Writers of Pro Football Prospectus 2008

Most Recent FO Features

McGloinMat13.jpg

» Week 4 Quick Reads

Our look at which quarterbacks were "streakiest" in Week 4 includes more discussion of Raiders quarterbacks than we ever anticipated.

08 Nov 2011

Week 9 DVOA Ratings

by Aaron Schatz

A big, dominating win over Buffalo propels the New York Jets past Green Bay and into the number one spot in this week's Football Outsiders DVOA ratings. Obviously, this will be a bit controversial. It's hard to find anyone out there who doesn't rank undefeated Green Bay as the top team in the league right now, and even harder to find anyone willing to consider the Jets as being in that class.

Normally, the various advanced ratings systems on the Internets are mostly in agreement about which teams are overrated and underrated. Not this year. DVOA is going out on a limb with its Jets love. The Jets are tenth in the league in point differential. Pro-football-reference's Simple Ratings System has the Jets ranked 11th. Jeff Sagarin has the Jets ranked eighth. Advanced NFL Stats doesn't have its power ratings up as I write this but last week they had the Jets a mediocre 14th.

The Jets' No. 1 ranking is powered by their performance on defense and special teams, and I don't think I'll get any argument from people that the Jets are very good in those two areas of the game. The surprise is their offense, which ranks 17th in DVOA this week. That's much higher than their simple rank in yards per play, which is 24th (5.0 yards per play). In fact, the Jets have allowed more yards per play on defense (5.1, eighth) than on offense!

The standard explanations for why FO has teams ranked higher than conventional wisdom don't seem to apply here. The Jets haven't recovered a suspiciously low percentage of fumbles (45 percent). Their losses have not been particularly close. They haven't played a particularly hard schedule (they rank 15th) although the difference between DVOA and VOA does seem to show that their rating has been raised a bit based on the specific strengths of their opponents. (The opponent adjustments aren't the same on every play -- they depend on the down as well as whether the offense is passing or running.)

It's silly to compare the Jets to the Packers, who have the exact opposite strengths. Every stat is going to say that the Packers offense blows the Jets offense out of the water. Instead, let's compare them to a similar team, the Baltimore Ravens. The Ravens are 22nd in offensive DVOA, five spots behind the Jets, even though they have averaged 0.1 more yards per play with one fewer turnover and played a slightly tougher schedule of defenses. Here are a few of the differences that cause the Jets to be ranked higher:

  • The Jets are much better on third downs. Baltimore ranks 15/24/15 on the three downs, while the Jets rank 28/13/7. One of the main reasons is that the Jets have avoided turnovers on third down, with only two. That means that even when they're going three-and-out, at least they're able to get a punt off to reverse field position instead of just giving the ball to the other team.
  • The Jets haven't had many super-long plays. As some readers know, the DVOA formula begins to discount additional yards gained once a play reaches 40 yards, because at a certain point a play's length becomes more about where the line of scrimmage was as opposed to a 60-yard touchdown pass really showing that much more offensive quality than a 40-yard touchdown pass. The Jets have had only two plays of 40+ yards, tied for the lowest total in the league. That depresses their yardage totals and averages. By comparison, Baltimore has seven such plays. Green Bay and Detroit lead the league with ten apiece.
  • However, the Jets simply don't lose yardage very often. They've lost yardage on just 6.9 percent of offensive plays, which ranks sixth-best in the NFL. Baltimore loses yards on 8.0 percent of plays (14th). New England is first (5.8 percent) while Chicago is last (12.1 percent).
  • In addition, the Jets have a lot of plays that we might consider "partial successes" -- plays that get yardage, but not much yardage. 22.2 percent of Jets plays have gotten 3-5 yards, the highest rate in the league. So even when the Jets aren't moving the ball in big chunks, they're at least getting a little bit of yardage. 41 percent of their 3-5 yard gains qualify as "successful" plays under our guidelines, which ranks 13th in the league. To compare this with Baltimore, the Ravens get 3-5 yards on 17.0 percent of plays (17th in the NFL) and their success rate on these plays is 35 percent (22nd).
  • The Jets offense has been surprisingly consistent. It doesn't look like it if you look at our offensive table, because they rank 29th in VARIANCE. But that's simply because of one game. The Jets were awful in Week 4, the game against Baltimore, with -69.3% DVOA. But the Jets offense has been positive in every other game, somewhere between 4.1% and 30.8%.

Hopefully, that does a good job of explaining why the Jets' offensive DVOA is so much higher than you might expect otherwise. Unfortunately, any other explanations for the Jets' super-high ranking seem to be hidden in all the layers of equations and baselines that I created originally to figure out the best teams over an extended 2002-2007 span. (Those are the years the baselines are currently based on.) The comparison of each play to the baseline seems to make sense play-by-play; it is only when you add them all together that it seems a bit high.

Believe it or not, this is only the second time in the 20-year history of DVOA that the New York Jets are number one. I went back through our spreadsheet of "DVOA as it developed each week" and the only other week the Jets would have been number one was Week 4 of 1993, when they were 2-1 with 40.2% DVOA. Here's an interesting table of every team that has ever ranked at No. 1 in DVOA during the regular season, going all the way back to 1992:

Weeks at No. 1 in DVOA, 1992-2011 Regular Season
Team Weeks at No. 1 Team Weeks at No. 1
PHI 41 NYG 11
NE 34 BAL 9
DAL 26 BUF 9
SF 25 NO 9
GB 22 TEN 9
PIT 22 SEA 7
DEN 19 CHI 4
TB 17 MIA 4
KC 16 JAC 2
STL 15 NYJ 2
IND 13 OAK 2
SD 13 CIN 1

Some more interesting tidbits about this chart:

  • San Francisco was No. 1 24 times between 1992 and 1997, and then only one time after that (Week 1 of 2003).
  • Although New England won the Super Bowl in 2001 and 2003, the Patriots never ranked No. 1 in DVOA during the regular season during those years. They were No. 1 twice early in 1997 and then not again until 2004.
  • Dallas has not been No. 1 since Week 1 of 1997.
  • The single week that Cincinnati was on top of the DVOA rankings was Week 3 of 2005.
  • The eight teams that have never been No. 1 in DVOA, along with their all-time high rankings: Arizona (6), Atlanta (2), Carolina (3), Cleveland (4), Detroit (2), Houston (3), Minnesota (2), and Washington (2).

* * * * *

This is the first week that the strength of a certain week drops below 95 percent in the WEIGHTED DVOA formula, and thus the first week you'll start to see some differences between regular DVOA and WEIGHTED DVOA. The main difference is that WEIGHTED DVOA has Pittsburgh a bit higher and Baltimore a bit lower, appropriate since Pittsburgh has been playing much better since Baltimore spanked them in Week 1. (Baltimore is still slightly higher than Pittsburgh, though, even in WEIGHTED DVOA.)

* * * * *

One last note: This week's Quick Reads goes into depth on the incredible season Aaron Rodgers is having, but we shouldn't ignore the really great season that LeSean McCoy is having as well. McCoy currently has 250 rushing DYAR, nearly 100 ahead of second-place Adrian Peterson. He's on pace for 500 rushing DYAR, which would be the third-highest rushing total in DYAR history behind Priest Holmes in 2002 and Terrell Davis in 1998. Surprisingly, considering what a great receiving back he is, McCoy is not on pace to come close to Marshall Faulk's 2000 record of 840 combined rushing/receiving DYAR, as he has "only" 59 receiving DYAR. 

 

* * * * *

These are the Football Outsiders team efficiency ratings through nine weeks of 2011, measured by our proprietary Defense-adjusted Value Over Average (DVOA) system that breaks down every single play and compares a team's performance to the league average based on situation in order to determine value over average. (Explained further here.)

OFFENSE and DEFENSE DVOA are adjusted to consider all fumbles, kept or lost, as equal value. SPECIAL TEAMS DVOA is adjusted for type of stadium (warm, cold, dome, Denver) and week of season. WEIGHTED DVOA represents an attempt to figure out how a team is playing right now, as opposed to over the season as a whole, by making recent games more important than earlier games.

