DVOA Analysis
Football Outsiders' revolutionary metrics that break down every single play of the NFL season

Final 2008 DVOA Ratings

by Aaron Schatz

They've spent the whole year bouncing in and out of the top spot, and we've spent the whole year trying to explain it. However, after yesterday's 44-6 crushing of Dallas, I don't think anybody is going to think it is completely weird to see Philadelphia end up as the top team in DVOA for the entire 2008 regular season.

Philadelphia is the only wild card team to finish the year number one in DVOA -- as currently constituted, anyway. An earlier version of DVOA had Denver number one in 1997 and future tweaks might put the Broncos back on top. Of course, that Denver team was 13-3 and only won a wild card because the Chiefs (who currently rank first for 1997) ended up 14-2. That's not quite the same thing as this year's Eagles. The Eagles are only the second team to lead the league in DVOA without a record of 12-4 or better. The other was the 1995 San Francisco 49ers at 11-5.

Not only do we have a wild card team at number one, we also have a wild card team at number two: Baltimore. However, neither one of these teams is the best team in the league right now according to weighted DVOA. That title goes to the Pittsburgh Steelers, who also moved past Baltimore this week to end the year as our top-rated defense. In fact, with both Baltimore and Pittsburgh having good final weeks, the on-again, off-again list of the "top defensive DVOA ratings ever" is back on again. Here is where the two teams from 2008 finally end up.

2002 TB -33.6%
2000 BAL -30.0%
2000 TEN -28.4%
2003 BAL -28.4%
2004 BUF -28.1%
2008 PIT -26.4%
2006 BAL -25.8%
1998 MIA -25.7%
1995 SF -24.7%
2008 BAL -24.5%

Pittsburgh has moved up to fourth in total DVOA, in a virtual tie with the Titans, who fall after ranking first last week. Yes, the annual problem of teams that rest their starters in the final game rears its annoying, ugly head again. For those who have not read us in past seasons, our research has shown that excluding those final "sit starters" games in DVOA actually doesn't improve its accuracy when it comes to correlation with playoff performance or the next season, so we leave every game in. However, there's no doubt, that's the reason Tennessee has fallen into fifth place, so make sure you use common sense when looking at the ratings. Honestly, the top five teams can be considered roughly equivalent, and when you look at weighted DVOA, Carolina and Indianapolis are getting pretty close to that pack. There are seven teams that have a reasonable shot at winning the title this year. I'm not talking about a longshot "upset" championship, like last year's Giants -- based on that, of course, all 12 teams could win it all. But honestly, if any of the top seven teams won it all, it would be hard to really call it an upset. The playoffs will be fun.

By the way, while the dregs of the AFC North (Cleveland and Cincinnati) ended up with the two hardest schedules based on average opponent DVOA, three of the top four teams ended up with schedules that ranked in the top eight: Pittsburgh (fourth), Baltimore (sixth), and the New York Giants (eighth).

Speaking of the dregs... Alas, the 0-16 Detroit Lions couldn't quite suck hard enough to reach the depths of the 2005 San Francisco 49ers, and end up with the second-lowest total DVOA since 1995. On defense, however, they are on their own terrible, horrible planet. The Lions are the worst defense in DVOA history by more than seven percentage points. They are worst than any previous defense by almost eight percentage points, because in the end the second-worst defense of the DVOA era belongs to... this year's Denver Broncos. St. Louis makes it a trio in the bottom five. Obviously, some of this is due to the league's overall offensive level -- we finished with a league DVOA of 4.4% compared to the baseline based on previous seasons -- but that doesn't even come close to making up the difference between the Lions and the other bad defenses of the past decade.

2005 SF -56.6%   2008 DET 33.3%
2008 DET -47.3%   2008 DEN 26.2%
2004 SF -46.1%   2000 MIN 25.6%
2000 ARI -44.3%   2004 MIN 23.8%
2000 CLE -43.9%   2008 STL 23.3%
2003 ARI -42.8%   2000 ARI 23.1%
2008 STL -42.1%   2004 SF 22.9%
2002 ARI -41.9%   2002 DET 21.6%
1999 CLE -41.7%   1998 CIN 21.4%
2000 CIN -41.7%   1996 ATL 21.3%

Looking forward to next season, two teams ended up in the DVOA top ten despite finishing at 8-8: New Orleans and San Diego. The Panthers may have edged out the Falcons for a division title, but that last place-to-first place thing in the NFC South looks alive and well for 2009. Past teams that finish in the top ten at 8-8 have a darn good record of improvement. The 2006 Giants and Jaguars did pretty well for themselves in 2007. The 1995 Vikings and Broncos both made the playoffs the next season. The 2004 Bengals finally had an actual good season in 2005. The 2002 Chiefs were 13-3 the next year, and the 1999 Raiders improved to 12-4. The main exception: the Dallas Cowboys of 1999, who were fifth in DVOA at 8-8 but collapsed in 2000 when Michael Irvin retired and a lot of their other veterans suddenly felt their age.

