Writers of Pro Football Prospectus 2008

Most Recent FO Features


» The Week In Quotes: October 21, 2016

We've got players and coaches who are cranky about losses, commentators who are cranky about players, and middle-aged men cranky at these damn millenials, and yes, Bill Belichick cranky about tablets.

28 Oct 2014

Week 8 DVOA Ratings

by Aaron Schatz

Over the last couple weeks, Denver has built a commanding lead in the Football Outsiders DVOA ratings. After the games of Week 8, the Broncos' lead is even more commanding. The Broncos' total DVOA actually goes down slightly compared to last week because of changes in opponent adjustments of the teams they've played so far this year. But most of the teams directly behind the Broncos lost this past weekend. Of the top eight teams in DVOA a week ago, only Denver and Seattle won in Week 8, and the Seahawks won just barely. The Broncos still rank sixth among the best teams in DVOA history. The playoff odds report now gives them over a 30 percent chance to win the Super Bowl with no other team any higher than 10 percent.

Best Total DVOA Through Week 8,
Year Team W-L DVOA
2007 NE 8-0 73.6%
1991 WAS 7-0 62.8%
1996 GB 6-1 54.0%
2007 IND 7-0 48.7%
1994 DAL 6-1 48.2%
2014 DEN 6-1 47.6%
2003 KC 8-0 45.2%
2006 CHI 7-0 44.3%
1999 STL 6-1 43.2%
2013 DEN 7-1 42.9%
1998 DEN 7-0 41.9%
1990 CHI 6-1 41.1%
2009 NE 5-2 41.1%

But while Denver is clearly the best team in the NFL this year, there are huge disagreements between the DVOA ratings and conventional wisdom -- or simple win-loss records -- when it comes to the next few teams. There are five teams in the NFL right now with six wins, but Denver is the only one of those teams currently in the DVOA top ten!

So, why doesn't DVOA rate these teams highly? Let's take a look at each one to see if we can figure out what's going on.

Dallas (8.5% DVOA, 11th): Between the DVOA commentary of two weeks ago and the article I wrote last week for ESPN Insider, I've written a lot about the Cowboys recently. They fell from seventh to 11th in DVOA after their loss to Washington last night.

New England (7.0% DVOA, 12th): In the last three weeks, the Patriots have somehow gone from 10th to 11th to 11th to 12th in DVOA despite winning every week. The Patriots' actual DVOA rating did go up slightly from last week (when it was 5.7%) but they drop one spot in the rankings because a) Pittsburgh scored a zillion points on Indianapolis and moved into the top ten and b) their DVOA rating for the rout of the Bears wasn't as impressive as you might expect. The Patriots only ended up with 17.4% DVOA for this game despite winning by four touchdowns. Some of this has to do with long Chicago drives that ended up with no points. The Bears went 55 yards on their first drive of the third quarter and then 54 yards on their last drive of the game, but then both drives stalled out because the Bears were so far behind that they were forced to go for it on fourth-and-10. Because the Patriots had such a huge lead, these drives only count for about half-strength in the DVOA system, but they still come out as the Patriots allowing the opposing offense to march the ball down the field with successful plays.

The real issue with the Patriots over the course of the entire season is their schedule. Early on, you might remember that the Patriots had the best defensive DVOA in the league... or rather, the best VOA because we had no opponent adjustments yet. And we all thought their schedule had been easy, but that early in the season, you can't fully tell. (Remember, after three weeks Kansas City looked bad and Atlanta looked like a juggernaut.) Well, now we know that the teams on New England's schedule were not the teams that were being misjudged early on. We judged those teams fine. Those offenses were terrible. The Patriots' schedule so far ranks 28th overall, but the real issue is that the defense has faced the easiest schedule of opposing offenses in the NFL. The average offense of the Patriots' eight opponents has -8.5% DVOA, which is equivalent to playing the Rams eight times. And the pass defense has had it absurdly easy. The Patriots have yet to face one of the league's top ten pass offenses, but they've played four of the bottom seven, plus a couple of teams (Miami and Chicago) which rank much higher running the ball than passing it. As a result, there's virtually no difference between the Patriots adjusted (DVOA) and non-adjusted (VOA) ratings for run defense. But if you took out the opponent adjustments for pass defense, the Patriots would go from No. 20 (10.1% DVOA) all the way up to No. 7 (-4.4% VOA).

It all changes now. Four of the Patriots' next five games come against teams ranked seventh or higher in DVOA pass offense, starting of course with the No. 1 Denver Broncos. And yet, it could be worse for the Pats. Their tough remaining schedule (the second hardest, trailing only Oakland) is almost matched by tough remaining schedules for the two teams fighting them for the AFC East title: Buffalo (fourth in remaining schedule) and Miami (sixth in remaining schedule).

Detroit (6.2% DVOA, 14th): I'm going to assume that this ranking doesn't really shock anyone. The Lions have played an even easier schedule than New England so far, ranked 29th this year. They've won two straight games by just a single point. The Lions are doing it entirely with a defense that ranks as the best of the league. They are just 26th on offense and 31st on special teams. On the other hand, you have to figure this team's performance is going to improve to get closer to its record over the next few weeks. There's no way the field-goal kicking can be as bad in the second half of the season, and the offense is likely to improve as they get skill players back healthy, especially Calvin Johnson. The schedule doesn't get any harder; in fact it gets easier, and only Washington has an easier remaining schedule by average DVOA of opponent.

Arizona (-1.0% DVOA, 18th): Yeah, this is the really confusing one. The fact is, most advanced stats methods for ranking the 32 NFL teams end up with somewhat similar results. For example, the Simple Rating System at Pro-Football-Reference usually comes out pretty close to DVOA. Not when it comes to Arizona. SRS has the Cardinals as the top team in the NFC and seventh in the NFL. We have them seventh in the NFC and 18th in the NFL.

I'll be honest -- going through the numbers, it is tough to figure out why the Cardinals are so low in DVOA. It certainly makes sense that the Cardinals aren't up high with Denver. They've been very consistent -- second in variance -- but their wins are consistently small. Their six wins have come by an average of 7.7 points, and then they're going to get dragged down by their big loss to Denver, even after incorporating opponent adjustments to reflect Denver's strength. The Cardinals only have 4.2 Pythagorean wins right now. Five of their seven games have a single-game DVOA rating between -10% and 10%. (The exceptions are the Week 2 win over the Giants and the Broncos loss.) Still, that reflects a team that maybe should be ranked eighth or ninth in DVOA. It doesn't reflect the idea that Arizona should be ranked only 18th, with a DVOA that is actually below average.

So what else is going on? It's not an issue of schedule, as the Cardinals have actually faced one of the toughest schedules in the league this year. The most obvious luck indicators aren't very strong: the Cardinals are average in "hidden" special teams value and while they've recovered more than half of fumbles, we're not talking about 80 percent. We're talking 5-of-8 (not counting muffed punts).

One big issue that DVOA sees is that the Cardinals have been remarkably good at converting third downs, probably in a way that is unsustainable. Only 36 percent of Arizona offensive plays count as successful by FO standards, the lowest rate in the NFL. However, the Cardinals keep drives going because right now the offense ranks 26th on first down and 31st on second down but sixth on third or fourth down. Including DPI gains, the Cardinals have gotten at least 15 yards on 12.6 percent of offensive plays, which ranks ninth in the league. They've gotten at least 15 yards on 20 percent of third- or fourth-down plays, which ranks fifth in the league. DVOA prefers teams that keep the offense on schedule and the chains moving with a steady mix of both small and large gains, bceause that's generally a better predictor of an offense which will continue to have success in the future.

The bottom of our ratings also has some discrepencies between win-loss records and DVOA. The last winless team, Oakland, is not dead last in DVOA or even close. Four teams are behind the Raiders, including the 3-5 Vikings and the 2-5 Rams. Instead of Oakland, the worst team in DVOA by far is Tampa Bay. However, the Bucs still don't get close to the worst DVOA teams ever. At -43.9% DVOA, they don't even rank among the 20 worst teams ever measured through Week 8.

* * * * *

Once again in 2014, we have teamed up with EA Sports to bring Football Outsiders-branded player content to Madden 15 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 8 are:

  • QB Tom Brady, NE (24-HOUR HERO): 288 DYAR, the 16th best passing performance since 1989 (30-of-35 with 4 dropped passes, 354 yards, 5 TD).
  • LOLB Paul Kruger, CLE: 3 sacks, PD, FF, and TFL on a running play.
  • ROLB Khalil Mack, OAK: 5 run tackles for a combined 0 yards, plus a QB hit.
  • SS Brandon Meriweather, WAS: 7 tackles, 2 sacks, 2 forced fumbles.
  • RT Zach Strief, NO: Saints had 7.0 yards per carry and 100 percent Success Rate running right against Green Bay.

Football Outsiders versions of Kruger and Mack leave four teams that have not gotten Football Outsiders players in Ultimate Team yet this season: Carolina, Chicago, Detroit, and Tennessee. We'll try to get to those teams in the next couple weeks.

* * * * *

All stats pages are now updated with Week 8 information -- or will be in the next few minutes -- including FO Premium, snap counts and playoff odds.

* * * * *

These are the Football Outsiders team efficiency ratings through eight weeks of 2014, 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 80 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. This week, we say goodbye to DAVE, our method which combines 2014 performance with our preseason projections. All numbers now represent 2014 only.

