Week 9 DVOA Ratings

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

by Aaron Schatz

A few big wins and losses stir up the Football Outsiders DVOA ratings this week, but the top three stay the same as last week. New England, Arizona, and Cincinnati are still the top three teams overall, each having one of the top three offenses in the league to go with a top-ten defense. Carolina is the big mover this week, going up from No. 7 to No. 4 on the strength of its big win over Green Bay. That means the three remaining undefeated teams now make up three of the top four teams in DVOA, along with the Cardinals. Our odds of at least one team going undefeated have gone up from 14.2 percent to 15.7 percent. The Broncos are no longer in the mix, but that is more than offset by the Panthers doubling their odds of going 16-0 after beating the Packers.

The change in circumstances for the Panthers and Falcons has been pretty severe. After four weeks, the Falcons and Panthers were both 4-0 and it was Carolina that looked like it was just skating by with close wins over bad opponents. After all, the Falcons at least beat the Eagles and Giants in their first two games. Since then, the Panthers have steadily improved against tougher competition while the Falcons have been mediocre against poor opponents. In fact, it's a bit surprising that the Falcons aren't lower than 17th in DVOA right now. Atlanta actually gets a positive DVOA for its loss to San Francisco this week, despite the opponent adjustment that comes from playing the worst team in the league by DVOA. Both the Falcons and the 49ers averaged the same 5.0 yards per play, but the Falcons picked off Blaine Gabbert twice without any turnovers of their own. Of course, DVOA gives the Falcons credit for a long drive that got down to fourth-and-goal from the 1, and sees that as more evidence of the Falcons being able to drive the ball in the future without too much of a penalty for Dan Quinn's awful decision to kick a field goal down four points in the last three minutes. DVOA also uses an opponent adjustment for San Francisco that's based on the whole season, without making the adjustment extra large to make up for the fact that the 49ers were starting Blaine Gabbert at quarterback and playing without Carlos Hyde or Anquan Boldin.

For your NFC South-judging entertainment, here's our first week-to-week DVOA graph of the season. This one has both Atlanta and Carolina's DVOA side by side, and you can see how the Panthers have surged past the Falcons since Week 5.

Still, in the end, DVOA doesn't say that Atlanta sucks. It just says Atlanta is blah: 14th on offense, 19th on defense and special teams.

On the other hand, speaking of southern divisions that suck... oh, AFC South, I just can't quit you. The Colts had significantly improved their playoff odds with the big win over the Broncos covered further in this week's Any Given Sunday column. That win had also reduced the chances that the AFC South will produce a champion with a losing record. But then came today's announcement that Andrew Luck will miss 2-6 weeks with a kidney laceration. That's a big blow to the Colts' chances of making the playoffs, right? Actually, maybe not.

When the Andrew Luck news came out, we ran a new, updated playoff odds simulation to try to account for what it would mean for Luck to be gone. Interestingly, the Colts have already been without Luck for two games this year and after Luck's big game against Denver, the difference in offensive DVOA is almost exactly equal to the 13.0% that we use as the standard difference between a starting quarterback and replacement level. The Colts have -1.7% offensive DVOA with Luck at quarterback and -14.0% offensive DVOA with Matt Hasselbeck.

We re-ran the playoff odds simulation, programming it to assume that Hasselbeck starts in Week 11 and Week 12. Then we gave Luck a 25 percent chance of returning each week until we considered him definitely back in Week 16 and Week 17. But this new simulation didn't really change the Colts' playoff odds much. In the old simulation, the Colts' playoff odds were at 60.8 percent. In the new simulation, their odds are 54.8 percent. Andrew Luck's injury doesn't change the fact that the Colts are the only team in the division with four wins. It doesn't change the fact that the team in second place, Houston, has a much harder schedule over the second half of the year. And it doesn't change the fact that the Colts are 3-0 in the division... and don't play another division opponent until Jacksonville in Week 14, giving Luck four weeks to heal before any game that might really put the Colts' playoff chances in jeopardy.

Here's a look at how the chances of various AFC South results changed between last week, this week's original simulation, and the new "Andrew Luck injury" simulation:

As the AFC South Turns
Result After Week 8 Original Week 9 Sim After Luck Injury
Indianapolis wins AFC South 52.5% 60.6% 54.6%
Jacksonville wins AFC South 24.3% 19.5% 22.7%
Houston wins AFC South 16.7% 12.7% 14.5%
Tennessee wins AFC South 6.4% 7.2% 8.2%
AFC South champion 9-7 or better 18.3% 28.1% 22.5%
AFC South champion 8-8 31.1% 37.1% 37.2%
AFC South champion 7-9 35.0% 28.8% 32.5%
AFC South champion 6-10 or worse 15.6% 6.0% 7.8%

* * * * *

Once again in 2015, we have teamed up with EA Sports to bring Football Outsiders-branded player content to Madden 16 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. We will also tweet out images of these players from the @fboutsiders Twitter account on most Fridays. One player each week will only be available for 24 hours from the point these players enter packs on Friday.

The Football Outsiders stars for Week 9 are:

  • RB DeMarco Murray, PHI (24-HOUR HERO): No. 4 RB of Week 9 with 41 DYAR; 18 carries for 83 yards and a touchdown, plus 6 receptions for 78 yards.
  • DT Linval Joseph, MIN: 10 combined tackles, half sack, 3 TFL, 2 QB hits.
  • RB Jeremy Langford, CHI: No. 8 RB of Week 9 with 28 DYAR; 18 carries for 72 yards and a touchdown, plus 3 receptions for 70 yards.
  • ROLB Lorenzo Mauldin, NYJ: 2 sacks, 4 QB hits.
  • LG Justin Pugh, NYG: Giants had 26 runs up the middle for 107 yards and a 65 percent success rate, plus no sacks allowed.

* * * * *

All stats pages are now updated with Week 9 information (or will be in the next few minutes) including FO Premium, snap counts and playoff odds.

