Week 16 DVOA Ratings

Week 16 DVOA Ratings

by Aaron Schatz

With one week to go in the 2008 regular season, the team that had the league's best record for most of the season finally moves into the top spot in the Football Outsiders DVOA ratings. Tennessee has been fourth or higher since Week 3, but this is the first time they've been at number one.

The thing people don't realize about the Titans is that they are not just winning with defense. The Titans only rank fifth in defense according to DVOA, but they now have a top ten offense, 12th in passing and sixth in rushing. Their nearest competitors in the AFC, Baltimore and Pittsburgh, have the top two defenses in the league but struggle to throw the ball (or, in the case of Pittsburgh, struggle to stay upright prior to throwing the ball). I actually noticed last week that the Ravens and Steelers had improved their defenses to the point where I would need to start comparing them to the top defenses of the DVOA Era. Then each defense gave up three touchdowns and a field goal on Sunday, so now each defense needs a dominant Week 17 to get into that all-time top ten.

Speaking of dominance... How about those New England Patriots? If Sunday's 47-7 win over the Arizona Sugar Plum Fairies looked like one of the most crushing victories in recent NFL history, that's because it was, in fact, one of the most crushing victories in recent NFL history. Right now, Sunday's Patriots victory ranks as the second-best single-game DVOA performance since 1995. The only team with a higher DVOA in one game was the Pittsburgh Steelers, who rudely welcomed the expansion Cleveland Browns (v2.0) to the NFL with a 43-0 shellacking back in Week 1 of 1999.

Best Single-Game DVOA, 1995-2008
Team vs. Year Week DVOA Score
PIT CLE 1999 1 143.1% 43-0
NE ARI 2008 16 142.3% 47-7
KC CIN 2005 17 141.4% 37-3
DEN CAR 1997 11 140.7% 34-0
OAK TB 1999 15 140.5% 45-0
KC ARI 2002 13 136.2% 49-0
ATL CAR 2002 12 133.8% 41-0
SF NO 2001 17 131.8% 38-0
PIT CLE 2005 16 127.5% 41-0
IND JAC 2000 4 127.1% 43-14

Some of these games, of course, have a bit of an explanation. The 1999 Browns were playing their first game ever (although the 2005 Browns, who lost to the Steelers by roughly the same score, don't have an excuse). The 2005 Bengals were resting their starters against Kansas City. The 2001 Saints weren't resting starters, but I'm guessing they were just mailing it in by the final week, and the 49ers needed a win for playoff position.

It's a bit shocking to see that this year's Patriots, with Tom Brady and roughly half the defensive starters on injured reserve, just had a win that was bigger than the biggest win of last year's perfect regular season. The best that last year's Patriots could do was 114.1% DVOA, in a Week 8 blowout of Washington. Actually, you'll notice that none of the best teams of the past dozen years show up on the list for the biggest wins except for the 1997 Broncos. The best win for the 1996 Packers was 104.1% (34-3 over Tampa Bay in Week 1). The best win for the 2001 Rams was 120.6% (42-10 over Miami in Week 3). The 1999 Rams do come close to the top ten, with 124.3% DVOA when they beat the expansion Browns (yep, them again) in Week 7, 34-3. The 2000 Ravens didn't have a game over 100%, and neither did the 2002 Bucs. The Dungy-era Colts have only gone over 100% once, with 104.2% DVOA in that big Sunday night win over the Eagles in Week 12 of 2006.

Depending on what happens next week, the Patriots could end up with the best game of all-time. Once each season is done, we consider the final rating for each game to be based on the opponent adjustments for the full season. This way, we have the most information possible about the quality of opponents. If Arizona plays well against Seattle next week, the Patriots game could get slightly moved up, past the Steelers-Browns game. (Note that if we compared each game to other games of the DVOA Era based on what opponent adjustments were in each specific week, the Patriots game wouldn't even show up near the top ten. The biggest games would be early in the season, when opponent adjustments aren't full strength. In Week 1, with no adjustments yet, Pittsburgh's DVOA would have been 183.4%.)

This win is big enough to bump the Patriots up five spots to 15th, and between the Arizona blowout and the fact that it has now been three months since the Miami debacle, the Pats now stand in the top 10 for weighted DVOA. This makes them yet another DVOA darling that will probably stay home for January. By the way, check out the colossal difference between DVOA and VOA for the AFC East teams. In particular, check out the difference for the quarterbacks. Remember this when you hear about how your favorite team should sign Matt Cassel as a free agent, or when you hear about what a huge mistake the Packers made by getting rid of Brett Favre.

* * * * *

Speaking of which -- Honestly, how often can one team lose by less than a touchdown? Actually, the record seems to be eight. Since the merger, three teams lost eight games by less than a touchdown: the 2001 Panthers, the 1993 Patriots, and the 1984 Browns. The Packers have now lost seven games by four points or less.

You won't be surprised to learn that the Packers have the biggest difference in the NFL between their actual wins and their Pythagorean wins (explained here, for those who don't know the concept). You don't see a lot of teams that go 5-10 despite outscoring their opponents. Right now, the Packers are on pace to underachieve their Pythagorean projection by .213, which would be more than any team since 1981. Standing in their way... the team with the second-biggest gap this year (.167), the winless Detroit Lions. It's likely that this week's game will move both teams' win totals closer to their Pythagorean projections... unless the Packers actually figure out a way to lose to Detroit. In which case, man, there is just no justice in this universe.