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 DEN 47.6% 1 47.7% 1 6-1 31.8% 1 -17.5% 2 -1.7% 22
2 BAL 26.0% 2 26.1% 2 5-3 11.3% 7 -10.7% 5 4.0% 6
3 SEA 18.8% 4 18.2% 3 4-3 8.3% 10 -10.2% 6 0.3% 17
4 KC 15.1% 9 15.2% 4 4-3 5.1% 13 -5.2% 9 4.8% 5
5 GB 15.0% 3 14.8% 5 5-3 14.9% 5 1.8% 18 1.9% 13
6 PHI 12.4% 6 12.4% 6 5-2 -3.7% 23 -6.0% 8 10.1% 1
7 IND 11.0% 5 11.0% 7 5-3 9.7% 8 3.7% 21 4.9% 3
8 PIT 10.0% 14 10.2% 8 5-3 15.9% 4 8.4% 26 2.5% 11
9 SD 8.8% 8 8.8% 10 5-3 16.6% 3 9.2% 27 1.4% 14
10 MIA 8.8% 10 9.0% 9 4-3 5.2% 12 -11.9% 4 -8.4% 32
11 DAL 8.5% 7 8.5% 11 6-2 12.9% 6 2.9% 19 -1.5% 21
12 NE 7.0% 11 7.3% 12 6-2 7.5% 11 5.4% 22 4.9% 4
13 CIN 6.7% 13 6.0% 14 4-2-1 0.3% 16 -3.1% 12 3.3% 9
14 DET 6.2% 12 6.1% 13 6-2 -10.8% 26 -23.6% 1 -6.6% 31
15 NO 5.5% 21 5.9% 15 3-4 18.0% 2 13.7% 30 1.2% 16
16 BUF 3.2% 16 3.1% 16 5-3 -16.6% 28 -13.7% 3 6.1% 2
17 CLE -0.3% 17 -0.5% 17 4-3 2.9% 15 3.4% 20 0.1% 18
18 ARI -1.0% 15 -1.2% 18 6-1 -10.3% 25 -8.1% 7 1.3% 15
19 ATL -2.4% 18 -2.8% 19 2-6 9.1% 9 15.4% 32 3.9% 7
20 SF -3.6% 20 -3.7% 20 4-3 -3.1% 21 -4.4% 10 -4.9% 28
21 NYG -4.6% 22 -4.9% 22 3-4 -2.3% 18 0.3% 16 -1.9% 23
22 HOU -4.8% 23 -4.8% 21 4-4 -2.5% 19 -2.5% 13 -4.8% 27
23 WAS -6.5% 24 -6.5% 23 3-5 -2.2% 17 -1.2% 14 -5.5% 29
24 CHI -8.6% 19 -9.0% 24 3-5 3.3% 14 5.8% 23 -6.1% 30
25 TEN -10.6% 26 -10.2% 25 2-6 -3.7% 22 5.9% 24 -1.0% 20
26 CAR -14.8% 25 -14.7% 26 3-4-1 -2.7% 20 11.1% 28 -1.0% 19
27 NYJ -19.2% 27 -19.2% 28 1-7 -21.3% 29 1.3% 17 3.3% 8
28 OAK -19.4% 29 -19.0% 27 0-7 -16.0% 27 6.4% 25 3.1% 10
29 MIN -24.4% 30 -24.5% 30 3-5 -27.7% 32 -0.9% 15 2.4% 12
30 STL -24.6% 28 -24.3% 29 2-5 -8.6% 24 14.0% 31 -2.0% 24
31 JAC -26.1% 31 -25.5% 31 1-7 -27.1% 31 -3.9% 11 -2.9% 25
32 TB -43.9% 32 -43.7% 32 1-6 -26.5% 30 12.8% 29 -4.6% 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 DEN 47.6% 6-1 47.9% 7.8 1 4.3% 8 -1.5% 21 8.0% 7
2 BAL 26.0% 5-3 29.4% 5.2 3 -3.4% 22 -1.1% 18 9.5% 9
3 SEA 18.8% 4-3 10.8% 5.2 5 4.8% 6 -3.3% 23 17.2% 19
4 KC 15.1% 4-3 13.5% 5.1 6 4.8% 7 3.3% 9 25.9% 29
5 GB 15.0% 5-3 15.1% 4.7 8 -3.5% 23 -6.3% 29 18.2% 22
6 PHI 12.4% 5-2 13.2% 5.2 4 -7.9% 32 1.0% 12 10.4% 11
7 IND 11.0% 5-3 11.8% 5.3 2 7.6% 1 -4.7% 25 15.2% 17
8 PIT 10.0% 5-3 11.0% 4.5 13 -6.7% 31 3.5% 7 17.2% 18
9 SD 8.8% 5-3 14.9% 4.6 10 2.4% 11 7.1% 3 5.6% 4
10 MIA 8.8% 4-3 16.3% 5.0 7 -2.0% 19 4.0% 6 7.8% 6
11 DAL 8.5% 6-2 8.0% 4.5 11 -3.8% 25 -1.4% 19 17.9% 21
12 NE 7.0% 6-2 15.0% 4.0 16 -4.7% 28 10.2% 2 14.7% 16
13 CIN 6.7% 4-2-1 4.5% 4.5 12 6.0% 4 -0.3% 15 27.8% 30
14 DET 6.2% 6-2 15.0% 4.6 9 -5.2% 29 -7.0% 31 9.0% 8
15 NO 5.5% 3-4 9.4% 4.1 15 -5.9% 30 -5.0% 26 20.6% 24
16 BUF 3.2% 5-3 5.8% 3.9 18 -3.3% 21 6.8% 4 6.5% 5
17 CLE -0.3% 4-3 4.2% 4.2 14 -0.7% 15 -1.4% 20 11.1% 13
18 ARI -1.0% 6-1 0.3% 3.9 17 5.0% 5 1.4% 11 3.4% 2
19 ATL -2.4% 2-6 0.9% 3.7 20 -4.6% 27 -5.5% 27 17.3% 20
20 SF -3.6% 4-3 -4.2% 3.7 21 7.1% 3 -0.5% 16 14.5% 15
21 NYG -4.6% 3-4 -9.1% 3.8 19 1.8% 12 -2.3% 22 33.0% 32
22 HOU -4.8% 4-4 -3.1% 3.1 26 -1.1% 18 -0.9% 17 1.8% 1
23 WAS -6.5% 3-5 -7.3% 3.4 23 -0.9% 17 -8.7% 32 20.8% 25
24 CHI -8.6% 3-5 -9.4% 3.6 22 -0.8% 16 -6.4% 30 10.4% 12
25 TEN -10.6% 2-6 -8.6% 3.3 25 0.4% 13 0.6% 13 21.5% 26
26 CAR -14.8% 3-4-1 -18.1% 3.4 24 3.8% 10 -6.2% 28 13.0% 14
27 NYJ -19.2% 1-7 -21.5% 2.8 27 7.5% 2 2.2% 10 9.6% 10
28 OAK -19.4% 0-7 -22.5% 2.0 30 -0.1% 14 14.2% 1 3.6% 3
29 MIN -24.4% 3-5 -18.6% 2.7 28 -4.2% 26 -3.5% 24 20.0% 23
30 STL -24.6% 2-5 -21.4% 2.1 29 -2.4% 20 4.4% 5 22.0% 27
31 JAC -26.1% 1-7 -25.9% 2.0 31 4.2% 9 3.4% 8 23.3% 28
32 TB -43.9% 1-6 -38.6% 1.8 32 -3.5% 24 0.1% 14 27.9% 31

Posted by: Aaron Schatz on 28 Oct 2014

139 comments, Last at 04 Nov 2014, 10:39am by mitch


by BaconAndWaffles :: Tue, 10/28/2014 - 7:01pm

In regards to Arizona's 3rd down conversion rate, if I understand correctly - a team that gains four yards every down will have a higher rating that a team that gains one yard on each first and second down and ten on third down?

I get that one of those scenarios is more sustainable, and therefore a better predictor of future success, but I didn't those two didn't net out to the same value.

by Alternator :: Tue, 10/28/2014 - 11:18pm

The results of gaining four yards each down: Good, Good, Good.

The results of gaining one, one, and then ten: Bad, Bad, Very Good.

It's really easy for that third pass to drop incomplete, or a defender makes a good play to plug the gaping hole and stuff the RB. It's a lot harder to shut down an offense that's just steady, methodical, and reliable.

Now, there's certainly a point where boom-and-bust comes out better than steady and reliable: imagine a QB who does nothing but launch bombs into the end zone, with enough accuracy that half the time they result in TDs, and the other half in interceptions. It's the extreme in boom-or-bust, but that's still going to be a respectable offense, and will put huge pressure on the opposing team to score almost every time they touch the ball. If the booms are big enough, the busts get outweighted - but it really does require big, steady booms.

by Pat :: Wed, 10/29/2014 - 11:39am

I don't understand the confusion - you're saying that you recognize that the second situation (1, 1, 10) is less sustainable than the first (4, 4, 4), right?

So why would they net out to the same value? You're trying* to identify teams who's previous performance will continue to be good - so why wouldn't the results of the first team be rated higher?

*: DVOA isn't entirely tuned to be predictive - weighted DVOA for instance would be more immediately predictive than DVOA, for one. But it's more predictive than descriptive.

by Will Allen :: Tue, 10/28/2014 - 7:02pm

After seeing that coaching/management fiasco in Dallas last night, I'm off the Cowboys' bandwagon. I guess I'd lean towards Green Bay in the NFC, in that they have the best qb, the defense is improved, if flawed, and the team isn't hobbled by management. I think Seattle is going to get more and more exposed on offense, and the defense isn't what it was last year. Really, however, I have no idea who is going to get hot in the NFC come January.

by theslothook :: Tue, 10/28/2014 - 7:14pm

49ers when they finally get healthy. Willis and Aldon should both be a major boost to the defense. O line should slso settle together. Them, Sea and Gb are the three fav imo

by commissionerleaf :: Tue, 10/28/2014 - 7:26pm

Seattle's defensive problems are down to three things: linebackers that are not playing well in coverage, the starting CB position not occupied by Richard Sherman, and splits happen. The first seems likely to improve a little bit, the second is likely to be fixed this week, and the third will happen when it happens. Expect the Seahawk defense to return to form (if not 2014 form). Of course, they did hold the opposition to 9 points this week.

I don't see San Francisco passing Arizona for the division; they're too flawed, particularly at quarterback and offensive line, to go on a tear like I expect Seattle to.

by bravehoptoad :: Wed, 10/29/2014 - 10:14am

SF isn't just getting a couple players back from injury. They're getting their top 3 linebackers, and 3 DBs, two of them starters, and 2 of their 1st-round-pick O-linemen, and then bonus special goodies like Marcus Lattimore.

Man, that's a lot of fresh grade-A meat.

I disagree with your assessment of Kapaernick--he's been an asset. The O-line has been terrible in pass-D, though, and I don't know see any particularly good reason why that'll change. They're going to be able to get J. Martin off the field--plus--but they're going to be starting a rookie center.

by jacobk :: Tue, 10/28/2014 - 8:21pm

Seattle's offense:
2012: 18.5% DVOA, 4th overall
2013: 9.4% DVOA, 7th overall
2014: 8.3% DVOA, 10th overall

I'm not sure what you think there is to expose. They're going to run it with Marshawn Lynch, take their shots down the field, and sprinkle in a little magic (and the occasional disaster) when they turn Russell Wilson loose. Now that the Percy Harvin experience is over I would expect the offense to drift back towards the 2012-13 numbers. Down a little bit for the loss of Tate, up a little bit with the improvement at QB.

I've given up the dream of every seeing that ridiculous three game run at the end of 2012 become the norm, but I don't see any reason to think the Hawks won't continue to slot into that 5-10 range at offense.

