Week 18 DVOA Ratings

Week 18 DVOA Ratings
Week 18 DVOA Ratings
Photo: USA Today Sports Images

by Aaron Schatz

Once again, it is time for postseason DVOA ratings. We're ranking all 32 teams, whether they are in the playoffs or not. Teams which did not play in the wild card round are treated as if they had a bye week. (That includes both the 20 non-playoff teams and the four teams with byes.) Based on weighted DVOA, the teams to beat right now are Baltimore and Dallas. The Cowboys are significantly ahead of the rest of the NFC. The Ravens may or may not be significantly ahead of the rest of the AFC, especially the Colts -- but it depends upon how you view the final two weeks of the season and the Curtis Painter Experience.

The Colts are 14th in the standard weighted DVOA ratings. That would be fine if an injury to Peyton Manning actually forced the Colts to start Curtis Painter in this week's playoff game against the Ravens, but we know that's the case, and we know that the Colts really aren't the 14th-best team in football right now. Usually, we don't subjectively screw around with the DVOA ratings. However, once we get to the postseason, we're only using weighted DVOA, and the goal of weighted DVOA is not only to figure out how well a team played over the course of the season but also to figure out how good they are right now. If we're going to do that, we need to adjust for the games in the final two weeks where some teams played their backup quarterbacks and said, "Screw it, we don't care." Of course, that inherently means we have to make subjective decisions about which plays should count and which plays should not. I know some readers will have trouble with exactly where I decided that teams "turned it off," but here's the list of plays that are not included in the weighted DVOA ratings below:

  • Week 16, Colts-Jets, after Donald Brown's touchdown with 10:20 left in the third quarter.
  • Week 17, Bengals-Jets, second half.
  • Week 17, Cardinals-Packers, final three quarters.
  • Week 17, Chargers-Redskins, after Philip Rivers throws a touchdown pass to Antonio Gates with 6:42 left in the first quarter.
  • Week 17, Colts-Bills, final three quarters.
  • Week 17, Panthers-Saints, the entire game.

These changes are made for offense and defense, but not for special teams. For those who disagree with these changes, I've also listed the "standard" weighted DVOA that includes these plays. The "LAST WEEK" listing is based on what weighted DVOA would have looked like a week ago with these changes made, so it will be different from the weighted DVOA actually listed on the site. And, of course, since weighted DVOA is meant to lower the strength of older games, these ratings do not include Weeks 1-4, and Weeks 5-10 are somewhat discounted. If you want to see the actual regular season final ratings, click here.

The playoff odds report is updated through the wild card games, and you will find DVOA matchup pages for the four second round games on the FO Premium page.

Scroll down below the weighted DVOA ratings for single-game ratings on the wild card round and some comments about why our preseason projections went wrong in 2009.

* * * * *

To save people some time, we remind everyone to put their angry troll hatred into the official zlionsfan angry troll hatred Mad Libs form:

<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>

If you are new to our website, you can read the explanation of how DVOA is figured here. As always, positive numbers represent more points so DEFENSE is better when it is NEGATIVE.

1 BAL 39.6% 2 10-7 16.1% 10 -19.0% 2 4.5% 4 39.6% 1
2 DAL 37.6% 3 12-5 26.5% 4 -8.9% 5 2.1% 13 37.6% 2
3 SD 28.5% 5 13-3 30.6% 2 2.3% 14 0.2% 21 24.3% 6
4 GB 26.9% 1 11-6 32.7% 1 -3.3% 8 -9.0% 32 27.1% 4
5 CAR 23.7% 6 8-8 11.0% 14 -14.0% 3 -1.4% 23 29.3% 3
6 IND 21.6% 7 14-2 23.8% 5 1.5% 13 -0.8% 22 8.7% 14
7 NYJ 20.3% 14 10-7 -3.0% 22 -21.4% 1 1.9% 14 26.6% 5
8 ARI 19.2% 12 11-6 22.6% 6 5.6% 19 2.2% 12 17.9% 9
9 HOU 18.8% 9 9-7 17.3% 8 0.1% 11 1.6% 15 18.8% 7
10 NE 18.3% 4 10-7 19.5% 7 2.5% 15 1.3% 17 18.3% 8
11 PHI 17.5% 8 11-6 11.1% 13 -2.1% 9 4.4% 5 17.5% 10
12 NO 17.3% 10 13-3 28.3% 3 8.3% 22 -2.7% 25 12.3% 13
13 PIT 16.4% 11 9-7 16.3% 9 -3.6% 7 -3.5% 28 16.4% 11
14 MIN 12.8% 13 12-4 15.5% 11 4.1% 17 1.4% 16 12.8% 12
15 SF 8.4% 15 8-8 -5.2% 24 -13.4% 4 0.2% 20 8.4% 15
16 TEN 8.1% 18 8-8 15.4% 12 9.6% 24 2.2% 11 8.1% 16
17 MIA 3.6% 17 7-9 3.7% 17 3.6% 16 3.5% 7 3.6% 17
18 DEN 1.8% 16 8-8 -1.6% 21 -0.4% 10 3.0% 8 1.8% 18
19 WAS -3.7% 21 4-12 0.2% 20 1.1% 12 -2.8% 26 -1.4% 20
20 CIN -4.2% 19 10-7 0.3% 19 5.6% 20 1.1% 18 -6.3% 22
21 ATL -5.0% 20 9-7 5.1% 16 7.0% 21 -3.1% 27 -5.0% 21
22 NYG -7.0% 22 8-8 9.2% 15 13.8% 26 -2.4% 24 -7.0% 23
23 BUF -11.1% 23 6-10 -20.6% 29 -5.9% 6 3.6% 6 0.1% 19
24 CLE -14.7% 25 5-11 -4.3% 23 19.1% 30 8.7% 1 -14.7% 24
25 JAC -15.9% 24 7-9 3.6% 18 15.7% 28 -3.8% 29 -15.9% 25
26 TB -17.1% 26 3-13 -18.8% 28 5.0% 18 6.7% 2 -17.1% 26
27 CHI -22.9% 27 7-9 -16.0% 27 9.5% 23 2.7% 9 -22.9% 27
28 OAK -24.9% 28 5-11 -11.9% 26 15.4% 27 2.4% 10 -24.9% 28
29 KC -30.9% 29 4-12 -11.3% 25 12.4% 25 -7.2% 31 -30.9% 29
30 SEA -40.8% 30 5-11 -23.6% 31 18.3% 29 1.1% 19 -40.8% 30
31 STL -45.2% 31 1-15 -31.2% 32 19.3% 31 5.3% 3 -45.2% 31
32 DET -47.4% 32 2-14 -22.5% 30 20.8% 32 -4.1% 30 -47.4% 32

Here are the one-game DVOA ratings for the first round of the playoffs. You'll notice that, even though the game went to overtime with a controversial ending, DVOA believes the Cardinals easily outplayed the Packers yesterday. The Cardinals averaged 1.5 yards more per play and also were much more efficient on third downs.

DVOA (with opponent adjustments)
NYJ 45% 40% -9% -4%
CIN -29% 25% 44% -10%
DAL 74% 47% -18% 9%
PHI -53% -21% 29% -3%
BAL 98% 24% -80% -6%
NE -30% -33% -4% -1%
ARI 69% 103% 25% -8%
GB -44% 43% 85% -2%
VOAf (no opponent adjustments)
NYJ 46% 42% -8% -4%
CIN -52% -5% 37% -10%
DAL 51% 35% -7% 9%
PHI -81% -27% 51% -3%
BAL 60% 15% -51% -6%
NE -57% -48% 8% -1%
ARI 37% 86% 40% -8%
GB -54% 40% 91% -2%

* * * * *

Now, a few words about this year's projections. There was a lot of discussion in last week's DVOA ratings thread about our poor 2009 preseason projections and how they compared to projections in years past. In addition, we know that the poor preseason projections were the reason why our premium picks against the spread were poor for the first half of the season -- those formulas take the preseason projections into account because in past years this has been a more accurate way to judge how good a team really is early in the season.

So what went wrong? The answer is, very clearly, defense. It was just a strange and unpredictable year for defenses, at least according to the methods we've developed, methods that have worked in the past.

