVOA Ratings

Final 2010 DVOA Ratings
Final 2010 DVOA Ratings
Photo: USA Today Sports Images

by Aaron Schatz

Last year, the final DVOA column of the regular season was all about how strange the 2009 stats were. Not this year. This year, the DVOA ratings make pretty good sense. The best teams are high, and the worst teams are low. Hard to imagine there's much of a disagreement about the Patriots at number one, and I doubt there's much disagreement about the Steelers at number two either.

So instead of trying to explain why various teams are ranked in various places, we can simply enjoy going through and seeing where the teams and players of 2010 fit historically -- at least as far as the regular season goes. Remember that the "DVOA Era" currently goes back to 1993.


Let's start with the question a lot of people were asking me on Twitter last night: Are the 2010 Seattle Seahawks the worst playoff team in NFL history? At least as far as DVOA is concerned, the answer turns out to be a surprising "no." The Seahawks were so good in their final win -- their third-highest DVOA of the year, 46% -- that they climb to -23.6% overall. That puts them ahead of the 2004 Rams, who had -26.5% DVOA. The Seahawks do pass the 1998 Cardinals as the second-worst playoff team in DVOA history. What's scary here is that the 2004 Rams and 1998 Cardinals each won their first playoff game. New Orleans beware!

New England ends the year as the second-best team in DVOA history. They have the highest weighted DVOA we've ever measured, and the highest offensive DVOA we've ever measured. Their weighted DVOA seriously laps the field, over ten percentage points higher than any other team.

If you've been following things for the last few weeks, those numbers won't surprise you. What may surprise you is that the Pittsburgh Steelers are also in the all-time DVOA top ten. The Steelers are only the third team since 2000 with both offense and defense in the top five, and both of those other teams -- the 2007 Colts and the 2009 Packers -- were dead last in special teams, while the Steelers are average in special teams. Coincidentally, they end up with basically the same ranks as the 1994 Pittsburgh Steelers, who were also fifth in offense and first in defense, plus 15th in special teams. But since standard deviation for DVOA is a lot higher these days, the 2010 Steelers end up with a much better total DVOA than the 1994 Steelers.

Year Team DVOA Year Team DVOA
2007 NE 53.1% x 2010 NE 54.0%
2010 NE 44.7% x 2007 NE 42.9%
1995 SF 43.3% x 1995 SF 41.8%
1996 GB 41.3% x 1994 SF 41.5%
2004 PIT 38.1% x 2004 BUF 41.4%
2010 PIT 37.3% x 2004 PIT 41.4%
2004 NE 35.9% x 2010 PIT 40.0%
1999 STL 34.3% x 1995 KC 38.1%
1994 DAL 33.7% x 2005 DEN 37.8%
2005 IND 33.6% x 2009 GB 37.3%

The Patriots' record-setting offense does not set a record for passing DVOA. That record still belongs to the 2007 Patriots. But this team was much better on the ground, finishing with the seventh-highest rushing DVOA ever. Yet they weren't the most efficient running team this season. The Philadelphia Eagles end up with the third-highest rushing DVOA ever. Obviously, a lot of that is Michael Vick, but LeSean McCoy was also excellent this season, 18.0% DVOA (fourth among qualifying backs).

Year Team DVOA Year Team DVOA Year Team DVOA
2010 NE 46.1% x 2007 NE 75.4% x 2000 STL 34.4%
2007 NE 45.2% x 2010 NE 72.6% x 2002 KC 32.0%
2002 KC 38.0% x 2004 IND 69.1% x 2010 PHI 29.2%
2004 IND 33.2% x 2009 SD 63.7% x 2003 KC 28.6%
2004 KC 32.9% x 2009 NE 57.0% x 2006 SD 27.5%
2003 KC 32.2% x 2006 IND 56.9% x 1998 DEN 27.1%
1998 DEN 30.7% x 2008 SD 54.9% x 2010 NE 27.0%
1995 DAL 30.1% x 1995 DAL 52.4% x 2004 KC 26.6%
2009 NE 29.6% x 2003 TEN 52.1% x 2005 DEN 26.3%
2006 IND 29.0% x 1994 SF 50.3% x 1993 SF 25.4%

On the other side of the table, the Carolina Panthers offense improved so that the Panthers don't rank among the absolute worst offensive DVOA ratings in history. We end up with the Panthers tied for the eighth-worst, and the Cardinals as tenth-worst. Somehow, the Carolina Panthers with DeAngelo Williams (for part of the year, at least) and Jonathan Stewart finish dead last in rushing DVOA.

Defense doesn't really set any records this year for any teams. At one point, it looked like the Steelers might have one of the top three or four run defense DVOA ratings, but they end up tenth.

And now, the moment you've all been waiting for...