Last year the defense was a core of very good players enjoying very good injury luck and decent bouncing ball luck. This year they're experiencing mildly bad luck on both the injury and bouncing ball front. I don't know if anybody tracks dropped interceptions, but there certainly seem to be a lot more of them this year. If the luck evens out the defense should be capable of getting back to the 2012 level. The 2013 level is probably out of reach.

If the Cardinals ever stop getting those 80 yard game winning bombs and fall back to the pack I expect the Seahawks will be right there to compete for at least a first round bye.

by theslothook :: Tue, 10/28/2014 - 9:03pm

I feel like Seattle's problems are being vastly overstated. In some sense, none of the three losses were really all that similar. In sd, they couldn't get off the field on third downs despite getting decent pressure. That's a testament to Rivers. Their second loss was a bad matchup as their d line go overwhelmed and the seahawks got too gimmick happy. Their third loss was driven mostly by special teams and some good execution from Austin Davis( a reasonable expectation for a pesky team at home).

by Will Allen :: Tue, 10/28/2014 - 11:01pm

I think Seattle's defense, given the less than dominating nature of their offense, would benefit mightily from getting HFA again. Last year it was pretty clear by now that they had a very good chance of playing at home in January, and beating them there was going to be a tall order Even so, the Niners came within a foot or so of getting it done. This year, HFA is going to be a fight, it would appear, and they just aren't going to be as good on offense, and they might not benefit from the opposing offense having to go on a silent count. The path isn't nearly as favorable.

by jacobk :: Wed, 10/29/2014 - 8:17am

Do you have any evidence either from the statistics or from the tape to suggest that the Seahawks won't be as good on offense? That they will be exposed?

If we're going to play the gut feeling game you may as well point out that the Seahawks are proven champions, and character is forged in the crucible of adversity, so these rough times early in the season will pay off in a battle hardened team by the time the playoffs begin. The Broncos, of course, are flying high after playing a bunch of cream puffs (and losing to the one True Champion they faced in their schedule) and will collapse as soon as they get punched in the mouth when the Real Games start.

by dbostedo :: Wed, 10/29/2014 - 8:32am

Well there is last year's offensive DVOA being higher than this year's; But it's not as big a difference as I would have expected. Also, Russell Wilson's DVOA is down something like 25%, which is very significant. You could easily assume that that will catch up with them.

But still - a gut feeling or off-hand analysis based on watching games is still a far cry from making statements like "character is forged in the crucible of adversity". That's hardly what he did. I could be mis-reading this, but you sound a bit too wound up about this; Nobody said "exposed" or that they're terrible - just that they won't be as good as last year and will have a tougher road to the Super Bowl.

by Will Allen :: Wed, 10/29/2014 - 10:04am

They've never been an offense which has physically dominated opponents, and have relied on deception and Russell Wilson's improvisation to succeed. Deception becomes harder and harder as time goes on, and improvisation is non predictive. Wilson's terrific, but they'd be better off if they had more receivers who get just flat out get the better of defenders on a regular basis.

by LyleNM :: Wed, 10/29/2014 - 11:30am

Except, of course, that Will did in fact say "exposed".

by Will Allen :: Wed, 10/29/2014 - 2:13pm

Of course, unless it is one of those very rare teams which is outstanding in every aspect, all teams are exposed.

by dbostedo :: Wed, 10/29/2014 - 5:37pm

D'oh! My bad, I was only looking at the previous post... forgetting the one before it!

by Will Allen :: Wed, 10/29/2014 - 9:06am

Yes, I have watched the team play. No, I'm not going to bother to link to video for you. If you think I'm wrong, that's o.k. by me.

(edit) If it makes you feel better, I still think they are a bad matchup for the Broncos.

by ammek :: Wed, 10/29/2014 - 6:49am

I don't know what you're seeing from the Packers. The lines are still weak points which the other NFC contenders can exploit — the Lions, Seahawks and Saints have already done so. Green Bay is not strong enough to win multiple playoff games on the road as it did in 2010, so it needs to win the North, which is far from certain (FO says it has a 42% chance) and get a bye. The games at Lambeau against Philadelphia and Detroit will help to uncomplicate the NFC picture, but I don't think either is a given for the Pack.

by Will Allen :: Wed, 10/29/2014 - 9:10am

Well, I'm mostly seeing that they aren't as soft as they were 8 weeks ago, which is damning with very faint praise. Like I said, I really don't have any strong opinion on the NFC winner at all. I liked Dallas, but seeing once again how badly managed that team is has really caused me to think that they will screw things up one way or another.

by justanothersteve :: Wed, 10/29/2014 - 11:30am

Will, I agree with you that nobody looks to be really stepping up in the NFC as the favorite. The Packers have serious problems along both lines and the only thing I can say positive about Bakhtiari is that he's not playing as badly as Bulaga is at the other OT spot. Or to use the zlionsfan template:

Green Bay is clearly ranked too f***ing high because it I can't see straight after last Sunday's game. Random amoeba races are way better than this. Maybe I'll feel better tonight when I get home and drink a half bottle of cheap anisette.

by TomC :: Wed, 10/29/2014 - 11:55pm

Between sips of wormwood, repeat to yourself: "Could be worse: I could be a Bears fan."

by tuluse :: Thu, 10/30/2014 - 12:38am

Yeah. Aaron Rodgers is almost good enough to hide any flaw.

by Silversmith :: Tue, 10/28/2014 - 7:32pm

Denver's balance on both sides of the ball is still absolutely mind blowing. With the Broncos' upcoming schedule after this weekend's game in Foxboro (@OAK, @STL, MIA, @KC, and BUF), there's a good possibility that the Denver defense can overtake Detroit for the #1 spot in defensive DVOA as well.

by JoeyHarringtonsPiano :: Tue, 10/28/2014 - 7:40pm

"there's a good possibility that the Denver defense can overtake Detroit for the #1 spot in defensive DVOA as well."

Wouldn't surprise me at all. Denver's defensive personnel is better than Detroit, except defensive line (even that's iffy now that Nick Fairley is out). I don't see Detroit sustaining this level of performance all season, especially with road games against New England, Arizona, and Green Bay still on the schedule.

by Otis Taylor89 :: Tue, 10/28/2014 - 8:04pm

Actually, the fact that they are going to be playing some bad offenses (BUF,OAK,STL) will probably hurt them and that they will also be playing some improving offenses on the road (NE,KC) will probably prevent them from taking the top spot.

by ChrisS :: Wed, 10/29/2014 - 12:28pm

As a Lions fan I expect the offense and special teams to continue to be awful and the defense to slowly get worse and that these close wins will turn into losses. Perhaps I am being too optimistic and I should expect armageddon as well. But then again they might get in the playoffs and even break the 20+ year playoff victory drought.

by lionsbob :: Wed, 10/29/2014 - 1:33pm

The Lions will get Calvin Johnson, Reggie Bush, Brandon Pettigrew, Eric Ebron, and Joseph Fauria back after the bye most likely. Oh plus LaAdrian Waddle...and rest of season Matt Prater. I expect the D to regress some, but I can't see the O and special teams continuing to be that bad.

by Alternator :: Tue, 10/28/2014 - 11:22pm

Denver would have to have the good fortune to continue with outstanding health, and there's no way to be confident in that. One or two good injuries and the defense can come crashing down.

Heck, one knee to Manning's neck and he could end up being Romo'd.

by shoutingloudly :: Wed, 10/29/2014 - 12:32am

My thoughts pretty much exactly, and I was raised in the Church of John Elway. My boys have been VERY lucky so far with injuries, pretty darned lucky with other teams' health as we played them, and even relatively-to-very lucky (as Thursday) with refereeing. I love how it's played out so far, but I take none of it for granted, and I'm glad to see the team seems not to, either.

Even the lofty playoff, conference title, and Super Bowl odds assume the current team and thus could take a big hit with the wrong injuries. It was a miracle — or, at least, the result of an equally horrifying injury report for the Pats — that they made the SB with the injuries last year. And let's just say there's a reason I'm constantly checking the Adjusted Sack Rates...

"We're Number One! Try Not to Die! Hope the Top Seed Isn't Settled By the Tie!"

by theslothook :: Wed, 10/29/2014 - 2:16am

I was afraid I'd jinx them last week when I brought this up - the only areas denver has suffered injuries have been the areas they can mask pretty well - running back and linebacker.

All of the areas denver is strong in are areas that are healthy and brimming with talent. You can't expect pristine health forever, but as long as they can avoid any long term injuries, they should be ok./

by Rick_and_Roll :: Wed, 10/29/2014 - 12:06pm

Injury luck is usually one of the key components to being a championship team, and by and large could explain the difference in last years elite NFC and AFC teams. SF and Seattle were healthy while Denver vs New England was a battle of attrition.

In 2013 both Denver and New England had, next to losing their QB, the worst injury luck.
1. They both lost elite or borderline elite players at key positions. (DEN: Miller, Clady and Harris) NE: Gronk, Wilfork, Vollmer)
2. Then each team had a slew of injuries all on the defensive side.

Hopefully this year everyone stays healthy for all teams and injuries have a minimal impact...

by RickD :: Wed, 10/29/2014 - 12:50pm

Don't forget Mayo. Lost him both last season and this season.

Every time Gronk gets hit in the open field, I cringe.

I share your hope that injuries have a minimal impact. Both Denver and NE had so many injuries last season that we all kind of knew that the NFC champion was going to win the Super Bowl with little difficulty. So far this year, the injury rate is regressing toward the mean.

by EricL :: Wed, 10/29/2014 - 2:18pm

I firmly believe that, given the depth the Seahawks had last year, the 2013 Seahawks could have withstood this year's injuries without the drop off the 2014 team has shown.

However, when they needed the depth to cover for injuries, it hasn't been there this year. Bobby Wagner, in particular, has been missed.

by BroncFan07 :: Wed, 10/29/2014 - 1:38am

Very true but remember that last year they were very unlucky with injuries. Hopefully this year will be the regression to the mean and luck will turn for the better.

by Alternator :: Wed, 10/29/2014 - 10:01am

Regression to the mean means that an essentially random and/or non-predictive outlier will tend to move toward the group average in subsequent iterations. While I have no argument with putting health among that group (some guys are chronic, most are one-off), regression to the mean would be average health going forward, not practically perfect health as some kind of cosmic balancing act.

by cjfarls :: Wed, 10/29/2014 - 1:25pm

Denver's injury luck has been far from perfect. They've lost their borderline probowl 3-down coverage LB, starting RB and Welker for numerous games, and had a couple of key guys limited in the first couple games as they recovered from last year (Harris, Miller, Clady).