Biggest Pythagorean Underachievers, 1970-2008
1981 NE 2-14 .500 322 370 6.7 .418 -.293 5-4
1971 CIN 4-10 .286 284 265 7.6 .541 -.255 8-6
2008 GB 5-10 .333 388 359 8.2* .546 -.213 --
2001 SD 5-11 .313 332 321 8.3 .520 -.207 8-8
1979 SF 2-14 .125 308 416 5.3 .329 -.204 6-10
1989 CIN 8-8 .500 404 285 11.1 .696 -.196 9-7
1990 SD 6-10 .375 315 281 9.1 .567 -.192 4-12
1987 LARD 5-10 .333 301 289 7.9 .524 -.191 7-9
1977 ATL 7-7 .500 179 129 9.6 .685 -.185 9-7
2004 TB 5-11 .313 301 304 7.9 .494 -.182 11-5
*based on 15 games

By the way, I've added Pythagorean wins to the second table, and the schedule rating is now for the entire season (including this upcoming week).

* * * * *

One more thing to hit on before we get to this week's numbers. From last week's DVOA comments:

BlueStarDude: Aaron, I don't understand why you still haven't discussed the Dallas DVOA with and without Romo. In 2005 when Roethlisberger missed a few games it was a major discussion point. We know that the drop-off from Romo to Brad Johnson was more than the drop-off from Big Ben to Maddox, so I don't get it. This is asked about in the comments every week. What gives?

First of all, as somebody mentioned a few comments down, the main reason I had not answered this question is that I don't have time to go through all the comments anymore. The best way to ask a question is definitely to send it to the mailbag, in which case Bill Barnwell or I will probably answer it. However, last week in a South Jersey motel I did have time to read some comments and I noticed this one, so let's take a look at the answer.

Tony Romo was out Weeks 7-9. No surprise, Weeks 7 and 9 were two of the Cowboys' three worst offensive games of the year. However, Week 7 was also the Cowboys' worst defensive game of the year. Can we blame the Romo injury for the bad performance by the defense? Were they forced to stay on the field way too long because their offense couldn't sustain drives? Subjectively, perhaps, but the goal of DVOA is to account for field position (if not yet time of possession) and properly separate the offense from the defense. So if we're going to ask how the Cowboys would be without those three games, I think it makes sense to only remove the offensive plays. We will take out both rushing and passing, though, because not having Romo affected the way the other team played the Cowboys and was partly responsible for a reduced ground game.

Without those three games, the Dallas offensive DVOA goes from 9.9%, which is 13th, to 21.2%, which would narrowly edge out Denver to rank third in the NFL. That's a significant jump. Their total DVOA would go to 18.3%, so they would jump from 16th in the NFL to seventh.

However, we should also note that two of the Cowboys' four best offensive games were Week 1 and Week 2. Therefore, removing the Brad Johnson games doesn't affect weighted DVOA quite as much as it affects total DVOA. The Cowboys' weighted offense would go from 9.9% (13th) to 17.3% (tied for sixth). In total weighted DVOA, they would go from 8.0% to 15.4% -- which actually only passes two teams, Minnesota and New England.

Of course, Johnson's bad play also has ramifications for the ratings of the Cowboys' opponents in those weeks. Without their defensive performance in Week 7, St. Louis would drop back into last place overall and would go from fifth to second on the list of the worst DVOA defenses ever. The Giants (Week 9) would drop from eighth to 13th in defense. Since the Cowboys were reasonable in Week 8, Tampa Bay wouldn't really be affected.

* * * * *

Housekeeping: Will Carroll is off again this week for the holidays, so there's no Black and Blue Report, but we'll link to his SI.com column in Extra Points when that goes up. All team and individual stats pages are now updated except ALY/ASR, which should be updated shortly. Playoff odds are now updated and you might notice that Sunday's San Diego win and Denver loss changed each team's chances of making the postseason by an astonishing 65.1 percent. Talk about your big losses.

* * * * *

These are the Football Outsiders team efficiency ratings through 16 weeks of 2008, measured by our proprietary Defense-adjusted Value Over Average (DVOA) system that breaks down every single play and compares a team's performance to the league average based on situation in order to determine value over average. (Explained further here.)

OFFENSE and DEFENSE DVOA are adjusted to consider all fumbles, kept or lost, as equal value. These ratings also include opponent adjustments. 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 the team is playing right now.