Here is a table showing the correlation coefficient of each year's preseason DVOA projections to the actual DVOA ratings of that season in offense, defense, special teams, and total. These are based on the projections as we ran them each year, using the formulas we were using at the time.

Year Offense Defense Special Teams Total DVOA
2005 .44 .27 .27 .30
2006 .45 .45 .28 .42
2007 .63 .25 .24 .62
2008 .36 .55 .27 .52
2009 .55 -.02 .42 .16

You may have noticed that 2009 really stands out. No, I'm not talking about the fact that this was our best year for projecting special teams performance. I'm talking about the fact that our defensive projections were completely, totally wrong in a way they never had been before. There was basically no correlation whatsoever between our projections for defense and how well defenses actually played. Those mistakes caused our overall projections to be way off compared to previous seasons, even though both offensive and special teams projections were better than the year before. (You can see those projections here.)

What happened? Was this just a hard year to project defense, or did my attempts to improve our projections before 2009 actually break the system? To check, I went back and calculated what the 2009 projections would have been if we used the same system used in 2008. The correlation between those projections and the actual defensive DVOA ratings was slightly worse, with a correlation coefficient of -.05. That suggests that this year's formula was really no worse than last year's.

What's odd is that team-by-team defensive performance didn't change that much more than usual between 2008 and 2009. From 2000-2008, the year-to-year correlation of defensive DVOA was .45. From 2008 to 2009, it was .37. It just so happens that the specific indicators that had always helped us to pick out which defenses would improve or decline didn't work this year. Either general defensive trends in the NFL and the way defenses are built has changed significantly, or 2009 was just kind of kooky. I'm guessing the latter.


116 comments, Last at 14 Jan 2010, 11:21am

1 Re: Week 18 DVOA Ratings

Typo for NYJ, showing weighted as 2.3% should be 20.3%

2 Re: Week 18 DVOA Ratings

91% defensive dvoa for the Packers, prior to opponent adjustment, yesterday. Yes, they sucked as bad as they looked, in case you were wondering.

It'll be interesting to see if the Vikings/Cowboys game becomes a pick'em, or even a Vikings home dog, prior to Sunday.

Let's see, Arizona's defense will not benefit from the homw field on Sunday. The Saints offensive line is better than the Packers'. The Saints' qb makes decisions more quickly. I think Warner better assume he needs to score in the forties again.

19 Re: Week 18 DVOA Ratings

In reply to by Will Allen (not verified)

Will, have you noticed if the Vikings have put two run-stoppers on the d-line at the same time? I have seen a Allen, Robison, Edwards and K.Williams line on passing downs, but do they every put Jimmy Kennedy with Pat Williams on obvious rushing downs? That might be a good idea against Dallas and their mammoth o-line.

36 Re: Week 18 DVOA Ratings

In reply to by peterplaysbass… (not verified)

Kevin Williams is better against the run, along with the pass, of course, than Jimmy Kennedy. Other than K Williams being gassed, I don't think there is ever a good reason to have Kennedy on the field instead. Kennedy has been a little better than I expected, but what the Vikings need defensively is for Pat Williams to return to what he was one or two yeras ago, instead of just being the good player he was this year. At his age, that may be a tall order. Short of that, it would be nice if Winfield had fully healed, and if Allen and Edward have good games. All is not lost; Flozell Adams in a noisy road environment could be made to look really bad by Allen, IF the Vikings can get some early points.

To do that, however, the Vikings offensive line is going to really have to step it's game up. The best way for them to slow down Ware and Spencer a little would be to run effectively, but this group, in good part due to the fall-off from Birk to Sullivan, just can't run block as well as they have in years past. Jay Ratliff is a great player, and Sullivan is going to have a lot of responsibility with regard to him. Who knows? It wouldn't be the first time that a team suddenly played better in the playoffs, but there isn't a lot of reason to think it will happen.

53 Re: Week 18 DVOA Ratings

Will, you certainly have been pessimistic all season. The team has obviously outplayed your projections, even with their end of season slide. Still, no one could argue with your assessments--this will be the worst possible opponent of the bunch for the Vikes. But I don't fear for them because of the D Line, what is undebateable is that the loss of the middle linebacker has really weakened them. The big chunks of rush yards that Arizona and Carolina took out of them are what the Cowboys are licking their chops over in the film room and mainly have to do with that missing piece of the Vikes. They need to do what you said, get early points to get Dallas throwing early. I'm sure Favre and the veterans won the showdown with Childress such that they will now, as they did vs NYG, throw to open the run game rather than vice-versa. So who would you rather have, a laser focused Favre or a red hot Romo yet to win a road playoff game ? The noise will discombobulate the Dallas Bigs and the pressure on the Cowboys, that I know a rested D Line will generate, will help offset the Viking secondary weakness. Nonetheless, this will be one seriously tough game--Dallas is playing the best defense of all right now. IF the Jones Boys get by this all the Wade haters will have to acknowledge the guy knows what he's doing defensively. I'll take Brett Favre, A.P. and company at home...

58 Re: Week 18 DVOA Ratings

I didn't think Favre would stay healthy, and I projected the Vikings at 9-10 wins in the DVOA projections thread back in early September. I was wrong about Favre, but I'd make the same assumption again about a 40 year old at qb. Absent Favre's play, this is an eight win team, assuming Tavaris Jackson could have achieved mediocrity.

Napoleon Harris, currently a dynamic playmaker for the Florida Tuskers, played a lot at middle linebacker for the Vikings last year, and they still had a top five defense, in good part because Pat Williams, prior to getting hurt, was playing at an All-Pro level, and Winfield was healthy. Pat is still good, but has slipped a bit, and Winfield still hasn't shown himself to be right, since getting hurt.

Oh, they can beat the Cowboys, but if they do it will because they will have departed from their standard of play since the 2nd Green Bay game. Maybe they will. If either you or I knew, we wouldn't be having this dialogue.

60 Re: Week 18 DVOA Ratings

Assuming you are some sort of long time fan of the Vikes, which you appear to be, it's certainly a mystery how one of your knowledge could say this was an 8 win team without no. 4. They had much more talent than a hypothetical 8 wins in that rotten division. Granted, Tavaris Jackson would've lost several games for them but, come on, that defensive front is truly superior. Just ask the teams that have to face Allen, Williams and company. As to the secondary, the week of rest surely had to help Mr Winfield.

My own view is that they were a 9/10 win team without a top flight QB. With Brett, once it was clear that his shoulder was, in fact, healed, they were going to win 11/13 games. The shoulder being 100% was pretty clear very early on. Worrying about him being 40 was an unfounded concern. By and large, the Bigs did a good job of protecting him. IF they can do it one more time I think he'll make everyone even happier yet that they signed him. But Dallas is generating a tremendous rush right now. I wouldn't want to be in the O Line's shoes...

71 Re: Week 18 DVOA Ratings

Tavaris Jackson has never had a good game against a good defense with something to play for. The Vikings lost four this year. Absent Favre, it is pretty obvious that they lose to San Francisco and Baltimore. Splitting with the Packers is pretty likely, and the possibility exists that they get swept. Toss is one real train wreck, which Jackson has shown himself capable of a few times, and you get to eight losses pretty easily. Especially once one considers that this team has not run blocked nearly as well this year, which means Jackson would have been asked to do a lot more this year than last. There isn't the tiniest shred of evidence to suggest that he would have been capable of it, and the fact that the defensive performance has declined significantly from last year still needs to be factored.

This is a team that easily might have won 14 games last year with this year's qb play, and a team that likely would have won eight, at best, this year, if forced to play with last year's qb play. If you think they have been anywhere close to as good as they were last year, on the line of scrimmage, you haven't been watching closely. Why do you think Peterson's rushing average has fallen from 4.8 yards per carry to 4.4, despite the fact that the passing attack has forced opponents to defend the run more honestly? They simply aren't as strong up front, offensively especially, but a 40 year old qb covered that up somewhat. If you wish to believe that the fact that the 40 year old didn't get hurt disproves the notion that a 40 year old, even this 40 year old, has a significantly higher chance of getting hurt, than a younger version, you just go right ahead.