So in general, while I don't know how their adjusted games lost compares to other teams, I don't think they are in some weird outlier of non-injury land this year. What they have been "lucky" at (more a testament to John Elway's GMing/Fox's player development) is having depth guys step up (the other Brandon Marshall and Corey Nelson at LB, Hillman and J.Thompson at RB, adding Ware to supplement Miller), and the losses they've had being relatively unimportant compared to losing their GOAT QB, etc.

by Alternator :: Wed, 10/29/2014 - 4:25pm

RBs are more fungible than most positions, and Denver didn't have an All-Pro there to begin with, and at this point Welker is what, the third receiver? Fourth? Denver has a better group of WR/TE than New England did, so Welker didn't have to be in as central a role, not to mention the wear and tear from advancing seasons building up.

Compare this to New England having lost their defensive leader for the season, plus several other key defenders for at least a month, and started the year without Gronk. Subjectively New England hasn't even had unusually bad health this year, either; Denver's just been lucky. It might continue and provide an easy road to a repeat Super Bowl appearance; the luck might end and they lose a few key guys.

by cjfarls :: Wed, 10/29/2014 - 6:55pm

I think you're oddly and somewhat defensively reading more into the post than need be, and I have no idea what New England's injuries have to do with any of it unless it is some persecution complex about the woes facing your team. All I said was that Denver has had injuries to numerous starters (factually true) and so claiming they've had "perfect health" was false... Also that good depth has limited the impact of the injuries they have had (you seemingly agree).

Why was Welker not a big deal?... good WR/TE depth with DT/Sanders/JT/Caldwell/Tamme. That doesn't mean that wasn't a significant injury.

Why was Ball not a huge loss?... small dropoff from Ball to his backups and improving oline play compared to early season (and fungibility of RBs, same as NE will likely show with Ridley)

Why has Trevethan's loss not been bigger?... Brandon Marshall's emergence and late round rookie draft picks playing well as reserves.

Why were Harris/Miller's limitations early not felt more?... drafting Roby and picking up Ware.

Like I said, I don't know what "average" injury is for a team or how it compares to Denver this year. I said before DEN has been relatively healthy, but not absurdly so. Denver has also been lucky in that it hasn't lost PM or a cluster of folks at any single position to really test their depth beyond the first line of backups. But it hasn't had "perfect" health as was claimed... it has factually had a number of injuries to starters. It just has had enough solid depth players behind them to replace them relatively seamlessly so far, and is also lucky that those guys all were/will be able to return in a few weeks.

Trevethan is probably the best illustration of my point: Going into the season, he seemed to many (me for one) to be the one guy that the defense couldn't afford to lose, being the best "true" LB on the team (Miller being a LB/DE-hybrid). At DT, DE/Passrusher, CB, and safety there were former starters/experienced rotational guys as backups/complements... Trevethan had Brandon Marshall (former Practice Squad afterthought) and a couple late round rookies behind him. It absolutely is a very significant injury, just like Mayo's was for NE. But Marshall and friends have played well, so you seemingly dismiss it as unimportant.

As always, if PM goes out for the year then DEN probably is only a good team that likely makes a quick playoff exit. But absent that (and this is similarly true for any starting QB), or maybe a few o-line losses (but with Chris Clark losing his starting role to Cornick, that means there is at least a somewhat solid 6th man on the bench now), its hard to see where just a few injuries would really test their depth beyond what they've had to already. Certainly there are guys that would be big losses (Miller, DT, JT, Harris, Talib), but nothing that looks like it would sink their season or take them out of the ranks of the league's elite teams.

Or PM goes down, Osweiler looks like the next Aaron Rodgers, destroys the league, and wins the Superbowl. Who knows at that point, but it is Denver's depth that makes it look like they are solid going forward, not some false claim of them facing no injuries to date and thus facing impending doom if they don't remain "lucky".

by theslothook :: Thu, 10/30/2014 - 12:53am

I think his point above still holds merit. Welker loss isn't nothing, but its mitigated by their qb and the quality at wide receiver. Travathan in a vacuum is a big loss, but linebackers are also covered up by playing more dime and using more coverage players. It might be a big loss in this particular matchup because of Gronk, but beyond a few tight end centric teams, its not prohibitive. In the grand scheme of things - which areas is it easiest to overcome injuries? I'd say rb and linebacker.

NE's injuries are similarly not too bad. Mayo is a loss sure, but they have plenty of depth there as well with Hightower and Collins. And Gronk has returned to full health so the receivers overall are in good shape and the o line looks to have started to gel.

Chandler jones is probably their most significant injury and he'll be back in a month.

Injuries, like anything else, need to be compared not to a static average, but to the rest of the league. Subjectively, I can't really think of another team that has had fewer injuries to overcome at this point in the year than these two.

by theslothook :: Thu, 10/30/2014 - 12:53am

I think his point above still holds merit. Welker loss isn't nothing, but its mitigated by their qb and the quality at wide receiver. Travathan in a vacuum is a big loss, but linebackers are also covered up by playing more dime and using more coverage players. It might be a big loss in this particular matchup because of Gronk, but beyond a few tight end centric teams, its not prohibitive. In the grand scheme of things - which areas is it easiest to overcome injuries? I'd say rb and linebacker.

NE's injuries are similarly not too bad. Mayo is a loss sure, but they have plenty of depth there as well with Hightower and Collins. And Gronk has returned to full health so the receivers overall are in good shape and the o line looks to have started to gel.

Chandler jones is probably their most significant injury and he'll be back in a month.

Injuries, like anything else, need to be compared not to a static average, but to the rest of the league. Subjectively, I can't really think of another team that has had fewer injuries to overcome at this point in the year than these two.

by JoeyHarringtonsPiano :: Tue, 10/28/2014 - 7:36pm

The last two wins by Detroit have been less than reassuring. Intuitively, this has the same feel as the 6-2 start in 2007 (finished 7-9) and 1999 (finished 8-8).

Of course, intellectually, I know this roster is several orders of magnitude better than those rosters. Or at least it will be once some of the injured skill players get healthy. Still very apprehensive about the rest of the season.

by LionInAZ :: Wed, 10/29/2014 - 10:59pm

They've won three games with Calvin Johnson on the sideline. That's reason for optimism.

I like to think of them as weirdly dangerous.

by JoeyHarringtonsPiano :: Thu, 10/30/2014 - 2:25pm

You're probably right, but following this franchise for the last 25 years has hard-wired routine pessimism into my brain.

My hope is that the offense cracks at least above average once Johnson comes back fully healthy and the players get used to the new system, and the defense regresses only minimally. Both of which would make them team to be reckoned with, indeed.

by LionInAZ :: Fri, 10/31/2014 - 6:39pm

I know where you're coming from.

The next three games (Dolphins, @ARI, @NE) will be telling. I'll be happy if they win one out the three. If they win two, we can call them a legit contender. If they lose all three, bleh.

by swgallagher :: Tue, 10/28/2014 - 7:46pm

Denver sure looks good.

Playing 5 of your first 7 games at home and having a bye in an 8 week stretch certainly helps the early season DVOA.

by Otis Taylor89 :: Tue, 10/28/2014 - 7:50pm

Denver may have played the 8th hardest schedule, but thru Week 8 they have only played 2 road games! They are also played several teams that had serve injury problems just in time for their games (ARI,SD,SF).
They also played the Jets....

by Sakic :: Wed, 10/29/2014 - 1:52pm

There's definitely been some strange scheduling this year. On the other side of the coin Green Bay has already played 5 road games with their bye week coming this week although considering the home field advantage improves for GB in the second half of the season I'm okay with that.

by Griffinmoos12 :: Tue, 10/28/2014 - 7:50pm

How does gronk not get a Football Outsiders card in Madden are you guys kidding me he put up the best receiving performance yet this NFL season and doesn't get a card. Are you kidding me smh.

by MarkV :: Tue, 10/28/2014 - 8:08pm

pick your reason:
1) DVOA is defensively adjusted and Chicago is terrible against TEs, as Aaron noted on Quick Reads comments. Not as impressive of a performance considering
2) FO cards TEND to emphasize disagreements between conventional stats (e.g. box score) and advance stats. Gronk was much worse in advanced stats than conventional ones
3) He may be unavailable because of other madden promotions (or because his madden stats are too high to be promotable)

by Theo :: Wed, 10/29/2014 - 4:25am

for everybody's sake, use the template!

by ChrisS :: Wed, 10/29/2014 - 12:33pm

Perhaps we need a separate, slightly different, template for Madden rants by disgruntled fans.

by RickD :: Wed, 10/29/2014 - 12:52pm

My guess is that since Brady got one, they didn't want to give two cards to Patriots.

Also not getting a card? Ben Roethlisberger, who had the best regular season passing game in decades.

by frievalt :: Tue, 10/28/2014 - 8:13pm

Perhaps DVOA doesn't like the Cardinals due to a higher variance deep passing attack? DVOA used to perennially overrate high percentage Andy Reid passing games, perhaps we're looking at the inverse.

If you don't have a particularly accurate QB and you DO have a great receiving corps perhaps swinging for the fences as hard as the Cardinals have is a good idea.

Also what happens when we look at ONLY the Carson Palmer games?

by cjfarls :: Wed, 10/29/2014 - 1:28pm

I'd love to see only the Palmer games rated. He's not great by any stretch, but that team is far more dangerous with him there than the other drek they were rolling out in his absence.

by dmstorm22 :: Tue, 10/28/2014 - 8:20pm

Is this a particularly low number of teams with a DVOA over 20% this late into the season?

Even if it is, it is hardly surprising given how truly up-and-down the whole league has looked aside from Denver. The entire AFC seems to be made of up a bunch of teams ranging from good to very good, with four below average and one dominant.

If Denver is able to beat NE in Foxboro, which will be tough, it will be interesting to see how they play the rest of the season. With a win, they'll essentially be two games up on the entire conference, in that to lose the #1 seed, they would have to drop two games (possible - given a road-heavy back half), AND have another team win out (seemingly inpossible, given how hard it is to quantify any other team). The wild card picture is going to have about 300 clinching scenarios the last few weeks.