Opponent adjustments are currently at 90 percent strength and will steadily grow stronger until Week 10. 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>

TEAM TOTAL
DVOA
LAST
WEEK
WEIGHTED
DVOA
RANK W-L OFFENSE
DVOA
OFF.
RANK
DEFENSE
DVOA
DEF.
RANK
S.T.
DVOA
S.T.
RANK
1 NYJ 34.1% 3 35.1% 1 5-3 6.3% 17 -19.8% 2 8.1% 3
2 GB 28.6% 1 28.2% 2 8-0 39.3% 1 10.7% 22 0.0% 16
3 HOU 25.6% 5 26.2% 3 6-3 20.3% 5 -3.7% 7 1.7% 9
4 SF 22.9% 4 22.5% 4 7-1 3.3% 21 -8.9% 4 10.7% 1
5 BAL 21.2% 6 18.8% 6 6-2 -0.4% 22 -26.4% 1 -4.7% 30
6 NYG 19.2% 7 21.6% 5 6-2 16.9% 8 -2.4% 9 -0.1% 17
7 ATL 18.3% 10 18.6% 7 5-3 7.6% 13 -8.2% 5 2.6% 6
8 DET 16.6% 9 16.7% 8 6-2 8.7% 12 -11.3% 3 -3.4% 28
9 NE 16.4% 8 15.6% 10 5-3 30.5% 2 13.8% 26 -0.3% 18
10 BUF 14.1% 2 12.6% 12 5-3 22.0% 4 7.2% 16 -0.7% 20
11 PIT 13.8% 12 16.1% 9 6-3 16.1% 9 2.6% 13 0.3% 13
12 NO 13.1% 11 12.9% 11 6-3 25.7% 3 11.5% 23 -1.0% 22
13 CHI 11.3% 15 11.1% 13 5-3 -0.7% 23 -3.1% 8 8.8% 2
14 PHI 9.3% 13 9.6% 14 3-5 18.3% 7 8.5% 18 -0.5% 19
15 CIN 8.3% 14 8.8% 15 6-2 5.6% 19 -0.2% 10 2.5% 7
16 DAL 8.0% 16 7.8% 16 4-4 9.5% 11 0.4% 12 -1.1% 23
TEAM TOTAL
DVOA
LAST
WEEK
WEIGHTED
DVOA
RANK W-L OFFENSE
DVOA
OFF.
RANK
DEFENSE
DVOA
DEF.
RANK
S.T.
DVOA
S.T.
RANK
17 TEN 2.9% 17 2.3% 17 4-4 12.0% 10 9.6% 21 0.5% 12
18 MIN -2.0% 20 -2.2% 18 2-6 7.0% 15 9.0% 20 0.1% 15
19 TB -4.9% 19 -5.5% 19 4-4 5.2% 20 14.0% 27 3.9% 5
20 SD -7.1% 21 -7.5% 20 4-4 6.5% 16 12.2% 24 -1.5% 25
21 OAK -8.9% 18 -9.9% 23 4-4 7.2% 14 14.4% 28 -1.7% 26
22 CAR -9.1% 23 -7.6% 21 2-6 20.1% 6 21.0% 31 -8.2% 32
23 MIA -9.2% 27 -8.0% 22 1-7 5.6% 18 15.7% 29 0.8% 11
24 WAS -13.7% 25 -15.8% 25 3-5 -13.0% 28 -0.1% 11 -0.8% 21
25 DEN -13.8% 28 -13.2% 24 3-5 -7.8% 24 8.2% 17 2.2% 8
26 CLE -19.1% 24 -19.0% 26 3-5 -15.2% 30 4.2% 14 0.2% 14
27 SEA -19.7% 26 -19.8% 28 2-6 -10.8% 26 5.0% 15 -3.9% 29
28 KC -21.4% 22 -19.5% 27 4-4 -14.2% 29 8.7% 19 1.5% 10
29 JAC -22.2% 29 -22.7% 29 2-6 -28.3% 32 -7.3% 6 -1.2% 24
30 ARI -23.7% 30 -24.0% 30 2-6 -11.3% 27 19.3% 30 6.9% 4
31 STL -35.8% 31 -35.4% 31 1-7 -19.6% 31 13.4% 25 -2.8% 27
32 IND -41.4% 32 -41.3% 32 0-9 -9.6% 25 24.0% 32 -7.8% 31
  • NON-ADJUSTED TOTAL DVOA does not include the adjustments for opponent strength or the adjustments for weather and altitude in special teams, and only penalizes offenses for lost fumbles rather than all fumbles.
  • ESTIMATED WINS uses a statistic known as "Forest Index" that emphasizes consistency as well as DVOA in the most important specific situations: red zone defense, first quarter offense, and performance in the second half when the score is close. It then projects a number of wins adjusted to a league-average schedule and a league-average rate of recovering fumbles. Teams that have had their bye week are projected as if they had played one game per week.
  • PAST SCHEDULE lists average DVOA of opponents played this season, ranked from hardest schedule (#1, most positive) to easiest schedule (#32, most negative). It is not adjusted for which games are home or road.
  • FUTURE SCHEDULE lists average DVOA of opponents still left to play this season, ranked from hardest schedule (#1, most positive) to easiest schedule (#32, most negative). It is not adjusted for which games are home or road.
  • VARIANCE measures the statistical variance of the team's weekly DVOA performance. Teams are ranked from most consistent (#1, lowest variance) to least consistent (#32, highest variance).



TEAM TOTAL
DVOA
W-L NON-ADJ
TOT VOA
ESTIM.
WINS
RANK PAST
SCHED
RANK FUTURE
SCHED
RANK VAR. RANK
1 NYJ 34.1% 5-3 29.2% 6.1 3 1.5% 15 0.1% 20 15.0% 22
2 GB 28.6% 8-0 30.6% 7.8 1 -3.1% 24 3.6% 9 1.9% 2
3 HOU 25.6% 6-3 34.1% 6.0 6 -5.8% 30 -6.9% 28 18.7% 28
4 SF 22.9% 7-1 24.5% 6.1 4 -0.2% 19 -10.6% 32 7.5% 8
5 BAL 21.2% 6-2 22.4% 6.4 2 4.0% 7 -8.4% 30 25.9% 30
6 NYG 19.2% 6-2 24.6% 5.9 8 -6.6% 32 13.8% 1 14.1% 21
7 ATL 18.3% 5-3 12.3% 5.9 7 -1.2% 21 2.1% 16 4.3% 3
8 DET 16.6% 6-2 26.7% 5.7 9 0.0% 18 6.8% 5 6.5% 5
9 NE 16.4% 5-3 10.7% 6.1 5 3.7% 9 -5.2% 26 8.9% 9
10 BUF 14.1% 5-3 14.2% 5.5 13 3.0% 11 2.8% 12 16.3% 25
11 PIT 13.8% 6-3 15.9% 5.5 12 -2.5% 23 -8.0% 29 15.4% 24
12 NO 13.1% 6-3 16.8% 5.7 10 -5.9% 31 6.6% 6 12.3% 19
13 CHI 11.3% 5-3 7.2% 5.2 15 8.7% 4 -3.4% 25 10.4% 16
14 PHI 9.3% 3-5 5.2% 5.0 16 4.5% 5 1.4% 18 10.4% 15
15 CIN 8.3% 6-2 22.0% 5.6 11 -4.4% 26 2.1% 14 1.4% 1
16 DAL 8.0% 4-4 3.0% 5.2 14 1.7% 14 1.3% 19 16.7% 26
TEAM TOTAL
DVOA
W-L NON-ADJ
TOT VOA
ESTIM.
WINS
RANK PAST
SCHED
RANK FUTURE
SCHED
RANK VAR. RANK
17 TEN 2.9% 4-4 7.2% 4.6 17 -4.5% 27 -0.8% 23 23.1% 29
18 MIN -2.0% 2-6 2.4% 4.4 18 1.5% 16 6.5% 7 16.9% 27
19 TB -4.9% 4-4 -7.3% 4.3 19 11.7% 2 5.4% 8 15.2% 23
20 SD -7.1% 4-4 -0.7% 3.6 25 1.4% 17 2.3% 13 5.4% 4
21 OAK -8.9% 4-4 -4.1% 4.1 21 2.8% 12 2.1% 15 27.9% 31
22 CAR -9.1% 2-6 -6.1% 4.0 22 -1.1% 20 3.2% 10 9.0% 10
23 MIA -9.2% 1-7 -8.5% 4.2 20 4.2% 6 9.2% 2 10.4% 14
24 WAS -13.7% 3-5 -17.3% 3.7 24 -2.2% 22 7.0% 4 14.0% 20
25 DEN -13.8% 3-5 -11.8% 3.8 23 3.9% 8 3.0% 11 9.6% 13
26 CLE -19.1% 3-5 -10.5% 3.2 27 -5.3% 28 -0.4% 22 6.5% 6
27 SEA -19.7% 2-6 -21.3% 3.4 26 3.6% 10 -5.5% 27 11.2% 17
28 KC -21.4% 4-4 -20.4% 3.0 28 -5.4% 29 8.5% 3 30.3% 32
29 JAC -22.2% 2-6 -27.6% 2.9 29 13.8% 1 -8.4% 31 9.1% 12
30 ARI -23.7% 2-6 -16.5% 2.9 30 -3.3% 25 -0.4% 21 6.8% 7
31 STL -35.8% 1-7 -39.8% 2.2 31 10.7% 3 -1.8% 24 9.1% 11
32 IND -41.4% 0-9 -46.3% 1.7 32 2.6% 13 1.8% 17 11.9% 18