* * * * *

[ad placeholder 4]

These are the Football Outsiders team efficiency ratings through nine weeks of 2015, 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 for strength of schedule and to consider all fumbles, kept or lost, as equal value. SPECIAL TEAMS DVOA is adjusted for type of stadium (warm, cold, dome, Denver) and week of season. Because it is early in the season, opponent adjustments are only at 90 percent strength; they will increase 10 percent every week through Week 10. 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 NE 40.0% 1 39.7% 1 8-0 25.2% 1 -7.3% 8 7.5% 1
2 ARI 36.2% 2 34.7% 2 6-2 19.3% 3 -18.4% 3 -1.4% 17
3 CIN 30.0% 3 28.5% 3 8-0 23.2% 2 -3.7% 10 3.1% 9
4 CAR 21.6% 7 22.4% 4 8-0 5.4% 10 -21.1% 2 -4.8% 28
5 NYJ 19.7% 5 19.1% 5 5-3 4.9% 11 -17.6% 4 -2.9% 25
6 GB 18.7% 4 17.0% 8 6-2 14.3% 5 -3.1% 11 1.3% 12
7 PIT 16.8% 10 17.1% 7 5-4 15.6% 4 -2.7% 12 -1.5% 18
8 DEN 16.4% 6 17.4% 6 7-1 -17.5% 31 -29.1% 1 4.8% 6
9 SEA 14.0% 9 15.5% 9 4-4 -0.9% 15 -8.4% 7 6.4% 4
10 PHI 13.0% 12 14.0% 10 4-4 -4.5% 22 -17.2% 6 0.3% 15
11 KC 12.3% 11 12.9% 12 3-5 7.7% 9 0.5% 13 5.0% 5
12 OAK 10.7% 8 13.3% 11 4-4 12.9% 6 4.1% 18 1.9% 11
13 BUF 9.2% 14 8.0% 13 4-4 9.7% 8 2.7% 16 2.2% 10
14 STL 3.6% 13 3.3% 14 4-4 -15.0% 28 -17.5% 5 1.0% 14
15 NYG 0.5% 15 0.6% 15 5-4 -1.9% 17 4.6% 20 7.0% 2
16 BAL -3.3% 16 -3.7% 16 2-6 -1.6% 16 8.2% 25 6.5% 3
17 ATL -5.2% 17 -5.6% 17 6-3 0.8% 14 4.4% 19 -1.6% 19
18 IND -7.4% 21 -6.8% 18 4-5 -4.5% 23 1.7% 15 -1.2% 16
19 WAS -8.8% 18 -8.1% 19 3-5 -4.2% 20 5.8% 23 1.2% 13
20 JAC -9.1% 24 -9.0% 21 2-6 -2.8% 18 0.8% 14 -5.5% 29
21 MIA -9.1% 20 -8.9% 20 3-5 2.7% 13 10.0% 27 -1.8% 22
22 NO -11.7% 19 -11.3% 23 4-5 11.4% 7 19.4% 32 -3.8% 27
23 DAL -12.9% 25 -12.9% 25 2-6 -4.2% 21 7.0% 24 -1.6% 20
24 TEN -13.7% 28 -17.0% 28 2-6 -16.8% 30 -5.0% 9 -2.0% 23
25 TB -14.1% 26 -10.7% 22 3-5 -8.9% 24 2.9% 17 -2.3% 24
26 MIN -14.9% 27 -12.4% 24 6-2 -13.2% 27 5.4% 22 3.7% 8
27 SD -15.3% 23 -15.9% 26 2-7 4.5% 12 14.0% 29 -5.8% 30
28 CLE -16.8% 22 -16.1% 27 2-7 -11.8% 25 9.6% 26 4.5% 7
29 CHI -21.6% 29 -21.2% 29 3-5 -3.2% 19 10.8% 28 -7.6% 32
30 HOU -23.5% 30 -22.9% 30 3-5 -12.6% 26 4.6% 21 -6.3% 31
31 DET -32.5% 31 -32.7% 31 1-7 -16.1% 29 14.5% 31 -1.8% 21
32 SF -36.3% 32 -37.9% 32 3-6 -19.1% 32 14.1% 30 -3.2% 26
  • 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 NE 40.0% 8-0 40.2% 8.9 1 -0.2% 18 1.6% 17 4.3% 4
2 ARI 36.2% 6-2 41.5% 7.3 3 -12.7% 31 5.3% 9 12.7% 19
3 CIN 30.0% 8-0 29.7% 7.5 2 3.4% 10 -0.9% 21 10.1% 16
4 CAR 21.6% 8-0 22.1% 6.7 5 -2.5% 23 -8.9% 29 2.2% 2
5 NYJ 19.7% 5-3 21.6% 7.0 4 1.6% 15 0.0% 18 7.2% 8
6 GB 18.7% 6-2 18.6% 6.4 6 -0.7% 19 -10.3% 31 23.2% 30
7 PIT 16.8% 5-4 10.6% 5.8 8 8.7% 2 2.3% 15 11.2% 17
8 DEN 16.4% 7-1 26.1% 5.9 7 -4.2% 26 7.2% 4 17.8% 27
9 SEA 14.0% 4-4 13.9% 5.3 10 -3.7% 25 2.7% 14 9.1% 12
10 PHI 13.0% 4-4 12.7% 5.3 9 -1.2% 20 2.7% 13 12.9% 21
11 KC 12.3% 3-5 14.3% 4.9 12 -1.3% 21 -0.5% 19 20.7% 28
12 OAK 10.7% 4-4 9.6% 5.0 11 3.2% 11 -2.1% 24 13.9% 22
13 BUF 9.2% 4-4 16.1% 4.3 16 2.8% 13 7.4% 3 14.5% 24
14 STL 3.6% 4-4 6.7% 4.5 14 1.1% 17 -3.4% 26 10.0% 14
15 NYG 0.5% 5-4 5.9% 4.8 13 -8.8% 28 8.8% 2 4.9% 6
16 BAL -3.3% 2-6 -4.0% 4.2 17 5.2% 7 5.2% 10 2.1% 1
17 ATL -5.2% 6-3 4.5% 4.3 15 -11.9% 30 -2.0% 23 3.7% 3
18 IND -7.4% 4-5 -13.8% 3.0 26 5.4% 6 -8.3% 27 4.7% 5
19 WAS -8.8% 3-5 -15.0% 3.8 19 6.1% 5 -1.8% 22 10.0% 15
20 JAC -9.1% 2-6 -13.5% 3.6 22 4.6% 8 -11.7% 32 8.1% 10
21 MIA -9.1% 3-5 -11.5% 3.8 18 2.9% 12 4.3% 11 28.6% 32
22 NO -11.7% 4-5 -12.5% 3.1 25 2.0% 14 -10.2% 30 8.0% 9
23 DAL -12.9% 2-6 -16.5% 2.9 27 8.0% 3 3.6% 12 5.2% 7
24 TEN -13.7% 2-6 -14.1% 3.7 21 -9.8% 29 5.4% 8 25.9% 31
25 TB -14.1% 3-5 -14.6% 3.4 24 -6.2% 27 -2.6% 25 22.5% 29
26 MIN -14.9% 6-2 0.9% 3.7 20 -13.2% 32 9.0% 1 14.3% 23
27 SD -15.3% 2-7 -13.5% 3.5 23 -1.4% 22 7.1% 7 8.4% 11
28 CLE -16.8% 2-7 -23.0% 2.9 28 9.4% 1 7.2% 5 9.8% 13
29 CHI -21.6% 3-5 -21.8% 2.5 30 3.7% 9 -8.5% 28 11.3% 18
30 HOU -23.5% 3-5 -17.0% 2.8 29 -3.1% 24 7.1% 6 15.2% 26
31 DET -32.5% 1-7 -32.0% 1.3 32 1.5% 16 -0.6% 20 12.7% 20
32 SF -36.3% 3-6 -36.3% 1.8 31 7.4% 4 1.8% 16 14.7% 25