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

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

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

1 TEN 31.1% 3 30.8% 1 13-2 10.6% 10 -19.0% 5 1.5% 12
2 PHI 30.0% 1 28.2% 2 8-6-1 9.4% 14 -19.4% 4 1.3% 13
3 NYG 27.6% 4 25.1% 4 12-3 21.5% 2 -3.6% 8 2.4% 10
4 BAL 27.1% 2 25.6% 3 10-5 3.1% 19 -24.2% 1 -0.1% 19
5 PIT 21.5% 5 22.5% 6 11-4 -1.1% 21 -23.7% 2 -1.1% 23
6 CAR 20.2% 6 23.4% 5 11-4 17.1% 7 -0.8% 9 2.3% 11
7 NO 13.0% 10 18.4% 7 8-7 22.5% 1 9.4% 21 -0.1% 18
8 TB 12.1% 7 4.5% 17 9-6 0.5% 20 -11.9% 6 -0.2% 20
9 IND 11.8% 8 16.1% 8 11-4 19.6% 4 5.7% 16 -2.2% 24
10 ATL 11.3% 12 15.5% 9 10-5 17.6% 5 9.0% 19 2.7% 9
11 GB 9.0% 11 7.9% 14 5-10 10.4% 12 0.8% 12 -0.6% 22
12 SD 8.4% 19 7.4% 15 7-8 17.3% 6 10.1% 22 1.1% 15
13 MIN 8.2% 9 8.7% 11 9-6 -4.2% 23 -19.8% 3 -7.4% 32
14 CHI 7.6% 13 2.9% 18 9-6 -4.8% 24 -8.5% 7 3.9% 5
15 NE 7.3% 20 11.7% 10 10-5 14.1% 9 10.6% 24 3.8% 6
16 DAL 7.0% 14 8.0% 12 9-6 9.9% 13 -0.4% 10 -3.4% 27
17 MIA 6.8% 15 8.0% 13 10-5 17.0% 8 5.1% 14 -5.1% 29
18 WAS 5.3% 16 -1.5% 19 8-7 7.7% 16 0.1% 11 -2.3% 25
19 NYJ 3.9% 17 5.2% 16 9-6 4.5% 18 4.1% 13 3.5% 7
20 ARI -3.4% 18 -9.9% 23 8-7 10.6% 11 10.4% 23 -3.7% 28
21 JAC -6.0% 22 -7.1% 21 5-10 7.6% 17 13.1% 25 -0.5% 21
22 DEN -6.0% 23 -8.1% 22 8-7 21.1% 3 21.7% 30 -5.4% 30
23 HOU -8.9% 21 -5.1% 20 7-8 8.5% 15 18.1% 29 0.6% 16
24 BUF -13.3% 25 -21.1% 28 7-8 -5.7% 25 14.3% 26 6.7% 1
25 CLE -18.3% 24 -18.7% 26 4-11 -15.7% 27 6.9% 17 4.3% 4
26 SF -21.8% 26 -25.4% 29 6-9 -17.5% 29 9.0% 20 4.7% 3
27 SEA -21.9% 27 -18.3% 25 4-11 -9.2% 26 15.6% 27 2.9% 8
28 OAK -22.8% 30 -26.2% 30 4-11 -20.5% 31 7.7% 18 5.4% 2
29 CIN -25.4% 28 -19.0% 27 3-11-1 -17.6% 30 5.2% 15 -2.6% 26
30 KC -26.7% 29 -17.8% 24 2-13 -4.1% 22 16.4% 28 -6.2% 31
31 STL -45.6% 32 -41.0% 31 2-13 -22.3% 32 23.4% 31 0.2% 17
32 DET -48.0% 31 -45.3% 32 0-15 -15.9% 28 33.3% 32 1.2% 14

  • NON-ADJUSTED TOTAL DVOA gives performance without adjustments for schedule strength, fumble recovery luck, and weather/altitude on special teams.
  • ESTIMATED WINS uses a statistic known as "Forest Index" that emphasizes consistency as well as DVOA in the most important specific situations: red zone defense, first quarter offense, and performance in the second half when the score is close.  It then projects a number of wins adjusted to a league-average schedule and a league-average rate of recovering fumbles. Teams that have had their bye week are projected as if they had played one game per week.
  • 2008 SCHEDULE lists average DVOA of opponents played this season (including this upcoming week), ranked from hardest schedule (#1, most positive) to easiest schedule (#32, most negative). It is not adjusted for which games are home or road.
  • PYTHAGOREAN WINS represent the number of wins projected from the team's points scored and allowed, as described in this article.
  • 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 2008
1 TEN 31.1% 13-2 32.4% 11.2 1 -3.0% 26 11.9 1 9.8% 5
2 PHI 30.0% 8-6-1 27.7% 10.5 4 0.7% 15 9.8 5 14.2% 9
3 NYG 27.6% 12-3 27.3% 10.5 3 1.6% 14 10.8 3 22.0% 27
4 BAL 27.1% 10-5 26.6% 10.7 2 2.3% 8 10.9 2 16.4% 15
5 PIT 21.5% 11-4 16.9% 10.0 5 5.0% 4 10.4 4 7.7% 3
6 CAR 20.2% 11-4 18.5% 9.6 6 1.7% 12 9.6 7 15.8% 14
7 NO 13.0% 8-7 14.2% 9.1 8 0.0% 21 9.1 9 13.5% 7
8 TB 12.1% 9-6 9.9% 8.4 13 0.2% 20 8.8 13 19.1% 23
9 IND 11.8% 11-4 14.2% 9.4 7 1.9% 11 9.0 10 17.1% 19
10 ATL 11.3% 10-5 8.5% 8.8 11 0.3% 19 9.2 8 21.0% 26
11 GB 9.0% 5-10 3.8% 8.2 15 0.3% 18 8.2 17 15.6% 12
12 SD 8.4% 7-8 9.7% 8.8 10 -1.0% 23 9.0 11 19.3% 24
13 MIN 8.2% 9-6 5.6% 8.1 16 2.9% 6 8.7 14 15.6% 13
14 CHI 7.6% 9-6 4.2% 8.9 9 0.5% 16 8.3 16 10.2% 6
15 NE 7.3% 10-5 11.1% 8.1 17 -7.0% 29 9.7 6 31.8% 31
16 DAL 7.0% 9-6 7.0% 7.9 19 3.7% 5 8.4 15 31.1% 30
RANK 2008
17 MIA 6.8% 10-5 21.0% 8.4 12 -7.9% 30 8.1 18 17.3% 20
18 WAS 5.3% 8-7 1.8% 7.9 18 0.4% 17 6.5 22 5.8% 1
19 NYJ 3.9% 9-6 12.4% 8.3 14 -8.3% 32 8.9 12 16.8% 17
20 ARI -3.4% 8-7 -0.7% 6.9 21 -4.7% 27 7.2 20 30.0% 29
21 JAC -6.0% 5-10 -5.6% 7.3 20 1.9% 10 6.3 25 7.8% 4
22 DEN -6.0% 8-7 -5.4% 6.6 22 -2.8% 24 6.4 23 20.6% 25
23 HOU -8.9% 7-8 -12.0% 5.8 25 2.5% 7 6.6 21 14.2% 10
24 BUF -13.3% 7-8 -5.0% 6.0 23 -8.0% 31 7.7 19 13.8% 8
25 CLE -18.3% 4-11 -24.3% 5.8 24 7.8% 2 4.8 27 17.0% 18
26 SF -21.8% 6-9 -14.7% 5.5 26 -6.4% 28 6.3 24 6.8% 2
27 SEA -21.9% 4-11 -19.3% 5.2 27 -2.9% 25 5.2 26 18.2% 22
28 OAK -22.8% 4-11 -24.1% 5.0 28 1.9% 9 3.8 29 33.2% 32
29 CIN -25.4% 3-11-1 -34.4% 4.3 30 8.5% 1 2.7 30 17.4% 21
30 KC -26.7% 2-13 -27.3% 4.3 29 1.6% 13 4.2 28 15.1% 11
31 STL -45.6% 2-13 -43.6% 2.0 32 -0.1% 22 2.2 32 25.2% 28
32 DET -48.0% 0-15 -50.2% 2.5 31 7.3% 3 2.5 31 16.6% 16