Posted by: Aaron Schatz on 08 Nov 2011

168 comments, Last at 12 Nov 2011, 9:12am by Mr Shush

Comments

1
by QQ (not verified) :: Tue, 11/08/2011 - 3:55pm

Still amazed that FOAA 11 predicted Mia to have a better Offense than GB this year. Granted nobody would have guessed Rodgers would be THIS good so far but even at simply last year's level it wouldn't even be close.

14
by Drunken5yearold :: Tue, 11/08/2011 - 4:40pm

Didn't FOAA11 predict that Henne was going to have a "Brees in '04" season where the QB improves in conjunction with an increase in offensive talent? I remember the numbers before the '04 season predicting how good the Chargers offense would be that year, and everyone denounced it as an anomaly, and it reached the point where Aaron was fiddling with the projection. I didn't think Miami was going to be as good as predicted, but I thought the logic made sense and I wouldn't have been surprised if the Miami offense had been quite good this year.

There's also no way that FOA could have predicted that Rodgers would turn into Robo-QB and have one of the best years by a QB ever. Rodgers has been so amazing that it doesn't even matter that the rest of the Packers are pretty sucky. Average running game, average special teams, below average defense ... and an amazing passing attack. The Packers literally have only two strengths:

1) Rodgers
2) WRs that seem to catch any ball thrown near them.

Their OL isn't even any good. The Chargers DL sucks and they were putting pressure on Rodgers all game.

I wouldn't be surprised if the Packers end up losing a playoff game because Rodgers has an off-night. Their performance is tied to him in the same way that Indy depended on Manning.

15
by bigtencrazy (not verified) :: Tue, 11/08/2011 - 4:47pm

The Packers offensive line is fairly solid despite early injuries first to Belaga and now Clifton. I wouldn't let one game classify that group as 'not any good'.

Though it is concerning that Sitton has taken a step back this year from last year's excellence.

In fact, two guys key to last year's success, Shields and Sitton, are playing well below 2010 levels. And for no apparent reason.

17
by Independent George :: Tue, 11/08/2011 - 4:51pm

Clifton is 35, so I think the injuries are a valid reflection of him. It's not to say he's not good when he's healthy, but that at that age, you really shouldn't expect him to be healthy.

19
by bigtencrazy (not verified) :: Tue, 11/08/2011 - 4:58pm

I don't understand your point.

I was merely stating that despite losing the starting tackles at different times the line has held up fairly well. Marshall Newhouse has certainly distinguished himself.

86
by Aaron Brooks Go... :: Wed, 11/09/2011 - 9:20am

Your point was "fairly solid despite" which the implication of "pretty good otherwise"

The other poster's point was that you should expect Clifton to be injured, which means GB is really only "fairly solid" as a default as opposed to "pretty good", which you were suggesting.

18
by ammek :: Tue, 11/08/2011 - 4:54pm

I wouldn't say the line was any better than average. It is quite good at adjusting — Newhouse was getting manhandled in the first half against San Diego, but had a much cleaner second half.

After Rodgers and the WRs (who did have a plague of the drops in the two games prior to the bye week), the only other above-average units at the moment are kicker and punter.

22
by Part One (not verified) :: Tue, 11/08/2011 - 5:03pm

"2) WRs that seem to catch any ball thrown near them."

This is exceedingly disingenuous.

Over his last two games, Rodgers has more incompletions due to drops and one spike than he has due to your standard ball-falls-to-the-ground.

He has 3 interceptions this year. Two were drops by his receivers, Jennings and Jones, that went right into the defenders' hands.

How many huge drops did Rodgers' receivers have last year in the playoffs? Nelson had 3 alone in the Super Bowl. Jones dropped two TD passes of over 60 yards in the playoffs: the Philly game and the Super Bowl.

The "thrown near them" is just an awful qualifier. Rodgers isn't merely throwing the ball "near them". He's hitting receivers in stride or putting it in the only spot they can catch the ball (See Jones' TD Sunday for an example of the former, and Finley's comeback route and crossing route for examples of the latter), or, as his strike to Jennings to seal the Super Bowl last year showed, he's hitting the receiver in stride which is also the only place the receiver can catch the ball.

His receivers' YAC numbers are just as good after the catch as Brady's 2007 receivers were AND his average pass travels further than Brady's.

His WRs aren't catching "any ball thrown near them". Rodgers is just playing at a level completely unprecedented.

51
by RickD :: Tue, 11/08/2011 - 6:55pm

It may be inaccurate, but it hardly seems "disingenuous."

31
by Aloysius Mephis... :: Tue, 11/08/2011 - 5:39pm

I would amend "catch any ball thrown near them" to "get open whenever they feel like it."

25
by Part Two (not verified) :: Tue, 11/08/2011 - 5:27pm

"There's also no way that FOA could have predicted that Rodgers would turn into Robo-QB and have one of the best years by a QB ever"

This is also completely ridiculous.

This is how someone could not have predicted Rodgers doing this:

1) Treat all completions as equal.
2) Disregard WR drops.
3) Treat all interceptions as equal.

After missing the Patriots' game last year due to a concussion, Rodgers went:

134 for 197 (68%), 1727 yards (288 ypg), 8.8 YPA, 14 TD, 3 INT (One dropped by Driver that bounced off his foot into a Bears player's hands, another tipped pass against the Bears, and finally a throw straight to Urlacher like his only legitimate interception this year).

Two of those games were home, in Green Bay, in freezing temperatures. One of those games was away, in Chicago, in freezing temperatures, on the worst field imaginable. One of those games was away, in Philadelphia, in freezing temperatures. Two were in domes.

As we know, not all completions are equal. Some plays the receiver makes a wonderful catch, some plays the quarterback throws behind a receiver, makes him jump, etc etc...

...Tom Brady is accurate. He rarely misses horribly, but he also doesn't often hit his receiver right in stride, even on the crossing/under/flat hungry route trees he often works through. He gets the receiver the ball still in a position to get upfield, but they aren't "perfect" throws. Their ability to still get upfield is more of a function of 'a nickelback on Welker' or a 'linebacker on a running back/TE' than it is a perfectly in stride pass. Brady doesn't take YAC off the table, at all, but he doesn't put more YAC on the table, given the scheme the Patriots run.