79 comments, Last at 16 Nov 2015, 1:09pm

1 Re: Week 9 DVOA Ratings

Ya' know, Pat Williams didn't reach his prime until pretty deep into his career, so I think there is chance that Linval Joseph still has gigantic upside at age 27, and Joseph is now in the midst of a stretch of games which are All-Pro in quality. Now, Sharif Floyd is no Kevin Williams, but if he could manage to stay healthy for a season, he might be really, really, good, and give Zimmer a tackle tandem which may not exactly replicate the Williams Wall, but is also really, really, good, for many years. A creative defensive guy like Zimmer, with confidence that the middle of his front was always secure, would probably dial up some fun stuff to see.

3 Re: Week 9 DVOA Ratings

I've been impressed with Joseph. I'd still like to see the Vikes add another solid DT and DE (1 each) to build a 'rotation' and also allow them to get creative with personnel depending on the situation (3 DEs in 3rd-and-long, for example).

But first, the O-line...

2 Re: Week 9 DVOA Ratings

Out of curiousity, when is the difference between the Titans offensive DVOA with and without Mariota?

4 Re: Week 9 DVOA Ratings

I can't answer that, but the fact that they are 31st in variance, before only the Miami Dolphins and their Jeckyll-Hyde coaching carousel, is kinda telling.

5 Re: Week 9 DVOA Ratings

MIN is clearly ranked too low because cheese curds and walleye. Phil Simms team rankings is way better than this. You lot of pointy-headed intellectuals know nothing about mansport, raiderjoe was unavailable to assist with editing.

44 Re: Week 9 DVOA Ratings

It's just not the same until you show up, RJ. I'll see your Sierra Nevada and raise you a New Belgium 1554.

Also, your shiny new QB seems to be performing well. Raiders ready to start challenging for AFC West supremacy as of next year?

48 Re: Week 9 DVOA Ratings

I've noticed that when you're called upon to be 'Raiderjoe' you appear as 'Raiderjoe'. Otherwise, not so much.

Nice gimmick, FO, who I assume is the culprit, but it's old and obvious.

6 Re: Week 9 DVOA Ratings

I am curious, while I understand that you do not adjust your playoff sims for every injury, about how many percentage points of DVOA do you expect Dion Lewis' injury to cost the Patriots?

14 Re: Week 9 DVOA Ratings

Well they lost Lewis, but just got LaFell back, so that might largely cancel. I'm more worried about the fact that their o-line is currently a LT manned by a practice squad RT, a LG who was terrible last year, an Undrafted Rookie Free Agent at center, a rookie RG, and a C playing RT.

15 Re: Week 9 DVOA Ratings

When you put it that way, it sounds like the O-line is doomed. But actually they're playing pretty well. The rookies have been playing very well. Stork has experience playing off-center. Cannon and Vollmer will both eventually be healthy.

But I really thing somebody needs to pay off that old gypsy woman who put a curse on the line. Pretty soon they'll be playing Bruce Armstrong at tackle and John Hannah at guard.

32 Re: Week 9 DVOA Ratings

I still don't get how you can watch their games and come out with 'pretty well'.

The only thing saving that offense from being an abject disaster is the fact that Brady's release speed is historically good, and making the offensive line almost irrelevant. (Scheme is helping too)

CBS sports released an article earlier this year (week 3) with data that I believe is gleaned from PFF showing release times for each QB - going into this year the fastest ever was P. Manning's 2014 - at 2.24 seconds post-snap (2nd was Favre at 2.41) . As of week 3, Brady was at 2.08. He also was one of the fastest QBs in the league to get sacked.

From my observation - it seems like he's only gotten faster as the season has moved on - I don't have PFF, so I can't check, but I'd guess he's still in "Fastest Ever" territory.

34 Re: Week 9 DVOA Ratings

Yeah, Brady is carrying a great deal of weight right now. You could see how strained he was this past week to get the ball out immediately on every throw. I bet the coaches were grateful that they were facing a team that gave them the luxury to run the ball so effectively because I don't think this group could have held up for one of those 80/20 pass/run splits.

35 Re: Week 9 DVOA Ratings

It would appear to me that they are most vulnerable to a team with a good rotation of interior defensive linemen, who can get fast pressure, while defending the run, and a good defensive backfield. Of course, that a pretty good generic definition of a good defense, period, but I'll have to think about who best meets that standard.

8 Re: Week 9 DVOA Ratings

I'd be surprised if they don't. As a Vikes fan, I can tell you that even without looking at DVOA, their 6-2 records seems like a hoax. They haven't had any real convincing games. I am surprised that the defense is ranked so low, because the Vikings' D looks fairly decent. But overall...yeah, I'm not buying it.

The strange part is that, had they actually played their 5th preseason game (San Fran) like an actual season game (it was), they'd be 7-1. And feel even more like a hoax.

10 Re: Week 9 DVOA Ratings

Then you consider they were about two plays away from beating Denver, and an 8-0 record, which would be as fraudulent as the 2000 team that made it to the Conference Championship, with one of the worst defenses seen by humankind.

