Week 9 DVOA Ratings

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

by Aaron Schatz

As I've detailed numerous times over the past two months, the NFL just doesn't seem to have any really great teams this season. That's a big reason why the Philadelphia Eagles are still No. 1 in DVOA despite yet another loss in Week 9.

Yes, I know it seems absurd to have a 4-4 team at No. 1, but it makes more sense when you look closer at all of Philadelphia's individual games as well as the performance of the rest of the league in 2016.

Last week, I detailed some of the reasons why the Eagles are at No. 1 right now. Essentially, they have mixed three big wins and one gigantic win (over Pittsburgh) with four very close losses which have positive DVOA ratings because they came against the other good teams of the NFC East. Yes, the Eagles are in last place, but this is a last-place team that has outscored its opponents by a combined 57 points. Also, contrary to all the September stories about the Wentz Wagon, the Eagles are doing this entirely with defense and special teams. They lead the league in both categories, with a big gap between the Eagles and every other team. Meanwhile, the offense ranks 23rd.

The Eagles drop from 28.6% to 27.5% in total DVOA this week, but that's a small drop. They weren't going to drop much more after a very close loss to another good team. Both the Giants (11.3%) and the Eagles (18.3%) had good DVOA ratings for Sunday's game. The Eagles gained 6.0 yards per play compared to 5.0 yards per play for the Giants, with an equal number of turnovers. Plus, the system is going to reward the Eagles' defense for an interception that put them in prime scoring position (the Giants' 34), even though they didn't score from that position because the game situation forced them to go for it on fourth-and-10. With a gap of a few percentage points between the Eagles and the rest of the league, they weren't going to fall into second place based on this game. Meanwhile, below them, the DVOA system was surprisingly nonchalant about the No. 2 Cowboys dismantling Cleveland (23.2% DVOA after opponent adjustments), while the No. 3 Broncos got their asses whupped and dropped three spots. So Philadelphia is still No. 1.

Someone asked me on Twitter if having Philadelphia ranked first in DVOA shows that DVOA is flawed. The answer is that of course DVOA is flawed. All statistical ratings systems are flawed. We will never perfectly measure what the true quality of every NFL team is in the current week. There are things DVOA captures too much, or not enough. There is also value to subjective analysis that helps us consider the issues that DVOA does not incorporate, such as the importance of current injuries or the knowledge of which ratings should be discounted because circumstances have changed since September. They programmed a computer to be the greatest chess player in the world, only to discover that the greatest chess player is not a computer or a human being but rather a human being and a computer working in tandem. Figuring out the best teams in the NFL is no different.

Normally, a team such as the 2016 Eagles would be ranked third or fourth, and we would all be sitting here saying, "Wow, isn't it interesting that DVOA likes the Eagles so much even though they are just 4-4." But the Eagles look more out of place because there just aren't any great teams to be ranked higher than them this season. The Cowboys would probably be ranked higher if they had one dominant win on their record that matched the Eagles' 34-3 stomping of Pittsburgh. The Seahawks would probably be ranked higher if they had not lost to the Rams in Week 2. The Patriots would probably be ranked higher if not for the Week 4 game where they started an injured third-string quarterback and got shut out by Buffalo. But none of those things are true.

The Eagles are the worst team ever to be ranked No. 1 in DVOA at midseason. Only two teams come close.

  • The 2011 Jets were a similar team to these Eagles: 5-3 at midseason with 30.2% DVOA despite having a poor offense. That Jets team was 20th on offense, second on defense, and third on special teams after Week 9. The Jets actually collapsed somewhat over the second half of the season, going 3-5 over the final eight games and dropping to 10th in DVOA by the end of the year. Just like it is weird to see the Eagles ahead of the Cowboys right now, it was strange to see the 5-3 Jets ahead of the undefeated (and No. 2 in DVOA) Green Bay Packers at midseason. However, that Packers team excelled in part by winning close games, and they weren't historically great despite the eventual 15-1 record. 2011 ended up as one of only six seasons since 1989 (and the most recent) where no team ended up with DVOA over 30% at the end of the year.
  • The year before, the 2010 Giants ranked No. 1 in DVOA at midseason with 30.9% DVOA and a 6-2 record. The 5-3 Eagles were right behind them at 30.5% DVOA. However, 2010 is a strange year where the eventual two best teams (New England and Pittsburgh) really surged in the second half of the year. Like the 2011 Jets, the 2010 Giants declined at midseason, finishing 4-4 over the last eight games and ninth in DVOA. However, the 2010 Eagles did not decline. They also finished 10-6, so they were 5-3 in their last eight games, and they finished the year third in DVOA behind New England and Pittsburgh.