Aaron Rodgers puts more YAC on the table. Watch Rodgers' six games last year, as I did. As Trent Dilfer did, who is saying Rodgers is playing better than anyone else ever right now. Rodgers isn't making his receivers have to make any adjustments to the ball. No half jumps, no twisting and turning, no having to slow down to catch a ball on their hip. His precision has been better than who is hailed as the most precise passer ever, Tom Brady.

Since that Patriots game concussion, Rodgers has 6 interceptions. 2 were bad throws to Brian Urlacher. 4 weren't remotely his fault. That's a 38-2 TD-INT ratio since that concussion. Playing for a coach who takes his foot off the gas pedal when the team gets up.

The average Rodgers pass this year travels 8.3 yards in the air. That leads the league. Eli is second at 6.7. Eli is as close to first (1.6 yards) as he is to 18th (Joe Flacco, 5.1 yards).

Rodgers is as close to his peer in vertical passing as second place is to Joe Freaking Flacco. And he's doing it more precisely than Tom Brady. And this wasn't out of nowhere, if you only paid attention to HOW Rodgers was putting up his incredible numbers at the end of last season.

26
by QQ (not verified) :: Tue, 11/08/2011 - 5:28pm

Even without Robo QB 2011 Rodgers, it was pretty clear that GB's Offense would be Spectacular:

1-Pro Bowl caliber QB
2-Best WR Corps in the NFL + New WR Cobb
3-Return of Finley (Stud TE) and Grant (Decent RB)
4-More O Line Depth/Improvement

Even if you just compare to the SB where GB's Offense was great, this year they were able to add in 2-4 more playmakers (Finley, Cobb, Driver, Grant) plus a 1st Round OT for depth and the probable improvement by Bulaga.

-GB's receivers don't have Great Hands but they are awesome at YAC when they do catch it

29
by Part Three (not verified) :: Tue, 11/08/2011 - 5:37pm

Haha, whoops.

Thanks for mentioning the return of Finley. Rodgers did all of that work at the end of last season WITHOUT who the Packers' pass offense was designed for.

87
by Aaron Brooks Go... :: Wed, 11/09/2011 - 9:36am

Well, let's not start sucking Finley's dick quite yet. GB's offense got a lot better once he got hurt last year, and isn't working much better than the 2nd half of last year now that he's back. I'd say his returns are fairly diminishing.

41
by Sound_Automatic (not verified) :: Tue, 11/08/2011 - 6:25pm

GBs o line must not be that terrible considering they consistently leave only five guys in protection and still are able to keep Rodgers upright. Not a GB fan, btw.

52
by Arkaein :: Tue, 11/08/2011 - 6:56pm

GB's line is better than most people give them credit for.

A lot of people look at the sacks and short scrambles due to pressure and think that Rodgers must be under heavy pressure, but that hasn't really been the case, except for maybe in the Minnesota game.

If there is one criticism of Rodgers' game is that he will hold the ball longer than a QB like P. Manning. However, the flip side of the coin is that his patience allows a lot of WRs to get open deep, and he never makes high risk throws this year.

Against SD he was sacked 4 times and had a handful of scrambles for a couple of yards when no one was open. However, there was only one sack which was due to fast pressure and lost more than 2-3 yards. In all other cases Rodgers held the ball for about 3 seconds before trying to run. Early in the game he had a lot of success running the ball. Later SD adjusted and did a better job of contain, and Rodgers scrambles turned into short gains or short sacks.

Even considering these facts GB is only slightly below average in Adjusted Sack Rate. Not great, but far from terrible. I'm looking forward to an Under Pressure article that looks at offensive lines. I think it will show that GB's O-line does a decent job.

47
by Kal :: Tue, 11/08/2011 - 6:47pm

I think it's actually pretty predictable. Rodgers played well last year despite a massive amount of injuries to his offensive players. No injuries and improving the next year? Why wouldn't that signal Rodgers having a great year?

85
by Tom W (not verified) :: Wed, 11/09/2011 - 9:13am

I agree with most of what was said in the previous posts, except the one by the guy who isn't impressed by GB's receivers, but I'd like to talk about the Packers' defense. Ranked #22 in defensive DVOA, after being near the top last yr., despite the rash of injuries. Last yr., their secondary play was arguably the best in the league. This yr., it's been well below average. Yes, it would help if they had a pass rush, and injuries have been a factor. Losing Nick Collins was a huge blow, and Morgan Burnett and Tramon Williams have played hurt. But, that doesn't explain the number of blown assignments on a near weekly basis. It's been suggested that Williams's bad shoulder has forced Dom Capers to play more zone than usual. OK, that's maybe an excuse for a week. But, these are well-compensated professionals. Giving up big plays continually because somebody doesn't know what coverage he's supposed to be in, is simply unacceptable. GB has compensated somewhat for the mind-boggling number of passing yds. they've allowed by leading the league in picks, 2 of which they returned for touchdowns last week, but in each of their last two games their defense has turned blowouts into nail biters in the time it takes to grab a beverage and a snack. If this trend isn't reversed, I see a repeat of '09, when Warner put up 50 on them in the playoffs and they lost despite a brilliant performance by Rodgers.

115
by DisplacedPackerFan :: Wed, 11/09/2011 - 1:34pm

The defensive players have started to talk about stuff too I've seen a few interviews. Raji was quoted as saying that there were 14 missed assignments vs the Chargers. He also echo'd Woodson who openly questioned how Matthews was being used. Raji then went on to basically say that the line is often in put in a run oriented defensive set, not a passing one, he didn't give numbers but reading between the lines seemed to me that he was saying he's had maybe 1/2 to 2/3 as many chances this year compared to last year to be in a alignment that allows him to freely rush.

There was discussion about how the defense reacts to motion and such and that most of the mistakes are because a shift is called but not everyone gets it so you have 10 guys playing the proper, shifted, call and 1 guy playing the base call. I agree it's still completely unacceptable. It's even more frustrating because the talent clearly displays itself at times this year and has in the past as well.

Shields drop in play was also discussed. Some of it is that he is being asked to do more than he was last year. This is in part because they aren't asking Woodson to do as much and in part because they think he is able to, but he isn't learning it as quickly as hoped. This is still only his 3rd year ever playing corner, I understand that, but it's still exploitable as shown.

What's worse is that the errors have been in all aspects. The d-line, the linebackers, and the secondary. It's not just one unit that keeps having the breakdowns and it's not just one player. The defensive coaching hasn't really changed which is what I generally think of when issues like this happen.

This team can still win it all, but it's not like last year where you knew either unit could keep the team in the game and the losses generally came because the special teams screwed up big time.

The issues this year feel like they might be more fixable. The better running game and addition of Finley and getting a successor to Driver has made it easier for the offense to prevent itself from disappearing like it did in stretches, even during the 2nd half and playoffs last year. So the issues are covered better than last years offense would have been able to. The special teams have also improved to help cover for some of the other mistakes as well.

All that said I still have worries that there could be a 2009 repeat, if things don't change. I realize that every team is giving Green Bay it's best shot that happens to every defending champion, and that is a very small part of this too. But the defense needs to clean this stuff up.

138
by Tom W (not verified) :: Wed, 11/09/2011 - 11:18pm

With Cullen Jenkins' putative replacement, Mike Neal, still out and Frank Zombo back only recently, I've also wondered why the Packers seemingly haven't done more to put Raji in position to take advantage of his pass-rushing ability. During the pre-season, when it became clear that Neal was going to be out until at least mid-season, I even thought they might move Raji outside in their base defense and return Pickett to nose, since he's a total non-factor against the pass. It will be interesting to see how Neal plays when/if he ever gets healthy.

103
by bravehoptoad :: Wed, 11/09/2011 - 12:32pm

Great year? Sure. Best year of all time? Come on.

2
by RickD :: Tue, 11/08/2011 - 3:55pm

We're still scratching our heads at why DVOA vastly preferred the Jets over the Patriots during their first meeting in Foxborough, a game that was seemingly dominated by the Pats.