Well, this bunch is a long way from being one of my least favorite Vikings team, and not because they have a shot at a winning record. They have their obvious limitations, but they are well coached, play as physically as their talents allow, with a lot of intensity and intelligence. I could do worse (and have) with a team I watch after Thanksgiving. If they are still in the chase for a tourney berth at kickoff for the week 17 game in Lambeau, that'll be a successful season

26 Re: Week 9 DVOA Ratings

I can't get behind their rank here at all. If we just made four broad categories (1-8, 9-16, 17-24, 25-32), I think they're firmly in the 9-16 bracket. SOS might be masking that they're closer to 16 than to 9, but the Vikings are well better than the bottom quarter of the league. And 22nd in Defensive DVOA? That doesn't even come close to passing the smell test.

28 Re: Week 9 DVOA Ratings

There'll always be an argument. If they play well against the Raiders, that could just be the limitations of a Musgrave offense against a Zimmer-style defense (they did struggle against the Bengals, too). OR it's the week-to-week inconsistency of a young quarterback, etc. If the Raiders get the upper hand but the Vikings have key players out, does that speak to the defense as a whole or to the depth in particular? OR is it that traveling to the West Coast is kind of, ahem, problematic sometimes?

I devised a crude metric just now for scoring defense. Nothing revolutionary, just took our opponents' average points in games they didn't play us and compared it to what they scored on us. On average, they scored a field goal less per game against our defense that expected. (If you remove that looking-worse-every-week San Francisco game, it improves to 4.5 points per game. The Niners are the only team the Vikings have faced so far this year that have outperformed their average.) I suppose now I should do that for the rest of the league to see how meaningful it actually is.

29 Re: Week 9 DVOA Ratings

Well, of course, the reason to watch the games is because you can't really know for sure what you are going to get. The next three offenses they play are good, period, and they'll be playing two of them on the offense's home field. If the Vikings defense does well against those 3, or even 2, and the Vikings defensive rank remains in the bottom third of the league by DVOA, we can be fairly confident that DVOA is missing something about the Vikings defense. I suspect that if the Vikings defense has a good three weeks, they'll jump to the top half of the DVOA ranks.

41 Re: Week 9 DVOA Ratings

I find myself disagreeing with you less and less. Hopefully that's a compliment.

I did that scoring analysis for the rest of the league and then repeated it for points scored. Overall, the Vikings allow about a field goal less than expected per game, good for 9th in the league. On the other hand, the score about a field goal less than expected, which rates 20th in the league. Summing the two puts them at 16th overall, which feels right.

21 Re: Week 9 DVOA Ratings

4th game in a row winning while losing the turnover battle. That is very odd. They just don't seem like a good enough team for that to happen.

I'm pretty much with you, but I think they are better than DVOA suggests, but not a lot. I'd put them middle of the pack.

16 Re: Week 9 DVOA Ratings

Packers are coming off consecutive road games against undefeated teams. It'll get easier for them, too. And they should be happy to have faced real challenges in the middle of the season, while there's still time for the coaching staff to fix the flaws that have been exposed.

9 Re: Week 9 DVOA Ratings

Curious why the Vikes/Rams only moved one spot up/down respectively from last week to this. It would seem like the algorithm would have a larger swing when a team that it considers much worse than the Rams beats them.

11 Re: Week 9 DVOA Ratings

DVOA is a per play metric, which only roughly correlates with winning and losing, especially with regard to a very close score. A lot of random events determine the winner and loser of a close game.

42 Re: Week 9 DVOA Ratings

What Will said - also, even though the Vikes have a worse DVOA than the Rams, and just beat them (in a close game, evenly matched, in OT), it was also in Minnesota, and DVOA accounts for home-field advantage (or travel 'disadvantage') as well.

Per recollection.

13 Re: Week 9 DVOA Ratings

The fact that the Steelers are 4th in offense is amazing. For half the season, Ben has been out and their backup is on IR so they've played their third and fourth string QBs. Their team MVP and starting RB has missed more than three games, their Pro Bowl center hasn't played, their #2 WR and top deep threat has missed 5 games and their starting LT is on IR.

17 Re: Week 9 DVOA Ratings

DeAngelo Williams has been excellent when he's been called to fill in for Bell. He's far from a typical back-up RB. Oh, and Antonio Brown is playing like the best WR in the NFL. The Steelers are really loaded with talent on offense.

And really, the injury to Vick didn't hurt them in the slightest. Landry Jones has been better.

18 Re: Week 9 DVOA Ratings

Bengals are No. 1 in the Week 9 BES Rankings - http://besreport.com/week-9-bes-rankings/

They have quite a few impressive wins over the last month and a half....slightly more impressive than the Pats over that span.

Wacky stuff with the Saints at No. 4 despite their loss to the Titans. Maybe the saints could be on the verge of a run. The BES had the Colts as high as No. 10 a couple weeks back. They gave the Pats and Panthers hell but finished against the Broncos.

The BES was likely also the only rankings out there that had the Bears higher than the Chargers going into MNF. Bears have been competing, just unable to close games out. Nice win over the Bolts though.The Bucs are also getting an arguably deserved amount of love from the BES as well. They've been a real problem for opponents.

25 Re: Week 9 DVOA Ratings

Are stats adjusted for time and score of the game?

For example, Texans were down 0-41 in the 4th quarter against Atlanta. They put up 21 points in the 4th quarter and had some long yardage drives that increased there yards per play quite drastically when Atlanta was playing a different style of defense; less effort, less creative play calling, more back-ups, etc. Those yards should count, but should not hold the same weight as team that drives 80 yards, at a good yard per play average, in the 4th quarter when the game is tied.

33 Re: Week 9 DVOA Ratings

Your understanding is correct - but they refuse to say what their definition of "Garbage Time" is - so there's no way to tell if it's a reasonable definition.

I remember this being a big complaint in 2007 with the Patriots where a 9 minute 14 play drive by an opponent who was down 40 points would heavily knock their defense. FO refused to define what the cutoffs were.

51 Re: Week 9 DVOA Ratings

Wait, what?

I had no idea. I've been 'getting into' DVOA because I thought it specifically addressed plays most influential to wins (WPA) and certainly not garbage time stats. I thought it disregarded those stats almost entirely. If this is true how could DVOA possibly be predictive? Garbage time numbers comprise a good majority of the overall, but they're useless vis-a-vis wins.

53 Re: Week 9 DVOA Ratings

Except that the data FO has and uses shows that they're (garbage time stats) not useless at all. Counter-intuitive could be good sometimes.