New Orleans and San Diego also ended up with two of the top three offenses, sandwiching the Giants at number two. Buffalo led the league in special teams for the third time in five years.

Finally, updating two stats I discussed in last week's commentary:

  • Green Bay ends up with 8.9 Pythagorean wins, which means the difference between their actual win-loss percentage and the Pythagorean projection is -.183. The Packers end up ranking as the ninth most "unlucky" team since the 1970 merger. As you might expect, eight of the other nine teams in the top ten improved the following year, most by three wins or more. On the other hand, Green Bay's DVOA rating has dropped from fifth after Week 11 to 16th at the end of the season. Yikes.
  • Arizona's final win tweaked the opponent adjustments enough that last week's 47-7 Patriots blowout over the Cardinals now stands as the greatest single DVOA game of all-time, at 143.8%.

When all is said and done, after all that up and down and weird teams ranking high and low, the DVOA ratings ended up pretty close to the actual final standings, at least when it comes to the playoff picture. Ten of the top 12 teams ended up in the playoffs, including the entire top eight. The exceptions were Minnesota (14th) and Arizona (20th), and I doubt anyone is shocked to see Arizona that low after the last few weeks.

* * * * *

All 2008 stats pages are now updated with final 2008 regular-season stats, except drive stats, which should be updated Tuesday or Wednesday. Playoff odds are also updated. Our leading Super Bowl contender is the defending champion Giants, with a 24 percent chance of winning it all.

* * * * *

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

OFFENSE and DEFENSE DVOA are adjusted to consider all fumbles, kept or lost, as equal value. These ratings also include opponent adjustments. SPECIAL TEAMS DVOA is adjusted for type of stadium (warm, cold, dome, Denver) and week of season. WEIGHTED DVOA is adjusted so that earlier games in the season become gradually less important. It better reflects how the team is playing right now. LAST YEAR represents 2007 rank, while LAST WEEK represents rank in Week 16 of 2008.

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

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

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

1 PHI 31.7% 11 2 30.5% 2 9-6-1 9.3% 12 -20.7% 3 1.6% 13
2 BAL 30.3% 22 4 29.7% 3 11-5 5.6% 18 -24.5% 2 0.2% 17
3 NYG 29.4% 15 3 23.8% 5 12-4 23.4% 2 -4.0% 8 2.0% 11
4 PIT 27.0% 8 5 30.8% 1 12-4 1.7% 20 -26.4% 1 -1.1% 23
5 TEN 27.0% 10 1 24.4% 4 13-3 8.8% 14 -16.9% 5 1.3% 15
6 CAR 19.8% 27 6 21.9% 7 12-4 18.6% 6 1.3% 10 2.5% 10
7 IND 15.4% 2 9 22.6% 6 12-4 21.3% 4 3.7% 14 -2.2% 26
8 SD 12.3% 6 12 14.7% 10 8-8 21.3% 3 10.7% 22 1.7% 12
9 NO 11.5% 21 7 16.7% 9 8-8 23.7% 1 11.6% 24 -0.6% 22
10 MIA 9.2% 29 17 7.5% 13 11-5 17.1% 7 3.0% 13 -4.9% 29
11 NE 9.2% 1 15 17.5% 8 11-5 14.6% 9 8.9% 20 3.5% 6
12 ATL 8.4% 28 10 9.8% 11 11-5 16.5% 8 11.4% 23 3.3% 7
13 TB 7.7% 7 8 2.4% 17 9-7 0.3% 21 -7.6% 6 -0.3% 19
14 MIN 6.8% 14 13 5.3% 14 10-6 -4.9% 23 -18.2% 4 -6.5% 32
15 DAL 6.3% 3 16 8.8% 12 9-7 8.5% 16 -0.7% 9 -2.9% 27
16 GB 6.2% 5 11 3.7% 15 6-10 9.2% 13 2.7% 12 -0.3% 20
17 CHI 5.2% 20 14 -0.7% 18 9-7 -4.3% 22 -5.5% 7 4.1% 5
18 WAS 4.6% 12 18 -4.4% 20 8-8 8.5% 15 1.9% 11 -2.1% 25
19 NYJ 2.2% 25 19 2.6% 16 9-7 3.2% 19 3.9% 16 2.8% 8
20 ARI -0.5% 23 20 -7.9% 21 9-7 11.9% 10 9.3% 21 -3.1% 28
21 HOU -7.2% 17 23 -2.5% 19 8-8 11.4% 11 19.1% 29 0.5% 16
22 JAC -8.5% 4 21 -10.5% 22 5-11 7.0% 17 15.0% 26 -0.5% 21
23 DEN -11.0% 18 22 -19.2% 26 8-8 21.0% 5 26.2% 31 -5.8% 31
24 BUF -13.6% 19 24 -18.4% 24 7-9 -6.2% 24 13.5% 25 6.0% 1
25 SF -18.8% 32 26 -23.1% 29 7-9 -14.6% 27 8.8% 19 4.6% 3
26 OAK -20.5% 30 28 -20.6% 28 5-11 -17.7% 31 8.6% 18 5.8% 2
27 CIN -21.2% 16 29 -14.4% 23 4-11-1 -15.4% 29 3.7% 15 -2.1% 24
28 CLE -21.6% 13 25 -25.0% 30 4-12 -17.5% 30 8.4% 17 4.3% 4
29 SEA -23.0% 9 27 -18.7% 25 4-12 -9.5% 26 16.3% 27 2.8% 9
30 KC -29.6% 26 30 -19.3% 27 2-14 -6.3% 25 17.5% 28 -5.7% 30
31 STL -42.1% 31 31 -35.0% 31 2-14 -18.6% 32 23.3% 30 -0.2% 18
32 DET -47.3% 24 32 -42.1% 32 0-16 -15.3% 28 33.3% 32 1.4% 14