If DVOA is going to love the Jets for that game, it's not a surprise that they love them overall. But we still suspect that something is amiss with that game.

3
by Staubach12 :: Tue, 11/08/2011 - 3:57pm

Are you royalty?

4
by RickD :: Tue, 11/08/2011 - 3:59pm

"We" refers to several of us who made this comment last week.

5
by Independent George :: Tue, 11/08/2011 - 4:15pm

I thought we established that DVOA just likes teams whose green-and-white color scheme matches the FO website.

13
by Aaron Brooks Go... :: Tue, 11/08/2011 - 4:30pm

FEI doesn't have that problem. It hates the Spartans.

43
by Yinka Double Dare :: Tue, 11/08/2011 - 6:34pm

So do road games, apparently.

67
by NWebster :: Tue, 11/08/2011 - 9:51pm

I just thought it was a chubby chaser for head coaches.

32
by Staubach12 :: Tue, 11/08/2011 - 5:43pm

Okay. Now it makes sense.

10
by nat :: Tue, 11/08/2011 - 4:24pm

No, he's just not alone. By a longshot. Aaron himself confessed to not understanding the Jets/Pats game's DVOA, and still can't explain it.

I am normally a big fan. But Aaron has really dropped the ball on this one.

Aaron, if you're going out on that Jets limb, please go back and re-check the Pats/Jets game, and tell us what you find out about the Jets and about DVOA.

Either DVOA saw something real - in which case you come out looking like a genius by telling us exactly what that real thing was. Or the Jets inadvertently hit a corner-case where DVOA doesn't work so well - in which case you come out looking like someone with strong personal integrity.

Just keeping quiet and hoping we forget doesn't really work.

20
by Aloysius Mephis... :: Tue, 11/08/2011 - 4:58pm

As I recall, for that Pats game the Jets had a weirdly super-high DVOA on offense. Looking at the PBP the Jets did have 11- and 13-play drives for touchdowns, and they didn't turn the ball over. They had seven three-and-outs, which is bad, but I notice that some of those featured more successful plays than unsuccessful ones. For example, they'd have a five yard pass completion, then a three yard run, then an incomplete pass. That's two good plays to one bad one, even if it results in a three and out. That could explain...

You know what? Now that I'm writing this it doesn't quite make it. Two long touchdown drives and then a bunch of flailing around against a lousy defense shouldn't equate to a *great* offensive performance, which is what DVOA says the Jets turned in. Something's still screwy.

50
by Kal :: Tue, 11/08/2011 - 6:50pm

If you think of it on a per-play basis it makes perfect sense.

Let's say 7 3 and outs with each 1st and 2nd down being successful and 3rds all sucking. That gives you, right there, 14 successful plays and 7 non-successful plays.

Then we'll assume all the plays from the 11 and 13-play drives are successful. Now you have 38 successful plays, 7 non-successful plays.

Doesn't that sound AMAZING? I don't guarantee that's what happened, but I bet it's close.

106
by The Powers That Be :: Wed, 11/09/2011 - 12:47pm

Not really. The Jets had seven 3-and-outs. Two of them followed your script (two successful plays followed by failure to convert). Four of them consisted of three consecutive unsuccessful plays.

Overall, the Jets ran 52 plays (a spike brings the official count to 53). Of those, 24 were successful (46%). Is that a particularly good number? I don't know, but it doesn't seem crazy-good. I should note that some of those unsuccessful plays were close (4 yards on 1st-and-10, that sort of thing). By way of comparison, the Patriots were around 54% success rate in that game (and their offensive DVOA was quite a bit better than the Jets').

127
by Arkaein :: Wed, 11/09/2011 - 3:19pm

I just took a look at every Jets play on offense, and my guess is that the high offensive DVOA for the game is a result of consistency that DVOA likes for projecting future success.

We know that DVOA is part descriptive and part predictive. The descriptive part works best with explosive plays and third down and red zone success, since these lead to sustained drives and TDs. However, I think that the predictive part values consistent production the most, since it tends to be more repeatable and reliable than a mix of explosive and dud plays, even though in any one particular game it may not lead to great success.

Therefore, I'm thinking that the DVOA in question is more a result of a prediction of future success than a description of success in that game itself. It still seems a little off, but if the Jets improve on offense over the rest of the season then we can say that the DVOA got the predictive part right at the expense of the descriptive part.

16
by ammek :: Tue, 11/08/2011 - 4:49pm

This has now been brought up every week on this thread since the Pats-Jets game. Please.

21
by nat :: Tue, 11/08/2011 - 5:00pm

And it would go away the second Aaron posted a clear explanation of either how the Jets were truly dominating that game, or how that game was a corner-case that hit a weak spot in DVOA.

Until then, we have no reason to trust this year's DVOA. Really. I am about the biggest Aaron-fan here, and have been for years. But this stubbornness is uncalled for. It just can't be that hard to look at the high YAR plays in that game and explain what happened.

83
by OmrothLol (not verified) :: Wed, 11/09/2011 - 7:08am

Oh come on - give him a bloody break. So ONE GAME has a weird result? FO over-reacting to that would be far far worse than "being stubborn" and ignoring that abberation.

89
by Aaron Brooks Go... :: Wed, 11/09/2011 - 9:41am

Not when it's also spuriously arguing that the Jets are the best team in the league, and that one of their large losses should have been a blowout win.

96
by nat :: Wed, 11/09/2011 - 10:21am

You're probably new here. DVOA game results have been questioned in the past many times. Every time, or so it seems, the FO fans could find a good clear explanation of why DVOA would say what it did. Not everyone agreed, but pretty much everyone could understand the process. For examples of that, see the many arguments about whether DVOA undervalues the 70-yard TD or overvalues turnovers.

This Jets game seemed different. Yes, the Jets avoided turnovers. But their results were otherwise pretty poor. It was stunning to hear that their offense was one of the best for that week at almost +30% DVOA.

The fans tried, but could not come up with a reasonable theory as to why DVOA would rate the Jets offense so highly. Aaron, the world's expert on the subject, cannot explain it. It's possible there is a bug in the program. It's more likely that there is a weak spot in DVOA that the Jets game happened to hit. It's even possible that the Jets offense actually played really well, and that our impressions of that game are an optical illusion.

We'd like to know. Because we would learn something about DVOA, about the Jets offense, or about football. And because absent an explanation, we can only doubt the recent changes in the DVOA system.

112
by OmrothLol (not verified) :: Wed, 11/09/2011 - 1:15pm

I'm not new here - I remember in 05 when Aaron had to put the number 6 Colts at the top of the ESPN DVOA page to stop America burning down his house.

You're trying to sound reasonable, but everyone here is basically saying "I've lost all respect for DVOA until this is resolved/explained!" which is completely ridiculous for ONE GAME.

114
by nat :: Wed, 11/09/2011 - 1:31pm

Since you've been here a while, you realize that it is very, very, very unusual for a game's DVOA to be this strange without there being some good clear reason. DVOA is almost always quite defensible, or at least understandable if you disagree with it. Aaron's "it's a mystery" response may be unprecedented in DVOA history. It's at least very rare, and worth some investigation.

BTW, using All Caps for emphasis makes you look demented. The tags for emphasis (em) and strong work quite well.

125
by alaano (not verified) :: Wed, 11/09/2011 - 2:58pm

Have to say I agree with the comments about ONE GAME. For crying out loud, as my grandma would say.

Let me ask you this? Having seen the first game and the two teams in subsequent weeks, who you expect to win the rematch? And how did the playoff go last year? Personally (as a Pats fan in NY, so I see every Jets game and due to schedule have seen a lot of Pats), I expect the Jets to win. So actually, what DVOA says about their first game, as a predictor, is pretty much the same as our eyes, i.e. if they play 100 times the Jets will win most.

131
by nat :: Wed, 11/09/2011 - 4:07pm

Your questions are off-topic. DVOA didn't, couldn't, and shouldn't rework its answers to fit the future. Heck, even you and I can predict last week's scores 100% of the time!