However, these Eagles do not have the best DVOA ever for a 4-4 team. That honor belongs to... another Philadelphia team. The 2006 Eagles are the only other team to ever lead the NFL in DVOA with a 4-4 record. That team was even better, with 36.1% DVOA after Week 9. And that Eagles team showed that DVOA was right about them, as they went 6-2 over the last eight games of the year.

The other 4-4 team with a higher DVOA than these Eagles was the 2005 San Diego Chargers, who ranked third with 34.3% DVOA when they were 4-4. That was a really weird year for the Chargers. They outscored opponents by 106 points but finished only 9-7. The early part of the year looked a lot like the first two months of the year for the current Eagles. Their first four losses all came by four points or less, while their wins came by scores of 45-23, 41-17 (over the defending champion Patriots!), 27-14, and 28-20.

[ad placeholder 3]

By the way, the whole "DVOA loves the Eagles" thing is very strange. We've always assumed this is based on the way that the Andy Reid offense is built to make consistent small gains that move the chains. However, that assumption makes no sense if you look at this year's ratings. The Eagles are No. 1 entirely because of defense and special teams, not because of Doug Pedersen bringing the Reid offense back to Philadelphia. Meanwhile, the actual Andy Reid offense seems to be somewhat underrated by DVOA, as the Chiefs are 11th (and 21st in offense) despite a 6-2 record. All this Eagles talk also may be for naught because the Eagles' remaining schedule is the hardest in the league. Don't be surprised if the Eagles are still ranked No. 1 in DVOA at the end of the year but finish 8-8 or 9-7 and out of the playoffs. (More likely, given how the Patriots have played since Tom Brady's return, the Eagles will end the year No. 2 in DVOA but out of the playoffs.)

Anyway, the whole "no great teams in 2016" issue goes beyond just the Eagles. Dallas also has the lowest DVOA for any team ever ranked No. 2 at midseason. Atlanta is No. 1 in offensive DVOA, but only two teams have ever been No. 1 in offense at this point with a less impressive rating (2008 Giants, 1997 Broncos). On defense, the Eagles don't stand out as unimpressive compared to past No. 1 defenses, but the rest of this year's great defenses have all taken a step backwards. With Denver, Seattle, and Minnesota all giving up big offensive games this week, the Eagles are now the only defense below -20% for the year.

Meanwhile, Houston stands out for having a winning record despite ranking 30th in DVOA. (Detroit stands out to a lesser extent, ranked No. 27 with a 5-4 record.) The Texans have the lowest DVOA ever for a team with a winning record through Week 9, while Detroit ranks 10th.

Worst DVOA by Team with Winning Record Through Week 9, 1989-2016
Year Team W-L DVOA Rank Final
2016 HOU 5-3 -26.6% 30 -- -- -- --
2012 IND 5-3 -22.5% 27 11-5 -16.0% 25 Wild Card
1990 CIN 5-4 -20.3% 22 9-7 -12.3% 21 Won AFC Central
2010 CHI 5-3 -19.0% 25 11-5 2.4% 14 Won NFC North
2007 DET 6-2 -17.1% 25 7-9 -29.0% 29 Missed
2015 MIN 6-2 -14.9% 26 11-5 5.7% 11 Won NFC North
2002 BUF 5-4 -14.1% 26 8-8 -8.0% 23 Missed
1996 IND 5-3 -13.3% 23 9-7 -12.2% 22 Wild Card
2006 SEA 5-3 -12.7% 24 9-7 -13.0% 24 Won NFC West
2016 DET 5-4 -12.4% 27 -- -- -- --

* * * * *

Once again this season, we have teamed up with EA Sports to bring Football Outsiders-branded player content to Madden 17 Ultimate Team. Each week, we'll be picking out a handful of players who starred in that week's games. Some of them will be well-known players who stood out in DVOA and DYAR. Others will be under-the-radar players who only stood out with advanced stats. We'll announce the players each Tuesday in the DVOA commentary article, and the players will be available in Madden Ultimate Team packs the following weekend, beginning at 11am Eastern on Friday. We will also tweet out images of these players from the @fboutsiders Twitter account on most Fridays. The best player of each week, the Football Outsiders Hero, will require you to collect a set of the other four Football Outsiders players that week, plus a certain number of Football Outsiders collectibles available in Madden Ultimate Team packs.

The Football Outsiders stars for Week 9 are:

  • LG Kelechi Osemele, OAK (FOOTBALL OUTSIDERS HERO): Oakland RB had 29 runs, 126 yards, 3 TD on runs going left or up the middle.
  • LOLB Mark Barron, LARM: Led team with 9 tackles including a sack and three run tackles for a yard or less.
  • DT Maliek Collins, DAL: 2 sacks for 16 lost yards.
  • HB Mark Ingram, NO: No. 2 among Week 9 RB with 64 DYAR (158 rushing yards, 13 receiving yards, 2 TD).
  • WR Robert Woods, BUF: Led all Week 9 WR with 67 DYAR (caught 10 of 13 passes for 162 yards, 8 first downs).