Year Team DVOA
2000 BUF -12.9%
1997 SEA -9.8%
1997 CHI -8.2%
2002 CIN -8.1%
2010 SD -8.1%
1998 OAK -7.7%
2004 STL -7.5%
1996 NYJ -6.8%
2002 WAS -6.7%
1993 MIN -6.7%

The Chargers, suddenly realizing they were playing reasonable special teams, desperately tried to return to their record-setting ways in Week 17, but it was no go. They simply can't beat the squib-tastic 2000 Buffalo Bills.


Tom Brady ends the year with the fifth-highest passing DYAR in NFL history (2,141 DYAR) and the fourth-highest DVOA for a quarterback with more than 100 passes (53.4%). The top six DYAR seasons all belong to Brady (2007, 2009, and 2010) and Peyton Manning (2004, 2006, and 2009). Boomer Esiason's ridiculous 1997 actually sneaks into the Brady-Manning stratosphere for top DVOA seasons. Manning finishes second this year with 1,674 DYAR, a total that was topped by six different quarterbacks a year ago.

For the second straight year, Brady gets a nice bump from opponent adjustments, although he would be number one either way and the bump isn't as large as the one he got in 2009.

Rookie Jimmy Clausen, at -607 DYAR, is the clear winner for the title of "worst starting quarterback of 2010," but that total barely makes the all-time worst 20 quarterback seasons. Michael Vick's 210 rushing DYAR finishes as the third most valuable quarterback rushing season in DYAR behind Vick in 2006 and Vick in 2004.

Jamaal Charles narrowly passes Marshall Faulk to set the all-time record for rushing DVOA by a running back with a minimum of 100 carries. Charles finishes with 35.8% DVOA. Faulk had 35.0% DVOA in 2000, the year the Rams defense broke down and Kurt Warner only started 11 games (although Trent Green was pretty good as his backup). Only one other running back had 30% DVOA or higher in a year with at least 150 carries: Priest Holmes with 31.1% DVOA in 2002. Charles doesn't make it near the top of the DYAR lists, however, because of the Kansas City committee setup. Thomas Jones has -90 DYAR, ranking him next to last, ahead of only Chester Taylor. You can't just find this kind of quality running back management anywhere, you know.

Mike Wallace ends the year as the DYAR leader among wide receivers. Steve Smith actually finishes dead last in DYAR, and it is hard to use "the quarterbacks were awful" as an excuse given how good he was in 2005, when the Panthers had Jake Delhomme and no other receivers and forced the ball to Smith non-stop. Neither player comes anywhere near the all-time best and worst in DYAR, although Wallace's 48.6% DVOA is the fourth-highest for a wide receiver with at least 50 passes. Oakland rookie Jacoby Ford has 110 DYAR rushing, which tentatively sets a record for wide receivers who don't take snaps from a Wildcat or option package. However, it is possible we'll change some of those from rushing to passing in the next couple weeks if it turns out they were slightly-backwards wide receiver screens.

I explored Antonio Gates' astonishing record-breaking season in an Extra Point commentary a couple weeks ago. Gates didn't play another game after that, so he ends up establishing a new tight end DYAR record in just 10 games. I could probably learn five or six languages and still not have enough adjectives to talk about how awesome Antonio Gates is. Jermichael Finley also finishes in the all-time top five for DVOA by tight ends with at least 25 passes -- barely qualifying with 26 passes in an excellent partial season. Daniel Graham of the Broncos is dead last this year with -133 DYAR, the third-lowest tight end total in history behind Boo Williams in 2004 and Keith Cash in 1996. Graham had a catch rate of 49 percent, and tight ends shouldn't have catch rates below 50 percent. Ever.

All team and individual stats pages should be updated in the next few minutes. FO Premium is already updated with final 2010 ratings. We'll get all the 2010 stats onto the player pages sometime in the next couple weeks, as well as updated similarity scores based on 2008-2010 rather than 2007-2009. Loser League results will be announced in Scramble for the Ball on Wednesday.

* * * * *

These are the Football Outsiders team efficiency ratings through the end of the 2010 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 2009.

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

To save people some time, please use the following format for all complaints: <team> is clearly ranked <too high/too low> because <reason unrelated to DVOA>. <subjective ranking system> is way better than this. <unrelated team-supporting or -denigrating comment, preferably with poor spelling and/or chat-acceptable spelling>