The on-topic questions have to do with what things happened in that one game, and how DVOA measured them. To put the question differently, if the Jets offense were to play as badly/well this week as they did in that game, would you consider it a great day for the Jets offense? A day that would bode well for the offense's future production?

This isn't about trash-talking. This is about what went wrong or right when DVOA claimed the Jets offense was dominating on a day when they went three-and-out seven times in ten tries. (Not counting trash time)

If DVOA were known to be flaky, and included Aaron's opinions injected every week, this wouldn't be a discussion worth having. But it's objective and rather advanced in its analysis. So when it comes up with something truly weird, it's natural to ask 'how did it come up with that?' and 'is it broken, or is it detecting something that we can't?'

'It's a mystery' is just a hair short of 'it's broken' as an answer.

I truly hope the answer is either 'Here's what the Jets did really well and why it didn't help that time' or 'Here's the unusual thing the Jets did, and why it fooled DVOA into thinking they were playing well.'

But if the answer is 'it's broken - and here's when to expect it to be right, and the situations when you should doubt it' that's still vastly better than 'It's a mystery' or 'It's a secret'.

141
by Andrew Potter :: Thu, 11/10/2011 - 2:58am

I'm trying to crunch the numbers on this. I've just reached the end of the second half, and so far it looks like the answer is "second down". Here's the breakdown:

Kicking/Punting

NE kicked off 3 times in the first half for a net of 108 yards, or an average Jets starting position of the NYJ 29. They also punted 2 times for a net of 84 yards, averaging 42 net yards per punt.

NYJ kicked off 1 time in the first half for a net of 51 yards, not counting an extra 7 yards on a NE penalty. They punted 4 times for a net of 156 yards, averaging 39 net yards per punt. They also gained 10 yards on one NE penalty, but lost 15 yards on one penalty of their own. Most importantly, however, NE also fumbled one of those punts - though NE did recover the fumble.

ST advantage: Jets. Better kicking net, good kick returns, slightly lower punting net more than offset by the NE fumble.

Jets Offense vs. Patriots Defense

1st down: 10 plays for 38 total yards (avg 3.8). 6/10 plays successful by DVOA formula.
2nd down: 10 plays for 62 total yards (avg 6.2). 8/10 plays successful. 4 first downs, 1 touchdown.
3rd down: 5 plays for 16 total yards (avg 3.2). 1/5 plays successful. 1 first down, 1 sack for 0 yards.

Overall: 25 plays for 116 total yards (avg 4.6). 15/25 plays successful. 5 first downs, 1 touchdown, 1 sack.

Patriots Offense vs. Jets Defense

1st down: 15 plays for 68 total yards. (avg 4.5). 6/15 successful. 2 first downs, 2 sacks for 14 yards.
2nd down: 13 plays for 111 total yards (avg 8.5). 7/13 successful. 4 first downs, 1 touchdown, 1 sack, 1 INT, 1 fumble (0 lost).
3rd down: 6 plays for 13 yards (avg 2.2). 2 successful, 2 first downs. (Plus one pass interference penalty for 38 yards, not counted in the numbers.)

Overall: 34 plays for 192 total yards (avg 5.6). 15/34 successful. 8 first downs, 1 touchdown, 3 sacks, 1 INT, 1 fumble (0 lost).

Conclusion

Based on only the first half numbers, it's easy to see what DVOA sees.
- The Jets ST was comfortably superior.
- The Jets had fewer yards per play, but a much higher success rate. (60% to 45%)
- The Jets had fewer first downs, but not by much on a per-play basis. (20% to 23.5%)
- Mark Sanchez was sacked once. Tom Brady was sacked three times. I didn't count the run/pass numbers to see how often per dropback, but based on the numerators I think it's safe to say Brady was also sacked more often per pass play.
- Each team had one touchdown.
- The Patriots had one turnover AND two fumbles which did not result in a turnover. (One on offense, one on special teams.) The Jets had 0 turnovers and 0 fumbles.

From the box score it looks like the Jets were awful. They were three-and-out four times in five drives, and averaging a full yard fewer per play than the Patriots. However, they were only poor by DVOA on third down. The problem is even their successes on first and second down left them facing third downs, which they then failed to convert.

I'll work through the second half numbers later and see how that holds up. It's 7am, so time for more caffeine.

142
by nat :: Thu, 11/10/2011 - 8:52am

Special teams aren't the issue. You can focus on offensive plays if it saves time.

No surprise that the numbers look close or even in the Jets favor for the first half. The Pats only had a three point lead at the half. Neither offense looked good in that half.

These numbers support the idea of "less bad than the Patriots" but not of "great day of offense".

I'd be careful about using just the "success" cutoff - even though I've done it myself. VOA would be better, if only Aaron would publish those breakdowns.

Good work. Can't wait to see the second half numbers.

146
by Andrew Potter :: Thu, 11/10/2011 - 11:40am

Special Teams were only about 10 of around 75 total plays, including penalties. Excluding them doesn't really save time.

I don't think those numbers look close or even. I think they're comfortably in the Jets' favour, except yards per play and first downs.

Also, the difficulty of using VOA or similar is that I would need league average figures against the Jets defense, and against the Patriots defense. I'm afraid I'm not about to go work that out, so unless somebody can provide it for comparison we'll just have to work with what we have from that one game. I think, however, that 60% success rate is very high, whereas 45% success rate is high but not outstanding. Most teams who achieve a 60% success rate are going to score a lot of points, because most of those successes won't do what the Jets did ie. get the successful yardage and barely a single yard more.

I'll work on the second half figures this afternoon.

147
by nat :: Thu, 11/10/2011 - 12:32pm

You can't fairly include sacks (an issue of lost yardage) without including yardage in general.

Success isn't an all or nothing thing - although first downs are. On any down but third, a success by an inch isn't much more valuable (in VOA terms) than a failure by an inch. That's why I shy away from treating success as a 1 or 0 kind of thing.

What you're pointing to is a Jets team that got a lot of "successes" that were barely successful and didn't lead to anything useful. Counting successes won't show that, but VOA should.

But I wasn't asking you to do the VOA stuff. Aaron should do that. My guess: he already has but doesn't know how to explain the result. With the data he has, getting VOA by down is close to trivial.

148
by Andrew Potter :: Thu, 11/10/2011 - 1:17pm

You can't fairly include sacks (an issue of lost yardage) without including yardage in general.

Sure I can. Sacks aren't just an issue of yardage. They typically cost more yards than any other play, they cost downs, and they increase the chances of failing to gain a first down. They're very negative plays.

Look at it this way:
- If the offense gains ten yards on first down, they now have three more attempts to gain another ten yards and their likelihood of scoring increases.
- The the offense loses ten yards on first down, they now have two more attempts to gain another twenty yards and their likelihood of scoring decreases markedly.

There isn't any other play (not even penalties, which generally don't cost downs) that will consistently lose the offense 5-15 yards and a down, and has such high potential for a turnover. The only thing worse for an offense than a sack is turning the ball over.

On any down but third, a success by an inch isn't much more valuable (in VOA terms) than a failure by an inch.

We're not talking about inches here though. We're talking about the difference between consistently having 2nd and 5, 3rd and 2, vs. the difference between bipolarly having either a new set of downs or 2nd and 10, 3rd and 10. Steady gains rather than boom-or-bust.

And we haven't even mentioned that New England turned the ball over in the red zone once, and fumbled two other times. The Jets played a clean game in that respect. I expect that makes up a chunk of the difference.

Which leads us to the other factor: red zone offense. The Jets were in the red zone three times, and scored a touchdown every time. The Patriots were in the red zone five times: two touchdowns, one interception, and two field goals. That's surely a huge single game VOA difference.