* * * * *

All stats pages are now updated through Week 9 of 2016. There may be some errors because of technical problems with the play-by-play feeds this week, so feel free to let us know if you find any of them. Snap counts, playoff odds, and the premium DVOA database are also fully updated.

Update: Unfortunately, the technical issues with NFL feeds mean we will not be able to update drive stats or pace stats for an unknown period of time. We are working on a solution to the problem, but until then, those pages will not be updated.

* * * * *

[ad placeholder 4]

These are the Football Outsiders team efficiency ratings through nine weeks of 2016, measured by our proprietary Defense-adjusted Value Over Average (DVOA) system that breaks down every single play and compares a team's performance to the league average based on situation in order to determine value over average. (Explained further here.)

OFFENSE and DEFENSE DVOA are adjusted to consider all fumbles, kept or lost, as equal value. SPECIAL TEAMS DVOA is adjusted for type of stadium (warm, cold, dome, Denver) and week of season. As always, positive numbers represent more points so DEFENSE is better when it is NEGATIVE. Opponent adjustments are currently at 90 percent strength; they will hit 100 percent next week. As always, positive numbers represent more points so DEFENSE is better when it is NEGATIVE.

WEIGHTED DVOA represents an attempt to figure out how a team is playing right now, as opposed to over the season as a whole, by making recent games more important than earlier games.

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 PHI 27.5% 1 27.8% 1 4-4 -7.8% 23 -24.5% 1 10.8% 1
2 DAL 21.7% 2 22.3% 2 7-1 20.8% 2 -1.2% 14 -0.3% 17
3 SEA 20.6% 6 19.8% 5 5-2-1 1.6% 14 -17.7% 6 1.3% 10
4 ATL 20.4% 5 22.1% 3 6-3 24.3% 1 8.0% 24 4.1% 6
5 NE 19.3% 4 21.0% 4 7-1 18.4% 3 1.7% 18 2.6% 8
6 DEN 14.4% 3 15.2% 6 6-3 -6.0% 22 -19.8% 2 0.7% 13
7 GB 9.4% 7 9.0% 7 4-4 4.3% 11 -7.7% 7 -2.6% 22
8 MIN 8.9% 8 8.5% 9 5-3 -10.3% 25 -18.1% 4 1.1% 12
9 OAK 8.5% 11 8.6% 8 7-2 17.6% 4 10.3% 27 1.2% 11
10 BUF 5.6% 9 6.3% 11 4-5 13.2% 6 6.3% 22 -1.3% 20
11 KC 5.5% 10 5.4% 12 6-2 -5.3% 21 -6.2% 11 4.6% 5
12 MIA 4.7% 18 6.9% 10 4-4 -4.3% 20 -6.3% 10 2.8% 7
13 SD 4.7% 14 4.4% 13 4-5 0.9% 15 -7.6% 8 -3.8% 25
14 NYG 3.8% 17 3.8% 15 5-3 -2.7% 18 -7.2% 9 -0.7% 19
15 WAS 2.8% 16 4.1% 14 4-3-1 6.1% 9 3.8% 21 0.5% 15
16 CHI 2.5% 15 3.7% 16 2-6 1.8% 13 -0.1% 16 0.6% 14
17 PIT 2.1% 12 0.2% 18 4-4 5.6% 10 3.0% 19 -0.6% 18
18 ARI 1.8% 13 1.4% 17 3-4-1 -10.5% 26 -18.2% 3 -5.9% 30
19 CIN -0.9% 19 -0.4% 19 3-4-1 12.5% 7 7.6% 23 -5.8% 29
20 NO -1.9% 23 -1.4% 20 4-4 17.3% 5 16.2% 29 -2.9% 23
21 TEN -4.0% 22 -4.1% 21 4-5 10.1% 8 8.9% 26 -5.1% 27
22 BAL -4.1% 20 -5.0% 22 4-4 -24.5% 32 -17.8% 5 2.6% 9
23 CAR -6.8% 21 -7.3% 23 3-5 -1.0% 17 0.6% 17 -5.3% 28
24 LARM -10.3% 25 -7.9% 24 3-5 -21.3% 30 -6.2% 12 4.9% 4
25 TB -11.4% 26 -12.9% 26 3-5 -3.8% 19 3.6% 20 -4.0% 26
26 IND -12.4% 28 -12.0% 25 4-5 0.5% 16 19.8% 30 6.9% 2
27 DET -12.4% 24 -13.2% 27 5-4 3.8% 12 21.3% 32 5.0% 3
28 JAC -13.5% 27 -13.7% 28 2-6 -12.2% 28 -0.5% 15 -1.8% 21
29 SF -22.3% 29 -24.6% 29 1-7 -11.6% 27 10.4% 28 -0.3% 16
30 HOU -26.6% 30 -27.6% 30 5-3 -23.5% 31 -3.4% 13 -6.6% 31
31 CLE -32.5% 32 -32.2% 31 0-9 -8.1% 24 20.7% 31 -3.8% 24
32 NYJ -32.6% 31 -33.1% 32 3-6 -17.2% 29 8.0% 24 -7.4% 32
  • 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.
  • ESTIMATED WINS uses a statistic known as "Forest Index" that emphasizes consistency as well as DVOA in the most important specific situations: red zone defense, first quarter offense, and performance in the second half when the score is close. It then projects a number of wins adjusted to a league-average schedule and a league-average rate of recovering fumbles. Teams that have had their bye week are projected as if they had played one game per week.
  • PAST SCHEDULE lists average DVOA of opponents played this season, ranked from hardest schedule (#1, most positive) to easiest schedule (#32, most negative). It is not adjusted for which games are home or road.
  • FUTURE SCHEDULE lists average DVOA of opponents still left to play this season, ranked from hardest schedule (#1, most positive) to easiest schedule (#32, most negative). It is not adjusted for which games are home or road.
  • VARIANCE measures the statistical variance of the team's weekly DVOA performance. Teams are ranked from most consistent (#1, lowest variance) to least consistent (#32, highest variance).