I'm not positive it is in this case, but they've yet to find a way to make the formulas more predictive by throwing out garbage time.

57 Re: Week 9 DVOA Ratings

There was a discussion of a similar situation last year:


The final Pats/Bills game was basically 100% garbage time because the Pats had nothing on the line and didn't really try. At least in theory, not including/discounting that game should improve the predictive power of DVOA, but the Outsiders have not found a way to do so.

55 Re: Week 9 DVOA Ratings

You've missed the point of DVOA entirely, and probably WPA, too.

WPA isn't a predictive stat primarily. It's a backward-looking, narrative, "story of the game" stat. It measures the importance of each play to that one game. While it's true that playing skillfully gets you higher WPA, getting the same result in a tie game early and in a close game late can have wildly different WPA impact, despite demonstrating the same skill against an opponent trying just as hard.

DVOA is about measuring the predictive skill shown in each play, as picked up in the play by play. It turns out that "garbage time" plays are predictive. Sure, if you run the ball all the time when up by two scores, you aren't calling optimum plays for advancing the ball. But it still says something about your future prospects if you gain six each time or get stuffed each time. Even back ups playing against back ups (which seldom really happens) says something about your roster depth.

FO could turn DVOA into a WPA-like "story of the game" stat by weighting each situation with a leverage or "clutch" factor. But that would weaken it as a predictive stat.

37 Re: Week 9 DVOA Ratings

Eagles at the midway point:
The bad: based on cost & performance Bradford, Murray & Maxwell have proven to be three of worst off-season moves in the past two decades for the Eagles. Somewhere along the line, expectations for all of them plummeted, so now fans seem to be ok with them and under the impression they are improving. They are not. Bradford dropped this week to 28th in both DYAR & DVOA, Murray has been hanging around 26th all season and Maxwell receives safety or LB help on nearly every play.

What was true and obvious about all of them before the season remains true and obvious: Bradford is among the worst starters in the league and not an improvement on Foles or Sanchez, Murray was in no small part a product of the Dallas offense (and offensive line), Maxwell is a penalty-generating machine who is not a top CB or a notable upgrade on Cary Williams.

The really bad: the WR's are terrible. Miles Austin might be the best one (DVOA thinks so), but it's really a crapshoot (in the sense of shooting at crap.) Jordan Matthews is no way, shape or form able to carry the load of a #1 and his lack of football skills (catching, route-running, body control) are miserable. He can get wide open on crossing routes 5 yards past the LOS, but that's about it and half the time he'll drop the ball anyway (catch rate 64%.) Cooper and Huff are spare parts at absolute best and really, truly just terrible at worst. Agholor gets a mulligan - the rookie did nothing of note in limited action but he's a rookie and now injured. The WR corps is a real tire-fire.

TE is just as bad. Celek is too old to be particularly productive in the passing game. Ertz has not developed at all since his rookie year: he can't block, he runs terrible routes and as a consequence he disappears for games at a time. He's 38th in DYAR, one of the very worst pass-catchers in the league amongst qualifying TE's. That he can't block is just such a disaster: if you block worse than Jimmy Graham, you better be worth something in the neighborhood of Jimmy Graham to the passing game. This off-season, TE has got to be addressed - they can't go on like this.

The predictable: the o-line is a mess. Peters got injured, which anyone with a brain knew was a huge risk going into this season. There's no depth (which again, was plain as day), so Kelce and Johnson are having to do all the work. There's almost no inside running game whatsoever because the G's get pushed around. Pass protection is ok. The whole thing is chaotic and when they faced a good defense in Carolina, they looked utter hapless. It's hard to imagine what their plan is for the future here. They have one above average player (Kelce), one average player (Johnson) and a bunch of question marks (and initial returns are that they're not-starting-caliber.)

The secondary is a mess. Malcolm Jenkins (despite a stinker in Dallas) is playing like an All-Pro. Walter Thurmond has been an improvement at S over Nate Allen. Carroll & Maxwell for Fletcher & Williams has been a wash. Buuuut... there's no depth. Rowe looks as much like a bust as a rookie can after 8 games. So, when the they have to play nickel or dime, they're bringing in journeyman special teamer Chris Maragos to take over at S for Thurmond and switching Thurmond down into the slot (where he has been less effective) or bringing on journeyman E.J. Biggers. Biggers has been pretty ok, but neither Thurmond or Bigger is anywhere near as good as Boykin was for the Eagles - overall, the secondary is a touch worse than last year: their base 4 DB set is improved, but their nickel and dime sets are far worse.

The good: the LB corps has been a mess, but there's so much depth and so many talented players there that it hasn't mattered. Alonso, Kendricks, Ryans and now Hicks have all suffered injuries but I haven't seen the unit miss a beat. Barwin has slowed down in terms of sacks, but they've asked him to pick up a lot of the slack for other players during the reshuffling. He's been very. very solid. Graham seemed to take a little time adjusting to being a full-time starter, but he's come on recently including his humiliation of the best T in the game last week in Dallas. He abused Tyron Smith and hopefully that's a sign of things to come.

The great: This d-line. Any one of Cox, Logan and Thornton would be an acceptable Pro Bowl selection and if Logan doesn't make the All-Pro team, then the system is broken. (spoiler: the system is broken.) It almost doesn't matter who is back there in the secondary or who lines up at LB because these three are going to do 90% of the work anyway. When teams try to punish the Eagles for their lack of depth at secondary by lining up 4 or 5 wide, the d-line just eats them alive. It is unwise to try and pass block against these three guys without a TE and RB staying in to block. They're amazingly good, certainly the best group the Eagles have had since Reggie White's day - that you can even speak of them in the same sentence as the White, Simmons, Brown group is really astounding. If Cox, Logan, Thornton did a Pro Bowl trifecta like those guys in 1991, it wouldn't be an outrage.

Jordan Hicks is a stud. It's a killer that he is out for the year because he was having the best rookie season I can recall an Eagles having in a long time - I'd place him above even DJax in terms of immediate impact: he was the difference in the wins over the Jets and Cowboys and made at least once spectacular play every single game. Just a fun, exciting player.

Ryan Mathews is amazing. If he had another 3 carries (and assuming each of those carries was average) he'd qualify as a "starting back" by DVOA... and he'd be #1 in both DYAR and DVOA. He'd be leading in DVOA by a wiiiide margin, to boot. Every carry the Eagles give to Murray and not Mathews is a wasted carry - Mathews was a true #1 in SD, so we know he can handle an increased load. He should be getting the ball 15 times a game and Murray should be the change of pace back. It's plain as day Mathew's quickness and ability to get to the outside is a major asset in this offense. His only downside is that he fumbles too frequently.