1 NE 44.7% 14-2 1 3 54.0% 1 46.1% 1 5.5% 19 4.1% 8
2 PIT 37.3% 12-4 2 10 40.0% 2 17.9% 5 -18.5% 1 0.9% 16
3 GB 23.3% 10-6 5 2 25.2% 4 14.7% 7 -10.8% 2 -2.2% 27
4 BAL 23.0% 12-4 3 1 27.8% 3 9.4% 12 -7.9% 4 5.6% 4
5 PHI 22.0% 10-6 4 4 21.4% 6 21.3% 3 1.4% 14 2.1% 13
6 NYJ 18.7% 11-5 10 9 14.9% 7 5.6% 16 -7.8% 6 5.3% 5
7 SD 17.6% 9-7 6 11 21.5% 5 19.3% 4 -6.4% 7 -8.1% 32
8 ATL 15.9% 13-3 7 18 13.5% 9 10.9% 10 1.3% 12 6.3% 2
9 NYG 14.5% 10-6 8 17 11.6% 10 11.4% 9 -8.1% 3 -5.0% 30
10 NO 11.0% 11-5 9 6 13.6% 8 9.7% 11 -2.6% 9 -1.3% 22
11 TEN 8.0% 6-10 11 22 1.1% 17 -0.7% 20 -4.1% 8 4.6% 6
12 TB 4.2% 10-6 17 27 9.4% 14 11.5% 8 6.9% 23 -0.4% 18
13 HOU 4.0% 6-10 18 15 9.8% 13 26.0% 2 20.9% 31 -1.2% 21
14 MIA 3.8% 7-9 12 16 2.3% 15 4.6% 18 -1.9% 10 -2.7% 29
15 IND 3.0% 10-6 14 8 -1.0% 18 16.6% 6 8.2% 24 -5.3% 31
16 CHI 3.0% 11-5 15 25 11.5% 11 -12.0% 28 -7.8% 5 7.2% 1
17 KC 0.1% 10-6 13 28 -11.3% 25 7.8% 13 6.1% 21 -1.6% 25
18 DET -1.4% 6-10 19 32 1.5% 16 2.9% 19 6.6% 22 2.3% 11
19 CIN -2.4% 4-12 20 19 -5.5% 21 5.0% 17 4.7% 18 -2.6% 28
20 CLE -3.3% 5-11 16 26 -6.2% 22 -1.4% 22 4.1% 17 2.2% 12
21 OAK -3.6% 8-8 22 30 11.1% 12 -4.5% 23 1.5% 15 2.4% 10
22 JAC -8.9% 8-8 21 23 -4.6% 19 7.7% 14 21.2% 32 4.5% 7
23 DAL -9.8% 6-10 24 5 -9.6% 23 -0.8% 21 10.0% 27 0.9% 15
24 SF -10.0% 6-10 26 20 -4.8% 20 -7.3% 24 1.4% 13 -1.3% 23
25 MIN -13.6% 6-10 23 7 -17.2% 28 -11.8% 27 0.9% 11 -1.0% 20
26 DEN -15.0% 4-12 28 12 -10.6% 24 6.2% 15 19.6% 30 -1.6% 24
27 WAS -18.3% 6-10 29 21 -23.1% 29 -7.5% 25 9.2% 26 -1.7% 26
28 BUF -19.4% 4-12 25 24 -16.2% 27 -9.3% 26 10.2% 28 0.2% 17
29 STL -20.0% 7-9 27 31 -15.9% 26 -15.2% 30 5.9% 20 1.1% 14
30 SEA -23.6% 7-9 30 29 -31.7% 31 -14.3% 29 15.0% 29 5.8% 3
31 CAR -34.3% 2-14 32 14 -31.5% 30 -31.9% 32 1.8% 16 -0.6% 19
32 ARI -36.2% 5-11 31 13 -34.7% 32 -31.3% 31 8.3% 25 3.4% 9
  • 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.
  • NON-ADJUSTED TOTAL DVOA does not include the adjustments for opponent strength or the adjustments for weather and altitude in special teams, and only penalizes offenses for lost fumbles rather than all fumbles.
  • 2010 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.
  • 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 2010
1 NE 44.7% 14-2 39.9% 14.1 1 5.8% 7 12.3 1 18.6% 17
2 PIT 37.3% 12-4 34.0% 12.3 2 5.2% 8 12.1 3 13.9% 12
3 GB 23.3% 10-6 22.8% 10.9 4 2.4% 16 12.1 2 15.3% 15
4 BAL 23.0% 12-4 15.8% 11.5 3 6.1% 6 10.6 6 5.1% 2
5 PHI 22.0% 10-6 22.9% 10.8 5 -0.3% 20 9.4 10 9.3% 6
6 NYJ 18.7% 11-5 18.9% 9.8 8 7.1% 5 9.8 8 23.9% 28
7 SD 17.6% 9-7 23.1% 9.7 10 -4.9% 26 10.9 5 23.2% 26
8 ATL 15.9% 13-3 17.9% 10.5 6 -1.8% 22 11.2 4 4.9% 1
9 NYG 14.5% 10-6 17.0% 10.2 7 -3.3% 24 10.1 7 22.6% 24
10 NO 11.0% 11-5 15.2% 9.8 9 -5.4% 27 9.2 11 10.9% 8
11 TEN 8.0% 6-10 5.0% 8.9 12 2.8% 14 8.5 16 20.4% 21
12 TB 4.2% 10-6 8.5% 8.6 14 -4.4% 25 8.7 15 14.7% 14
13 HOU 4.0% 6-10 3.3% 8.2 16 3.3% 13 7.1 18 18.5% 16
14 MIA 3.8% 7-9 -0.8% 9.0 11 9.9% 3 6.2 23 13.8% 11
15 IND 3.0% 10-6 3.4% 8.8 13 3.5% 11 9.1 13 7.4% 3
16 CHI 3.0% 11-5 0.3% 8.3 15 0.9% 18 9.5 9 25.4% 30
RANK 2010
17 KC 0.1% 10-6 7.3% 8.0 17 -6.8% 28 9.1 12 22.8% 25
18 DET -1.4% 6-10 -2.5% 7.8 18 4.1% 10 7.8 17 9.0% 5
19 CIN -2.4% 4-12 -12.0% 7.3 20 11.2% 2 6.1 25 12.0% 10
20 CLE -3.3% 5-11 -6.5% 7.5 19 9.5% 4 6.1 24 19.4% 18
21 OAK -3.6% 8-8 2.2% 7.0 22 -2.3% 23 8.9 14 33.9% 32
22 JAC -8.9% 8-8 -7.9% 6.7 25 0.9% 19 6.4 22 19.9% 20
23 DAL -9.8% 6-10 -9.0% 6.8 23 1.8% 17 7.0 19 22.3% 23
24 SF -10.0% 6-10 -4.0% 7.1 21 -7.4% 29 6.8 20 24.4% 29
25 MIN -13.6% 6-10 -16.9% 6.7 24 5.0% 9 6.0 26 14.0% 13
26 DEN -15.0% 4-12 -13.9% 6.2 27 -0.9% 21 5.2 29 20.6% 22
27 WAS -18.3% 6-10 -19.0% 5.9 28 3.4% 12 5.9 27 8.9% 4
28 BUF -19.4% 4-12 -27.9% 5.7 30 12.0% 1 4.4 31 19.6% 19
29 STL -20.0% 7-9 -11.4% 5.8 29 -10.2% 32 6.8 21 11.5% 9
30 SEA -23.6% 7-9 -22.8% 6.2 26 -7.4% 30 5.5 28 28.6% 31
31 CAR -34.3% 2-14 -37.7% 3.9 31 2.8% 15 2.4 32 9.5% 7
32 ARI -36.2% 5-11 -27.1% 3.7 32 -8.4% 31 4.4 30 23.6% 27