At least the ESPN Playoff Machine will be fun.

by jdiko :: Tue, 10/28/2014 - 8:56pm

Indianapolis special teams have been ranked 7th or higher since week 3. Is this coaching or better personnel management?

by theslothook :: Tue, 10/28/2014 - 9:00pm

Hard to say but I suspect it started with a different philosophical approach to specail teams - namely not pretending its just an extension of your practice squad. Plus i really love Pat McAfee

by Mike B. In Va :: Wed, 10/29/2014 - 6:20am

I think the approach is it. Buffalo was consistently near the bottom over the last few years and they made a big deal out of prioritizing getting some quality ST players in the off season. It seems to have worked, since it's the same ST coach from last year's disaster.

by Ian Chapman :: Tue, 10/28/2014 - 9:11pm

Concerning Arizona, it could be that your system (DVOA) is telling you something important that the record and other metrics don't show yet. While not perfect, DVOA has been a pretty good indicator of how well teams are performing. It might be that Arizona really isn't as good as their record indicates.


by Sporran :: Tue, 10/28/2014 - 9:59pm

Yes, but the question is now "Why aren't they as good as their record indicates?" DVOA isn't magical -- it's using actual data. Using the same data, humans should be able to figure out what DVOA is seeing that isn't otherwise obvious.

by Ian Chapman :: Wed, 10/29/2014 - 12:14am

I agree. I think the heart of it is that the Cardinals seem to be converting low percentage third down plays (as compared with the rest of their offense) at an unsustainable rate. Or to put it another way (based on what I've seen w/r/t to the Card games I've seen), the Cards always seem to come up with an unlikely big play (long bomb, key pick, or even....in week one...a seeming loss of sanity by the opposing QB) and get a win after being dominated most of the game. I consider the Eagles game to be a prime example of this. The Eagles controlled most of that game at least according to my Mark-1 eyeball, but the Cards were able to get a couple of key (and low percentage) plays to keep it close and then win...and the Eagles weren't able to do the same.

I consider how well (or rather how NOT well)Arizona did against Denver to be a good indicator of how good they really are. Sure Denver is better than anyone right now, but how well a team does against Denver does seem to be a good measure right now.


by dmstorm22 :: Wed, 10/29/2014 - 12:21am

You picked the three games the Cards were anything but in control. Yes, they were lucky to win against Philadelphia and San Diego, but at least the Eagles they were in the game the entire time.

The Broncos blew them out, but the Broncos are by far the best team in the NFL so far.

But what about the other four games? They won their other games by 11, 9, 10 and 11. Sure they weren't all good teams, but they aren't lucking into wins the way they did in 2011, and in their fluke 4-0 start in 2012.

by BJR :: Wed, 10/29/2014 - 7:03am

I too watched the Eagles/Cardinals game and came away with the same impression. You only have to look at the yardage totals to realize Philly really screwed that one up.

Over on Brian Burke's site they also have Arizona pegged as a below average team. Of course their numbers will be weighted down by the games Palmer did not play, and they are a significantly better team with him in the lineup. Still, I have to believe these metrics are telling us something here.

by cjfarls :: Wed, 10/29/2014 - 1:42pm

I have a theories, that in combo may help explain things:

1) Offensive DVOA being weighed down by games with drek at QB, and a boom/bust philosophy. Palmer being back should help that regress to more their actual level, and the boom/bust philosophy is just that... they can play with anyone, but it may not be consistent sustainable. If they are hot/lucky at the right time they can win, but you'd expect a stinker occasionally.

2) A DEF scheme highly dependent on big-blitzing. Being a Denver fan, I saw this a lot in the 2005 team. Can work really well for a long time against weak and average, but you get a good QB in there and it can fall apart fast. At some point you have to be able to beat teams straight up, and I'm not convinced AZ has the D-line/LB talent to do that against the best teams (Though a top-flight cornerback can hide things well for a bit (see Champ Bailey '05 and Patrick Peterson).

So overall I expect that they are better than DVOA suggests, but probably not in the top-10.

by herewegobrownie... :: Tue, 10/28/2014 - 9:48pm

Hard to believe the Browns now rank ahead of where they were much of the last few years on defense.

The defense has been strong on turnovers and that's mostly it - looking forward to Tashaun Gipson making the Pro Bowl.

by Mr.Morden :: Tue, 10/28/2014 - 9:58pm

I'd still be interested to see you add Det-Ind (Caldwell reunion special) and Det-Den (Caldwell-Manning reunion special) to your special Super Bowls list on the playoff odds page.

Also, what's the worst ever DVOA for a one loss team in week 8 or later? How does Arizona stack up on that list?

by icecon :: Tue, 10/28/2014 - 10:36pm

I really don't think Arizona is such an enigma. This line here pretty much sums it up,

"DVOA prefers teams that keep the offense on schedule and the chains moving with a steady mix of both small and large gains, bceause that's generally a better predictor of an offense which will continue to have success in the future."

Arizona's style contradicts DVOA in a sense. Bruce Arians loves two things: going deep and feeding Andre Ellington. A tough schedule has limited the run game somewhat. And the QB merry go round and schedule has added incompletions to the deep passing game (and with Carson Palmer beign solid but not elite). Arians is a good coach and so he calls the right plays to get the 3rd down conversion.

So you have a team that stutters in between screams on offense, not something DVOA likes. They're 6-1 largely because of their defense, and thus winning it close. That shouldn't take anything away from them though.

by Will Allen :: Tue, 10/28/2014 - 11:03pm

Winning close games isn't nearly a predictive as kicking the hell out of opponents.

by icecon :: Wed, 10/29/2014 - 1:11am

This is absolutely true but also pretty useless. The Broncos and Ravens are the only teams that have really blown teams out multiple times and the Ravens are 5-3. So we know the Broncos are really good and will win a lot and the ravens are pretty good and will win a decent amount.

That's not very helpful. That said I do fancy the Ravens to beat the Steelers and the Broncos to beat the Patriots this weekend. Will it play out that way? We'll see, but nothing is a given in this league.

by Will Allen :: Wed, 10/29/2014 - 10:01am

It's not helpful to know that winning close games is in good measure based on random events, and that it is unwise to assume that the random events will continue to help a 6-1 team to win more games?

by theslothook :: Tue, 10/28/2014 - 11:56pm

Also I don't think there's anything magical about Arizona when they make a habit of throwing deep vs others. There's a reason its not predictive.

by tuluse :: Wed, 10/29/2014 - 1:13am

When Drew Stanton was in it was clear that Arians went the route of long passes to the sideline as a method of hiding his QB instead of quick screens and smoke routes (this is the same strategy Ron Turner used with rookie Kyle Orton).

Long passes to the side line will usually fall incomplete, but when you connect, it's a very good play.

So yeah, Arizona is probably better than DVOA thinks they are because Palmer is better than Stanton.

by dbostedo :: Tue, 10/28/2014 - 11:56pm

Also, in general teams that are worse on 1st and 2nd down compared to 3rd down will have their performance on 3rd down tend to get worse over time.

Can Arizona buck that trend? Possibly, but it's not likely as that would make them an outlier. Arians being a good coach is probably not a good enough reason for them to be an outlier, though "luck" might be.

by icecon :: Wed, 10/29/2014 - 1:03am

I'm aware that 3rd down teams generally get worse. It's pretty harsh to write any team's ability to convert on 3rd down as just luck.

There's a lot that goes into that. For example, their offensive line's surprisingly good pass protection has helped them deal with 3rd and long blitzing. The play calling matters. The players preparation and ability matters. The defensive scheme matters which could fall into "luck."

Btw, I'm not a Cardinals fan, but I do think the fundamentals of what they're doing over there is good.

by Bright Blue Shorts :: Wed, 10/29/2014 - 6:04am

If it can be repeatedly reproduced then it is a skill, otherwise it's luck.

Time will tell which one it is for the Cardinals.

by Pat :: Wed, 10/29/2014 - 12:09pm

If the play calling on 3rd down is so much better than on 1st and 2nd down, either they need to start using some of those 3rd down plays on 1st and 2nd (if it's *their* playcalling that matters) or other teams need to start using their 1st and 2nd down strategies on 3rd down (if it's the defense's that matters).

It doesn't have to be luck to be unsustainable. If 3rd and long blitzing is the problem, eventually someone will realize that, and stop blitzing. And then everyone else will start doing it too. And then their performance on 3rd down will regress to the average performance on 1st and 2nd.

In case it wasn't obvious, I don't seriously think the idea that their playcalling is just way better on 3rd down makes any sense. That would suggest they're a good team with a bad coach, which I don't think they are.

by icecon :: Sat, 11/01/2014 - 6:48pm

Why wouldn't the playcalling be better on third down? Isn't a team better off using its best plays for the most critical situations? On first and second down, you are more able to take risks and try different things. I'm not saying it IS better, because I am not going to go back and watch all those games and try to rate the plays but it's not implausible.

Also, one interesting related fact: the 2008 Super Bowl winning Steelers' offense were equally hated by DVOA (-1.5%, 21st ranked). The offensive coordinator of that team? Bruce Arians.

That team did, however, have six wins of 15+ points in the regular season.

by tuluse :: Sat, 11/01/2014 - 7:36pm

If you use your best plays on 1st down, you don't have to face a 3rd down and the possibility of punting.

by EorrFU :: Wed, 10/29/2014 - 12:59am

That goes back to the old overrated running back problem... Boom or Bust backs are not as valuable unless their name is Sanders.

by ClavisRa :: Wed, 10/29/2014 - 1:43am

What's the Patriots' DVOA for the first half of the NE-CHI game? Playing with a huge lead like that changes so many things about your approach to the game offensively and defensively, and the Bears did nothing to reduce that massive lead, so it's hard to give the 2nd half much weight at all once the Patriots opened it with yet another scoring drive.

by nat :: Wed, 10/29/2014 - 6:54am

While there are reasons to trust DVOA less late in blowouts, you can still learn from it. One thing we learned on a Sunday was that the Patriots' prevent defense gives up yards and first downs quickly. That's a problem that could cost them in future games. It almost cost them the Jets game.

by Dr. Bill :: Wed, 10/29/2014 - 8:32am

I think this is a problem. The victory condition for the football is a win, with no consideration for on-field production. If I'm up by X points, and I bleed the clock out so that no matter what you do, you end up with X-1 or less points, then I was successful operationally. It's what we mean in the Marine Corps by "trading space for time." The goal isn't necessarily to hold ground, or kill the enemy, but rather to achieve the particular victory conditions for that conflict: to get a win. So trading ground (space) for another operational asset (like time) may not look good from a civilian perspective--it might look like failure--from a military professional standpoint it might be the height of wisdom.