The coaching is what it is: the offensive play-calling is super repetitive and relies on superior conditioning for successful drives. They run the same 8 or so plays over and over and over, they seriously seem to have only 6 running plays in their entire playbook (which is why I think a Mathews for Murray switch is so crucial - his great numbers aren't the product of fooling teams with a change-up/3rd down scatback, he's beating them with the exact same play run on the exact same down and distance as Murray, the modified Lombardi sweep that is the running play they call 15 times a game.) Their conditioning and pace are key: they rely on defenses getting worn out in order to get bursts of points in bunches on consecutive drives. It really works: you'll see them shut down for long stretches every game and then suddenly able to score almost effortlessly on several drives in a row.

Their quick pace also neutralizes the opponents' blitzing. You could see it in the Dallas game: they came out blitzing, Bradford wilted as per usual and the offense stunk up the joint for the first half. Later in the game, as the defense was winded and taking longer to line up, they stopped blitzing altogether (despite its major and obvious success) because the coaches couldn't get in calls quick enough. The fast pace really takes away one of the strongest weapons against the offense: the blitz causes Bradford's total panic and confusion on the hastily assembled o-line and its crummy interior.

You see the strategy stumble when the opponent is able to control the ball (several games early on, also vs. Carolina) when the defense doesn't get winded. The offense spends those entire games sputtering and there's no individual player on offense who can make a great individual play and spark them (apart from Mathews who won't be given the opportunity.) Games like the one versus Carolina or the first one versus Dallas are just painful to watch - when the pace isn't working to their advantage, they look like the worst offense in the league.

I think my preseason assessment that they'll end up somewhere around 10-6 seems very reasonable. Their schedule gets super soft for the next three games, versus teams with bad-to-awful defenses (Detroit, Miami, Tampa Bay) - I've said a few times I think they'll be 7-4 before running into the buzzsaw stretch New England, Buffalo, Arizona. They'll close out the season versus divisional opponents who play them tough and be something like 10-6. Their weaknesses are more than enough to send them spiraling down to 6-10, but their strengths are not enough to get them to 11-5. I just don't see. They could win the division, of course - it wouldn't shock me if they were able to make some noise on the strength of that front seven (of course, injuries have plagued LB so there's no reason to think a fully healthy Alonso, Kendricks and Ryans are a sure thing for the remainder of the year.)

59 Re: Week 9 DVOA Ratings

I think Thurmond has been good - significantly better than "an improvement over Nate Allen". And I think Mathews has merely been good - though still significantly better than Murray. Otherwise, I agree with all.

60 Re: Week 9 DVOA Ratings

Check Mathew's DVOA and the difference in production - Mathews has inarguably been among the very best RB's in the league. Again, Mathews and Ryans have alternated series (as opposed to downs within a single series) so you can see Mathews being productive in the exact same context as Ryans - there's no way to waive off his production. Seriously, with 3 more carries to qualify he'd be #1 in both DYAR and DVOA by a healthy margin. That's remarkable.

Also, I think Thurmond's performance at S has been getting overrated by Eagles' fans because he's such a ballhawk. He's frequently very late in getting there when he has over the top responsibilities - it's the entire reason Dallas was picking on Nolan Carroll. They clearly noticed the problem. He's good, but he's not anything special at safety. And he's only average in the slot - Biggers has played better. Still, I LOVE having him back there, there's reasons to think he'll keep improving and he gives them a versatility that is absolutely essential because of their lack of depth. He's one of Kelly's few roster moves to be really proud of. I like him of only BECAUSE he's not Nate Allen!

61 Re: Week 9 DVOA Ratings

Ok, yes I misphrased my Mathews comment. Obviously, by the numbers he has been having an amazing season. I think it's a mistake to say that he *is* amazing. He's a good running back who should clearly be the lead back, but I am very skeptical he would put up the same numbers with as many carries as Murray. I think Mathews' 2011 and 2013 seasons are more what one should expect of him over the run of this season.

As for Thurmond, I never claimed he was anything special. He has flaws, as you point out, but unlike Nate Allen, there's a lot he's good at! The Eagles haven't had two effective safeties since Dawkins and Mikell were together.

62 Re: Week 9 DVOA Ratings

Yeah, I don't want to crap on Thurmond who has been a pleasant surprise in a season with almost none of them. He's on the border of being interesting - I think it's going to be a tough question about what to do with him this off-season. Depending on what happens at QB, they could be very strapped for cap space and I'm not sure he's done quite enough to warrant getting real paid so the market for S/CB's and their cap will make his status in the future a real question...

And I think we agree on Mathews - his numbers would likely drop with more carries, but he'd have to utterly collapse to not be a significant improvement on Murray. I'm not convinced the numbers would drop THAT much though - really, if they switched it so he was getting 15-18 carries a game and Murray 8-10, I'm not sure that's a big enough of an increase in workload for him to have any discernible effect. And as I mentioned, they run the same play over and over with him or Murray in there so it's not like there's anything for defenses to wise about with more tape. His quickness/acceleration is just so much better of a fit for this offense. I really wish Kelly hadn't panicked and signed Murray - Sproles, Mathews, Polk would've been a perfect acceptable back-field. (But hell, if we're rewriting the past, Kelly shouldn't have cut Dion Lewis.)

64 Re: Week 9 DVOA Ratings

I like your analysis. I hope that Kelly can rebuild that O-line before getting launched because when he gets his inside/outside running game going with the tempo it is a thing of true beauty. At times it gets to the point where the DBs or LBs won't even get lined up properly after a big run and a WR will be running all by themselves on a packaged play. With the Defense that Phi is currently running it would be brutal combo.

I remember when they were the Sea-chickens.

69 Re: Week 9 DVOA Ratings

Yeah, it's really astounding to see in real time the effect the system and pace have. Every game this season, they've had big stretches where the offense looks utterly incompetent... and then they get rolling and the defense suddenly looks entirely helpless to stop them.