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


202 comments, Last at 14 Jan 2011, 11:27pm

139 Re: Final 2010 DVOA Ratings

Maybe my statement was too strong.

Yes, its good to take 1 when 1 is there.

I just think, very often, BJGE takes 3 when 20+ is there. We're talking about a guy who through 330 career carries, has had 5 carries of 20+, and no carries of 40+.

I wouldn't mind seeing him as a short yardage guy, or something along those lines, but its clearly a position that could be upgraded. The Patriots rushing offense has been ranked by DVOA 2, 9 , 4, 2, the last 4 years. I would guess that the rushing DVOA correlatees quite highly with the passing DVOA in the Pats case.

Take this statement : "(A)BJGE has x DVOA while rushing behind the Patriots (B)line, with the Patriot's (C) TEs blocking, (D) Tom Brady setting up the playaction, and the patriots (E)WRs blocking downfield"

I feel like part A is the least important part of that statement, and by a very wide margin.

144 Re: Final 2010 DVOA Ratings

There's something to be said for consistency though. Given your statements of his lack of 20+ and 40+ gains, he still has managed to average 4.4 YPC this year. That's a testament to his consistency. He's not relying on huge runs to pump up his YPC.

162 Re: Final 2010 DVOA Ratings

I agree that consistency is valuable, and I absolutely understand why DVOA thinks BJGE is good.

I just don't think it has much to do with BJGE, and has a lot to do with the other factors in the equation. I'm not a huge fan of ALY, but NE's is the highest in the league, and by quite a large margin. Don't forget, Woodhead rushed for 5.6ypc, and actually was used as a traditional RB quite a big (getting carries on 1st and 2nd down out of non-draws). Woodhead also had a higher dvoa than BJGE. Even Maroney had a very high DVOA behind this line last year.

I agree that BJGE's performance is valuable, I just don't think its his talent that is the determining factor in that performance. I think its the environment.

164 Re: Final 2010 DVOA Ratings

Even Maroney had a very high DVOA behind this line last year.

Since when is 1.2% (good for 29th in the league, out of 50 backs with at least 100 carries) a "very high DVOA"?

If you want to say he had a very high DVOA three years ago, sure. Last year? Not so much.

158 Re: Final 2010 DVOA Ratings

Agree, but you're leaving out the most important thing (although you imply it, as do others). BJGE does FOUR things well: he SEES the hole, he hits it hard and goes through it, he holds onto the ball, and he gets the extra yard.

Seeing the hole is something that a relatively small subset of RB's actually seem to do (and sometimes the great ones don't...Chris Johnson certainly doesn't always seem to see it). I'd rather have a guy that always gets what's there than one who sometimes does amazing but often leaves yards on the field.

81 Eagles' Rushing DVOA

This has probably been asked and answered before, but given Philly's more-overrated-than-usual rushing DVOA, but does DVOA distinguish between Vick's sacks on "traditional" dropbacks and "sacks" on scrambles where he doesn't reach the line of scrimmage?