Did the Chicago Bears in that second half exploit a bad prevent defense? Or did a brilliantly executed prevent defense kill the half and seal a victory?

by Will Allen :: Wed, 10/29/2014 - 9:15am

That's true, bit from a metrics standpoint it doesn't correlate. Better teams tend to do better when down by 30 in the 2nd half, and better teams tend to do better when up by 30 in the 2nd half, so it doesn't make sense to simply toss the data. In other words the perfect metric doesn't exist, and all you can do is try to weight things in the best way possible.

by jonnyblazin :: Wed, 10/29/2014 - 11:28am

Well this happens for other teams too. For instance the Ravens completely stuffed the Bucs offense when they rolled up a 38-0 lead, but ended up with a positive defensive DVOA for that game (below average) because the Bucs moved the ball in the 2nd half and scored 17 points.

by Dr. Bill :: Wed, 10/29/2014 - 11:29am

In behavioral research we always have to account for human agency. So while there may be population-level correlations that are useful, we can often find intention behavioral within a given context that isn't explained by the correlation. What I mean is that data may accurately reflect that good teams can mount furious comebacks/comeback attempts, and good teams with large leads may sagely allow yardage/time trades that increases win probability. Both can be simultaneously true.

by Will Allen :: Wed, 10/29/2014 - 2:19pm

I never stated that it couldn't. Sometimes reality is hard to measure.

by ChrisS :: Wed, 10/29/2014 - 12:42pm

I agree to some extent. I think that the definition of a successful play with a big lead should change. Besides the usual FO definition of success, any play that keeps the clock running and does not lose significant yardage (sacks are still bad, though not as bad), and does not result in a turnover is a successful play.

by Will Allen :: Wed, 10/29/2014 - 3:45pm

Yeah, I'd be surprised if Aaron hasn't played around with such ideas, in order to make DVOA more predictive. I used to try, the first couple years I was visiting this site, to suggest ways in which the metric doesn't measure true team value. Then I came to more greatly appreciate the limits of any attempt to measure very complex human behaviors, and that Aaron likely spends a lot of time experimenting with ways to make his measurements better. Now, I mostly use DVOA to give me some sense where my eyes might be fooling me, so as to cause me to question my judgement. Sometimes I reconsider my judgement, sometimes I don't. I thought my eyes had greater insight than DVOA with regard to Dallas going into their game with Seattle, and had my eyesight confirmed, but now that I've seen once again how badly managed that organization is, I'm starting to question my original judgement. In other words, soothsaying is hard.

by RickD :: Wed, 10/29/2014 - 12:57pm

The problem in the Jets game wasn't the Pats' pass defense so much as it was the run defense. A moderately competent Jets' QB could have won that game.

Excuse my giggling.

Another thing to compare and contrast is how well the defense has played since Brandon Browner started playing. That seems to mark the dividing point between the dismal first quarter and the high-functioning second quarter of the season.

And now we get the brutal third quarter of the season. Can you say "November sweeps"?

by ammek :: Wed, 10/29/2014 - 6:40am

The playoff odds page no longer lists Oakland's chances of finishing 0-16. I thought the Browns game might have been its best shot at a win. The remaining schedule is nasty: home to Buffalo on week 16 is the only one that looks even remotely winnable. I had the Chiefs games as opportunities, but they're now 4th in DVOA — an eye-popping ranking that seems to have escaped comment so far.

The ignominy of a winless season would be unjust, really, as Oakland isn't *that* bad. The front seven on defense is fine, and Derek Carr has higher DVOA than Russell Wilson. Whoever gets to run the team after Reggie McKenzie will inherit the nucleus of a good roster. It may not be a career-ending opportunity, for once. If the Raiders move to LA, perhaps they could ask to join one of the South divisions.

by BJR :: Wed, 10/29/2014 - 7:11am

Agree. Oakland actually has some good young talent at receiver, and Carr has done ok all things considered. That division is savage this year - but it won't be forever. Manning and Rivers don't have many seasons left.

Amusingly, their best shot at a win could come in week 17 if Denver has already sealed HFA and starts the scrubs.

by herewegobrownie... :: Wed, 10/29/2014 - 10:11am

I'm still trying to figure out where the nearly 3 projected wins for the Raidas are coming from (granted, they have been close a few times.)

by JoeyHarringtonsPiano :: Wed, 10/29/2014 - 10:16am

Mike Tanier once called the 2008 Lions "The Cadillac of winless teams" based on DVOA, which was better than a good number of 2 and 3 win teams. They had horrible luck and lost quite few close games in addition to some blowout losses. This Oakland team may end up the same way and take the dubious honor of "best winless team."

by JoeyHarringtonsPiano :: Wed, 10/29/2014 - 10:20am

I've watched a couple of Chiefs games, and DVOA has confirmed what my eyes told me...they're pretty good. They have close road losses to Denver and San Fran to go with a road upset of San Diego. People have already forgotten about them stomping the Patriots at home. They're the very definition of "under the radar." Still cant't figure out what happened in week 1 against Tennessee, though.

by LionInAZ :: Wed, 10/29/2014 - 11:27pm

I tend to discount week 1 results -- teams have too many new players, coaches, and schemes. I suspect it takes a week or two of film in games that count to figure opponent strengths and weaknesses.

by Mugsy :: Wed, 10/29/2014 - 9:48am

The Jets managed to keep it pretty close against the Broncos (down by 7 in the 4th qtr).
So I don't think Miami and Buffalo are going to be push-overs.. anything is possible, they both have pretty good defenses.

by nat :: Wed, 10/29/2014 - 11:02am

Because the Patriots had such a huge lead, these drives only count for about half-strength in the DVOA system...

Somehow I had missed this feature of DVOA. When was it added?

Regardless, it's not the weighting that's wrong with DVOA in blowouts. It's the baselines. DVOA baselines assume that you can represent an average team playing against an average opponent with data collected for an average play in each situation. That's good enough for most situations. But in blowouts it's flat out wrong. Average play results late in blowouts are more representative of bad (or depleted) teams facing much better opponents. The result is an easy DVOA grading curve for teams getting blown out.

In this case, DVOA is right to knock New England for weak prevent defense. But it's almost certainly off - perhaps by a lot - on how bad that play was.

Giving this game a 17.4% DVOA seems strange. I bet if you ran the numbers again, swapping the first and second half, you'd see this game in the 30s at least.

by dank067 :: Wed, 10/29/2014 - 12:57pm

Wouldn't the blowout situation work both ways, though? If defenses in general tend to give up a higher rate of successful plays in prevent situations, the defense would also not be judged as harshly for giving up more yards and points, or it might just offset the "easier grading curve" for the offense, as you say.

by nat :: Wed, 10/29/2014 - 3:04pm

DVOA compares these plays in blowouts to the right kind of situations. But it compares them to an average of a skewed mix of teams.

Getting the first right can't offset getting the second wrong.

So, no, it wouldn't work both ways.

by Pat :: Wed, 10/29/2014 - 12:01pm

Meh. The Eagles rating just reinforces what I've thought for a while now. They're not going anywhere this year - they wouldn't have a chance against a top-tier team in the league.

Look at their DVOA splits: they're a league-average team on offense+defense, combined with the best special teams in the league, by *far*. They're better than the #2 team by a gap about equal to the gap between #2 and #13.

What's the problem with that? It's all in punt and kick returns, and that's just not sustainable - not the least of which because their returner is dinged up. Plus they're just getting ridiculously lucky from opponents' special teams: best in the league by far.

It's a bit surprising because if you just glance at them, you say "oh, yeah, 5-2 team, #6 overall, that seems about right, OK, they're a decent team." And even their Super Bowl win % isn't that bad - 6.2%. But the problem is that you just don't get fantastic special teams like this that stick around. Only 2 teams since 1996 have had special teams over 10% for the season, and those teams both had their success spread more across the entire unit.

by young curmudgeon :: Wed, 10/29/2014 - 12:40pm

Pittsburgh is clearly ranked too high because I root for them and have very little faith in their defense holding a lead. My own personal anguish at watching them try to fritter away big leads against Cleveland and Indianapolis, and play an inexplicably drab game against Tampa Bay is way better than this. Go Stillers LOL YOLO FUBAR TSOL.

I realize that DVOA is a measurement rather than a predictive instrument (or so I think, correct me if I'm wrong), but looking at the teams that the Steelers are currently "better than," I would have no confidence in them beating New England or San Diego, and be in a toss up against Arizona, Cincinnati, New Orleans (in New Orleans), San Francisco, Dallas, and perhaps Miami. It seems to my subjective judgement as if there are a whole lot of teams bunched together; weighted DVOA suggests this with Philly at 12.4 to Buffalo at 3.1 being the end points of a group of 11; I'd include Arizona and San Francisco in what seems to be a large group of teams a "have a chance to be halfway decent, have an equal chance to be mediocre" category.

by RickD :: Wed, 10/29/2014 - 1:00pm

The Steelers team that beat the Colts could be anybody, Broncos included. And then there's the Steelers team that lost to the Bucs. Variance!

by MilkmanDanimal :: Wed, 10/29/2014 - 1:24pm

Tampa is clearly ranked too low because *collapses sobbing, drinks a bucket of vodka, sets self on fire*

by PatsFan :: Wed, 10/29/2014 - 1:46pm

Whatever happened to zlionsfan, anyways?

by LionInAZ :: Wed, 10/29/2014 - 11:37pm

He still contributes occasionally, but my recollection is that he's involved in game charting. So he's probably too busy for the comment boards.

by Steve in WI :: Wed, 10/29/2014 - 3:04pm

So the Bears now are 14th in offense, 23rd in defense, and 30th in special teams. (I'm surprised by how much worse their defensive DVOA got after NE, although opponent adjustments may have also played a part?)

Is it bad that I still think DVOA overrates their defense?

Edited to add: also, the playoff odds page now says it's more likely that the Bears will end up with the top pick in the draft than that they'll win the Super Bowl. (Granted, it's a 0.1% chance of the top pick). Is there anyone who would have predicted that before the season started?

by Eddo :: Wed, 10/29/2014 - 3:33pm

Yeah, simultaneously every Bears fan and no Bears fans.

I swear, fans expect every Bears team to have the 1985 Bears defense combined with the 2007 Patriots offense. And then, when they don't, they're "terrible" or "in dire straits" or the worst team ever. Why can't they just be a mediocre team?

Basically what I'm saying is I grow more and more tired of my fellow Bears fans every day.

by Will Allen :: Wed, 10/29/2014 - 3:43pm

Vikings fans expect every ray of hope and promising development to, just as it appears to reach potential, be dealt a crushing blow. With this belief, Vikings fans prove to be most excellent prognosticators.

by Alternator :: Wed, 10/29/2014 - 4:32pm

To a true pessimist, most surprises are pleasant ones.

by Steve in WI :: Wed, 10/29/2014 - 5:49pm

Really? Consensus on the Bears going into the season had them anywhere from 7-11 wins, I'd think, with an outside shot at a deep playoff run if the offense took another step forward and the defense was at least average. I don't think anyone predicted that the Bears would even be in the conversation for a really good draft pick.

by Jerry :: Wed, 10/29/2014 - 6:45pm

I swear, fans expect every Bears team to have the 1985 Bears defense combined with the 2007 Patriots offense. And then, when they don't, they're "terrible" or "in dire straits" or the worst team ever. Why can't they just be a mediocre team?