Worst DVOA Ever Watch

Yes, we now have three different teams from 2008 listed among the ten worst defensive DVOA ratings since 1995. At least one of those teams can play offense. The Lions are now on another planet when it comes to bad defense. The difference between the 2008 Detroit Lions and the second-worst defense through Week 16 (the 2000 Vikings) is equal to the difference between the Vikings and the 14th worst defense, which belongs to the 2004 Rams. Our playoff odds report now gives Detroit a 95.4 percent chance of going 0-16.

2005 SF -55.6%   2008 DET 33.3%
2004 SF -49.7%   2000 MIN 26.5%
2008 DET -48.0%   2000 ARI 24.6%
2008 STL -45.6%   2004 SF 23.9%
2000 CLE -44.8%   2008 STL 23.4%
2000 ARI -43.5%   2004 MIN 23.1%
2000 CIN -42.6%   2005 HOU 22.6%
1999 CLE -42.3%   1998 CIN 21.8%
2003 ARI -41.2%   2008 DEN 21.7%
2002 ARI -40.0%   2002 DET 21.0%

Happy whatever holiday you celebrate, and good luck to your favorite team in Week 17.


65 comments, Last at 30 Dec 2008, 12:13pm

1 Re: Week 16 DVOA Ratings

I think the "Last Year" column should read "Last Week".

2 Re: Week 16 DVOA Ratings

hope this quiets down the Cowboys fans' angst. As a Cowboy fan, I don't care so much about our DVOA as about actually getting in the playoffs. If we do, we're talented enough to follow the Colts' and Giants' footsteps.

3 Re: Week 16 DVOA Ratings

We're also dumb enough from top to bottom to lay a complete egg on the field on Sunday and have Ed Werder pee his pants in glee.

4 Re: Week 16 DVOA Ratings

I wonder how bad things are going to be for the Cowboys players this Sunday. Not only is it in Philly (always a - ahem - 'spirited' fan base), but the Eagles players and fans will also be out to avenge the early-season loss, as well as fight for their own playoff berth (subject to outcomes of earlier games).

I don't want to know what they'll be throwing at the Cowboys bench...

61 Re: Week 16 DVOA Ratings

blinking red noses, maybe?


armchair journeyman quarterback

5 Re: Week 16 DVOA Ratings

All I can say is...I hope the Giants rest their starters. Not that the Vikes don't (obviously) need the practice against top-flight competition.

And the Houston-Chicago game might be worth watching. Looking at their weighted DVOA, they're closely matched. And it's in Houston, where they may not be playing for the postseason, but they'd probably like to earn a .500 season for once.

All in all, it's a welcome change to still have so many meaningful games in Week 17. The only NFC slots that are certain are the #1 and #4 seed.

7 Re: Week 16 DVOA Ratings

NE and SD defense now fixed; that was a problem because of a search-and-replace gone horribly wrong. All stats pages are now updated except OL/DL which will come shortly. Enjoy!

57 Re: Week 16 DVOA Ratings

Thanks for answering the Romo-BJ /Bollinger question. (I actually think it was me who noted a few comments later that if I really cared that much I would email you because you don't usually have time to read through the huge comment streams anymore, which is understandable.)

The other discussion I vaguely recall from 2005 was how Maddox's QBing somehow affected the defense's DVOA. I think there was even some further research that showed similar trends in prior seasons. I don't remember though if there was a solid conclusion reached though.

8 Re: Week 16 DVOA Ratings

Some other "Special Super Bowl" matchups that could happen this year (as seen on the Playoff Odds report):

Pittsburgh vs. Dallas (4th Super Bowl matchup)
New England vs. New York Giants (rematch)
New England vs. Carolina (rematch of a few years ago)
Baltimore vs. New York Giants (rematch)
Atlanta vs. San Diego (Michael Turner bowl)

9 Re: Week 16 DVOA Ratings

Am I the only one sophomoric enough to laugh at the abbreviation for the Los Angeles Raiders?

16 Re: Week 16 DVOA Ratings

Rest assured, you are not alone. In fact, I am thinking of frying something... and cannot imagine why.

With that abbreviation, it's a wonder Seabass did not play for them back then.

10 Re: Week 16 DVOA Ratings

At some point, Aaron is going to have to built in a feature in the DVOA formula that accounts for teams running up the score. NE seems to be consistently overrated ever since they started doing this last year. They haven't been able to run up the score much this year until the game against the cardinals though, so I guess their DVOA is pretty much in line with their ability this year.

My bad if such a "garbage points" feature already exists.

15 Re: Week 16 DVOA Ratings

Um, the Patriots had 31 points at halftime. Those points are decidedly not "running up the score", as they occurred in the first half. The Patriots' first play of the second half was a little screen pass to Moss - definitely not a "running up the score" play - that went for a TD because of terrible Arizona defense. From that point on, the Patriots only scored 3 FG - should they have punted from deep within Arizona territory? If they had gone for it on fourth down those times, even more people would have accused them of "running up the score".

I, too, greatly dislike the Patriots, but they did not run up the score against Arizona.