149
by Andrew Potter :: Thu, 11/10/2011 - 1:22pm

Replying to myself, sorry. Here are the second half stats:

Kicking/Punting
NE: 2/2 FG, 2/2 XP
NE to NYJ: 3 KO, 130 return, 85 net (+7 yards on 1 penalty)
NE to NYJ: 2 punts, 101 yards, 20 return, 81 net
NYJ: 2/2 XP
NYJ to NE: 3 KO, 1 TB, 53 return, 147 net
NYJ to NE: 3 punts, 146 yards, 27 return, 119 net

Patriots Offense/Jets Defense
1st down: 17 plays, 179 yards, 13 successful, 6 first downs
2nd down: 11 plays, 25 yards, 6 successful, 2 first downs, 1 touchdown, 1 sack (+10 yards on 1 penalty)
3rd down: 8 plays, 35 yards, 4 successful, 3 first downs, 1 touchdown

Jets Offense/Patriots Defense
1st down: 12 plays, 55 yards, 6 successful, 1 first down, 1 touchdown
2nd down: 10 plays, 62 yards, 4 successful, 4 first downs
3rd down: 6 plays, 46 yards, 3 successful, 2 first downs, 1 touchdown, 1 sack

At this point, it looks less like what I posted before is the case and more like the interception, two fumbles, and red zone performance are the keys to the split. Even without that though, the Jets performed a lot better than their drive outcomes suggest.

Also, the reason I'm using success rate is it's the closest I can get to any components of the DVOA formula. Since it's DVOA's assessment of the game we're trying to figure, rather than the scoreboard's, it seems reasonable to look at DVOA components.

151
by nat :: Thu, 11/10/2011 - 1:56pm

Funny that you didn't finish the analysis:

Second half
Patriots: 23/36 success rate = 64%
Jets: 13/28 success rate = 46%

FYI: only one fumble. The other was overturned.

Overall:
Pats: 54% success rate
Jets: 53% success rate

Now add in all the extra yardage, both positive (getting more than just enough for "success") and negative (sacks) for a total of 190 extra yards, and that 1 int and 1 fumble - worth about 75 yards on average. Take away about 40 yards to value Welker's catch at just 33 yards, because DVOA hates long plays.

The Patriots end up with a better success rate, and an extra 85 or so yards of extra value.

Now factor in strength of defense.

The Patriots should have ended up with a much better offensive DVOA, using your methods and adding in the usual values for turnovers, and adding extra yards at a discount.

Thanks for doing all that data work.

152
by Andrew Potter :: Thu, 11/10/2011 - 2:17pm

Funny that you didn't finish the analysis:

I don't appreciate the implication here, thanks. I'd intended to tidy it up, but I've been staring at numbers for the past two hours and I had dinner to cook for my kids, so I quickly got the data up before doing that. I admitted farther down the post that the second half didn't match my first half conclusions.

FYI: only one fumble. The other was overturned.

FYI: Two fumbles.

1Q 10:16 - T.J. Conley punts for 36 yards to NE46. Wes Welker - no return. FUMBLE, recovered by NE (Antwaun Molden). Penalty: Holding on New England (Jermaine Cunningham) -10 yards.

1Q 2:12 - Tom Brady pass to Aaron Hernandez for 11 yards to the NYJ43. Tackled by Jim Leonhard. FUMBLE, recovered by NE (Deion Branch). Penalty: Holding on New England (Nate Solder) -10 yards.

I don't know how DVOA addresses the second, but it did happen and is in the play-by-play. If the Jets recover that, the holding penalty is declined and the Jets take over so it's every bit as bad as any other fumble.

The overturned fumble was 3Q 13:27.

DVOA hates long plays.

An oft-cited myth. DVOA doesn't hate long plays. One successful long play is less predictive than ten successful short ones, sure, but the long play is still worth more than any of the individual short ones.

I think you're underestimating the negative value of sacks, underestimating the value of ball security, and underestimating the value of red zone performance. Of course, you could equally say that you think DVOA is overestimating those so we're not going to get any conclusions from that.

Like you, I hope one of the Outsiders takes this on because it's a very interesting test case. I've certainly learned a thing or two from digging up the data. (The Jets special teams are fantastic!)

155
by nat :: Thu, 11/10/2011 - 5:45pm

Sorry. Didn't mean to imply anything. It was just a funny thing to omit, seeing as how prominently it figured in your earlier post.

1 fumble. But you are right. There was a "muff" on a punt return. I searched for "fumble" on the nfl site's play by play. Still, the issue is the Jets offense as compared to the Patriots offense. So special teams muff is beside the main point. But you are right. A fumble and a muff.

"DVOA hates long plays" - I was using the oft-cited short hand. I know that DVOA values Welker's catch a lot. But I'm assuming there is some discount on the values after about 40 yards, because that's what FO has always said.

50 yards of field position as roughly equivalent to an interception is taken from several core FO articles. It's approximate, but not far off. Fumbles are worth about half that, depending on where they happen, because DVOA factors out "fumble luck".

Red zone performance has been shown to be non-predictive - by Aaron. Right now it still gets a 25% bonus, because, well because of no reason that makes much sense any more. But the Jets had just 7 red zone plays. So pretend those were 9 plays instead to factor in the red zone bonus. It doesn't really change much.

Negative value of sacks - here you may have something. DVOA certainly values a yard behind the line of scrimmage more than one leading up to the arbitrary "success" line more than one leading up to the first down marker, more than one past the first down marker, etc. Yards are of diminishing value the more you get on a play. This leads to interesting questions, like "How many sack yards does Welker's 73 yard catch offset?"

So here we go:
The Pats and Jets offenses had about the same "success rate" ignoring the actual yardage.
The Pats offense had an interception and a fumble, possibly worth about 75 yards of fields position, perhaps a little bit more.
The Pats offense had 190 more yards of production, none in garbage time, which needs to be discounted because some of it was sack yardage and other parts of it were "excess" yards.

Do we really think there are 190 yards of discount owed on that yardage? Really? There were just 19 excess sack yards - the number of sacks is already counted in the success rate so no need to double count them here. There was only one play longer than 30 yards, leaving 43 excess yards. Even with the extreme theory of sack yards counting x2 and long yardage counting x0, that's still a discount of just 62 yards.

But supposing there were 190 yards of discount to be had. That brings the offenses even - but only before adjusting for defenses.

That's the rub. It's not that these discounts and turnovers are to be ignored. It's just that there isn't any where near enough room in them to turn the Pats' statistical advantages - higher success and more yards - into a stomping by the Jets offense.

==>
This is where Aaron and company come in. They can tell us how VOA valued those sack yards, or how steeply the value of yardage dropped on Welker's play. They can tell us whether that interception was worth more than a 73 yard completion and by how much.

And I concur. The Jets' special teams are amazing.

I learned one other thing: The Jets DSR was higher than I would have guessed from watching the game. By concentrating all their success into two long and one short drive, they looked worse than they were, while scoring more points than they normally would have with those DSR numbers. It's a paradoxical effect.

158
by tuluse :: Thu, 11/10/2011 - 5:58pm

Red zone performance isn't weighted for no reason. It's weighted because it correlates well with winning the game it happens in. It doesn't correlate well with future performance. Playing well in the red zone is part of winning that game, but not necessarily a repeatable skill.

144
by nat :: Thu, 11/10/2011 - 9:45am

A similar, but easier and perhaps more useful analysis from drive stats:

Jets whole game DSR: .667 - would rank 18th in league
Jets whole game TO/drive: .000 - would rank first

Patriots whole game DSR: .758 - would rank 4th in the league
Patriots whole game TO/Drive: .091 - would rank 7th in the league
(Note: if we count the one fumble as .5 turnover ala VOA, this goes to .136, good for 16th in the league.)

The Jets are credited for a great game despite being below average in sustaining drives against a horrible defense. To their credit, 0 turnovers is better than 1 or 1.5 turnover.

For what it's worth, 18th in offensive DVOA is 5.6%, so you would expect the Jets offensive DVOA to be 5.6% + a bonus for avoiding 1.5 turnover - a penalty for playing a horrible defense. DVOA gave them a 29.7%.

Similarly, you would expect the Patriots offensive DVOA to be 22.0% + a bonus for playing a great defense.