74 Re: Week 18 DVOA Ratings

I have been watching and it is true that their O Line play seems less sharp in recent weeks than earlier in the season. I saw them live both this year and last and I can assure you that they were just as crisp earlier this year as they were a year ago. Why have they fallen off, at least on the road, at the end ? Not sure but they do look fatigued. The week off should work wonders. I do think the tension convention that was going on over which M.O. should be the choice, Childress or Favre's, was a contributing factor. I can tell you that teams do not like confusion over what to prepare for, especially late in a meaningful season.

Revisiting games and imagining what would have happened with someone else playing is not something that is based in reality. Everything would have been different, as you know. So it's impossible to know what would've happened.

As to 40 yr olds and their actuarial percentages, etc.--I want to thank you for giving me the clearance to think what it is I think. Appreciate it. And something tells me, gee I wonder what, that you are convinced your team is about to lose. Perhaps they will, of course, with this very tough opponent and the problems they have manifested. But they should show up and play the game anyway, shouldn't they ??

76 Re: Week 18 DVOA Ratings

You're welcome. Say, when you read the words....

"Oh, they can beat the Cowboys, but if they do it will because they will have departed from their standard of play since the 2nd Green Bay game."

....is there some reason that you interpret the statement as saying I am convinced that they are going to lose?

Yes, everything would have been different if Jackson had been their starting qb this year. They likely would have been playing from behind more often, in good part due to a defense which wasn't playing as well, and a running game which was not as potent as last year's, with a qb who hasn't yet learned to read the entire field, is less accurate, and who doesn't manipulate opposing defensive backs with anything close to the skill that the other guy displays.

Look, if you think age is a meaningless factor in terms of the chances of injury, fine, then your assertion that concern over a 40 year old's age being unfounded makes sense. Otherwise it is just a fine display of 20-20 hindsight, and of limited value.

90 Re: Week 18 DVOA Ratings

Tavaris Jackson was pretty awful allright. But your view into the past was only that, the past. We will never know if he would've grown. Unfortunately for him, he hit the probable closing of his window of opportunity in pro athletics. It's gotten smaller and smaller over the years. When you and I were first starting Pee Wee league the pro's used to say "it takes five years to make a quarterback". Can you imagine everybody? Today's greedy owners want people to win by their second season--everyone, not just QB's.

As for the particular 39/40 yr old that started the season for Minny--he was, afterall, the all-time iron man. Just because a load of fans, threaders on websites, talking heads and various other pundits, thought Favre and the Vikes were both nuts for bringing it into reality that he would suit up, particulary after his first serious injury, well they need something new to talk about.
But anyway, I'm not looking to pick a cyberfight with you. All of your concerns are totally understandable. You're a fan of the other 4 time loser, besides Buffalo, and you're preparing yourself, by looking at all the negatives, for the very possible barrage of a focused Cowboy team who, right now anyway, looks like the certain equal of the Vikes. This week I'll be a fan. I'm going to hope for Minny to beat dem Cowboys. It'll be good for football if Minny can get into another SB and their fans are truly deserving. Besides, I don't like Jerry Jones. Minny has the most talent, I hope they rise. Good luck to your team, may they not lay another egg like it was in Carolina and Arizona...

94 Re: Week 18 DVOA Ratings

Well, we are certainly in agreement regarding the distasteful nature of one Jerrel Jones. In any case, I thought bringing in Favre was a no-brainer, even with the significant chance of injury, because the alternative was so unacceptable.

If the Vikings can regain, on the line of scrimmage, some of their form from the not too distant past, they can win. Will they? I dunno, but there wouldn't be trends if what happened recently didn't very often continue to happen.

39 Re: Week 18 DVOA Ratings

In reply to by peterplaysbass… (not verified)

But do the Cowboys really employ their line as a mammoth run-blocking, downfield unit? If so, then I'd agree that having the Williamses, or Williams/Kennedy, in there would be a good move. But in general, I think sacrificing the size, for more speed and ability to get pressure, is probably a good move.

Given that they've had 2 weeks to rest up, I'm hoping that Allen and Company look more like their early-season form. When that defense doesn't get good pressure, they look horrid.

115 Re: Week 18 DVOA Ratings

..."Walkthrough: Secret Identity" describes the Dallas "Freeze Draw" and how it meshes with the short-passing game and Romo's quick release...and something else that has been implemented since the first Eagles game back in November, namely, the flanker-screen and phantom screen...the counter to all the double A gap blitzing by Philadelphia and any other team that will dare to try it...when the Cowboys deploy in "22" they sometimes will send Austin deep if the running game is really humming along...between the phantom/flanker-screen, freeze draw, and the "22" package they toyed with the Eagles like a cat with a ball of yarn...it was beautiful to watch...McDermott was told he must blitz and so he did...but Dallas was ready...Barber being injured may be better for the Dallas rushing attack because Barber has been much too tentative this year and is'nt hitting the gaps allowing opposing defenses to bottle him up sometimes...Jones and Choice are faster through the gaps, like Carolina's Stewart/Williams combo...Jones gets to the second level pretty fast and he can get outside as well...he got the Eagles two weeks consecutive with TD runs...

116 Re: Week 18 DVOA Ratings

Yeah they sure are humming along with all this for the last month. They've been able to defeat the blitz but can their Bigs defeat some of the top pass rushers from the D Line in a really, really loud environment ? We're about to find out. There were no Jared Allen's & Co. on N.O. and Philly. Barber is a dancer allright, but he's a pounder vs the heavy D Line rush and you have to have that. Is he really out for sure ? You have a meddling owner there that wants all his toys in place no matter what...

3 Re: Week 18 DVOA Ratings

Maybe it's just me, but there seemed to be A LOT of defensive injuries this year. I think something like 4 of the top 5 3-4 Nose Tackles went down. Two of the best defenses lost coordinators. One of those, the Eagles, lost a couple of very good linebackers. The Titans lost Haynesworth to the free agent wasteland known as the Redskins

I think it really HAS been a very strange year for defenses and much of it could be due to injuries.

32 Re: Week 18 DVOA Ratings

Ed Reed was injured and out of a number of games and Polamalu missed about 3/4 of the season. Two top Safeties as well.

38 Re: Week 18 DVOA Ratings

The Bears defense that was projected be to great by by the FOA, had 5 of it's opening day starters playing in week 16.

45 Re: Week 18 DVOA Ratings

Who are these top 5 nose tackles you speak of? Shaun Rogers missed 4 games, Vince Wilfork missed 3, Casey Hampton played every game, Jay Ratliff played every game, Aubrayo Franklin played every game, Haloti Ngata played every game...

All I can think of are Jamaal Williams and Kris Jenkins, but 2 NT's missing most of the season doesan't sound so out of the ordinary to me.

I don't think it was injuries, I think it was just a freak year. I keep thinking of how the Niners were projected to be 30th (apprx.) in defense and finished 4th, they just finally lived up to their talent. Their front 7 was very healthy, but the secondary wasn't. It was just something you couldn't really predict.

79 Re: Week 18 DVOA Ratings

I thought Shaun Rogers went out for the year a while ago. I also thought Wilfork missed more games.

I guess I was a bit confused, although I may have been thinking about that Panthers tackle that went down before the year. I know they aren't a 3-4 but everyone was saying how important he was to their run D. But given their final DVOA, I guess they were OK.

99 Re: Week 18 DVOA Ratings

You'd have to look at a week-by-week breakdown, but the Panthers spotted the field a 3-game head start, in large part to the loss of Maake, and other DT's.

4 Re: Week 18 DVOA Ratings

In the cases of GB, DEN and NO, I don't think conventional wisdom would've fared much better than -.02.

7 Re: Week 18 DVOA Ratings

I would disagree with regards to Denver and NO. Both teams signed a bunch of vets that were good for other teams.

NYJ you didn't mention and they improved a lot. They also signed a couple guys, plus Rex Ryan.

6 Re: Week 18 DVOA Ratings

I'm not the sharpest knife in the drawer, but that Jet's Wei. DVOA of 2.3% looks like a typo to me.