51 Re: Week 16 DVOA Ratings

Why respond with reasoned arguments? Haters know the facts without needing to be bothered with details.

I have not heard anybody gripe about the Saints "running up the score" by 42-7. As always, the Pats enjoy more scrutiny from people who think that, at a certain point in the game (presumably the first play in the second half?) a Patriot who has a clear path to the end zone should simply step out of bounds instead.

Because somehow that way of humiliating a defense is better than simply playing the game normally.

19 Re: Week 16 DVOA Ratings

I *think* the DVOA calculations throw out "garbage time", but I don't remember what exactly constitutes "garbage time". I think it only applies in the 4th quarter, so a game that turns into a blowout at the 10 minute mark (as the NE-ARI game did).

I don't know why you think NE has been consistenntly overrated since last year. Last year DVOA thought they had the best offense ever (in the DVOA era) and a decent but not great defense, and that they were almost unbeatable. Subjectively, I think this was absolutely true, and I'd be interested to hear why you think saying this was "overrating" them. No one WAS able to beat them until the Superbowl, and only a couple of teams managed to make it close (after they had suffered some injuries). The only game that they won last year that they probably shouldn't have was the Baltimore game, but it was so close it could have gone either way. Even in the Superbowl, I think most people agree that the Patriots, or at least their offensive line, played their crappiest game all season, while the Giants played their absolute best (especially their front four and Eli Manning), and even then the Giants needed a dropped INT and a magical helmet catch in the final two minutes to win the game. If that game is played ten times in 10 parallel universes, the Patriots probably win the majority of them. How was DVOA overrating them?

This year DVOA has said they started out as worst teams in the league, with an average or slightly above average offense and a bad defense, but have been steadily improving all year, primarily due to offensive improvement. Again, I think you'd be hard pressed to find someone who disagrees subjectively with that. In the early weeks of the season they WERE one of the worst eams in the league, but I don't think it's unreasonable to say, on average, that they're roughly the 15th best team in the league and playing the 10th best of any other teams right now... Can you think of more than 10 teams that are playing better football than New England right now?

24 Re: Week 16 DVOA Ratings

I do think that last year's Patriots were probably the best NFL team of my lifetime (since 1981), but saying that the Giants played "their absolute best" is pretty wrong. Their D ... you could make that argument. But if the best their offense could do was 17 points against a mediocre to above average defense ... yeah ...

And 10 teams playing better football than them right now ... would you have said that after the Pittsburgh game? Has that much changed since then?

31 Re: Week 16 DVOA Ratings

The Pittsburgh game was a bit fluky in that the Patriots made a couple of really bad plays that snowballed. Under normal circumstances, if the two teams played another game at a similar performance level, you'd expect something like a 27-17 Pittsburgh win, not a blowout.

35 Re: Week 16 DVOA Ratings

The Pittsburgh game was a weird one. The score makes it look like a blowout, but there are blowouts where one team dominates the entire time, and there are blowouts that happen because one team gets the upper hand in a close game forcing the other team to take risks, which backfire and turn it into a blowout. The NE-PIT game was the latter.

NE was right in that game, going hit for hit with Pittsburgh until the tail end of the 3rd quarter. Pittsburgh had played slightly better, but it was anyone's game--after 40 minutes of football, the score was 13-10 Pittsburgh, and NE had just made a good red zone stop to keep the score close. Then Hobbs got sick, so the backup kick returner came in, muffed the kickoff, and the Steelers got a quick cheap TD. Next NE series, Cassel fumbles on the first play and the Steelers recovers and get a quick FG. Then on the next series, Cassel fumbles AGAIN. In a 2 minute span, bad luck, some select horrible individual failures, and a couple of really goood plays by the Steelers cost the Patriots three possessions and gave the Steelers 10 points. All of a sudden, the Patriots are down by 13 late against the best defense in the league. And from there desparation (two Cassel passes thrown desparately into coverage and intercepted) turned it into a Steelers blowout.

Pittsburgh obviously was better--I think eight times out of ten they win that game--and indeed they're definitely one of the 10 playing better football than the Patriots right now. However, I wouldn't use that one game to say that the Patriots are playing badly right now. They played horribly for 3 crucial minutes in a close game, and lost big, but sandwiched around that game were a close loss, a close win, and two dominating wins. And they've won three straight since then, two games dominantly (albeit against suspect competition).

59 Re: Week 16 DVOA Ratings

The NE defense was ranked 8th last year, but that was probably somewhat artifically inflated because the offense was so good that other teams had to abandon their running game early and allowed the Patriots to key in on the pass.

25 Re: Week 16 DVOA Ratings

"I *think* the DVOA calculations throw out "garbage time", but I don't remember what exactly constitutes "garbage time". I think it only applies in the 4th quarter, so a game that turns into a blowout at the 10 minute mark (as the NE-ARI game did)."

Actually, I believe Aaron has said that removing garbage time plays decreases accuracy. However, garbage time plays are compared to similar plays, so a team running three times up the middle at one yard a pop is not penalized as harshly as if this occurred in the first quarter of a game.

22 Re: Week 16 DVOA Ratings

Oh, one other point. In regard to your comment "they haven't been able to run up the score much this year". The only difference between "running up the score" and "winning (or losing) a shootout" is how good your defense is. The Patriots have actually scored 30 or more points six times this season, but have only "run up the score" (i.e. scored more than 30 points while allowing fewer than, say, 10 or fifteen) twice (Arizona and Denver). They lost one of six those games. Their defense has been so bad that the thinking this year is to take as many points as you can get.

53 Re: Week 16 DVOA Ratings

Indeed. And we are talking about the Arizona freaking Cardinals after all. They could score 3 TDs in five minutes if the defense took a brief nap.