  • NON-ADJUSTED TOTAL DVOA gives performance without adjustments for schedule strength, fumble recovery luck, and weather/altitude on special teams.
  • ESTIMATED WINS uses a statistic known as "Forest Index" that emphasizes consistency as well as DVOA in the most important specific situations: red zone defense, first quarter offense, and performance in the second half when the score is close.  It then projects a number of wins adjusted to a league-average schedule and a league-average rate of recovering fumbles. Teams that have had their bye week are projected as if they had played one game per week.
  • 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.
  • PYTHAGOREAN WINS represent the number of wins projected from the team's points scored and allowed, as described in this article.
  • VARIANCE measures the statistical variance of the team's weekly DVOA performance. Teams are ranked from most consistent (#1, lowest variance, Washington) to least consistent (#32, highest variance, Oakland).

RANK 2008
1 PHI 31.7% 9-6-1 29.5% 11.4 4 1.5% 13 11.3 5 13.7% 10
2 BAL 30.3% 11-5 29.3% 11.8 1 3.5% 6 11.9 2 16.1% 15
3 NYG 29.4% 12-4 27.5% 11.6 2 2.6% 8 11.3 4 21.7% 27
4 PIT 27.0% 12-4 22.2% 11.1 5 5.9% 4 11.8 3 12.0% 5
5 TEN 27.0% 13-3 27.2% 11.5 3 -2.7% 25 12.1 1 12.9% 7
6 CAR 19.8% 12-4 18.8% 10.2 7 0.4% 17 10.1 9 14.3% 11
7 IND 15.4% 12-4 16.5% 10.4 6 1.8% 11 10.2 8 21.2% 25
8 SD 12.3% 8-8 14.1% 9.9 8 -1.6% 23 10.2 7 20.2% 22
9 NO 11.5% 8-8 12.3% 9.6 9 -1.3% 22 9.5 11 13.0% 8
10 MIA 9.2% 11-5 21.1% 9.2 10 -7.2% 30 8.8 16 16.6% 17
11 NE 9.2% 11-5 12.5% 9.0 13 -6.0% 29 10.6 6 28.7% 30
12 ATL 8.4% 11-5 8.8% 9.0 12 -0.6% 19 9.7 10 21.0% 24
13 TB 7.7% 9-7 8.8% 8.4 16 -0.9% 21 9.0 14 19.0% 21
14 MIN 6.8% 10-6 3.5% 8.5 15 2.0% 9 9.2 12 15.4% 14
15 DAL 6.3% 9-7 2.9% 8.4 17 4.7% 5 7.9 19 28.6% 29
16 GB 6.2% 6-10 6.1% 8.4 18 -0.8% 20 8.9 15 15.0% 13
RANK 2008
17 CHI 5.2% 9-7 3.8% 9.0 11 -0.3% 18 8.7 17 10.4% 4
18 WAS 4.6% 8-8 -0.2% 8.3 19 1.8% 12 7.0 22 5.5% 1
19 NYJ 2.2% 9-7 9.3% 8.5 14 -7.4% 31 9.2 13 16.8% 18
20 ARI -0.5% 9-7 2.6% 7.8 20 -3.9% 27 8.0 18 29.1% 31
21 HOU -7.2% 8-8 -11.0% 6.5 22 2.6% 7 7.3 21 13.3% 9
22 JAC -8.5% 5-11 -9.6% 7.6 21 1.9% 10 6.2 25 7.9% 3
23 DEN -11.0% 8-8 -9.9% 6.5 23 -3.2% 26 6.2 24 21.4% 26
24 BUF -13.6% 7-9 -7.6% 6.3 24 -7.6% 32 7.8 20 12.7% 6
25 SF -18.8% 7-9 -12.1% 6.3 25 -5.5% 28 6.9 23 7.2% 2
26 OAK -20.5% 5-11 -21.7% 5.4 27 1.3% 15 4.6 27 31.1% 32
27 CIN -21.2% 4-11-1 -28.4% 5.0 29 8.9% 2 3.2 30 17.4% 19
28 CLE -21.6% 4-12 -29.5% 5.9 26 9.1% 1 4.4 28 20.2% 23
29 SEA -23.0% 4-12 -21.0% 5.3 28 -1.9% 24 5.4 26 17.7% 20
30 KC -29.6% 2-14 -29.1% 4.3 30 1.4% 14 4.4 29 15.0% 12
31 STL -42.1% 2-14 -40.6% 2.7 31 0.5% 16 2.6 32 24.4% 28
32 DET -47.3% 0-16 -49.2% 2.5 32 6.0% 3 2.8 31 16.4% 16