8 Re: Week 18 DVOA Ratings

Wow, you throw out some early Jake Delhomme games, and the Panthers start to look pretty good.

9 Re: Week 18 DVOA Ratings

You'll notice that, even though the game went to overtime with a controversial ending, DVOA believes the Cardinals easily outplayed the Packers yesterday.

Technically, I think you meant VOA.

Edit: Also, on the table with Wild Card games, both ST VOA and DVOA are identical.

18 Re: Week 18 DVOA Ratings

Thank you, Temo. This DVOA vs VOA issue drives me nuts.

Use VOA to compare teams in the same game, or to compare results on the field. Use DVOA to compare the quality shown by teams playing different opponents.

For example: VOA says the Jets dominated their opponent more than the Cardinals dominated theirs. (no surprise there) DVOA says that given the opponents involved the Cardinals' play indicates a higher quality team.

10 Re: Week 18 DVOA Ratings

Some strangeness with the weighted offense & defense rankings. Shouldn't the Ravens be #2 in weighted defense, not #3?

Also, Chargers Offense DVOA has similar typo to the Jets overall DVOA before fixing, off by a decimal point.

Sorry for the nitpicks, thanks for getting these up.

11 Re: Week 18 DVOA Ratings

One of the bigger correlations this year (and not on other years) is that of the coaching changes; many defenses changed either coordinators or coaches, had a big turnover - and those appeared to give big dividends. It's hard to accurately predict this using a model no matter what. In the future you should likely look at these sorts of outliers and use them as a way to give a range. For instance, Rex Ryan: you could use the upper bound as 'as good as the Ravens D' and as bad as ' as good as the Raven's O', and then evaluate appropriately.

Same goes for Gregg Williams and Josh McDaniels, who are likely the big outliers.

Another observation is that FO's correlation is based largely around how good it predicts things in the past; that might be too closely tied to specific teams and how well they did. One of the reasons I think the Eagles get consistently overrated is due to how they play; the short passing game was what the Pats used and what the Eagles use, and both showed success early. But that doesn't mean that's the right way to predict future performance.

12 Re: Week 18 DVOA Ratings

Hey Aaron.

Have you done any calcutions comparing the correlation of your pre-seasons predictions compared to the predicitions of other "experts".

And btw, what was the DYAR and DVOA of Kurt Warner in that game? I dont think i've seen any QB play that well the whole season (except maybe from Brady against Titans). But did he not breake any kind of record? Like the highest DVOA in a game having 25+ attempts or something?

13 Re: Week 18 DVOA Ratings

DVOA believes the Cardinals easily outplayed the Packers yesterday. The Cardinals averaged 1.5 yards more per play and also were much more efficient on third downs

I'm not surprised at the overall differential, nor at the Pack's horrendous 91 defensive DVOA. I am surprised that you choose to highlight the third-down numbers. The Cardinals ran 57 offensive plays, but only five of them were third downs. They converted three (60%); only one was a third-and long, a 13-yard pass to Urban on 3rd&8; the others both came on the Cards' last touchdown drive, a 7-yard completion on 3rd&4 and a one-yard run on 3rd&1.

The Packers, meanwhile, converted 5-of-11 third downs (45%), plus a pair of fourth downs. I guess they did have a turnover too. Even so, I imagined the big difference would be in the second-down numbers: by my count, Arizona ran 19 second-down plays, and an astonishing 14 of them went for a first down or a touchdown.

14 Re: Week 18 DVOA Ratings

Well, they appear to have outperformed the Packers on both 2nd and 3rd down, and even if they do even better on 2nd than 3rd, no one goes around saying 'boy, they were really good on 2nd down in that game', so the more natural inclination is to highlight 3rd down performance.

30 Re: Week 18 DVOA Ratings

Well they weren't particularly great on third down; that's why it seems strange to mention it. Of course, they didn't need to be: they were lights-out everywhere else.

Has any team ever had a lower percentage of third-down plays on offense, while committing only one turnover? Perhaps back in the days when teams punted on third down.

16 Re: Week 18 DVOA Ratings

This alignes precisely with my unscientific perspective on the game. Although both teams drove the ball up and down the field, it felt much easier for the Cardinals. The Packers needed a lot more downs to cover the same yardage.

33 Re: Week 18 DVOA Ratings

That probably has at least as much to do with the fact that GB was playing from far behind for most of the game, it gives more of a desperation feel to each play, since the Cardinals could afford to not make every play count more than GB could.

Of course, the fact that GB needed to go for two 4th downs was also certainly a factor that indicated better efficiency by Arizona.

24 Re: Week 18 DVOA Ratings

Just to pile on………

Cardinals' successful plays from scrimmage (in DVOA terms) by quarter:

1Q: 59% (10/17)
2Q: 70 or 80% (7 or 8/10)*
3Q: 100% (9/9)
4Q: 67% (14/21)

* Second-quarter Fitz catch-and-fumble was initially successful in gaining 6 yards, but became a turnover.

Total: 40/57, or 70%.

What's the league average? About 44%?

Stats exclude penalties: Cards gained two first downs on Packer defensive penalties, plus five yards leading to a 1st&5.

15 Re: Week 18 DVOA Ratings

I remember being so excited to get the FOA Almanac, and then to see the prediction that the Ravens would be mediocre this year. What surprised me so much at the time was the abysmal rankings FO gave the Ravens offense. They had them ranked worse that the Raiders at 29th. I simply still can't what logic (or statistical evidence) there was for such a prediction. The Ravens had a solid young offensive line, a sophomore quarterback who had a reasonably promising perfomance his rookie year, and a stable of running backs, at least one of which (McClain) was highly ranked by DVOA. Admittedly, no-one saw Ray Rice coming outside of the Ravens organization, but still, there just didn't seem to be reason the Ravens offense would absolutely implode like that. Was this a holdover from the history of abysmal Ravens offenses?

22 Re: Week 18 DVOA Ratings

Without looking it up, since I don't have it here, I can think of a few points against them.

Second-year QBs usually regress (I think), and Joe Flacco wasn't really gunning it around in year one as it was.
Their receivers were nothing to write home about, and I think Derrick Mason retired at one point, though I don't know if this was during that time or not.
Their offensive line is young, and I think young O-lines have more problems than older ones.

Though this is all from spotty memory about a team I don't follow closely, so I could be completely wrong.

34 Re: Week 18 DVOA Ratings

Second-year QBs usually regress... Really? That seems kind of amazing to me. I would think that QBs improve over time from experience alone. Does other teams getting tape on the player really make more of a difference than the player's personal improvement?

I fully expect large Derrick Mason regression was a big factor in what caused them to scale back the Ravens, and he certainly did regress somewhat. I actually thought that would be counterbalanced by an improvement in Joe Flacco and a drafting of a right tackle in the first round (I'm sure draft picks impact projections).

But if you're right that secondary QBs usually regress, the projection makes a lot more sense.

41 Re: Week 18 DVOA Ratings

In addition, the O-line was largely an unknown. They would be (at the beginning of the season) losing a long-time stalwart tackle, and starting a rookie and a bunch of guys with 1 year of experience.

Combined with a 2nd-year QB, I can certainly see the reasons for pessimism. That the Ravens performed so well is a testament to both the scouting/front office (no surprise there, Ballmer's been good there for years) and the coaching/training. They look dangerous, and the Steelers should be worried about their hegemony.

54 Re: Week 18 DVOA Ratings

All players typically regress on average. So someone like Flacco who comes in and lights it up his rookie year is expected to be worse. This is probably one of the biggest misconceptions NFL fans have about young players. Yes 1 of every 5 guys who has a great rookie year goes on to improve and become a star, but 4 in 5 regress a bit the next year.

Yes older players regress more quickly than young ones, but young players generally regress to the mean too. I know you can then claim there is something different about Flacco that makes him a special case, but of course anyone with a good rokie QB will claim that about them.

65 Re: Week 18 DVOA Ratings

I think you see this with QBs in particular because of the lack-of-game-tape effect. As the league gets more information about a quarterback's tendencies, defenses are more effective at limiting his production. I think it's especially pronounced in quarterbacks that have some rookie success, as they might be too content with going with what worked their rookie year, thus playing into the adjusting defenses' hands.