56 Re: Week 16 DVOA Ratings

Any team in the NFL is capable of three TDs in five minutes, which is the problem with whining about running up the score. Heck, the record for single-quarter scoring against Superbowl-quality teams is 35 points. I suspect the regular season record is at least as high.

I propose these guidelines for running-up-the-score whiners:

If it's not a drive that started in the fourth quarter, it's not running up the score. Nobody kneels on a third quarter drive. That would be insulting. Treat every team as if it were capable of one quarter of excellent football. And be humble: no matter how good you think you are, remember that even the best team is capable of one quarter of terrible football.

If it's the first drive for the backup QB, it's not running up the score. The guy needs realistic game experience.

If it's a defensive score, it's not running up the score.

If it's a field goal, it's not running up the score.

If it's conservative, ball-control tactics, it's not running up the score. This includes runs and safe passes. That's how today's teams get first downs and keep the ball.

If the defense brings their safeties up, it's not running up the score to go deep to keep them "honest". It's a two-way street: If you expect me to go easy on you, you shouldn't take unfair advantage.

I'll go further: if the defense blitzes, anything goes. It's not safe for one team to play aggressively while the other team goes easy.

If the previous score was by the losing team, answering that with a TD is not running up the score. It's maintaining the lead.

If the losing team "wins" or "ties" the second half, it's not running up the score. It's delivering a quality product to the fans. Besides, the lead didn't increase, so who's being embarrassed? Can it ever be embarrassing to "lose" the second half by seven points? These are professionals, for Pete's sake.

If the lead is 19 points or less when you score, it's not running up the score. (Two TDs and a last second field goal might be amazing, but it's only one play over a "not-quite safe enough" lead) Some people might argue that the limit should be 16 points. Until last week I might have agreed. Two offensive TDs and a field goal in a couple of minutes seems attainable now, don't you think?

No team is expected to kneel and then punt. You kneel to end the game and win. Otherwise, you run plays that have a chance of getting the first down.

No team is expected to "go easy" on defense. You are never expected to "let" the other team score to avoid hurting their feelings.

11 Re: Week 16 DVOA Ratings

Strange fact

1998 - Gary Anderson misses 38 yarder to seal NFC Champ - Morten Andersen hits 38 yarder in OT
2008 - GB kicker has game sealing 38 yarder blocked, Chic kicks 38 yarder in OT

13 Re: Week 16 DVOA Ratings

So my prediction made 8 weeks back is looking ominously accurate - I predicted the Vikings would go on a 5-1 stretch to get to 8-5 getting our hopes up only to dash them by losing 2 or 3 or their last 3.

After the win in Arz I thought maybe they won't fall to pieces this time around but after watching this weekends loss to Atlanta and GB's loss to Chic I know that once again a Viking team will crumble down the stretch.

14 Re: Week 16 DVOA Ratings

"However, last week in a South Jersey motel" You can't imagine how unlikely it is for the sentence that starts with those words to be continued with a discussion of football.

18 Re: Week 16 DVOA Ratings

Okay--based on Aaron's note about not being able to read the comments much anymore, I may send these DVOA related questions to him.

#1. What is the best DVOA and overall finish in the list for a non-playoff team? (Right now, the Saints & Bucs from the NFCS are 7 & 8, respectively.)

#2. What is the best DVOA and overall finish in the list for a LAST-PLACE team?
(Has to be this year's Saints, right?)

#3. What is the best DVOA for an entire division (at least since the regrouping in 2002)? (All NFCS teams are in the TOP TEN!!!--best division ever?)

#4. When a team's record fairly underachieves its DVOA/Est. Wins/Pyth. Wins, what is the projection for the following year? (Saints & Packers really stand out)
(I would assume that your record being higher than your DVOA means you won some games by fumble luck/random return TD's/stupid coaching by opponent/Ed Hochuli inadvertant whistle/etc.--thus meaning that you aren't as good as your record, and you may have some regression the next year.)

26 Re: Week 16 DVOA Ratings

Regarding #1, Philly would be out right now, and are at #2.

Regarding #2, if Dallas beats Philly and Washington wins, then the answer won't be this year's Saints but rather this year's Eagles.

And as for the best division ever, DVOA notwithstanding, the NFC East will have all four teams at 500 or better, and went 4-0 versus the NFCS this year.

45 Re: Week 16 DVOA Ratings

And as for the best division ever, DVOA notwithstanding, the NFC East will have all four teams at 500 or better, and went 4-0 versus the NFCS this year.

The NFC South will also have each team at least .500. All of the games between the 2 divisions were at the NFCE stadium, and only the Falcons-Eagles game was decided by more than a touchdown.

The two divisions were awfully close this year.

54 Re: Week 16 DVOA Ratings

True, the divisions are awfully close this year, and I already knew that all NFCS teams were at 500 or better (I brought up the NFCE and that just to point out that they did too).

I'd say the NFCS has been more consistent and balanced, even if I think that overall the NFCE has been slightly better. Total division DVOA: NFCE 69.9, NFCS 56.6.

But both divisions have been money this year.

20 Re: Week 16 DVOA Ratings

Wow, look at the top 10 teams by DVOA, and then look at the Weighted DVOA rankings for those ten teams. Hint: one of them sticks out like a sore thumb, and is clearly in a nosedive. Harsh!

21 Re: Week 16 DVOA Ratings

"Our playoff odds report now gives Detroit a 95.4 percent chance of going 0-16."

Huh? This is woefully out of line with both the Vegas moneylines and plain common sense. 80 percent chance, sure. 95.4, no way. There appears to be something wrong with your simulation.

23 Re: Week 16 DVOA Ratings

Doesn't this just mean that there is a 95.4% chance that Green Bay will beat Detroit? GB's wDVOA is +7.9%, Detroit's is -45.3%, and they're playing at Lambeau, right?