79 comments, Last at 02 Jan 2009, 3:58pm

1 Re: Final 2008 DVOA Ratings

Now that the Eagles have finally shown that being in the top spot really might not be a fluke against a team with just as much at stake as them while in the spotlight, it would be awfully nice for them to make some noise in the playoffs so that people who have swung around to being reasonably convinced today won't just write it off as another fluke next week.

It's not unreasonable to think that all 4 wild card teams might come into next weekend as the favorite.

6 Re: Final 2008 DVOA Ratings

Take Balt over New England. Sure thing.

Pitt played 7 games against playoff foes this year. All 4 of their losses came against playoff teams. (Phila, Giants, Indy, Tenn) Why do I worry most about the one team in the playoffs they beat 2x (Balt)?

27 Re: Final 2008 DVOA Ratings

I feel the same way- both of those Ravens games could have gone either way. The fact that Pittsburgh won them both doesn't make me feel better- if they play another tight game can they scrape another one out?

As a Steeler fan, there's no team I truly fear or want to face- this team can beat anyone or lose to anyone just depending on how inept the offense is that week. If they don't shoot themselves in the foot repeatedly I like them against anyone in the AFC. If the o-line forgets to block and BRoeth is throwing it up for grabs they'll lose to anyone.

31 Re: Final 2008 DVOA Ratings

Ok those lines are way off. Here's the real lines as of right now:

Falcons -2.5 @ Cards
Colts -1 @ Chargers
Baltimore -3 @ Dolphins
Eagles -3 @ Chargers

*based on the actual payouts, Ravens & Eagles may increase in their spread totals

these playoffs will test whether the iggles are actually legit. they'll beat minnesota but then we'll see what they're made of.

i'm actually wondering if this is the first time since the wildcard was brung in that all the road teams are favored.

2 Re: Final 2008 DVOA Ratings

I think it's quite the testament to (1) the Denver O-Line Awesomeness and (2) Jay Cutler's improvement to see Denver's offensive DVOA at 21% when their defensive DVOA is at 26%. For their offense to be so good even when their defense isn't good at anything other thank keeping Jay Cutler off the field says a lot.

Does DVOA grade an offense higher if their defense consistently puts them in worse situations? Like, if a team has a terrible defensive DVOA, and their offense overcomes it, does their offense get graded higher? Just saying, it's a lot harder for the Denver offense to look good when they've got to travel 80 yards on EVERY drive.

13 Re: Final 2008 DVOA Ratings

DVOA compares each play based on similar plays according to down + distance, field position, and score + time.

There is no direct bonus to the offense based on their starting position. There is also no direct penalty to the offense based on their starting position.

It is possible that a team might be good between the 20s, but not in the red zone. Maybe the play calling in certain situations is ugly. *cough* eagles *cough*. In that case, if they started every drive in the red zone, their DVOA would be horrible, but if they started every drive after a touchback, their DVOA would look pretty good.

At least, that's my understanding of how the formula works. Am I wrong?

Over the course of a full season, we expect these differences to even out, but it's possible that they don't. Maybe that's why DVOA loves the Eagles. Maybe they get in situations they do well in compared to other teams a disproportionate amount of the time.

32 Re: Final 2008 DVOA Ratings

I believe that is correct, but let me try to explain the same thing a little differently.

There's essentially two factors that go into a team's total DVOA (or total offensive or defensive DVOA)--how good they are at each kind of play, and how often they find themselves doing that kind of play.