129 Re: Eagles' Rushing DVOA

In reply to by Stats are for losers (not verified)

No it does not. QB runs in the play-by-play are usually just listed as runs, not designated as designed runs or scrambles. We do collect that information in the game charting project, but won't have that until after the season.

Because of this, yes, on any of Vick's teams (or any other teams with running quarterbacks), the passing DVOA will be slightly deflated (because a lot of sacks will be failed scrambles/runs, not failed passes), and the rushing DVOA will be slightly inflated (because those scrambles will only count for rushing DVOA when they gain yards).

83 Re: Final 2010 DVOA Ratings

i'm looking forward to the 1991 DVOA recap, as that might skew the top ten. The '91 Skins were the best team i feel i ever watched (until the 07 Pats). I'm interested to see in a few months where they end up vs the '10 Pats.

85 Re: Final 2010 DVOA Ratings

" Daniel Graham of the Broncos is dead last this year with -133 DYAR, the third-lowest tight end total in history behind Boo Williams in 2004 and Keith Cash in 1996. Graham had a catch rate of 49 percent, and tight ends shouldn't have catch rates below 50 percent. Ever.

Aaron, was there any split between Graham's numbers with Orton throwing to him vs Tebow.

89 Re: Final 2010 DVOA Ratings

I don't quite grok why the playoff odds report gives PIT a better chance of reaching the AFC championship game (and winning the AFC) than NE. Presumably this is because NE is more likely to face BAL in the second round, but presumably DVOA picks NYJ as the likely winner of the WC game, which gives NE NYJ and PIT BAL.

94 Re: Final 2010 DVOA Ratings

Presumably that's a mistake. Maybe those numbers are reversed (switch NE and PIT numbers for those two games) I can't make sense of it otherwise.

124 Re: Final 2010 DVOA Ratings

I'm not so sure. If you have a situation where there are two equally good 5/6 seeds and two equally bad 3/4 seeds, the 2 seed gets an advantage because it gets the bad team whenever one (but not both) of the 5/6 seeds win. I guess that's what's going on here.

130 Re: Final 2010 DVOA Ratings

Even more pronounced in the NFC. If Seattle and Green Bay happen to advance, the Falcons have a much tougher road to the Super Bowl than the Bears.

118 Re: Final 2010 DVOA Ratings

Assuming no errors in the playoff odds report, it comes down to PIT having a big edge in the divisional round match-ups because of their #2 seed. The AFC #5 and #6 seeds are about 15-23 DVOA higher than the #3 and #4 seeds. If one wildcard wins and one loses, PIT plays a much weaker team, while being only 7 DVOA behind NE. If both wildcards win or both lose, NE should have better odds than PIT, but not enough to overcome the difference.

95 Re: Final 2010 DVOA Ratings


AVG 1.07%
SD 18.97%


1. 8/10 ELITE or GOOD teams made the playoffs.
2. Did Special Teams really keep SD and NYG out of the playoffs?
3. Did SEA's special teams get them to the playoffs?
4. IND is GOOD on OFF, BAD on DEF, and HORRID on S.T. How'd they make it and SD didn't.
5. KC is AVG on OFF and DEF and merely bad on S.T., maybe that's why SD didn't make it.


AVG 2.07%
SD 20.38%


1. 9/12 GOOD or ELITE teams made the playoffs.
2. Thoughts about KC being bad at the end of the season and playoff implications?
3. SEA is HORRID recently.


MEAN 3.33%
SD 15.89%


2. W/O NE in the mix, HOU is ELITE and ATL, NO, BAL are GOOD.
3. Alternatively, the level of suck at the bottom is greater than 2 SD from the mean, and removing them from the mix also moves HOU into the ELITE range and pulls STL into the HORRID range, CLE and OAK become BAD, and the NYG and TB become AVG.

MEAN 2.96%
SD 9.26%


1. PIT is almost a full SD above GB
2. DEN, HOU, JAC are 2 SD from the mean, removing them from the mix pushes NO into the AVG range, pushes NE STL KC DET and TB into the BAD range and makes SEA horrid.


SD 3.65%


1. 6/9 ELITE/GOOD S.T. teams made the playoffs, 5/9 ELITE/GOOD OFF teams made the playoffs, and 6/9 ELITE/GOOD DEF teams made the playoffs.

103 Re: Final 2010 DVOA Ratings

I wonder if the NE offensive performances against GB and Pitt (and the NYJ and Chi) show that great offense beats good to great defense in the modern NFL. If so, as much of a cliche as it is, to beat NE you must defeat them in a high scoring game.

Reminds me a bit of UNLV-Indiana in the 1987 final 4.

105 Re: Final 2010 DVOA Ratings

On the contrary, I think Green Bay had the formula. Keep converting 3rd downs and keep the clock running. Try to get a 2-1 possession edge. If a team can manage that, score touchdowns instead of field goals, and not turn the ball over, they stand a good shot of winning. Kansas City is obviously the weakest team on the AFC side of the tournament, but they could possibly follow the blueprint to a 21-17 win.