I agree, 5.6% does seem a little low for an NFL calibur team to beat another NFL team, but not crazy...after all, does Detroit really qualify as an NFL calibur team?

Incidentally, does anyone know what IS the formula for predicting the probability one team will beat another based on their DVOA? I remember someone worked it out once, and it must be used somewhere in the playoff odds simulation...

36 Re: Week 16 DVOA Ratings

Well, as mentioned, Green Bay is not at all your typical 5-10 team, and some people probably think they are. FO stats see Green Bay as an average team with excruciating bad luck. I consider this a reasonable claim. You can feel free to argue with that if you like, though.

If you accept that the Packers are an average team, the question becomes "what are the Lions' chances of defeating an average team on the road?" The Lions have lost 22 of their last 23 games, with the lone win coming against the Chiefs at home last season. That's a winning percentage of .043. This season, they're .000, and in road games they're .000 for the last season and a half. Given that, I don't think the playoff odds estimate is being too unfair to Detroit.

I do think that, like those late-season games for New England last year, this game will have some unquantifiable emotional aspects because Detroit is chasing history. I wouldn't be as certain as the playoff odds simulator, but I think 90% or even 93% is reasonable.

55 Re: Week 16 DVOA Ratings

Yeah, 95.4% must be an error. I just don't see that 4.6% chance of a win.

Did you watch the meltdown in Week 2? Indoors? Have you seen what passes for the Lions' defense these days?

I don't know what you consider common sense to be. To me, it's writing "0-16" in pen and moving on to 2009.

And waiting to find out who the Lions' Week 11 opponent will be next season.

27 Re: Week 16 DVOA Ratings

Green Bay's numbers don't add up. They say TDVOA 9%, OFF 1.4%, DEF 0.8%, ST -0.6%. Obviously these don't add up to 9%.

28 Re: Week 16 DVOA Ratings

Green Bay's numbers don't add up. They say TDVOA 9%, OFF 1.4%, DEF 0.8%, ST -0.6%. Obviously these don't add up to 9%.

30 Re: Week 16 DVOA Ratings

RyanC is right: Based on the rankings, it looks like GB's offensive DVOA should be 10.4% instead of 1.4%...

33 Re: Week 16 DVOA Ratings

excuse the bitterness, but i am so digusted by the play of the Cardinals since they charged me for fucking playoff tickets. damn sumuvabitches....

Mailing comes in right before week 12, these jokers are 5th in dvoa. bastards...

34 Re: Week 16 DVOA Ratings

It's likely that this week's game will move both teams' win totals closer to their Pythagorean projections... unless the Packers actually figure out a way to lose to Detroit.

That can't possibly be right. Both Green Bay and Detroit are currently below their Pythagorean predictions (I don't think even Pythagoras would predict Detroit with negative wins), so regardless of who wins or loses, only one team will move closer to their projection.

42 Re: Week 16 DVOA Ratings

Current situation:
GB Actual Wins < GB Pythag Wins
DET Actual Wins < DET Pythag Wins

If GB defeats DET:
GB Actual Wins increase; GB Pythag Wins increase, but not by more than one full win (difference is now less)
DET Actual Wins stays same; DET Pythag Wins decrease (difference is now less)

If DET defeats GB:
GB Actual Wins stays same; GB Pythang Wins decrease (difference is now less)
DET Actual Wins increase; DET Pythag Wins increase, but not by more than one full win (difference is now less)

Actually, no matter the outcome, both teams move towards their pythagorean wins.

49 Re: Week 16 DVOA Ratings

You're both wrong. You can absolutely gain more than one pythag win if you really destroy the other team. Also, Detroit's pythag would improve in a close game, as long as they score some points - even 14-10 would do it.

58 Re: Week 16 DVOA Ratings

You're right; if the Packers win by a score more dominant that 28-0, their Pythagorean wins will increase by more than one win. Likewise, if the Lions win by a score more dominant than 44-0, their Pythagorean wins will increase by more than one.

Additionally, a close Lions win will result in the Packers moving further from their Pythagorean win percentage.

So, FO was right in the sense that, if the Packers win, both teams will move towards their Pythagorean win percentage. It would take a total Packer blowout (not out of the question, but they'd bascially need to hold the Lions to under 13 points and score nearly 40 themselves) to contradict Aaron's statement.

37 Re: Week 16 DVOA Ratings

Interesting matchup not mentioned:

Dallas vs. Miami (The Parcells Reclamation Bowl, or possibly the Dallas Versus Team Made Up of 30% of 2007 Cowboys Players/Coaches/Executives Bowl).

39 Re: Week 16 DVOA Ratings

Playoff Odds page is broken - Indy Colts are locked in the 5 seed already but says
79.8% 20.3% for 5 and 6 seed respectively.

As such I doubt the #'s for Baltimore and the other teams competing for the 6 seed are accurate ?

It also says, for example, that Minn is 78% to win Division and be 3 seed, but every site I have seen says they can somehow get a WC as well under some situations. Surely those get at least a 0.1% occurrence?

43 Re: Week 16 DVOA Ratings

Re: Colts & the Five Seed
I know in the past, tiebreakers were not considered. It appears they are still not considered.

Re: Minnesota & the Wild Card
The only way for Minnesota to get the wild card is for them to tie the Giants and for (1) the Bears to win, (2) the Cowboys to lose, and (3) the Buccaneers to lose. That way, the Bears win the division, and the Vikings tie with the Eagles for the last wild card; the Vikings would be ahead of the Eagles based on conference record, I believe. I imagine the Playoff Odds Report does not take a chance of a tie into account; how would you model that?

47 Re: Week 16 DVOA Ratings

how is tenn offence ranked 10th when on offence drive chart they are 24th?