Of course, you will get a player like Peyton Manning who is equally good at adjusting his own game. And those are the ones who become stars.

78 Re: Week 18 DVOA Ratings

I believe Lewin's research suggested that quarterbacks, on average, improve significantly year on year up until year five, after which they roughly plateau until around year nine before starting to decline significantly. The biggest jumps in performance were between years one and two and years four and five. Now, of course regression to the mean is a factor here: Flacco's not a randomly selected quarterback, he's a quarterback selected specifically because he had a good rookie year, and therefore a candidate for negative regression towards the mean. Then again, looking at his conventional statistics, Flacco has thrown more passes for a higher completion percentage, higher YPA, higher touchdown percentage, lower interception percentage and lower sack percentage this year than he did last, presumably while being asked to do more. I find it hard to buy any argument that says he hasn't genuinely improved.

82 Re: Week 18 DVOA Ratings

Using p-f-r's ANY/A stat, here is a list of QBs since 1990 who threw at least 150 passes in both the first and second seasons in which they threw an NFL pass. The first list contains only those QBs who, like Flacco, managed ANY/A of 5 or more in year one:

QB / change / yr1 / yr2

Palmer +2.3 / 5.0 / 7.3
Garcia +1.7 / 5.6 / 7.3
Flacco +0.8 / 5.3 / 6.1
Roethlisberger +0.6 / 6.9 / 7.5
Leftwich +0.5 / 5.2 / 5.7
O'Donnell +0.5 / 5.2 / 5.7
Ramsey +0.0 / 5.0 / 5.0
Batch +0.0 / 5.6 / 5.6
Campbell -0.2 / 5.5 / 5.3
Brooks -1.4 / 6.4 / 5.0
Ryan -1.4 / 7.0 / 5.6
Bulger -2.4 / 7.7 / 5.3

That's a very small sample size. Although there is some downward movement, every QB on the list had ANY/A of at least 5.0 in both seasons.

Reducing the barrier for entry to 4.0 ANY/A in Season One adds:

P Manning +2.5 / 4.8 / 7.3
Collins +1.4 / 4.4 / 5.8
Edwards +1.3 / 4.6 / 5.9
Couch +1.1 / 4.0 / 5.1
Bledsoe +0.7 / 4.5 / 5.2
Plummer +0.5 / 4.4 / 4.9
Carter +0.3 / 4.3 / 4.6
Young -0.1 / 4.5 / 4.4
Harrington -0.1 / 4.0 / 3.9
George -0.3 / 4.2 / 3.9
McNown -0.3 / 4.3 / 4.0

Conclusion: most QBs improve in their second year. However, it's rare for a QB to be over 5.0 in year one, and then over 6.0 in year two. Only Flacco, Palmer, Big Ben and Garcia — who was 30 years old — have managed it.

NB Eli and McNabb don't make the list because their first season was too feeble. Manning improved to 5.6 in year two before regressing in 2006 and 2007, while McNabb did not top 5.0 until year three, though he has been above 5.5 every year since.

25 Re: Week 18 DVOA Ratings

I think the really obvious explanation for the whiff on the ravens O is that there were so many question marks on the ravens going into the season.

We know now that the offensive line is freaking amazing and has come together fantastically, but back in july what did we know? They lost the best center in football and replaced him with an aging matt birk. They had no depth at tackle whatsoever with adam terry a gimp, jared gaither still largely developmental, marshall yanda a gimp, and a rookie who might have a learning disability.

There is also the massive resurgence of the raven passing game. This team has gotten more out of Todd Heap and Derrick Mason and Kelly Washington than anyone expected plus ray rice is looking like marshall faulk 2.0.

The only place where the ravens really suffered from injuries all year was in the defensive secondary, but they're holding on with one good corner and one good safety and not much else. Everyone talks about the crazy number of penalties the ravens have, and it's the street-free agent secondary that is responsible for most of them. It's going to be downright fascinating to see what pierre garcon and company do this weekend.

47 Re: Week 18 DVOA Ratings

Ray Rice was unexpectedly good, and the Ravens still don't have any WRs that scare me. They were lucky to get such a huge lead yesterday. The way they are constituted would make it hard for them to come from behind.

But no, I don't see how that would rate them as low as 29th.

72 Re: Week 18 DVOA Ratings

In reply to by Still Alive (not verified)

When you guys make statements like this--"was expected to be way worse" and Rice "was unexpectedly good"--who are you talking about that found it so unexpected ? Fans, talk radio, writers ?
The Ravens brass knew what they were doing. Granted, the guy really grew rapidly. But his growth was not unexpected. And he, along with their defense which still has two of the great players in the NFL, along with some very good other ones, make them very, very dangerous in these playoffs. Rice busting that long one yesterday was not a fluke, boys. The Pats have been giving up chunks all year. They gave up 400 yds total in 4 of their last 8 regular season games. They were 3-6 vs greater than .450 teams. What was expected was that the Ravens would, indeed, run on them very well...

88 Re: Week 18 DVOA Ratings

In reply to by Rick A. (not verified)

If you are honestly telling me they expected him to be a top 5 back this year you are kidding yourself. That is all we were saying. There were a lot of questions about whether he would even get a sizeable minority of the carries, much less a minority.

I know August was a while ago, but I remember it clearly. No one was saying "Ray Rice will be a top back in the NFL this year". A few were predicting he would make an impact, and many that he would be good eventually, but there was not any more hype for him than there was for Donald Brown, Shonn Green, etc.

91 Re: Week 18 DVOA Ratings

In reply to by Still Alive (not verified)

Well, you have answered the question. You say there wasn't alot of "hype" regarding him. In other words, the media. But the Ravens knew he could be a possible impact player in time.
Teams don't think in terms of "top 5" rankings and so on. Agents and fans do.
People handling the fortunes of the draft think of finding "impact" players.
He flowered very quickly and will continue to grow. That's a scary thought right now for the opposition. No matter how the game has changed from the manipulations of the owners, and no matter how much a school of thought like DVOA seems to want to call some of the basic tenets of football "myths", the fact is that December/January will always be about who can run and who can play defense, at least in the case of the non-dome teams...

20 Re: Week 18 DVOA Ratings

I know there were turnovers, but what did the Cardinals' offense do, 45 points in 51 plays? As a guy who like defense, all I can say is "ooooo, grooooosssssss".

56 Re: Week 18 DVOA Ratings

It was a clear illustration of what the League has wanted. Everything has moved more and more in the direction of Arena Football. You'll see more and more games like this in the playoffs in the future. As much as I marvel at the incredible talents of Kurt Warner and the rapid growth of a coming super talent, Aaron Rodgers, a game like that leaves me personally bored. It's just for the fans of today...

77 Re: Week 18 DVOA Ratings

I've been hearing that a lot since the weekend. And I find the timing ironic, since two of the winning wildcard teams threw only 25 passes between them. When did that last happen? I'd guess about 1976.

All four games followed the same early pattern: one team built a sizeable lead. However, this did not result in four 'arena football' experiences. The Packers-Cards game was the exception, because 1) Warner was otherworldly accurate; 2) the Packers have a very good offense which is, alas, used to playing from behind; 3) both head coaches had outstanding offensive gameplans.

The Jets and Bengals could have played 50 quarters of overtime without reaching 800 passing yards. If you were bored by the Sunday evening game, and not by the Cincy game, you should probably stop watching football altogether; or invest in some 1940s mud-and-cleats classics.

81 Re: Week 18 DVOA Ratings

All 4 games were boring, Cards-Pack was the least boring because it was close. I spent the whole 2nd half wondering if either team would get a defensive stop. Which finally happened in OT.

Of course that happens when the officials refuse to call holding. If you can't get to QB, you can't defend in today's NFL. While that was always true to some degree (and by always I mean since 1978), I'm not sure the "coverage sack" exists anymore.

83 Re: Week 18 DVOA Ratings

It was only close because the Packers had the passing attack to make it close. In fact, four minutes into the third quarter, at 31-10, it was the least close of all of the games.