1 PHI 27.5% 4-4 28.5% 6.6 2 -0.4% 19 9.2% 1 10.1% 22
2 DAL 21.7% 7-1 25.1% 6.9 1 -1.2% 22 2.2% 12 4.1% 5
3 SEA 20.6% 5-2-1 22.6% 6.0 5 -4.3% 27 0.9% 16 9.7% 19
4 ATL 20.4% 6-3 16.8% 6.5 3 2.9% 6 -0.9% 19 7.7% 11
5 NE 19.3% 7-1 21.1% 6.4 4 -5.0% 29 -7.8% 29 15.4% 29
6 DEN 14.4% 6-3 19.6% 6.0 6 -2.2% 23 2.8% 10 9.4% 17
7 GB 9.4% 4-4 7.8% 4.9 11 2.4% 10 2.4% 11 9.9% 21
8 MIN 8.9% 5-3 11.1% 5.8 7 -0.8% 21 0.0% 18 11.4% 23
9 OAK 8.5% 7-2 8.6% 4.9 12 1.1% 12 -2.2% 21 7.8% 12
10 BUF 5.6% 4-5 7.4% 4.7 14 -0.4% 20 -9.2% 32 14.1% 25
11 KC 5.5% 6-2 8.0% 5.3 8 -9.0% 32 5.0% 6 26.5% 31
12 MIA 4.7% 4-4 3.1% 5.1 9 -2.8% 26 -4.7% 26 16.1% 30
13 SD 4.7% 4-5 4.7% 4.5 16 3.5% 4 -8.4% 31 3.3% 4
14 NYG 3.8% 5-3 -5.8% 4.7 13 6.8% 1 1.3% 14 0.8% 1
15 WAS 2.8% 4-3-1 4.7% 5.1 10 0.7% 13 8.6% 2 8.2% 14
16 CHI 2.5% 2-6 2.6% 4.2 21 0.3% 14 -3.1% 24 15.2% 28
17 PIT 2.1% 4-4 1.4% 4.0 23 2.8% 7 -6.4% 27 28.7% 32
18 ARI 1.8% 3-4-1 9.7% 4.3 18 -4.7% 28 2.9% 9 9.4% 18
19 CIN -0.9% 3-4-1 1.1% 4.2 19 0.0% 16 -3.5% 25 8.9% 15
20 NO -1.9% 4-4 1.0% 4.5 15 4.3% 2 -2.0% 20 1.7% 2
21 TEN -4.0% 4-5 3.2% 4.4 17 -7.8% 31 -2.9% 23 8.2% 13
22 BAL -4.1% 4-4 4.6% 4.2 20 -7.0% 30 5.1% 5 5.6% 8
23 CAR -6.8% 3-5 -6.2% 3.3 24 0.0% 17 6.2% 3 5.7% 9
24 LARM -10.3% 3-5 -3.4% 4.1 22 -2.6% 25 1.2% 15 15.0% 26
25 TB -11.4% 3-5 -14.5% 3.2 26 3.3% 5 5.6% 4 15.1% 27
26 IND -12.4% 4-5 -10.2% 3.2 27 -2.2% 24 -8.2% 30 5.1% 7
27 DET -12.4% 5-4 -10.4% 3.2 25 -0.2% 18 4.4% 7 7.7% 10
28 JAC -13.5% 2-6 -10.3% 3.1 28 1.3% 11 -6.6% 28 4.5% 6
29 SF -22.3% 1-7 -26.2% 2.2 31 2.4% 9 3.3% 8 12.5% 24
30 HOU -26.6% 5-3 -23.5% 2.7 29 2.7% 8 -2.7% 22 9.3% 16
31 CLE -32.5% 0-9 -32.9% 0.1 32 3.8% 3 1.9% 13 2.6% 3
32 NYJ -32.6% 3-6 -30.7% 2.4 30 0.3% 15 0.6% 17 9.7% 20