To me, as a coach and not a GM, Kelly's biggest issue is his inability to develop players. Every QB he's fielded has been mediocre-to-awful (without what's now clearly a bizarre anomaly in Foles' 7 game stretch in 2013), every receiver he's developed has the same major deficiencies several years running (Ertz, Huff, Matthews), in terms of o-line he's still dependent on Reid and his well-known ability to develop players on the line (Peters, Kelce, Kelly), he's overlooked homegrown RB's who have gone on to be productive elsewhere (Polk, Dion Lewis) in favor of FA's, his CB's are consistently below average and his best players on defense where all FA's or Reid players (Cox, Graham, Alonso, Barwin, Kendricks, Jenkins, Thurmond, Curry, Ryans.) Even his special teams ace, Chris Maragos was developed by Pete Carroll. Basically, he's got Bennie Logan to his credit and a whole lot of underperforming players and outright busts.

There's still reason to think Hicks and Agholor could be extremely productive, but right now players like Jordan Matthews, Zach Ertz and Lane Johnson look overwhelmed in the roles Kelly has cut out for them. And those are the best ones. The problem with building a team around FA's is that you've got to nail most of them. If you whiff on pricey acquisitions like Maxwell and Murray, you've going to find yourself in cap trouble quickly.

72 Re: Week 9 DVOA Ratings

"I think my preseason assessment that they'll end up somewhere around 10-6 seems very reasonable. Their schedule gets super soft for the next three games, versus teams with bad-to-awful defenses (Detroit, Miami, Tampa Bay) - I've said a few times I think they'll be 7-4 before running into the buzzsaw stretch New England, Buffalo, Arizona. They'll close out the season versus divisional opponents who play them tough and be something like 10-6. "

Still insane to me, because they still *look* like the team I thought they would be, which I had no idea how they were going to get to 8-8. The difference, I think, is that their schedule turned a lot easier. I didn't expect the Redskins to stay crap, or the Lions to fall apart, and definitely didn't expect Dallas to fall apart.

Looking at the rest of their schedule, though, I think they've got 3 likely losses (New England, Arizona, and Buffalo), which pretty much puts 9-7 as the upper end. They don't have a chance in hell of beating New England or Arizona, in my mind, so 10-6 is an absolute top. But Buffalo with Taylor, I think, is going to be a real problem.

73 Re: Week 9 DVOA Ratings

It certainly seems like the Eagles have a top defense even with Maxwell being disappointing (or exactly what you thought he'd be if you're a certain type of fan). That's how they're winning games.

And like chemicalburn points out gaining an advantage in plays run.

74 Re: Week 9 DVOA Ratings

Yeah, I agree that their schedule is really what's propping them up. I mean, in the preseason you couldn't know they'd face Brandon Weeden without Dez Bryant (or Hardy & McLain on defense) and Matt Cassell (without Sean Lee on defense), Washington without RGIII or DJax, the Giants without Victor Cruz and a hobbled OBD & Randle, that Miami would be a total joke and that Detroit would collapse. Things have really, really fallen their way in terms of schedule. It was a soft schedule that became ice-cream soft. They only have 3 more away games: one is in Detroit while the another is the NYG who the Eagles have owned for a decade and a half. And it could keep getting softer - I mean, you'd have to be crazy to rule out Taylor & McCoy missing the Buffalo game or Carson Palmer being hurt by the time they play Arizona in late December.

I agree that it's fair to assume losses in that New England through Arizona stretch... but I gotta say I'm interested to see how they look versus the Patriots. Their d-line is exactly the kind of group that gives New England, they get pressure without the blitz and collapse the interior. The strength of their secondary is covering short passes and they have a some of the best cover LB's in the entire league plus a safety in Malcolm Jenkins I wouldn't rule out being able to handle Gronkowski. Oustide of JJ Watt, no one bats more balls than Barwin and Cox is not a slouch in that department either - they're both players who are acutely aware of getting into short passing lanes and disrupting quick passes. They're totally prepped to deal with the hurry-up and don't get worn down. They're just an abnormally good match-up for stopping the New England offense.

On offense, they're interesting because Kelly's system isn't difficult to understand, it doesn't win by complexity. It's super-repetitive. It doesn't fool a defense, it stretches them to the limit vertically and horizontally on every single snap and wears them down. In that respect, I'm not sure any of that peerless Belichick preparation and scheming will be useful. It's generally simple defense with a massive amount of individual talent (Seattle, Carolina) that shut it down, not clever ones. I'm very curious to see how they look against New England because I don't think you can out-coach a Kelly offense, you're either better than it or not. And at that point, if Mathews has been handed the reigns, I think they could have a real ballgame.

Of course, Buffalo is a horrible, horrible match-up for them, so even if they pull of a victor in New England, I'd expect a "letdown game" the following week where they get stomped at home by Buffalo. It's actually harder for me to imagine them beating Buffalo than New England.

75 Re: Week 9 DVOA Ratings

Thinking of your analysis about the PHI D and what Kelly does on offense I begin to wonder if they're a really bad match-up for NE.

PHI's D is built inside out, front to back, which is to say they are strong in the area of the field where you run high percentage/high effeciency plays and bad on the area of the field where you give up 2 min scoring drives. They may encourage Brady to repeatedly bomb the field on them.

Then on offense they're going to just run what they run, and run it fast. If NE gets tired and can't make play calls, combined with NE's offense either punching in for a TD or a punt in under 2 minutes. I wonder what that does to NE's D over the course of the game?

I remember when they were the Sea-chickens.

76 Re: Week 9 DVOA Ratings

Giving up quick scores is actually, in a perverse way, to the Eagles' advantage. Like most teams built from the run'n'shoot model, getting into shootouts works in their favor. There's almost no game where they score at least 30 points and lose - even if their defense gives up a lot of points as well. I mean, they just tire the opposing defense out and don't give them time to get reset in between plays. I'd take the Eagles 9 times out of 10 in any mildly high-scoring OT game just because of the conditioning factor. An end like the one we just saw versus Dallas is extremely likely.

Also, Brady has not been able to bomb teams as well as he used to for going on two seasons. I think if they can hang with Gronkowski - and they are better suited to do so than just about any team in the league - then Brandon LaFell and Amendola on deep routes aren't going to beat them.

Keep in mind that almost every team they faced has had to keep their TE in to block in order to handle the d-line. If you neutralize Gronk (easier said than done) New England is an entirely different team. Philly is 6th in the league vs. TE's (and 2nd against RB's) and that high defensive efficiency is combined with really low production because so much blocking help is required - TE's just don't get that many targets versus the Eagles. Plus, the defense is so used to seeing TE/RB's stretch the field and slip out on tricky plays.