The Packers managed only 10 points in the first half, during which there were three turnovers, a punt, and a quixotic field goal attempt. It only became a shootout because the Packers had the offense to make it one (and because McCarthy tried the onside kick).

Did you find the legendary Bills-Oilers comeback game boring too? And Superbowl XIII?

95 Re: Week 18 DVOA Ratings

As a point of contrast, some of my favorite all time games were the Titans/Rams, Rams/Patriots, and Patriots/Giants Super Bowls. Games that featured really hard hitting defense along with some really exciting big offensive plays.

To me, the Packers/Cardinals game was no better than what a 6-3 contest between the Jets and Ravens would look like, and perhaps worse. What I really hated about the game on Sunday was the crappy tackling; it's like watching a low scoring game that comes about because wide open receivers dropped 16 passes.

103 Re: Week 18 DVOA Ratings

To be fair, the crappy tackling was all on one side. These early numbers charge the Arizona defense with only 3 missed tackles, while crediting them with 30 QB sacks-hits-pressures. For the Packers' offense to put up the production (and stage the comeback) that it did in those circumstances should not, I think, offend the discerning viewer.

On the other side of the ball, your criticisms are valid. Even so, the Cards' receivers dropped zero of 32 passes (one other was thrown away) and, try as he might, Dom Capers couldn't find a way to stop Warner. Although the Packers' defense was undoubtedly overrated during the regular season, it was effective enough at times (as Cowboys fans will confirm) to suppose that its total disintegration Sunday had something to do with the opponent.

As for your analogy, well, as perhaps the only neutral who sat through all of this year's Bills-Browns crapfest, let me assure you, watching terrible quarterbacking and offensive lineplay is definitely two layers of hell lower than bad tackling. At least it has felt that way since Marquand Manuel left Green Bay.

104 Re: Week 18 DVOA Ratings

Perhaps I'm remembering wrong, or perhaps their definition of a missed tackle is different than I, but I seem to recall more misses by the Cards. Maybe I'm throwing into that category those instances where the Cards defender took such a horrid angle that he precluded actually touching the guy with the ball. I saw a lot of that as well.

It's really important to not conflate points scored, high or low, with the quality of play. A 38-35, even a 41-38 game can be decently played defensively, and a 10-7, or even a 7-3 game can have decent offensive play. The you have, as you put it, the crapfests.

107 Re: Week 18 DVOA Ratings

I definitely remember a few times when Cardinals DBs went for a pass break up, missed badly due to a late reaction, and allowed what would have been a routine completion to go for a big gain.

The James Jones TD on 4th and 5 was one example, he caught a simple curl right at the marker but spun away from the defender and went 25 more yards. Probably not a missed tackle by the strict definition, but not a good looking play for the defense either.

92 Re: Week 18 DVOA Ratings

I don't know where you have been "hearing that alot" because, the point is, the public nowadays, overall, loves that kind of game. By the way, I don't recall saying in that thread anything about the Jet game. Although I will say, even though it was a quite predictable snorer, it was good to see the Jets do their thing after reading all the negativity here regards them, along with the article that was posted last week mocking Ryan and ridiculing the team.

As to Kurt Warner--he's been "otherworldy accurate" his whole career, especially in the playoffs. As Casey Stengel used to say "you can look it up". You weren't surprised were you ? Kurt is still perfect for the NFL, afterall he came from Arena Football. Can read the defense in an instant, and can deliver the most catchable ball in history, on a dime at 50 yds...

And, don't worry, after the SB the Arena League will soon crank it up again and you can get some different jersey for yourself to sit in front of the TV screen with. Perhaps someone will invent a ratings system for that League and go public with it too, creating a website where everyone can all post all their words of wisdom. Then it can all be pretty much year around...

93 Re: Week 18 DVOA Ratings

So an enjoyable game for you is one that proves an FO article wrong? Aren't you taking the "me vs the world" thing a bit too far?

I think the public, overall, enjoys that kind of shootout because it's quite rare. If it happened every week, the game would become like basketball, the first three quarters basically irrelevant.

97 Re: Week 18 DVOA Ratings

Man, you got basketball pegged. But, don't you see, this kind of shootout is going to become less rare. Especially if the owners and TV can help it. Too bad few teams will ever have the kind of talents that Kurt and Aaron Rodgers are gifted with. Still, this is the wave of the future.

As to what kind of games I enjoy--well, look at Will's post above you. That, too, are the kinds of games I enjoy. I was at that Ram/Tenn SB and it was every bit a "classic" game with unbelievable hard hitting, exceeding the regular season standards. Another one I saw live was the Denver/G.B game, whatever number that was, and wow ! what a great all around game it was. The reason I enjoyed those, actually turning me into a fan for 60 minutes, is that you saw the highest level of performance on offense and defense. The absolute best in pro football playing at levels that are hard to conceive. The Az/G.B. game the other day was a pitiful showing of non-defense, horrible tackling and fundamentals, complete breakdowns continually. I assure you if you were there live you would've seen lots more of the atrocious mistakes that TV did not bring you. And this was supposed to be the elite levels of the NFL and the Pack was allegedly right up there as one of the top two or three defenses. Kurt Warner is one of the greatest QB's in history, and he can do that to anyone, but still there is no excuse for G.B.. Rodgers had one of the great playoff performances himself, obviously, with no help from his vaunted D at all. But this is what the public loves and Chas Woodson just got voted the Defensive Player of the year award. I'm sure he'll be very happy with his award while he sits at home watching with all the rest of us...

98 Re: Week 18 DVOA Ratings

"But, don't you see, this kind of shootout is going to become less rare. Especially if the owners and TV can help it."

You keep saying this, but scoring has remained fairly constant through the years. Yes, more rules have been implemented to aid offenses than defenses, but that's usually in response to defenses starting to dominate more.

Mean points-per-game-per-team, since the merger:
1970: 19.3
1971: 19.4
1972: 20.3
1973: 19.5
1974: 18.2
1975: 20.6
1976: 19.2
1977: 17.2
1978: 18.3
1979: 20.1
1980: 20.5
1981: 20.7
1982: 20.2
1983: 21.8
1984: 21.2
1985: 21.5
1986: 20.5
1987: 21.6
1988: 20.3
1989: 20.6
1990: 20.1
1991: 19.0
1992: 18.7
1993: 18.7
1994: 20.3
1995: 21.5
1996: 20.4
1997: 20.7
1998: 21.3
1999: 20.8
2000: 20.7
2001: 20.2
2002: 21.7
2003: 20.8
2004: 21.5
2005: 20.6
2006: 20.7
2007: 21.7
2008: 22.0
2009: 21.5

100 Re: Week 18 DVOA Ratings

Yeah, the field remains the same size, while the players get bigger and faster. If the rules weren't continually tweaked, defenses would eventually drive down scoring significantly. At some point they might need to figure out a way to increase the size of the field, which means taking seats out of some stadiums.

Actually, I think a bigger field, with rules which inhibit defenses less, would produce a better game. Let's see 'dem linebackers run and hit!

109 Re: Week 18 DVOA Ratings

In reply to by Will Allen (not verified)

Larger field = less seats = less revenue? For this group of benevolent owners?

Try the same size field with 9 on 9 or 7 on 7. Same revenue, less employee expense, more passing and scoring. Note: historical FO stats may need some recailbrating.

113 Re: Week 18 DVOA Ratings

Um, "this union" was broken 20+ years ago. NFL contract battles are now fought between "large market" and "small market" owners. The union just took advantage of some battle lines to increase their share of the revenue last time; and they went up to 2/3, despite the fact that they produce 100% of the value.

114 Re: Week 18 DVOA Ratings

I'd say the union is still rather influential; by all accounts, the NFLPA is what's preventing the season from being expanded to 17 or 18 games. And that doesn't take away jobs, like the 7- or 9-man proposals would.