137 Re: Final 2010 DVOA Ratings

Green Bay did a good job of getting extra possesions by recovering an onside kick, and giving up a kickreturn (almost) TD, and a pick 6. Thats all the TOP difference was.

When you have twice as many actual drives as your opponent, having twice as much TOP isn't surprising.

140 Re: Final 2010 DVOA Ratings

Well, that's a bit dismissive. Sure, the Pats had a 1:09 drive after the Connolly kick return. On the other hand, the Packers drive that ended in the TD pass to Jones took less time than that. The Packers got the extra possession on the onside kick, but what matters is what they did with those possessions, which was to convert 1st downs and keep Brady on the sideline. Green Bay had 4 time-consuming drives that ended with points, plus another one at the end when they were (wrongheadedly) trying to kill clock:

11 plays, 6:15 elapsed, 3 first down conversions
13 plays, 6:26, 5
13, 6:56, 5
11, 5:05, 4
12, 4:22, 4

The Patriots' saving grace was that they forced GB to kick two field goals at the end of those drives. The huge plays were the defense holding GB to field goals after getting to 1st and goal from the 2 and 1st and 10 from the 11.

117 Re: Final 2010 DVOA Ratings

NE didn't have great offense against GB. They got one defensive TD, and another TD needed only a 4 yard drive (requiring all 3 downs) after the Connolly kick return.

GB did a good job pressuring Brady all night. If Woodson just hangs on to a routine INT it wipes out a TD and NE's best drive of the game.

201 Re: Final 2010 DVOA Ratings

NE's big problem was the DL and LB corps were terribly depleted for that game allowing GB easily run the ball.

This is still an issue, although the return of Spikes helps a lot.

125 Re: Final 2010 DVOA Ratings

I don't think 2 games from a season allow any firm conclusion. My impression is that when elite offense and elite defense meet, it could just as easily be a near shut-out as a blowout or an even match. Weak offense versus weak defense seems more often to be an even match-- is there any study to confirm or refute this?

GB's defense did plenty well enough against the NE offense for NE to be defeated. PIT's defense will have to be better if they meet NE again, but PIT is capable of scoring plenty of points in a hurry, as they've done several times this season. If they'd recovered either of two onside kicks against NE, they would have been driving for a potential game-winning TD rather than just driving to soften the final score.

I think NE comes out with a firm lead in DVOA because they are that good and managed to win their regular season games against top opponents, not because elite offense is any kind of silver bullet against other NFL strategies. After all, NE lost 33-14 last year in the playoffs.

138 Re: Final 2010 DVOA Ratings

There is a feeling among Steelers fans that the Patriots offense is a nightmare matchup for the Steelers defense. The priorities of the Steelers defense are, in order: Stop the run, don't give up big plays, tackle the receivers after short receptions. The attitude is that if you can drive down the field on nothing but short passes, then we'll let you, but eventually we will get a sack, turnover, or you'll just drop some passes. The Patriots offense appears to be designed to drive down the field with short passes, and they are good enough to never do the bad things.

In the matchup earlier this year, the Steelers offense started slow, so they got behind, and I think the defense had a bad day tackling -- obviously the Patriots played really well, too, and are just a better team generally. Things could work out differently in a rematch, but I think it would take the Patriots having a bit of an off day for the Steelers to win.

151 Re: Final 2010 DVOA Ratings

Yeah Tom Brady is giving a lot of defenses fits these days. The Bears have similar priorities (although go about them complete differently), and they got lit up worse than they have in years.

Earlier this year, I thought the key was blitzing Brady, force him to make quick decisions and hope you get lucky with zone coverage behind the blitz. Now I'm not so sure. He has a complete grasp on what he's trying to do and how the defense is trying to stop him.

167 Re: Final 2010 DVOA Ratings

A glimmer of hope for PIT is that LeBeau revamped the pass D setup after the NE game, to cover short routes more aggressively at the cost of giving up more big plays. This has led to a fairly dramatic improvement in the overall pass D and, just as important, the ability of the D to get off the field. (Of course the flip side of the coin is that the NE D has also improved.)

96 Re: Final 2010 DVOA Ratings


What does this comment mean?

"But since standard deviation for DVOA is a lot higher these days, the 2010 Steelers end up with a much better total DVOA than the 1994 Steelers."

119 Re: Final 2010 DVOA Ratings

It means a team has to be better now to have a #1 defense and #5 offense than it did in 1994. According to DVOA, there was more parity in 1994, whereas now, the teams' values are more spread out. Standard deviation gives a single measure of how spread out data is, when you expect the data roughly follows a bell curve.

98 Pats impact on Strength of Schedule

The Patriots high DVOA this year also seems to have had a big impact on the schedule strength ratings of the entire league. As a telling anecdote, the Pats are the only team in the top 11 of Schedule Strength that did not play against the Patriots. Sure there are confounding factors in some cases like the AFC north teams also having to play BAL/PIT twice each which certainly contributes to why all those teams are in the top 10 in schedule strength. But when even teams like Detroit, Indy, and Minn are in the top 11 the schedule strength contribution from playing the pats has to be a significant factor.