57 comments, Last at 15 Nov 2016, 7:47pm

1 Re: Week 9 DVOA Ratings

Something that may just interest me: Seattle's offense jumped 5% after getting untracked against Buffalo's 17th ranked defense (1.5%) and Philly's defense dropped only 0.7% after allowing 28 points to NYG's 19th ranked offense (-3.4%). (Rankings and DVOA numbers from week 8 to avoid double counting this week's games).

I know there are other factors and that points isn't a great proxy for the things DVOA measures, I just found the disparity interesting given the surface level similarity.

19 Re: Week 9 DVOA Ratings

Personal pet peeve... it's "on track" (which has a positive connotation), not "untracked" (which is negative).

Further reading : http://www.nytimes.com/2015/12/20/sports/an-eggcorn-or-a-pet-peeve-part-of-the-sports-lexicon.html

25 Re: Week 9 DVOA Ratings

That was what I always thought, but I came across "untracked" as a dogsled term in a book, where the skis were caked with hardened ice and needed to be untracked before it could be moved. I figured that was the basis for the word describing faster motion after a slow start.

30 Re: Week 9 DVOA Ratings

Does that mean "more than $90,000" or "almost $90,000"?

It means "more than", or sometimes "at least".

And why not just say that instead?

Why have synonyms at all? Maybe we should kick out one of "big" and "large", one of "little" and "small", etc.

Synonyms make the language more vibrant, and that far outweighs any problems with having redundant words and phrases.

43 Re: Week 9 DVOA Ratings

Synonyms are fine. I just think the "upwards of" is not used consistently, so many times I honestly have no idea what the intent is.

Here's an example where they use it to mean "almost":

The headline says "upwards of 14,000" but the text says 9,000-14,000.

Here's a weird one:
"Reggie Jackson could miss upwards of 6 to 8 weeks"

Of course, my all time favorite are the commercials that say "save up to 40% or more!"

41 Re: Week 9 DVOA Ratings

I would guess most of the difference is the week 9 performance of all the defenses Seattle played weeks 1-8, changing their rankings and thus retroactively reassessing the performance of Seattle's offense against those teams, versus the performance of the offenses Philly played 1-8.

Plus the Giants did get wo short fields from turnovers on Sunday.

55 Re: Week 9 DVOA Ratings

I love football, and when my team didn't make it ways to final, I feel bad and stress. And because of this stress I have the problem of http://www.doctordarkcircles.com now and to reduce it, I am sure you will learn this website.

2 Re: Week 9 DVOA Ratings

The Week 9 BES Rankings are out - http://besreport.com/week-9-bes-rankings-2016/

Really surprised to still see the Eagles atop the DVOA ratings. They've plummeted all the way down to 24th in the BES. The Cowboys are No. 1 in the BES and that might be short-lived if the Pats (No. 2 in the BES) beat the Seahawks in Week 10.

To most, The oddity with the BES top-10 this week rests with the Redskins, Saints and Lions at N0. 8, 9 and 10 respectively. So the BES and DVOA definitely have some head-scratchers this week to say the least. haha! But there's ALWAYS a reason in the math...no matter how odd the results. That's the beauty of it all!

Love DVOA! Always a great read and top notch info. Thanks again!!

57 Re: Week 9 DVOA Ratings

Week 10 BES Rankings just went out - http://besreport.com/week-10-bes-rankings-2016/

More movement in the top-10. Saints stay didn't last long. lol! Chiefs highest ranked AFC team at No. 2.

3 Re: Week 9 DVOA Ratings

Jets in last! I have to admit, although this team is pretty bad, I can't seem them being really the worst team in the whole league.

I wondered if the Eagles would be still in first. If they were 6-2, as they probably 'should' be (since their four losses were all just about toss-up games), it wouldn't seem very strange at all. So I certainly think the people who believe the Eagles being first represents some sort of huge flaw in DVOA are off base.