105 Re: Week 18 DVOA Ratings

Once again, the crappy teams keep getting crappier due to con men owners not wanting to put out state of the market money in the free agency era, also the dumb owners that don't know what they're doing. Meantime, most of the top teams do attempt to follow the changes in pro football and field dynamic air oriented teams that can take advantage of the various rules tweaks etc. They score more points, as a group, than did their counterparts in earlier eras. Why don't you take a look at how many one-sided blowouts there are in recent times as compared to the past. And how many teams are getting way above the median in total scoring for the season as opposed to how many in the past. The total points record keeps getting broken or near broken. As for looking at the futile teams and the sorts of things that jump out, how about this--the Rams were shut out twice in the first four games. And Sunday Warner had 5 TD throws, I don't think the Rams had 5 TD throws all year--and one of theirs was from their holder on a fake field goal, which was the only way they were able to win even one game. Only a year ago Detroit couldn't win a game. So what I say the League and TV want is not able to be produced by the shit teams, but if you have seen the continuum of many years of pro football you cannot avoid the fact that offense, by the good teams, is alot easier to achieve--and they do, and the League wants it that way. Look how many guys are throwing 4K yards now. There will be more and more of this, just watch and see...

108 Re: Week 18 DVOA Ratings

Why don't you take a look at how many one-sided blowouts there are in recent times as compared to the past.

Depends what you mean by a blowout. Here are some figures for seasons since 1978:

Years with highest proportion of games decided by more than 14 points (FO's cuttoff point for gut/skate, etc):
1) 1992: 42.6%
2) 1990: 41.7%
3) 1985: 40.8%
4) 1984: 39.9%
5) 2009: 38.7%

Average of 2004-2009 (since rule changes): 36.5%
Average of 1978-2009: 35.9%

By more than 27 points:
1) 2009: 14.5%
2) 1986: 12.9%
3) 1996: 12.7%
3) 2003: 12.7%
5t) 1984/89: 12.4%

Average of 2004-2009: 10.8%
Average of 1978-2009: 10.4%

Splitting the difference, the years 2007-2009 all have above-average numbers of games decided by more than 20 points, but 2004-2006 are all below average.
Average of 2004-2009: 21.7%
Average of 1978-2009: 20.2%

And how many teams are getting way above the median in total scoring for the season as opposed to how many in the past.

Do you mean the mean?

Average annual median score by winning / losing teams, by six-year period, since 1980:
1980-1985: 26.3 / 14.0
1986-1991: 25.2 / 14.0
1992-1997: 25.3 / 14.2
1998-2003: 26.3 / 14.7
2004-2009: 26.8 / 14.4

And how to account for the small increase in total points illustrated by Eddo above?

Average no. of defensive and special-teams TDs over a six-year period, per team:
1980-1985: 16.8
1986-1991: 18.0
1992-1997: 21.3
1998-2003: 22.6
2004-2009: 21.7

This doesn't even begin to account for the increase in field goal range.

There's no doubt that offense is on a small upswing at the moment. But 'common-sense' interpretations of the way the game is changing are not borne out by the stats. In many ways, 2009 looks like an outlier, even compared with 2007 or 2008. It is too early to argue otherwise.

112 Re: Week 18 DVOA Ratings

Wow, that was alot of work. Very nice. But what are we supposed to say now, that the obvious is not really happening because of the presentation of a chart ?

Perhaps we are speaking too generally here. To be more specific, what people like me are speaking of is more air offense. In the 50's you had the Rams with Waterfield, Elroy Hirsch & Co. playing what was a harbinger of things to come and exciting the audience. So along came the Giants with Y.A Tittle, Rote, Gifford etc. for the 60's.(by the way, I think Tittle had 7 TD passes,or some such outlandish number,one day in 1962) There were also other guys and teams beginning to open it up like the Giants--you know, Jurgenson, Frank Ryan, Dandy Don, etc. In the meantime, the AFL was shocking the owners in the NFL with the commercial success of their wild and wooly games based primarily on the passing game. It went over so well that ABC actually flourished televising the Chargers, the Bills, Raiders, Chiefs etc. Tremendous QB's existed there, any of whom could've played in the NFL, Lamonica, Hadl, Dawson and, of course, along came Broadway Joe--who became the first guy, along with his fine Jet receivers, to throw for 4K yds. And then, as we all know, Joe & the J Men capped off the decade by pulling a so-called upset of the vastly over-rated Colts. They did it by playing NFL style but it was symbolic of the wild and wooly AFL. The next year the Chiefs did the same thing but with more of the AFL style. From the moment of the merger the owners were wanting more of what Tittle had done, and what the AFL had done, because it was certainly clear that there were increasingly greater numbers of folks out there that went for this rather than the bruising, running styles of the 30's, 40's, 50's. So in the 70's here came teams with more and more forays into the world of this--and the League began encouraging them right away. Of course, with a game where there is an inherent advantage to defense anyway, there had to be protections put in place for QB's and receivers, if teams were going to do this. So they tweaked on things right away. The Stingley thing happened and the tweaks became lots more invasive on the rules than "tweaking". The way was more clear to the ideal of exciting air attack if the athletes could be found to take advantage, and the coaches to pull it off--so here came Montana and Walsh in the 80's, here came Marino, Elway, Moon etc. They lasted into the 90's. By the 2000's there are more people that can pull off this aerial barrage game than ever. There are more 4000 yd seasons than ever before, and more teams that can aspire to that. This next decade will have more yet of the Brees, Mannings, Warners, Rodgers, Brady's, Schaub's, and on and on. This many of these QB's and teams, playing at the same time, simply did not exist in the past. People love it. The Az/G.B. playoff game will be proclaimed a "classic". I said it looked like Arena Football. It also looked like the AFL games did when I was a boy in the 60's. In any case, it is certainly what the League and their partner, TV, want. Because that's what more people yet now regard as their preference. Now, your numbers chart looks the way it does because there are also more crap teams than ever, who are more futile than ever...

21 Re: Week 18 DVOA Ratings

Is there something wrong with the O rankings or DVOAs?

BALT 16.1% 9
GB 32.7% 1
SD 3.6% 2
IND 23.8% 12

What am I missing?

23 Re: Week 18 DVOA Ratings

wow... amazing how colts offense fell off a cliff in the rankings.

57 Re: Week 18 DVOA Ratings

Don't worry, they're just the same as they ever were and you will see that quite clearly, as will the Ravens, next weekend...

26 Re: Week 18 DVOA Ratings

Sorry. The ranks were leftover from the version before I edited for the Week 17 "sit starters" games. Everything should be fixed now.

29 Re: Week 18 DVOA Ratings

Interesting. I never knew that. The special teams page (http://footballoutsiders.com/stats/teamst) has a heading calling "Non-Adj VOA"), which I guess is due to weather and "hidden" ST adjustments?

Edit: Which is explained clearly on the page I just linked. Nevermind.

67 Re: Week 18 DVOA Ratings

Thanks, boss, and thanks for the end of season missing starter adjustments (EOSMSAs? Not very catchy. Maybe just ESAa--End of Season Adjustments....)

31 Re: Week 18 DVOA Ratings

wow, only one team posted a positive ST DVOA (or VOA) for the first round

Re Packers/Cards, how much does a surprise onside kick count for DVOA? obviously not enough to cancel out a missed field goal

35 Re: Week 18 DVOA Ratings

You guys who post here wine so damn much. Until you get your scouting jobs, shut up, and enjoy the resource that these ratings are, instead of nitpicking every goddamn thing!

64 Re: Week 18 DVOA Ratings

You guys who post here wine [sic] so damn much. Until you get your scouting jobs, shut up, and enjoy the resource that these ratings are, instead of nitpicking every goddamn thing!

I think you meant to say "whinge".

85 Re: Week 18 DVOA Ratings

I think he's calling us boozers. I do enjoy a glass of red wine with dinner a few times a week. Maybe he's on to something.

37 Re: Week 18 DVOA Ratings

Dallas is clearly ranked too high because you're jinxing us. Anti-jinxing is way better than this. Eagirls are teh Sux.

42 Re: Week 18 DVOA Ratings

'Eagirls are teh Sux.'

I really hope you have a t-shirt with that emblazoned across the front. The back can say 'Sting they ass'.

40 Re: Week 18 DVOA Ratings

If your defensive projections did worse than last year's numbers at predicting this year's defensive DVOA, that suggests that the indicators that you use of improvement/decline were actually anti-indicators: indicators in the opposite direction from what you expected. Coaching changes are the obvious candidate for a factor that you thought would predict defensive decline, but which actually predicted improvement. I would look more closely at that, to see if there's any clue of when a coaching change is likely to lead to decline and when it is likely to lead to improvement (e.g. the new D coordinator's previous job & previous success).