For example, you could have a fantastic running DVOA and a lousy passing DVOA, but if for some mysterious reason your coach *cough* Andy Reid *cough* decides to call way more passes than runs, then your ofensive DVOA will be lower. Of course maybe the reason why your run DVOA is good is that you never do it, and if you did it more your run DVOA would drop, so there is justice in the system.

For the above question: Does a bad defense give your offense a DVOA bonus? There are really two possible things you could be asking: Does overcoming consistently bad field position give your offense a bonus, or Does consistently playing from behind give your offense a bonus.

In both cases, the answer is "it depends". If you are better than the average team behind your own 20 (maybe because you have a really good deep threat to stretch the defense and an o-line that you are comfortable protecting you from a safety), and you consistently get bad field position, then you would get a DVOA "bonus" for consistently overcoming bad field position, because you get to do what you're good at more often. If you have a really good passing offense, and you're consistently playing from behind, then your offensive DVOA benefits from having a bad defense because you're probably forced to pass a lot because you're often behind.

33 Re: Final 2008 DVOA Ratings

Oops...sorry for the double post, but I just had a thought. Perhaps this is an easy explanation of why DVOA loves West Coast and Ball Control offenses, and not flashy vertical passing attacks supported by grinding running.

If you tend to pick up yards in 8 yard chunks, you will have to run a lot of plays to move down the field, so you will get a lot of chances to run plays where you're trying to pick up yards in 8 yard chunks.

On the other hand, if you are better at deep passes and pick up yards in 30-40 yard chunks or not at all, then you're only going to get to do this twice on the average drive, and so you get to do what you're good at far less frequently...

44 Re: Final 2008 DVOA Ratings

DVOA does not tend to like West Coast offenses. There've been extremely poorly rated West Coast teams, and very highly rated vertical passing attacks. Tampa Bay's had bad seasons in DVOA. Philly's had bad seasons in DVOA. Moreover, from 2000-2003, the Eagles offense was pretty poorly rated in DVOA - lower than most people would put them in several years.

Having more plays doesn't help a West Coast team unless they're succeeding on all of those plays. And if they're succeeding on all of those plays, they're a good team.

Likewise, having fewer plays doesn't hurt a vertical team unless they're not succeeding. In which case they're not a good team.

3 Re: Final 2008 DVOA Ratings

All year I've been talking about how Dallas' DVOA seemed too low for how they played.

Now I feel like they are rated too high.

Can't wait for PFP09. And this time I'll actually believe their estimated wins. Maybe we'll be lucky and get like 7.5 or something. I can see somewhere around 5 though. Team just needs to be blown up.

19 Re: Final 2008 DVOA Ratings

Don't get me started. As a lifelong Cowboy fan who's been criticizing them harshly since the near-loss to Cincinatti (when they lost to St. Louis, I knew they had collectively planned to steal their paychecks for the remainder of the year), the week 17 loss to Philly came as no surprise, though the sheer brutality of the ass-kicking was unexpected.

Nobody -- not Romo, not T.O., not "Coach" Phillips -- seemed remotely upset when they lost to one of the Worst Teams Ever in week 7. Then after the season finale, Romo says something along the lines of, "Hey, if the worst thing that happens to me this year is that I lost a football game, I had a pretty good year." Thanks, Captain Effort.

This is a team with no direction and zero discipline, and as long as JJ retains head coaches that the players have no obligation to, the players will have no reason to put up pro-level effort. If Jones stays true to his word and keeps the coaching staff in place, 2009 will be another failed campaign. Step down, Jerry, and hire a strong GM to make the football decisions. (Jerry Jones reads this, right?)

55 Re: Final 2008 DVOA Ratings

Dallas fans need to chill out a little. This is a team that played really well at times and really badly at other times. They also suffered from injuries throughout, and when they were finally healthy, faced a brutal end of the year schedule... facing the top 4 teams consecutively to end the year. They split all their NFCE games and ended up a half game out of the playoffs.

Sure, the expectations were higher. But that doesn't mean the team is terrible.

8 Re: Final 2008 DVOA Ratings

I am curious why the Giants total DVOA is so much higher than their weighted DVOA, especially considering the Giants had a tough schedule.

14 Re: Final 2008 DVOA Ratings

Probably because some of their better games were early in the season (see week 1) and some of their poorer games were later in the season (see week 17).

15 Re: Final 2008 DVOA Ratings

Both ratings already include weighting for opponent (the D in DVOA is for "Defense-Adjusted"). The weighting in this sense is weighing recent games more heavily. So with the Giants dropping 3 of their last 4 games, their weighted DVOA is lower.

As Tom Petty might say, "The weighting is the hardest part...."