7 Re: Week 9 DVOA Ratings

They finally beat the Browns out for that honor in the BES Rankings - http://besreport.com/week-9-bes-rankings-2016/

That's an impressive feat. It is startling when you look at the demise of the 9ers over the last four or five years.

13 Re: Week 9 DVOA Ratings

SF has by far the best special teams among the four DVOA cellar-dwellers, which makes up most of its lead over the Jets and Texans. As with Philadelphia, in fact, I find it really hard to factor in special teams to the mythical eyeball test. The big plays all seem like one-offs, and everything else is just a few minor yards here and there.

Consistently awful special teams is easier to see.

4 Re: Week 9 DVOA Ratings

So, I've thought there was an FO Philly bias of some kind for many years, but this is a bit silly.

5 Re: Week 9 DVOA Ratings

Did you even bother to read what was written above? The idea that DVOA has a generic Philly bias doesn't make much sense, UNLESS all those Philadelphia teams play in a similar manner, which continues to trick DVOA into overrating them. But, as was explained, that doesn't apply, since this Eagles' team is actually weak on offense, making it different from the old Andy Reid teams that used to do well in DVOA. In general, DVOA having a bias in favor of a specific team over many years isn't a sensible theory.

11 Re: Week 9 DVOA Ratings

Philly has outscored their opponents by 57 points this year, which is 3rd behind only the Patriots & Cowboys. So it's not a surprise to see them near the top of the rankings.

24 Re: Week 9 DVOA Ratings

And Aaron yells at him: "You festering pile of parrot droppings! Your kind make me sick!"
GRAHAM: "I was hear to learn about American Football. Also, I'm dead."
AARON: "Oh, sorry, this is abuse. Silly Articles are the next door down. Also, sorry you're dead."
GRAHAM: "Oh, I'll head down there then. I've an excellent idea for a silly article. So sorry about the interruption."
AARON: "No problem... you stupid git."

And while this has nothing to do with FO or Aaron, it's always good to review some of Monty Python's greatest plays.

9 Record Vs. DVOA

Are there coaches whose teams regularly outplay or underplay their DVOA ratings? With relatively large sample sizes, that would seem to be a place one could find points that would help tweak the system.

15 Re: Record Vs. DVOA

In reply to by RobotBoy

There seems to be a disparity between Belichick defenses and DVOA ratings. For example, his teams often "trade space for time" with leads, giving DVOA the impression the defense is playing poorly, whereas from an operational perspective the defense is playing smart.

21 Re: Record Vs. DVOA

Except DVOA doesn't say their defense is playing poorly.

It says they are almost exactly average when compared with all defenses going back to 1989.

27 Re: Record Vs. DVOA

DVOA correlates to point differentials. Because of this, defensive DVOA, once corrected for starting field position, correlates pretty well for defensive points allowed.

Except that the Patriots consistently give up less points than DVOA and field position account for.

Currently the Patriots are second in the league in points allowed and points allowed per game (16.5) - behind the Minnesota Vikings (15.8), and slightly ahead of Arizona(17.5), Seattle(16.8), Philly (18.1), and Denver(18.4), teams that all have a roughly 20% defensive DVOA advantage on them. Field position doesn't even come close to compensating for that.

DVOA most certainly says the Patriots defense is playing poorer than is showing up on scoreboards. And it has said that for a very long time - so while we can't tell exactly what's happening, it's clearly not random.

31 Re: Record Vs. DVOA

Field position doesn't even come close to compensating for that.

Yes it absolutely can. New England's defense gives up 32.5 yards per drive through week 8 (week 9 drive stats aren't up yet, but New England had a bye anyway), good for 16th in the league. But they get given the best starting field position. If you add where their opponents start their drives to how long their drives go for, you end up with an average ending field position of 56.67, good for 8th in the league, which is almost the same ranking as their 6th-fewest points given up per drive.

32 Re: Record Vs. DVOA

Field position doesn't even come close to compensating for that.

Game situation is a factor, too. Patriots' opponents have turned the ball over on downs seven times in the last 20 minutes of the game, with the Pats leading by 20+ points on six of those occasions. (In the other, they were ahead by 11 with less than three minutes to go in the game.) All six stops took place in the Patriots' half; three were in field goal range. Let's consider that if the Patriots had not been so far ahead, they would have conceded an extra 9-15 points (field position on the following drive would have altered in the opponent's favor).

Now, this year's opponents have also missed five field goals against New England (joint most in the league). Opponents are 9 of 14 (63%) versus an NFL average of 83%. Granted, the average attempt has been a bit longer against the Pats, but still, with average luck, opponents ought to be at least 11 of 14. That's another 6 points.