43 Re: Week 18 DVOA Ratings

Fat lot of good it is now in practical terms, but it's still a surprisingly pleasant feeling to see a Texans defense ranked 11th in the league, albeit in weighted DVOA, after the years of suckitude Houston fans had endured up until Week 4 this season. Thank you Frank Bush, thank you Brian Cushing, thank you Bernard Pollard. Bring on 2010.

44 Re: Week 18 DVOA Ratings

I think the problem is predicting just one number, lets say the average, instead of thinking of the entire variety of outcomes (the whole distribution).

Sometimes you can get away with that. For example, take the statement "Rookie quarterbacks tend to struggle". Even in the rare cases that rookie QBs post adequate seasons, it is almost impossible for them to be truly outstanding. So, this statement is generally safe, even when you are wrong and a rookie QB is good, he will not be performing at an MVP level. Here the average is indicative of what usually happens.

That's not necessarily the case for other situations. I would think (although I cant say for sure) that results of hiring a new coordinator or head coach tend to be quite varied. So, even if the overall trend is negative (or positive), its not as useful in making specific predictions.

46 Re: Week 18 DVOA Ratings

Why was it a "strange and unpredictable" year for defenses if team-by-team performance didn't change much more than usual?

Because the WHY changed. Normally, some defenses stay the same, some improve, and some decline. There are certain trends which help to predict which defenses will fall into each category. This year, those trends didn't hold true. So overall, there was about the same change in defense from year to year as there usually is, but the specifics of it were much less predictable.

50 Re: Week 18 DVOA Ratings

I was thinking along these lines too. If the year by year correlation is that good, perhaps it's just a matter of dampening the specific trends more with the previous year's data?

96 Re: Week 18 DVOA Ratings

I didn't write that second paragraph. Looks like somebody (Aaron?) responded into my post instead of to my post, making it look like I said something I didn't accidentally...I think erasing a paragraph but I can't recall now! Seems like I had something in there about team to team stuff only being at .37.

We may be on the verge of debating definitions, which won't do anyone any good. I guess I just want it clear that it wasn't a strange or kooky year on the field. It was a fairly normal year on the field. The fact that it didn't reflect trends you believe in doesn't mean it was strange or kooky. If we could all work together to pin down the differences it might seem less kooky (or, skeptics might be more convinced that it was actually kooky after all).

There wasn't a sense from reading any of the FO authors in their audibles, previews, game analysis, interviews with Greg Cosell, chats with whoever, etc... that it was a kooky year in defense. I think it's dangerous to assume for your next set of preseason projections and next year's Premium selections against the spread that "kookiness" caused the misreads in 2009. There have to be other explanations that are just as reasonable. Maybe we could all put our heads together to try and come up with them.

Thanks for the response from whoever typed into my post (lol).

49 Re: Week 18 DVOA Ratings

Perhaps the game-charting project can shed some light on this. To use specific examples, how many snaps did teams play 5+ DBs in 2008 v. 2009? LBs? DLs? Can we specify for down and distance? (The idea being that if the defenses are reacting to the same shift in offensive efficiency versus the pass that FO's authors have mentioned from time to time, perhaps we're in a transition year and/or the pendulum may be swinging back.) Are teams running more on 3rd-and-short, or passing more on 2nd-and-short? (Perhaps the offensive efficiency is obsoleting some factors that were previously predictive.)

Another possibility is to drop each team in turn and re-run the correlation. Put another way, did your predictions for Denver and New Orleans dominate the prediction error?

Finally, was 2009 an outlying year in terms of new coaches or new schemes? I know much was made of the changes in Denver and Green Bay - how much is that playing into the picture?


I trust you won't just dismiss 2009 as kooky. It might just be one of those outliers that life throws up from time to time, but I'd want more evidence before concluding that. The real test of any method (or organization) is failure. How do you respond, and where do you go from here? In Aaron Sorkin's words, "Quo Vadimus?" :)

70 Re: Week 18 DVOA Ratings

Nat showed last week that scoring median was actually down a shade in 2009 from 2008. I went through to just check offensive scoring only (nat's study included all scoring).

I'm showing 19 more offensive TD's this year, but 89 fewer field goals (!). Hope somebody has a chance to doublecheck that. You know late night math can get iffy (lol). Anyway, if those numbers are correct, and you credit 7 for the TD's and 3's for the field goals (didn't want to try and track down PAT or two-point conversion stuff), you get

2009: 10,255 "offensive" points
2008: 10,389 "offensive" points

A decrease of about 1% this year.

Figured I'd throw that into the discussion to pick up from last week...

52 Re: Week 18 DVOA Ratings

no way shpuld Raiders be behind Tb Bu ccs. Somehting wirong with system. also jow come Caorlima 5th ranked but not even close to playoffs.?

61 Re: Week 18 DVOA Ratings

Can someone explain why the VOA ratings for the single games are not equal? Shouldn't it be a zero-sum system?

66 Re: Week 18 DVOA Ratings

Ok, but the numbers are still absurdly uneven.

Cincy 2/1 Jets 0/0
Phi 3/3 Dal 1/1
GB 3/2 Arizona 3/1
Bal 1/1 NE 1/1

It is assumed that the 1st is not random but the 2nd is. Thus the Bal game would be zero sum, I would assume.

68 Re: Week 18 DVOA Ratings

The newest version of DVOA (v6.0) actually does not count things equal for both teams. There are a set of plays that ONLY count to penalize offense, but there's no bonus for the defense (aborted snaps, false starts, delay of game).

73 Re: Week 18 DVOA Ratings

That clears up some of it, I would assume that the opposite is true for defense? Or that penalties are weighted?

Either way it is more clear.

87 Re: Week 18 DVOA Ratings

It seems like a lot of false starts are caused by offensive lineman being overmatched, and trying to get an early break. Is there no correlation to defenses in this?

69 Re: Week 18 DVOA Ratings

I found it interesting that the defensive dvoa average against wr1, however one would classify it, has improved since 2005. I presumed this was because of the increase in multiple wideout sets, where offenses could recoup larger passing value from better match-ups, but the average defense against wr2/3 has hovered around the same dvoa scale, with the only real defensive dvoa decline observed is against TEs.

The defensive passing dvoa against wr1 from '99 to '09:
7.4 -2005
0.5 -2009

I don't have a strong math background, so I'm not sure if it all washes out even after taking into account the weight of the outstanding defenses of the time, but it's really fascinating to mull over.
Sorry for length.

102 Re: Week 18 DVOA Ratings

I think that has to do with some TE's that are really WR's. Do you cover them with a LB (too slow?) or a CB (lose run help?)? Maybe the trick is to bring in a secon SS or a tweener CB/SS.

75 Playoff Odds

Hey, someone let Mike Harris know that it is 2010.

80 Re: Week 18 DVOA Ratings

I'm a little surprised that Baltimore's VOA would be so high for that game. It seems that much of their success stemmed from plays that wouldn't necessarily be predictive. There was the 80-yard Ray Rice run to open the game, followed by three Brady turnovers in the first quarter that gave the Ravens truly excellent field position, and allowed them to jump out to the quick lead. From there is seemed like the Ravens tried to just sit on the lead for an entire three quarters. True New England couldn't stop the run at any point, but Baltimore struck me as a team that looked a lot better than they actually played just because that first quarter was so overwhelming. I don't think the Ravens can count on the same events happening in consecutive weeks.

84 Re: Week 18 DVOA Ratings

At what point does "it was a weird year" become a Mad Lib? Every year is weird. It's written every year. The NFL is a reshuffle league, period, accept it.

86 Re: Week 18 DVOA Ratings

So, like the last couple of Patriots playoff losses, it has been the offense that has shit the bed.

101 Re: Week 18 DVOA Ratings

It's official - the Panthers are better than the Colts!!!!!!!!!!!!!!!!!!!!!

OK - maybe not really, but they did start playing well towards the end of the season.