50 Re: Week 16 DVOA Ratings

Several reasons. First, drive stats and DVOA use somewhat different methodologies, so you shouldn't expect a perfect correlation between the two. Furthermore, drive stats are not adjusted for strength of schedule or opponent. Also, you only cited their yards/drive; other drive stats for TEN are higher than 24th.

60 Re: Week 16 DVOA Ratings

So this comment got me thinking.

Speaking of which -- Honestly, how often can one team lose by less than a touchdown? Actually, the record seems to be eight. Since the merger, three teams lost eight games by less than a touchdown: the 2001 Panthers, the 1993 Patriots, and the 1984 Browns. The Packers have now lost seven games by four points or less.

1. What is the best DVOA for a team with a losing record?
2. What is the best DVOA for a team with 10 or more losses?
3. Is Ryan Grant the worst 1,000 rusher of the DVOA/DYAR era?
4. What is the worst record by a team that actually outscored it's opponents for the season?

I'm partially trying to judge just how often you get a team like 08 Green Bay or 01 San Diego (6.8% DVOA) in asking some of these questions.

And since as I'm typing this up I suddenly find myself with a bit more time on my hands than expected, I'll answer these myself.

4. What is the worst record by a team that actually outscored it's opponents for the season?
The Biggest Pythagorean Underachievers, 1970-2008 chart shows the 5-11 2001 edition of the Chargers and the 4-10 1971 edition of the Bengals outscoring their opponents, so I know that this years version of the Packers can't end up as the worst record (by percentage or by total losses) since they will end 5-11, 6-10, or 5-10-1. I can't really see how a team with a worse record that still outscored their opponents wouldn't already be on that chart, but I haven't run test numbers to verify.

1. What is the best DVOA for a team with a losing record?
Forgive the table format the code tags give a fixed width font but it still strips extra spaces so I'm doing some cheating to keep it formatted. :) I did a quick grab of the best losing team from each year just to get a quicker reference. I thought about only pulling them in if they had positive DVOA, but it was easier to script it this way. I'm including the next year column so that you can see what kind of affect under-performing has on performance next year. There have been some very good 8-8 teams as well but I'm looking for losers.

Year . Rank . Team . .Total . Last . Next . NON-ADJ . W-L
. . . . . . . . . . . DVOA . .Year . Year . TOT VOA

* I'm putting SD on the table since they could end up with a losing record and a better DVOA than GB.

So that seems to answer my question. The 2004 Chiefs had a losing record and still had an 11.4% DVOA. It's possible that GB or SD (though picking up 3% DVOA in a loss would be tricky for SD, GB can win and pick up 2.4% more easily to tie).

The rest of the table is not too surprising. You frequently get an average or just above average team with a losing record, though you are looking at 7-9 teams most of the time. I'm sure the worst DVOA with a winning record table would have a lot of 9-7 teams on it in the same situation.

2. What is the best DVOA for a team with 10 or more losses?
I think it would be better to just see how many teams have had 10 or more losses and still stayed positive on DVOA.
1997 . DAL . 0.6% . 6-10
1999 . DEN . 1.0% . 6-10
2003 . JAC . 1.1% . 5-11
2008 . GB . 9.0% . 5/6 - 11/10

So there have been 3 teams that managed to lose 10 or more games and still be a bit better than average, GB is on pace to be significantly better than these other teams though.

3. Is Ryan Grant the worst 1,000 rusher of the DVOA/DYAR era?
Heh, this one surprised me a bit. Grant isn't even close. I initially thought that having negative DYAR would be uncommon for a 1000 yard rusher. Turns out that Grant is actually interesting because he is the only 1000+ yard rusher this year with a negative DYAR and he is barely negative. 1996 and 2002 are the only years where there wasn't at least one 1,000 yard rusher with a negative DYAR. Most years see 2 or 3 and some years have had 5. I stopped looking closely after seeing that. There have been 3 running backs to get 1000 yards and have worse than a -100 DYAR. 1995 produced 2 of them. Edgar Bennett (GB) at -104 DYAR for his 1,064 yards and Garrison Hearst (ARI) -169 for his 1,067 yards. The other -100 DYAR performer who rushed for over 1,000 yards was Jamal Anderson (ATL) in 2000 with a -163 DYAR and 1,029 yards.

62 Re: Week 16 DVOA Ratings

Comparison of Bears and Vikings - Both 9-6.
Vikes Bears
DVOA 8.2 7.6
WT DVOA 8.7 2.9
Est Wins 8.1 8.9
Sched 2.9 0.5
Opp wk 17
Wt DVOA 25.1(NY) -5.1(HOU)

Giants lock up #1 seed but had success playing their best ball against the Pats last year in the final game. Houston has a five game winning streak. Vikes or Bears will win the division?

63 Re: Week 16 DVOA Ratings

Quick note: The "Special Super Bowl" pairings in the odds were because of a request from a journalist in New York -- we decided to just stick them in there for everyone. Perhaps we can add a few more for next week depending on who makes the playoffs.

64 Re: Week 16 DVOA Ratings

Speaking of which -- Honestly, how often can one team lose by less than a touchdown? Actually, the record seems to be eight. Since the merger, three teams lost eight games by less than a touchdown: the 2001 Panthers, the 1993 Patriots, and the 1984 Browns. The Packers have now lost seven games by four points or less.


Correct me if I'm wrong but all of those teams were in at least ther conference championship game within 2 years.

65 Re: Week 16 DVOA Ratings

The 1993 Patriots...if I recall correctly, that was the year that they featured the immortal Scott "Missin' Sisson", who made 53% of his FG's that year. Seriously. He missed 5 FG's of less than 30 yards over the course of the season! And as I recall, a number of those eight games lost by less than a TD were actually games lost by less than 3 or 6 points, where if he makes one or more FG's that he had missed, the Patriots win.

Thanks for reopening that old wound. :-)