16 Re: Final 2008 DVOA Ratings

The opponent adjustments are the same in DVOA as they are in WDVOA. The only difference is that the Weighted number puts more weight on recent games and less weight on the early part of the season (IIRC, the first couple weeks are dropped entirely).

10 Re: Final 2008 DVOA Ratings

Tennessee is clearly ranked too low because they played their scrubs*. A system in which someone were to rank teams informed by DVOA, but considering factors such as this** is way better than this. Teh FOMBC-inspired madlib is pwned.

* - technically, this is a reason unrelated to DVOA. I win.
** - technically, this would be subjective. I still win.

11 Re: Final 2008 DVOA Ratings

Tennessee is clearly ranked too low because they played their scrubs*. A system in which someone were to rank teams informed by DVOA, but considering factors such as this** is way better than this. Teh FOMBC-inspired madlib is pwned.

* - technically, this is a reason unrelated to DVOA. I win.
** - technically, this would be subjective. I still win.

12 Re: Final 2008 DVOA Ratings

Considering how close they finished to the top, I'm fairly certain a "Cheddar-Plax free" Giants team would be ranked 1st overall.

It's interesting to see a bunched leaderboard in offensive DVOA, while there are massive gaps between teams (just in the top 10) on the defensive DVOA leaderboard. Those gaps are why the Eagles and Ravens are ahead of the Giants in total DVOA.

Eli finished as a top 10 QB in 2008? That's not bad for a guy that previous FO articles declared was similar to Jay Schroeder.

22 Re: Final 2008 DVOA Ratings

Schroeder finished first in yards per attempt and sixth in QB rating for the 12-4 Raiders in 1990. I suspect Eli will distance himself from Schroeder at some point, but he isn't there yet.

34 Re: Final 2008 DVOA Ratings

By the time Schroeder had that season, he was 29 and had no previous seasons with a QB rating as high as any that Eli had in his first three full seasons.

Through age 27, Jay had thrown for about 10000 yards, two TDs more than Ints, had a QB rating around 70, and a sack percentage higher every year than Eli's career sack rate, excepting 1986 (where it was 0.1 lower). Eli's just below 15000, 24 more TDs than Ints, and avoids sacks much better.

Eli has already separated from Schroeder. Perhaps not as much as people would think, but still pretty clearly.

17 Re: Final 2008 DVOA Ratings


I said in the Audibles thread that my impression was that New England and Miami were pretty much exactly equal, and pretty much exactly equally deserving of the postseason.

DVOA seems to agree--they have (to the number of significant digits that Aaron lists) identical DVOA's. Not sure I understand how that is...Miami has a better offensive AND defensive DVOA, and I didn't think special teams could affect things enough so that New England's better ST could even things out...

But anyway, stupid tiebreakers...

28 Re: Final 2008 DVOA Ratings

And, as a Colt fan whose team might get lucky enough to advance in the playoffs and meet up with Baltimore, I can't express how much I was rooting for odd teams yesterday: wanted New England to win and Baltimore to lose so that NE could be in the mix, not Baltimore; and of course I rooted for Denver over SD.

52 Re: Final 2008 DVOA Ratings

That's interesting to think about. Symmetry worked out pretty well in 2006; first Dungy took out his disciple Herm Edwards, then beat bitter Baltimore, exorcised demons against the Patriots, and then beat the city that dominates much of the rest of the state of Indiana. This year, it looks like Indy could get revenge for recent playoff losses to San Diego and Pittsburgh, pay back the Titans for winning the AFC South crown, and then beat little brother in the Super Bowl. Just like in 2006, there would be storylines every step of the way.

18 Re: Final 2008 DVOA Ratings

And the QB DVOA Title goes to...David Carr!

Clearly, the Giants need to bench Eli Manning. He's holding them back!

20 Re: Final 2008 DVOA Ratings

Thanks to falling from a 9.0% DVOA to a 6.2% DVOA with the win vs Detroit (how bad is Detroit that beating them by 10 points means you actually end up statistically a worse team than before you played them), GB is not the best 10 loss team of the DVOA era. To pull the data from my comment last week of the 10+ loss teams that had a positive DVOA (OK last week this was off because while I talked about the 01 Chargers they got left off the chart).

1997 . DAL . 0.6% . 6-10
1999 . DEN . 1.0% . 6-10
2001 . SD . 6.8 . 5-11
2003 . JAC . 1.1% . 5-11
2008 . GB . 6.2% . 6-10

and to just rank all the losers that were "better than average"

Ten Best "Losers" of the DVOA era (OK 14)

Year . Team . DVOA . Record . NYDVOA . WChng

It's 14 teams since there have only been 14 "losers" with a positive DVOA so I cut it off there.