101 Re: Final 2010 DVOA Ratings

Ford did take a snap out of a wildcat-type package on 3rd and goal in one of the Raiders' losses to a Florida team, I forget which one. But it was his least successful rush of the year. None of his runs were designed passes.

104 Re: Final 2010 DVOA Ratings

Tampa jumped 15 spots from last year, Arizona dropped 18. Hooray and/or ouch. Let's call this "The Freeman Improvement And/Or Lack Of Kurt Warner Index".

106 Re: Final 2010 DVOA Ratings

The Pats alltime great DVOA is based in significant part on turnovers. An almost supernatural lack of turnovers makes their great offense super great. And a very high number of forced turnovers makes their otherwise horrid defense passable. (If you look at their defensive drive stats, they are 31st in yards per drive and 31st in defensive success rate. But, of course, turnovers matter a lot.) I do think that the turnover numbers on both sides of the ball are outliers to a certain extent -- particularly on offense. The risk for the Pats (and hope for fans of the other teams) is that the Pats regress toward the mean in terms of turnovers and/or hit a streak of bad turnover luck. Seems to be the only way they're losing.

141 Re: Final 2010 DVOA Ratings

I'm not buying the "outliers" comment. I'll agree that NE is in rarely charted territory, but it isn't like the ball's been on the ground 2-3 times a game and they keep snatching it up. Woodhead and Benny have fummbled a combined 1 time all season. And Brady definitely has had some luck with dropped ints, but no more than every other QB.

I'll admit that numbers like the 7 TOs they got from Buffalo and 4 from Chicago probably are higher than I expect in the playoffs, but only a handful of teams have turned it over once, so they clearly have a hand in opposing team's mistakes.

FTR, here are the teams who only lost the ball one time or less, with Baltimore and the Jets being the goose eggs.

NYJ (round one)
Green Bay

So, they beat 3 playoff teams without a significant TO advantage, and no advantage at all against Baltimore.

I suppose you could also say that NE would be in trouble if they turn it over multiple times and don't get any back, but to this I ask: What team isn't in trouble in that scenario? And why would NE be a more likely victim than any other team?

145 Re: Final 2010 DVOA Ratings

"And why would NE be a more likely victim than any other team?"

B/c if you flip a coin 20 times in a row and it comes up heads, the next time it's got to come up tails.

146 Re: Final 2010 DVOA Ratings

If I lose to my brother in 1:1 basketball 10 times in a row, the odds of him beating me again are significantly different than a similar coin example.

148 Re: Final 2010 DVOA Ratings

They are? I thought fumble recoveries were random. Seems to me that interceptions and fumbles are a lot more consistent and I've read nothing on FO to suggest otherwise.

Feel free to point me in the right direction if I am mistaken.

150 Re: Final 2010 DVOA Ratings

Getting interceptions and causing fumbles correlate to a certain extent from year to year and so are, to some extent, a repeatable skill whereas fumble recovery rates do not correlate at all. That said, just because there is some year-to-year correlation in causing interceptions and causing fumbles doesn't mean it is 100% skill and 0% luck or anything close to it. Moreover, even unquestionably skill-based performance will regress towards the mean if it is a far enough outlier. FO pointed this out with respect to the historic year Revis had in pass coverage last year. Brady's historic no-interception streak is like any other record -- by definition an outlier.

154 Re: Final 2010 DVOA Ratings

The streak is absolutely an outlier, but IIRC Brady has one of the lowest playoff interception rates in the history of the NFL. And BJGE has never fumbled, let alone lost a fumble, since high school.

At what point do we start to credit the player? Or are turnovers complete luck?

The luck angle strikes me as a "any given Sunday" storyline which limits discussion. If any team can lose to any other at any time, why even bother analyzing anything?

152 Re: Final 2010 DVOA Ratings

Gambler's fallacy. If a flipped coin comes up heads 20 times in a row, or 100 times in a row, it's got a 50/50 shot of being heads on the next flip. Assuming it's not a two-headed coin, of course.

155 Re: Final 2010 DVOA Ratings

No, if a coin comes up heads 100 times in a row, its coming up heads the next time because there's basically no chance its a fair coin at this point.

When your results are this far from the assumption, your assumption is wrong. In this case, the assumption that the Patriots turn the ball over at league average rates (and have been getting lucky) is the assumption.

159 Re: Final 2010 DVOA Ratings

"No, if a coin comes up heads 100 times in a row, its coming up heads the next time because there's basically no chance its a fair coin at this point."

There are a number of cases where that's not true. For example, if you tossed a fair coin a billion times, the odds that there's a 100-head streak in there somewhere are pretty good.