Add those together and New England is now conceding an extra 2 points a game – behind all those teams you mention. Throw in the advantage in field position and the numbers almost make sense.

DVOA most certainly says the Patriots defense is playing poorer than is showing up on scoreboards. And it has said that for a very long time

Defensive rankings:

DVOA - points conceded - yards per play - average starting LoS

2015: 12 - 11 - 8 - 2
2014: 12 - 8 - 12 - 1
2013: 20 - 10 - 7 - 3
2012: 15 - 9 - 24 - 7
2011: 30 - 15 - 30 - 2

35 Re: Record Vs. DVOA

"Game situation is a factor, too. Patriots' opponents have turned the ball over on downs seven times in the last 20 minutes of the game, with the Pats leading by 20+ points on six of those occasions"

OF COURSE IT IS. That's the whole point here - defense is not context independent - defense is entirely context sensitive.

"All six stops took place in the Patriots' half; three were in field goal range. Let's consider that if the Patriots had not been so far ahead, they would have conceded an extra 9-15 points (field position on the following drive would have altered in the opponent's favor)."

No, let's not concede that, because, again, defense is context sensitive. Getting into field goal range when you're down several scores has very little value, so your defense plays deeper and allows teams to make plays underneath, and inside, at the expense of time.

"Add those together and New England is now conceding an extra 2 points a game – behind all those teams you mention. Throw in the advantage in field position and the numbers almost make sense."

Even were we to do that, it puts them pretty much even with Denver, who has a defense 20% better than NE according to DVOA.

If you look at Atlanta, their defense is a bit worse (8% as opposed to 2%), their field position is 1 yard worse, and they give up a full point more per drive. If you look at the Rams, their field position is the same as Atlanta, their DVOA is -7% (9% better than the Pats), and they give up .3 more points per drive.

The biggest problem with looking at field position and average yards per drive is that they assume that the value of field position is linear, when we know that's not even close to true. Yards are less valuable in the middle of the field than they are closer to the endzone. Giving up yards is less valuable when you've got a big lead.

37 Re: Record Vs. DVOA

I was thinking Belichick when I wrote my comment but I wanted to expand the data base as widely as possible. So these particular stats in DVOA aren't at all context sensitive?

39 Re: Record Vs. DVOA

No, let's not concede that, because, again, defense is context sensitive.

I'll just mention one drive that's not context-sensitive to show you why DVOA thinks the Patriots have a worse defense than their points given up suggests. Cincinnati had a 17-play 89-yard drive against New England spanning the first and second quarters while down 3, so score doesn't come into play. Cincinnati only faced 4 third downs, and three of them were within 4 yards, so DVOA definitely dinged New England's defense for giving up this drive. But New England stopped them at the one-yard line, and Cincinnati scored zero points on this drive. Good for New England for stopping them, but Cincinnati could've kicked a field goal on fourth down and that's three more points given up by New England with hardly any change to their DVOA.

And even if you do take context into account, a defense that allows long drive when the team is up by lot only works if the offense keeps scoring or if the defense makes a few stops. If they don't, you get the Miami game, where New England nearly blew a 28-point lead.

Even were we to do that, it puts them pretty much even with Denver, who has a defense 20% better than NE according to DVOA.

And an opponent starting field position that's a whole 5 yards worse. Denver's defense also forces turnovers 55% more frequently than New England.

44 Re: Record Vs. DVOA

'And even if you do take context into account, a defense that allows long drive when the team is up by lot only works if the offense keeps scoring or if the defense makes a few stops.'
Even an English PhD like me knows that's only partially true. If you're up by twenty and there are five minutes left in the game, and you don't make a stop but force the other offense to take four minutes to score a touchdown, than that's a net positive for your team.
I think Belichick mismanages this strategy on occasion, and has made some seeming blowouts uncomfortably close, but overall it seems to have worked for him.

48 Re: Record Vs. DVOA

To be fair, I don't think anyone points to the Miami game as a good example of BBDB. In fact, I'm not sure it is an example of that strategy at all since Miami's drives contained a lot of big plays.

The Cincy drive is fair, as are the two against Ari that ended in a FG range sack and a missed FG.

45 Re: Record Vs. DVOA

DVOA most certainly says the Patriots defense is playing poorer than is showing up on scoreboards.

That is not what the person I was replying to said. He said that DVOA shows the Pats defense as playing poorly. They are not. They are playing exactly 1.7% below the historical average of all defenses going back to 1989. That's about as close to being average without being exactly average as you are going to get.

If you feel that New England has gotten above average results with its oh so average defense that is certainly a respectable opinion to have. However, as noted, that is not an opinion I was disagreeing with.