Greg Olsen's 3 catch 65 yard game is the best game BY FAR a TE has had against them all season. Jason Witten with 7 catches/56s yard is the only other TE to have more than 35 yards against them. The Giants really tried to attack them with the TE (in part because all their WR's were injured/crappy) and it blew up in their face.

It's very hard to use your TE as a weapon in the passing game versus the Eagles and even harder to use your RB. It's hard to throw quick passes, WR screens and run pick plays against them. I mean... New England is just a very good matchup for them. (In contrast to Buffalo: the Eagles are very vulnerable to deep passes and QB's who can keep the play alive with their feet and a power RB's like Williams. Buffalo's TE's are an insubstantial part of their passing offense and they like keeping them in to block. I expect the Buffalo game to really resemble the Carolina game.)

Who should your money be on? Your money should be on New England. New England is a very good team and the Eagles kinda stink. But of the handful of tough games the Eagles have remaining, I think the one versus New England is the one where they're most likely to pull off an upset.

77 Re: Week 9 DVOA Ratings

I'd put my money on NE. But if NE is favored by a TD or more it would be awfully tempting to take the points.

Is Buffalo's D going to be a good match-up for PHI? The college spread-option-no-huddle concepts seem to demolition 3-4 teams that like to blitz for disruption. Though as you point out Buffalo's O is likely going to be-able to exploit the PHI D.

I remember when they were the Sea-chickens.

78 Re: Week 9 DVOA Ratings

Buffalo has a penetrating d-line with a strong interior and fast, cover LB's - that's just kryptonite for this offense. You can see pretty clearly this season, the better/smarter the cover LB's (Carolina, Dallas with Sean Lee) the worse the offense does. You can compare that to how they look versus the Giants, Saints or Dallas (without Sean Lee) who have bad LB's. That's why a good defense with so-so cover LB's (but great CB's) like the Jets didn't look so hot against them, too. There's no premium to be had on shutting down the Eagles' outside receivers, it's all about being able to handle coverage within 7 yards of the LOS. Buffalo does that very well - you saw it all game long versus the Jets.

Similarly, it's also why I think Philly will have no shot against the excellent Arizona S-hybrid defense that puts an emphasis on short and interior coverage and was designed to handle the stretch/option offenses in Seattle and SF. They have bulkier coverage specialists coached to fight through screens/picks and hammer crossing/flat routes... and they can also run up the seam with TE's/slot guys and tackle on running plays.

79 Re: Week 9 DVOA Ratings

I can see that. I think the future when more Air-Raid style concepts and option style concepts come to the NFL will be 4-2-5 base defenses with two SSs.

I remember when they were the Sea-chickens.

39 Re: Week 9 DVOA Ratings

Why is Pittsburgh higher than Denver? Does the system not account for the Roethilsberger/ Bell injuries?

40 Re: Week 9 DVOA Ratings

The system sees that Landry jones played well in relief and that Williams played EXTREMELY well in relief. Plus, it sees that Manning and the Denver running game have stunk up the joint 7 out of 8 games. The level of drop-off from the starters to Williams and Jones in Pittsburgh has been far above the baseline in Denver...

You have a team with an excellent defense and terrible offense being rated lower than a team with a very good offense and a mediocre defense. Not that strange - DVOA is also saying they're pretty equal overall, which makes sense. Denver has superior ST, but Denver's unit is not notably brilliant nor Pittsburgh's notably awful (it gets dragged down by their early struggles with FG kicking.)

52 Re: Week 9 DVOA Ratings

Except that if PIT, without Ben, played DEN, home or abroad, PIT would likely be shut out. But even if not, would PIT ever be favored in either of those games? No. No way.

45 Re: Week 9 DVOA Ratings

There are tweaks in the playoff odds for injuries (not just Pittsburgh), but DVOA just looks at what happened. There's no allowance for which players are on the field. So the Steelers' rating includes plays with Roethlisberger, Vick, and Jones at quarterback, Le'Veon, DeAngelo, and Jordan Todman at running back, etc. and says this is what they've done so far. If Ben is healthy for the postseason, we'd expect the team to be better than a number that includes all three QBs, but DVOA is just based on what's actually happened, not what could've/would've/should've happened.

46 Re: Week 9 DVOA Ratings

Well, DVOA also has a predictive element, so to say isn't only descriptive isn't precisely true: it looks at what happened and uses that info to project what will happen going forward. So, it's not just saying Pittsburgh has played (ever so slightly) better than Denver, it's saying it thinks Pittsburgh will (ever so slightly) be the better team going forward as well.

66 Re: Week 9 DVOA Ratings

I'm not sure this is accurate. It tries to *be* predictive by describing accurately how each team has played and weighting certain things more than others, but it is all still based on what has actually occurred on the field. Only in the playoff odds do they manipulate the data to account for data that may change performance going forward.

68 Re: Week 9 DVOA Ratings

Well, as I said, it's a mix of descriptive and predictive. I'm not sure what a system based on stuff other than "what has actually occurred on the field" would look like - I suppose you're saying it would try harder to account for injuries? Anyway, I'm right because you can see it clearly in how they handle fumbles. If they were just trying to be descriptive, they wouldn't treat fumble recoveries they way they do and emphasize their "non-predictive" nature. That's just one of the clearest examples. Its love of constant short gains versus "boom or bust" offenses is another.

Truthfully, one of its bigger faults is that it sometimes gets a little neither fish nor fowl - the playoff rundowns try to compensate for that by leaning on prediction.

70 Re: Week 9 DVOA Ratings

Those are excellent examples of predicting that I overlooked, but I still think calling the system descriptive rather than predictive is accurate relating to the question that inspired this discussion. I agree fully with your last sentence, and while I know that irritates some people, I'm not bothered by it much. It's probably because I don't gamble or play fantasy; I can enjoy the insight without needing to know precisely what's coming next. :)

71 Re: Week 9 DVOA Ratings

There are many examples where DVOA either discounts or ignores something very important in a descriptive context because it isn't meaningful in a purely predictive context (like fumble recoveries).

It deliberately sacrifices descriptive ability in places for prescriptive power.

63 Re: Week 9 DVOA Ratings

DVOA calls PIT and DEN essentially equal. It also notes that DEN has played a much weaker schedule than PIT. If they were to play it should be a tossup with home field a factor.