So GB is the 4th best "unlucky" team based on DVOA (if you assume that a DVOA of 0% means you should be 8-8 which is not really a safe assumption I like the Pythagorean measure of unlucky better).

What does this mean for next year? That's hard to say on average the teams on that list were .54 wins better the next year. But there were teams that were 5 wins worse (2003 SF 7-9, 2004 SF 2-14) and teams that were 5 wins better (1999 DEN 6-10, 2000 DEN 11-5). I think I'll add the DVOA next year and change in actual wins to the chart so folks can draw their own conclusions. 6 teams had better records the next year, 6 teams had worse records, 1 team had the same record and we don't know about GB yet.

It does seem that these teams have down turns in performance though. Only 4 of them had a better DVOA the next year, though they had a significantly better DVOA all of them improving by 11 points so if they do get better they get significantly better. But that means that 9 of them had worse DVOA with 5 of them dropping by more than 11 points

I can't draw any real conclusions for that data. Being a "solid team that loses" doesn't seem to say what will happen next year There are some significantly better teams and some significantly worse teams.

And off of that track.

GB vs Detroit
I do find it interesting that putting up 31 points and 484 yards of offense vs Detroit lowered the Packers offensive DVOA by 1.2%, though really I'm sure that drop came more from opponent adjustments that trickled through all their opponents.

GB vs the numbers
Again, I just don't expect a team that had a QB who threw for 4,000+ yards, 28 TD with only 13 ints, had 2 receivers with over 1,000 yards each, a running back with 1,200 yards, a +7 turn over differential (thanks to 2 defenders with 7 ints each and another with 5 ints) to end up 6-10.

Those conventional stats just don't seem to mesh with the record and DVOA doesn't either as as a you don't often see a positive DVOA on a 10+ loss team. They were also better than than three 9-7 teams and three other 8-8 teams based on DVOA this year.

Unfortunately I still have no idea what that will mean for next year. The losers chart I made does seem to say that 10 loss teams that were positive in DVOA did show improvement as all of them had better records the next year (and the 6-9-1 Ravens had the same number of wins) and only the 01 Chargers had a drop off in DVOA, they just ended up 8-8 with a slightly negative DVOA. It's a small sample size, but when you add in the 10+ loss criteria things look a better for the Packers. We'll see in a year. :)

54 Re: Final 2008 DVOA Ratings

Re GB-Det: remember that the Lions' defense is the worst of the DVOA era by a mile and a half, so yeah, what the Packers were able to do is almost expected.

Besides, the win over Detroit wasn't the only factor affecting their DVOA. For one thing, well, they added another game against Detroit. For another, games like Oak-TB and Dal-Phi probably caused opponent adjustments. (Maybe even something like Hou-Chi, with GB having one game against the Texans and two against Chicago, would cause more of a drop than a rise in rough opponent strength.)

57 Re: Final 2008 DVOA Ratings

Yeah, I realize that my comments talking about the real why of this were very brief, but I did mention that it was more likely due to opponent adjustments. I don't have access to the premium database I can't actually look at the single game DVOA for the Packers/Lions game.

I still find it interesting though that stats like that are, as you mention, essentially expected vs Detroit. You could put a bunch of "replacement level players" on the field vs Detroit and still expect them to put up around 320 yards and 24 or so points. I figured that based on the replacement level being -13% DVOA (at least I seem to recall that is was is used here) and GBs offensive DVOA. So the "replacements" are about 20-25% worse than what GB has. Not exactly correct but should still show you a ballpark.

It's astounding how bad the Detroit defense really is. That is really what I was getting at but I wasn't exactly clear on it.

But I do have some hope for the Packers based on how other 10 loss teams with positive DVOA did the next season. Of course they should have a tougher schedule next year because Detroit can't be that bad again can they? :)

26 Re: Final 2008 DVOA Ratings

This has nothing to do with the DVOA figures, but here's a bizarre fact I just noticed:

In the NFC, All 4 NFC 2007 division-winners finished 3rd in 2008. Furthermore, all the 2007 2nd-place teams won the division in 2008.

23 Re: Final 2008 DVOA Ratings

It's official... "Defense-Adjusted Value Over Average LOVES Philadelphia!!!" But let us not be confused by the offensive prowess of the Iggles. That [offense] is a 50/50 proposition depending if McNabb can hit the side of a barn with a pass and whether Westbrook is healthy enough to be effective. The Philly D has been above average for many seasons under Coordinator Jim Johnson and it finally culminated in a dominant performance this past Sunday.

30 Re: Final 2008 DVOA Ratings

I think McNabb is a lot more accurate than most people think. He rarely throws interceptions to people not named Ronde Barber; if he were as bad as the Philly beat writers say he is, he'd be getting picked off like crazy.