That said, it would probably take years of game-charting results to determine if the lack of fumbles is due to some innate talent of the ballcarriers (they don't fumble when getting hit in approximately the same way), the offensive line (they protect the ball from fumble-inducing hits), a slate of weak defenses ... and you'd still have trouble telling that from a random slick patch of turf causing the defender to not strip-sack Brady when he had the chance.

161 Re: Final 2010 DVOA Ratings

That's true--in fact it's awfully close to 100%.

But what's the chance that a fair coin comes up heads for the first 100 throws?

And it's even more complicated with football statistics, where the "coin" is constantly trying to improve itself.

157 Re: Final 2010 DVOA Ratings

I think the Pats' turnover ratio is certainly partly due to some luck (i.e. Brady's dropped interceptions), but there are four factors that very clearly contribute to it that are demonstrably NOT luck:

1) Brady. I think everyone agrees he is an above-average QB, and above-average QB's generally have below-average interception rates. Maybe he's been extreme this year, but we certainly would expect him to throw fewer INT's than Joe Average QB.

2) The two RB's are extremely sure handed. Low (or high) tendency to fumble is a demonstrable skill (or lack thereof) of RB's. BJGE hasn't fumbled in a rush from scrimmage EVER (in 3 seasons, one as a full time starter), and Danny Woodhead has fumbled from scrimmage while playing for the Pats once (when his head was pile-driven into frozen turf this past Sunday).

3) Their defense has been playing with a "bend but don't break" philosophy this season, allowing many short and medium plays, but few big plays, and forcing teams to run lots of plays against them. More offense plays against obviously leads to more opportunities for defensive turnovers.

4) A big part of their document 3rd down defensive struggles have been because their defensive players, especially the LB's and the safeties, have been playing a bit of a gambling long as they know there's coverage behind them, they'll risk giving up the 3rd down conversion to try to jump the route. Usually this doesn't work and they give up the 3rd down conversion...but sometimes they get the INT. Hence they've gotten more INT's than in previous years. Interestingly, their CB's don't seem to gamble all that much, which could be why they give up so few big plays.

Granted, there is some randomness in each source of a turnover. But if you take two positive random variables that are nonuniform and weighted towards bigger numbers, and add them to two negative random variables that are weighted towards smaller values, then you'll obvioulsy expect the resulting distribution to be weighted strongly positively.

I'll agree that luck might have moved the Pats out into the very positive tail of that distriubtion instead of the merely fairly positive mean, but a fair part of their turnover ratio has been due to controllable factors.

166 Re: Final 2010 DVOA Ratings

At some point, an essay on "luck" (randomness) would be well warranted. I've had thoughts on it, but have never fully organized them to a significant extent. There are certainly many facets to the issue. For example, fumble recover is generally accepted on this site as being almost purely random in aggregrate, yet situational factors tend to benefit either the offense or defense for a particular fumble. Other elements of the game, which appear to be less random, often have a significant random component to them. For example, and this may not be a great example but serves for illustrative purposes, offensive holding penalties. Offensive holding is a function of the ability of the blocker and the rusher; however, whether or not the call is made is affected by position of the referee, his assessment, and his decisions. These factors are essentially random from the standpoint that they are not predictable. Additonally, there is what I call temporal randomness, which refers to the fact that there is no way to predict at what point or situation in the game the call will occur.

I think in #1 you begin to hint at the nature of what we're talking about. With interceptions, there are certain random factors that contribute to an interception, and Brady has a skill set that either reduces the effects of these factors or manages them better. Same for #2.

RE: #4. I don't know if your analysis is correct or not, but I think it was an awesome and clear analysis.

122 Re: Final 2010 DVOA Ratings

I'll note again how badly I underappreciated the Patriots offense when they simply controlled the game against the Vikings defense, and I said it was the worst Vikings defensive performance since Denny Green's last year. It just wasn't apparent to me how historically great this group was. It probably has soemthing to do with Brady being the only obvious HOFer in the bunch, and the offensive line being nicked up. There just isn't anything which screams greatness, but they are great nonetheless. I do also think it is underappreciated how having two tight ends who can run, catch, and block is a rare luxury, and creates match up nightmares for defenses.

136 Re: Final 2010 DVOA Ratings

You are rethinking your position too much. NE wasn't nearly as good prior to the Pitt game as they were afterward. Minny was pretty bad in the 2nd half of that game and probably would have allowed 40+ to the NE team that ended the season.

168 Re: Final 2010 DVOA Ratings

In retrospect that was a pretty darn good performance by the Vikings D against NE at least for the first half. The Pats had one scoring drive of 75 yards but that should have ended in an interception when one of the Vikings safeties let an interception go right through his hands.

I used to hate NE - because I thought they won some SuperBowls they shouldn't have and that they were getting more credit than they deserve. But I can't hate Bill Belichick because he's just so smart compared to everyone else. He's 14-2 this year coming off some ridiculous string of double digit wins and he has 8 picks in the first 4 rounds - 2 in each. He has completely revamped the offence with no names and they are the best in DVOA history.

The man is a genius and I respect genius.