DVOA says New England's defense is pretty average. This is objective fact unless we'd like to argue what 0.0% DVOA means.

50 Re: Record Vs. DVOA

One well known shortcoming of DVOA (as compared to, for example, WPA) is that it doesn't look at time. For example, a 5 yard run on 1st down that ends in bounds is a failure if you're trailing by 9 with under two minutes to play--a defense will happily give that up. But DVOA thinks that is a success.

I suspect this is part of the issue with the Patriots. Their offense has been so good under Belichick that they're leading in the second half more often than not, and especially so late in the second half. In this situation, they are more than happy to give up lots of yards, and even a TD, if it takes so much time that their win probability actually goes up. (This is not unique to the Patriots; most good coaches do this). But a team that is consistently ahead late will give up a lot of successful plays.

The other factor that occurs to me is that a trailing team generally plays offense more aggressively, and we know that baseline NFL offenses aren't optimally aggressive, so a team like the Patriots that is consistently ahead late will face more optimal offenses and therefore look worse.

Of course, this can't be the whole story, or we would see every successful team that wins a lot undervalued by defensive DVOA. But it's probably part of it.

51 Re: Record Vs. DVOA

DVOA does have time buckets. So the Patriots are compared to other defenses that are playing ahead late. They're just not as good at it.

52 Re: Record Vs. DVOA

'Time Buckets', hmmm... They must come in handy when you have so much time on your hands it won't fit into a bottle.
It helps knowing that DVOA takes into account the time remaining and score. Are those 'buckets' particularized down to the actual point difference and, say, minutes of time remaining?
Belichick, with some notable exceptions, has done very well with a lead going into the second half: 'The Patriots have won 82 straight home games when leading at halftime...' Obviously that's due to many factors, from quality of offense to random luck, but coaching has something to do with it. To the fan part of my brain, it seems that the Patriots often go up big and then fritter away the lead through versions of prevent defense and offensive three-and-outs (Corey Dillon solved that problem during his first year but I digress). I'm left gnawing on my arm as the opponents' final drive falls short with a missed FG, interception, or other seemingly random act of a hoodie-wearing deity.
That's why I'd like to know if coaches have outperformed or underperformed DVOA over their careers, and if the analysis can extend to late-game situations, etc.

10 Re: Week 9 DVOA Ratings

When was the last time, in week 9, that the #30 DVOA team had a better record than the #1?

12 Re: Week 9 DVOA Ratings

I thought there was a template for saying "Philadelphia is ranked way too high because ..." :-D

14 Overall Quality Impressions and Parity

Does having no "great" teams cause the general viewership to think the games on average are worse than before? Does having some "great" teams improve the apparent quality of the league despite the increased chance of blowouts?

18 Re: Week 9 DVOA Ratings

Wondering what the odds of the Bears only winning one more game the rest of the season and grabbing a juicy top 5 pick.

That 24th ranked future schedule not going to help.

33 Re: Week 9 DVOA Ratings

Damn near zero unless Cutler gets hurt in the first quarter this weekend. Not getting a top-5 pick this year is actually OK with me, because if they had one I'd be terrified they'd use it on a QB who really should go in the 2nd or 3rd round.

36 Re: Week 9 DVOA Ratings

That's a really backwards way to think about it. If they wind up with the 10th-ish pick, that's where the risk of a reach is much greater. If they're in the top five, there would be much more attractive other options than reaching for a QB. I mean, just look at this year: you'd be much less likely to draft a reach QB over Joey Bosa than over Leonard Floyd.

The biggest reaches in recent memory - Ponder, Locker, Gabbert - all went outside the top five.

38 Re: Week 9 DVOA Ratings

I suppose Bortles and Goff could possibly be considered reaches, but I don't know really know how they were considered before the draft.

But if you don't think your team's management can take advantage of a top 5 pick. you're basically giving up on the whole organization anyways.

22 Re: Week 9 DVOA Ratings

-The Broncos- are clearly ranked -too high- because -their offense looked terrible against one of the worst defenses in the league-. -ESPN's Power Rankings- is way better than this. -Bring in Paxton Lynch!-

23 Re: Week 9 DVOA Ratings

The machine says the Patriots are a top-five team. My brain knows that this is despite playing four games without their best player (and arguably the best player in the league). Therefore, I conclude that the Patriots are the best team in the league. Centaur rankings!

29 Re: Week 9 DVOA Ratings

ERaiders should eb 2 oe 4. Not 9. No way are Packers or Broncos or Vikings better thabn the Raiders

46 Re: Week 9 DVOA Ratings

What your inquisitor in paragraph five probably asked, is whether DVOA is not merely "flawed", but a kind of illusory scam, like your entire website.