You can replace the first "Bears" with many other teams and describe the fans of every non-terrible team. (Fans of terrible teams expect 1944 Card-Pitt, and every good play by a young player is a step toward the Pro Bowl.)

by justanothersteve :: Wed, 10/29/2014 - 9:55pm

Agree. Packers fans expect the 2010 playoff offense combined with the 1961-62 Lombardi defense and the return game of 1996 featuring Desmond Howard. I heard on a Packers talk radio show earlier this year some idiot caller claiming Rodgers (pre-RELAX comment) was now a crappy, inaccurate QB and should be replaced. I didn't know a facepalm could be so painful.

by Eddo :: Wed, 10/29/2014 - 11:03pm

I don't doubt that. I do think the 1985 Bears, in particular, are possibly the worst thing that happened to the fandom in the long run. I heard complaints about how weak the 2012 defense was. And per FO, that was one of the best of the past twenty years, and by many other measures, the clear #1 that year.

Of course no one remembers that the 1985 team had to come from behind numerous times, to such strong teams as Tampa Bay and Green Bay. I'll actually be quite surprised if they're the #1 DVOA team of all time when it gets back that far, and I half-expect them to not even make the top ten.

by Will Allen :: Thu, 10/30/2014 - 12:18am

They really weren't all that great on offense. Competent, and more that enough when paired with a great defense, sure. There's a reason, however, that the defensive side of the ball had multiple HOFers, whereas the HOFer on the offense was well past his prime.

by tuluse :: Thu, 10/30/2014 - 1:11am

I expect if you split with McMahon apart from the rest, the 85 Bears are the best. Without him, they're probably just "very good".

by theslothook :: Thu, 10/30/2014 - 1:26am

I always wondered how McMahon was as a qb. I never watched him play, but even the general perception of his career feels weird. He gets praised for attitude, swagger and all that. But as a qb, I wonder if he was jake delhomme quality, dilfer quality or better?

by tuluse :: Thu, 10/30/2014 - 1:45am

He was far better than Dilfer who gets way to much credit for what he was.

Now, I mostly know the 80s Bears from highlights, but just that and looking at box scores, it's clear he was pretty good. However, it doesn't really matter how good McMahon actually was. He was so much better than his backups, he elevated the Bears from very good to possibly best of all time.

by JimZipCode :: Thu, 10/30/2014 - 2:31pm

>> He was far better than Dilfer who gets way too much credit

Time to climb back on my Dilfer-defense soapbox. Yes, Dilfer was an inefficient, borderline-terrible passer for the 2000 Ravens. But I would argue he gets way too LITTLE credit from advanced stat folks. He was the rare player for whom all the intangible stuff that people love to spew, really did make an impact.

Non Ravens fans may forget how slumped and defeated the Ravens offense looked toward the end of Tony Banks. Billick turned to Dilfer, and they had one more bad game and then sparked to life. Dilfer brought two things to the table. One, he played with that Brett Favre gusto, slamming into linebackers and then bouncing up smiling and clapping them on the back, that kind of thing. The same personality you see now on ESPN. It sounds stupid, and of course it's tough to tell as a fan, but it looked like some of that rubbed off on the squad. They didn't look as resigned and hangdog as they had looked for the last month playing with Banks.

The second thing Dilfer seemed to bring, was one deep strategic insight about what it means to play alongside a historically great defense. Most teams who are overbalanced with a strong D, they opt to play "mistake free, fundamentally sound" boring football. Run, run, run and punt. Win games 9 to 6 -- or lose by that score. Dilfer played as if he had made the opposite realization. "Hey, if the other team cannot do anything at all when they have the ball, ever, then we are free to take CRAZY risks on offense. Throw it deep into triple coverage, whatever." And so he took shots. Dilfer's stats are bad; but realize first that they are better than Banks', and second that his mistakes carried almost no negative impact to the team. (He threw one pick 6, in Tennessee week 11, but then came right back to throw the winning TD pass with 30 secs left.) So what if the defense has to go back onto the field? They'll get it right back; hell, we'll probably wind up with better field position after the exchange. And he'll take another shot.

In the decade+ since that Super Bowl, there have been persistent murmurings from some of the guys on that squad that Dilfer and Shannon Sharpe overruled calls from the sideline that they disagreed with, and substituted their own calls in the huddle. This has a ring of truth to it. Brian Billick went on to prove pretty conclusively over his 9-year head coaching career that he couldn't develop a QB to save his life; so Dilfer & Sharpe would seem to be justified in trusting their own judgement more. Sharpe already had two Super Bowl rings from his time in Denver. Dilfer had spent six years under the microscope in Tampa, and seemed to bring a lot of mental toughness to Baltimore. He was completely unfazed by a bad pass, or a bad possession, or a hard hit, or a criticizing headline.

It's easy to say that any replacement-level QB could have done what Dilfer did. But look over Brian Billick's coaching career and tell me exactly when he was able to find a replacement-level QB. Anyway, Dilfer brought some unique strengths to the situation that most "replacement level QBs" just don't have. Remember that Elvis Grbac, who was much better than replacement level, absolutely wilted in Baltimore the following season.

Trent Dilfer saved the Ravens season. Without him they would not have won the SB. Yes, he's the worst SB-winning QB of all time; yes, you could not build an offense around him; yes, he rode the defense to that win. But that's just it: he rode the hell out of that defense, in a way that no one else seemed to understand could be done.

Don't disrespect Dilfer on my watch. He was bad, but in an awesome way. :-)

by theslothook :: Thu, 10/30/2014 - 3:14pm

Its funny what has happened to Dilfer. IF he hadn't won that sb, he would probably not have the job he has now and be on a top 10 list of qb busts. As it is, he won the sb but now becomes the posterchild for the advanced stats community.

Anyways, I didn't follow the ravens or football all that closely that season, but I remember being fairly pessimistic about their chances all the way to sb. I guess in that sense, maybe Dilfer did bring something.

by young curmudgeon :: Fri, 10/31/2014 - 7:19pm

I always wanted to write an article entitled "Sympathy for the Dilfer." Now I don't have to!

by Will Allen :: Thu, 10/30/2014 - 3:14am

He was a good player who, once the injuries started, had a tough time staying on the field.

by theslothook :: Thu, 10/30/2014 - 3:32am

Got a contemporary player he reminds you of?

by ammek :: Thu, 10/30/2014 - 6:11am

It's really hard to imagine Chicago-era McMahon in a 21st century offense. Even when healthy he rarely attempted 30 passes per game, and on a good day that would be less than 20. Only about one-quarter of those passes were to wide receivers: the leading receivers on the mid-80s Bears were Walter Payton and the fullback Matt Suhey. Chicago's offense was quite old-fashioned even for the time: the similarity scores for McMahon at P-f-r are with players like Charley Johnson, Babe Parilli and Lamar McHan. McMahon was really a 1950s/60s quarterback: a tough guy with a big arm who was exceptional at soaking up pressure and delaying the release of screens to give Payton a more open field. He was Ditka-as-quarterback.

by Will Allen :: Thu, 10/30/2014 - 8:21am

I dunno, maybe Roethlisberger without the ability to extend plays into big passes downfield, and the ability to take the same pounding and keep playing. Ammek is right though, it was a really old fashioned offense.

by JimZipCode :: Thu, 10/30/2014 - 1:28pm

Maybe Alex Smith. Or no, Brad Johnson.

by Steve in WI :: Thu, 10/30/2014 - 10:40am

There were people who thought the 2012 Bears defense was weak? And I thought I hung around some ignorant Bears fans. (My personal all-time favorite comment came from a relative of mine who said after the NFC championship game loss to the Packers, the Bears should get rid of Cutler and make Caleb Hanie their starting quarterback).

by JimZipCode :: Thu, 10/30/2014 - 1:20pm

Baltimore is clearly ranked too high because – um, because they just don't look that good.

I'm a Ravens fan, and I don't necessarily have the best feel for what other teams are doing. So take this for what it's worth. But looking at the component rankings, one of those things is not like the others. The top-5 ranking for the Ravens D is believable, and the top-6 ranking for the spec teams. But the offense?

DVOA currently has the Ravens O better than Indy, New England, Cincinnati and others. Breaking that down, DVOA has them as the 5th-ranked pass offense and 16th-ranked rush offense. That ranking of the running game is very plausible; but DVOA places the Ravens pass offense better than Indy, Green Bay, New Orleans & Dallas (not to mention Atlanta, Arizona, Cincy). Really? That doesn't seem to pass the smell test.

There is more than one thing happening, as usual. One thing is that the Ravens O has faced the second-worst schedule of opposing D's by DVOA (according to the offense page). Probably the opponent adjustments are not yet at their most accurate, only halfway thru the season. That's fair. But I also wonder if DVOA consistently overrates teams that run a relatively "pure" West Coast offense.

Remember when Andy Reid's Eagles were always near the top of the rankings, even when they weren't very good? DVOA loved them. Jim Johnson's D was usually fabulous and effective; but DVOA also ranked the Eagles O very highly. And Reid's offenses were among the more "pure" WCO's since Bill Walsh retired, to the point of practically eschewing the running game. Now Gary Kubiak has taken over the Ravens offense, and he runs a fairly pure WCO, straight out of 1990's era Denver, with the Alex Gibbs zone stretch runs. And boom, Baltimore zips right to the top of the DVOA rankings, with most of the same personnel from last year's shitty offense. That seems very suggestive to me.
(I may have overplayed my argument just now, by saying "most of the same personnel." Obviously a few players have been high impact: Steve Smith, Zuttah, Wagner. Not trying to dismiss that.)

Aaron even says something to that effect in this post: "DVOA prefers teams that keep the offense on schedule and the chains moving with a steady mix of both small and large gains, because that's generally a better predictor of an offense which will continue to have success in the future." I think DVOA might be predisposed to a scheme bias. A good old fashioned West Coast dink and dunk offense will have a better chance of a high ranking than, say, an explosive Bruce Arians offense, which depends on hitting some big plays. If so, I think that's a flaw in DVOA. The high-variance schemes might be less likely to be successful on any one given play; but as you aggregate the shots, they become very likely to hit a big one; maybe more likely than DVOA admits.