I think the problem is that on close plays, he (wisely) aims for spots where it's a difficult play for a receiver, but a near-impossible play for a defender. This is fine when your receiver is Terrell Owens (who can get enough separation that he's not aiming away from the DB), and not so great when it's Reggie Brown, Todd Pinkston, Freddie Mitchell, Kevin Curtis, or LJ Smith.

75 Re: Final 2008 DVOA Ratings

It's unfortunate that most interceptions are blamed on the quarterback. There is much more to passing than just throwing towards the right player. It is good that McNabb has not thrown a bunch of picks based on the number of pass attempts Reid calls. But his [McNabb's] lack of interceptions does not account for his bad accuracy.

Not including fade routes, Iggles YAC is minimal due to the receiver having to worry more about catching the pass. A pass often thrown low, high and/or behind him. I'm amazed how Philly receivers actually make plays on throws so difficult. Let alone worrying if he'll [receiver] will lose a finger or gain a football print bruise in his abdomen from the laser throw #5 delivers.

How often have you seen McNabb throw a pass to a receiver in stride? Especially across the middle.

What is McNabb's completion percentage of swing or dump off passes where there is no defender within 5 yards? I've seen too many sail over the receiver's head or hit the ground 2 yards from his feet.

Is McNabb a better passer with a receiver that threatens a defense? Obviously, yes.

I'm critical of McNabb cause I believe he could have been much better. Donovan and Andy Reid are most responsible for the best football I have seen in 29 years of being a fanatic. I've settled with his horrendously slow starts, lackadaisical attitude during games and unwillingness to run or throw the ball away when a play breaks down. Granted since the ACL injury, Wilma's boy has learned to preserve the drive on the next play or offensive series.

DonnieMc still stirs the Iggles' cocktail. I hope he can "serve" the right drinks to a Super Bowl victory party.

35 Re: Final 2008 DVOA Ratings

The St. Paul paper is reporting that Pat Williams has been cleared to play Sunday. If this is true, and he is the normal Pat Williams, even for 25 snaps, the Vikings chances of upsetting the Eagles just improved enormously. Having both Williams playing like their normal selves will mean that Reid will likely go into his very pass-heavy scheme, and that Leslie Frazier will be able to play nickle defense a very large percentage of the time. At the Metrodome, this means the Vikings defense will likely be very formidable. If the Vikings offense doesn't turn the ball over more than once, and the Vikings special teams don't give up a touchdown (admittedly non-trivial ifs), I suspect the Vikings will win.

37 Re: Final 2008 DVOA Ratings

"Having both Williams playing like their normal selves will mean that Reid will likely go into his very pass-heavy scheme"

Having no Williams' playing would mean Reid would likely go into his very pass-heavy scheme.

The sun rising in the east and setting in the west would mean Reid would likely go into his very pass-heavy scheme.

43 MIN vs. PHI

Will - You are really drinking the Kool-Aid for a guy that thought the Purple & Gold would post 6 wins this year. I can't imagine that Pat Williams is anymore than 75% healthy. I hate to say it, but Philly will likely roll in that game - they are just too good with too few weaknesses (redzone Off and short-yardarge). The biggest matchup is on the other side of the ball. Philly'd defense will be too much for T.Jack/Gus.

36 DVOA across seasons

...we finished with a league DVOA of 4.4% compared to the baseline based on previous seasons...

How can you compare DVOA across seasons? DVOA is defense adjusted value over average, meaning it's normalized for the competition within a given season. Given that it's normalized, how does DVOA end up being anything other than 0 when collapsed across the whole league for a year? If it's value over average, then the bad should cancel the good when you take the whole year's worth of data and/or measurements into account.

The best I can do to make sense of the quoted statement is to compare it to an odds ratio. Assuming there's nothing goofy going on with non-zero league/season DVOA, saying this year's league DVOA is 4.4% relative to other years is just saying that this year's offense is, uh, more better than this years defense than the offense has been in previous years, right?

40 Re: DVOA across seasons

DVOA is not normed to the current year or even the previous year, but to a multi-year baseline period-I believe it's 2002-2006, but could be mistaken. I'd also point out DVOA now stands for Dependent Value Over Average. So, yes, offenses on the whole this year have been 4.4% more productive, and defenses 4.4% less productive, than they were in the baseline period.

39 Re: Final 2008 DVOA Ratings

I must not have been paying attention to the worst defense ever watch because I just noticed Minnesota previously had the 2 worst defenses ever! That seemed really odd to me until I looked at the rosters. I guess I had tried to suppress all those Wasswa Serwanga memories. I also noticed those 2 horrible Viking defenses were part of the most recent playoff teams. Ahhh, memories.