Of course, there could be a number of things going on. It's not like Kubiak doesn't let Flacco take any shots downfield; that's a weakness in my argument. Maybe any team that gets to play Tampa this year will pop up to the top of the offensive DVOA chart for a few weeks. Maybe other traditionally-strong passing offenses haven't played as well so far this year as usual, so the Ravens compare well right now.

And hell, maybe the Ravens are that good. That would be awesome. Unfamiliar territory for a Ravens fan, but awesome.

by theslothook :: Thu, 10/30/2014 - 3:20pm

I had a similar take years ago when the 2010 Pats offense, which seemed to go ultra short throws. But in thinking it through, it actually made sense. Throwing so many times short meant higher probabilities of converting and avoiding turnovers.

I think DVOA is based on year to year correlations and that ultimately makes sense. To make a long ball offense work, you need to keep your o line healthy to withstand the long blocks and your receivers need to be able to fight off dbs for the ball deep. I can see how so much more of that is harder to sustain year to year than a dink and dunk offense.

Now, I do think when it comes to praising the qb, it gets murkier. By scheme, Palmer has a harder job than most other qbs, but then can anyone really run a dink and dunk offense effectively? This is a question that I don't think gets asked enough.

by chemical burn :: Thu, 10/30/2014 - 7:59pm

Look man, gotta correct you on perpetuating the repeatedly disproven myth about Reid's Eagles and DVOA - DVOA loved them because they were a big play team. McNabb couldn't hit a slant or a quick out to save his life, but he had one of the best deep balls of the past 20 years. The Eagles would mix effective runs with BIG pass plays. They consistently had a notably low completion percentage. But they also avoided turnovers - McNabb's most underrated strength. As for Johnson's defenses being fabulous? That's true for 2001 & 2002 and those years only - after that DVOA never thinks much of them and I agree. It only ranks them in the Top 5 once during the rest of Johnson's tenure. They consistently lost playoff games because their defenses were mediocre and particularly bad against the run, like the pitiful 2006 loss to the Saints or even the Superbowl loss where they simply had no answer for the Pats' (truly) dink and dunk game plan. Same for the 2008 NFCCG loss where the Cardinals ran the ball down there throat in the 4th quarter and the defense simply couldn't get off the field - and that was against the one defense DVOA liked.

Also, for the record, the idea that Reid's Eagles teams "weren't actually good" is total bullshit and revisionist history. They made the NFCCG 4 years in a row and came within a hair of going back to the Superbowl in 2008. DVOA is one of the few measures that gets why those Reid teams were good and how they were competitive in a tough division with as much consistency and longevity as any team not led by Manning or Brady. They had a SINGLE LOSING SEASON in a span running from 2000 to 2011 and that took a major meltdown for that losing season to happen. You should be looking at what DVOA got right about those teams and how they were competitive (including their run/pass ratio presaging the current NFL standards) not dismissing DVOA's insight. DVOA is trying to explain to you that those Eagles teams (specifically their) offenses were notably productive for a few reasons: effective running game that picked up consistent yards (and has a home run threat), a deep passing attack, a QB that make plays with his legs (which allows the deep passes time to develop and also allows him to steal yards when fear of the running game and deep routes combine to clear everything out underneath) and a lack of turnovers. A lot of that fits the 2014 Ravens - Flacco has a slightly above average amount of INT's but no fumbles, DVOA likes Forsett (also who has only a single fumble) and Flacco hits the deep ball as good as anyone in the league (which you admit)...

Also DVOA doesn't think their offense is great, just ok. It also doesn't think their pass offense is amazing - (again as you point out) even in a very down year for passing, they're only 5th in the league. While it might not "pass the smell test" they are better the New Orleans or Indy, you need to remember that New Orleans looked awful in several games this season, borderline incompetent versus Cleveland and Detroit. That's 25% of their season where they looked absolutely terrible. As for Indy, Luck continues to turn the ball over and take a lot of sacks. DVOA hates that. I don't watch of GB to know what's up with their ranking, but it's easy to forget that they also started the season slow and that Rodgers has traditionally take a relatively high number of sacks. Finally, while GB is ranked 7th in passing DVOA and Baltimore 5th, the 8th through 4th slots (CLE, GB, INDY, BAL PIT) are clustered relatively close together, about 8% difference from the worst to best - a single significant excellent or awful game from all them could instantly reorder the ranking.

Anyway, I suspect the Ravens' offense is what it has been for years under Flacco: good, not great. Capable of being REALLY great, but inconsistent.

by jonnyblazin :: Thu, 10/30/2014 - 8:35pm

While the Ravens being the 5th ranked pass offense is shocking, I'm even more surprised with their 7th ranked pass defense given their sad sack secondary play outside of Jimmy Smith.

"And hell, maybe the Ravens are that good. That would be awesome. Unfamiliar territory for a Ravens fan, but awesome."

Well, they were about this good according to DVOA during the regular season from 2008 to 2010, so it can't be that unfamiliar. Each year they were the wild card but had unusually high DVOA for a wild card, and it wouldn't surprise me if this year would follow suit.

by JimZipCode :: Fri, 10/31/2014 - 2:27am

I'm definitely used to the Ravens being good as a team overall – expect it, even, with all the sense of entitlement that goes with being a fan of a good team. We Ravens fans are not quite as annoying as Patriots fans, but still pretty bad.

But extremely unfamiliar with the Ravens having a top-5 passing offense. That's an "up is down and black is white" sort of idea. 1996 was a long time ago.

by Pat :: Fri, 10/31/2014 - 4:03pm

Well, the Ravens have the highest offensive variance of any of the other top 5 teams, by a pretty large margin.

My guess? The problem is that the Ravens haven't faced a good variety of pass defenses yet. Nearly 1/3 of their pass attempts have come against one of the best pass defenses in the league (Cincinnati), but the other pass defenses they've faced are garbage. And my guess is that the Ravens' performance against Cincinnati was pretty darn good - while they struggled early, the pass offense did a great job in cases where normal teams struggle (losing by 2+ touchdowns). The end-of-half drive against the Bengals in the first game, for instance, was sadly a great performance that netted nothing: but a single play there can't eliminate the previous plays.

The 'DVOA overrates Philly' thing has really been said over and over, but as the other poster here mentioned, there's really no basis behind it. 2008 is a good example, since they were rated #1 by DVOA although only 9-6-1 - except a ton of other more conventional stats had them in similar positions (net drive success had them 2nd, net drive yards had them 4th, etc.). They were just underperforming, period, for some reason.

by intel_chris :: Thu, 10/30/2014 - 4:10pm

JimZipCode wrote:

I think DVOA might be predisposed to a scheme bias. A good old fashioned West Coast dink and dunk offense will have a better chance of a high ranking than, say, an explosive Bruce Arians offense, which depends on hitting some big plays. If so, I think that's a flaw in DVOA. The high-variance schemes might be less likely to be successful on any one given play; but as you aggregate the shots, they become very likely to hit a big one; maybe more likely than DVOA admits.


Actually, that is intentional. Way back when, DVOA was created to show that a particular Patriots team was better than the traditional analysts were admitting. That team was a dink and dunk team that however slowly managed to move the chains and win the "field position" game. As such, the metric is intentionally skewed to highlight such (dink and dunk) teams. There are even factors like capping long-runs and other irreproducible plays that de-emphasize boom-or-bust performances as unsustainable.

Of course, in the "random" nature of football, such plays do have an extreme impact on any given game. The recent last minute win of the Cardinals over the Eagles. Dink and dunk doesn't create that. More relevant to a Ravens fan is the melt-down of Rahim Moore in 2012, which effectively sent Flacco & Co to the SB. If you are trying, dink and dunk, the error isn't so crucial. The long ball and resulting TD is what made the difference.

I don't think there is a way to "fix" DVOA to properly account for that. DVOA is right, such plays aren't easily reproduced. Most of the time the Eagles beat the Cardinals and the Broncos play in the 2012 SB, whether they would have done as well as the Ravens is a separate question and they still might not have gotten a ring that year. However, that's why we play the game to see those efforts and to realize that the teams which exceeded our expectations of them are worthy of praise. The Ravens are the worthy 2012 SB champs, in part because they did the required high risk play and succeeded at it. This year the Cardinals are leading in the NFC west because they made enough of the plays that they needed to. You don't need DVOA when you are winning.

by The Hypno-Toad :: Thu, 10/30/2014 - 7:46pm

You don't need DVOA when you are winning.

This should really just be the default response when a slighted fan comes in, dancing with the FOMBC in the pale moonlight.

by chemical burn :: Thu, 10/30/2014 - 8:01pm

Yeah, it's pretty much perfect.

by Eddo :: Fri, 10/31/2014 - 11:09am

"Way back when, DVOA was created to show that a particular Patriots team was better than the traditional analysts were admitting."

I thought it was the opposite, that Aaron started developing DVOA to show that the 20014 Patriots were a fluky champion.

But regardless, your overall point is a good one.

by coremill :: Fri, 10/31/2014 - 11:52am

Whether DVOA underrates teams that have a more boom/bust offensive attack is actually a pretty interesting question. I don't think it's the case in the regular season, where the season is mostly long enough that a lot of the variance washes out and so you'd prefer the more consistent offense. But in the playoffs against good defenses, adopting a more boom/bust strategy could be optimal for underdogs who know they might struggle to consistently move the ball. Rather than try to string together consistent medium gains, which is what you would do if you were trying to optimize your DVOA, maybe you're better off taking more deep shots and hoping you hit the hot end of the variance swing.

by mitch :: Sun, 11/02/2014 - 9:32am

I don't see Ravens being the 2cd best team. And I don't see them having a strong combination of off and def in passing game.

They have a QBPR of 3.6 better than opponents, ave per pass before sacks is even, after sacks is only .4, They throw more INT's, they throw for a smaller completion % , the only area they win by any amount is TD passes which carries the least importance going forward, good luck with those Ravens, just not seeing it.

by mitch :: Tue, 11/04/2014 - 10:39am

Pitt 136.3 QBPR, Balt 91.3
Pitt 9.2 ave per pass, Balt 6.7 before sacks
Pitt 8 per pass, Balt 5.5 after sacks
Pitt completion % 67.7, Balt 66.7
Pitt 0 INT, Balt 1 INT

Pitt crushed Balt in the areas Balt was weak before the game.

Add in, Balt pounded teams with losing records as noted past couple of weeks making them a guts and stomp type team, pounding those teams scoring wise but being dead even in ave per pass before sacks is horrendously terrible against weak teams they pounded is not a good showing for the team considered 2cd best team.