It's as if the two teams were credited with each others plays.

145
by nat :: Thu, 11/10/2011 - 11:19am

Okay, the next question is how do we account for that 1st ranked TO/Drive?

What luck! We happen to have a team that this very day is 18th in yards/drive and 1st in TO/drive: The Detroit Lions.

What's their offensive DVOA? 8.7%

What? It's not 29.7% ?!? How could DVOA be so far off?!? Let's look at teams near this, like the Vikings (15th/5th), the 49ers (24th/4th). Oh, they're at 7.0% and 3.3%. I guess it's that one Jets game offensive DVOA that's screwy.

So the Jets drive stats point to a VOA of about 8-10% (because .000 is a tad better than .040, but the Lions DSR rank is 15th) and a DVOA somewhat lower than that to factor in the horrible Pats defense.

There's a reason Aaron can't find the hidden +30% VOA in the Jets offense of that day. It wasn't there. At this point, I'm guessing we're looking at a data transcription error.

154
by Jeff Fogle :: Thu, 11/10/2011 - 4:50pm

Great stuff Nat and BHA. Hope Aaron eventually does go play-by-play to solve the mystery. Important reminder in nat's posts I think that the Drive Stats get at reality without turning everything into Klingon first. Very informative, and not much controversy because nobody's going to say "what do you mean so-and-so is only 8th in the league in points-per-drive" or whatever. Readers should expect a passion for accuracy from "genius brands".

160
by nat :: Thu, 11/10/2011 - 9:57pm

Drive stats and VOA both have their place. The genius of VOA is that it's built up play by play. In a very real sense, VOA should explain the drive stats. Drive stats and special teams VOA should explain victories.

Then DVOA can predict future performance.

It doesn't always, even when it's working as expected, which is what makes football a game.

There's been talk of somehow factoring drive stats into DVOA. I sincerely hope Aaron doesn't do that. It would miss the point of DVOA.

161
by Jeff Fogle :: Thu, 11/10/2011 - 11:35pm

Not convinced that VOA represents genius, or that DVOA can predict future performance. Any more than common sense anyway. Haven't seen compelling evidence, and naysayers have made valid points the past few seasons. The Drive stats do so much heavy lifting...and at least they're in English.

Plus, I'm concerned in general about the words "expert," "genius," and the like being thrown around the field of analytics when there's so much disagreement amongst the "genius" at FO (as described by Michael Smith on NNL), the "expert" at the "Advanced" site (Burke was included on a "panel of experts" in a recent SI piece), and the MIT grad at USA Today who's been doing computer ratings forever. How good are the Cincinnati Bengals? It's widely accepted that they're probably not top five, and certainly not bottom five. One place has them 9th, another 15th, and the other 24th. 9-15-24 within a range of about 6-27.

Others like that heading into the weekend:
Houston 1-3-12
Pittsburgh 3-11-13
Philly 4-14-14
Dallas 6-15-16
NYG 6-9-16
San Diego 12-18-20
Chicago 4-13-13
San Francisco 2-4-14
NY Jets 1-8-15
Tampa Bay 17-19-27
Denver 21-25-30

Those interested in the field of football analytics who attend the Sloan Conference or are otherwise well-suited to evaluate have to wonder if there's a "there" there given those results. I'm sure each guy is convinced his "there" is there. We're eight weeks into the season...where everyone starts saying "this is where the numbers really become meaningful because it's the midway point of the season." Yet, the "meaning" has the Jets either 1st or 15th, the Niners either 2nd or 14th, the Bengals either 9th or 24th, Tampa Bay either league average or league doormat.

It's football feng shui, even if it's a formula telling you where the couch should go. Multiple formulas are putting the couch in multiple places. Just like every other year the places have been in existence. The branding is working given the increased (but disappointingly unquestioning) exposure. Is the analysis delivering anything significant beyond what "unadvanced stats" or accurately computed drive stats are already teaching? A passion for figuring out what went on in the Jets/Pats game would naturally come from a motivated analyst. Here's it's the readers providing the passion for solving that mystery.

166
by Jerry :: Fri, 11/11/2011 - 11:23pm

Maybe one of these systems is wrong. Maybe all of them are wrong. And, if they all agreed, two would be superfluous.

Obviously, each system measures something different. Not having paid much attention to it, I believe Aaron when he says that they normally are "in agreement about which teams are overrated and underrated." Apparently, things have unfolded in an abnormal way so far this season. We'll see if the rankings converge as the season continues. Meantime, I'm willing to accept that different systems will yield different results, regardless of where their creators got degrees.

156
by Andrew Potter :: Thu, 11/10/2011 - 5:49pm

That's not really what we're looking at here though, is it? We're looking at better than the first ranked TO/Drive. We're looking at not only 0 turnovers, but also 0 fumbles. We are, in effect, looking at perfect ball security.

So to analyse that, we really need to create a list of teams that had 0 interceptions and 0 fumbles on offense for a single game, then compare and contrast the teams' DVOA numbers. I can get a list of those teams, but I really don't have the tools to compare the numbers without week-by-week offensive DVOA numbers for individual teams. Does Premium Membership offer that? Even if so, it's an expensive question to answer.

Aaron's conclusion (week 6 DVOA thread) was that Red Zone Offense was the reason for the Jets' seemingly inflated DVOA. That, combined with a lack of turnovers, looks like it can go a fair chunk of the way toward explaining their single game DVOA against the Patriots. Beyond that? It becomes very difficult to figure out without access to the DVOA database and that is an expense I simply can't afford.

Fun conversation. It'd be neat to see somebody with access to the DVOA numbers take over from me here.

159
by nat :: Thu, 11/10/2011 - 9:42pm

You have a point about .000 TOs/drive being better than .040. The difference is less than half a turnover. Call it about 22 yards of value, spread over 53 plays. That's not nothing, but it's not earth-shattering. I did give the Jets an extra percent or two, even though the Lions are 15th in DSR, not 18th. I thought that was generous.

It's a bit tiring that you keep asking for concessions. Only 0 int and 0 fumble games? Why? Because we can't estimate the value of .44 turnover? It's worth 22 yards. You can argue for 25 yards if you want, but I'd already given them some extra to cover it, so why bother?

Aaron did not say that red zone caused the discrepancy. He said it was the biggest issue he could find. As you and I have discussed, for the Jets offense that amounts to treating 7 red zone plays as if they were 9 - actually 8.75, of which 7.5 were successful. It gives a slight increase in the overall success rate. So the biggest issue Aaron could explain was that instead of giving the Jets a 53% success rate, DVOA's red zone bonus gave them a 54% success rate. I guess it also increases the value of the excess yardage gained, too. But we're talking the excess yardage packed in 37 yards in 7 plays - up to 15 yards, if you're very generous. The added 25% is worth about 4 yards.

I think we've taken this as far as we can with success rates and drive stats. The next step is the VOA numbers. Aaron could easily give the Jets offense VOA numbers by down, by drive, by quarter, whatever. He could even do it without the red zone bonus. This is exactly what VOA and DVOA are good at. Then he could give the opponent adjustment factor he used.

If he choose, he could do the same for the Patriots, perhaps with and without the fumble and interception, so we could gauge the importance of going turnover-free. That way we'd have something to compare. As we've seen, the two teams had the same success rate, while the Patriots got 190 extra yards. Comparing the VOAs with the turnovers removed would be hugely instructive.

All we can do is ask, and hope.

162
by Andrew Potter :: Fri, 11/11/2011 - 2:31am

I'm not asking for concessions, and it's rather demotivating that you suggest I am. There's no hidden agenda here. I'm trying, to the best of my ability with the limited tools available to me, to analyse, with you, why DVOA might have produced the result it did in hopes of learning something about the game, and about DVOA. I'm not blindly defending DVOA, and I'm not going after preconceived ideas, I'm simply trying to crunch what numbers I have and see what it tells us about why DVOA and conventional wisdom differ in their view of that game.

I agree that we've probably taken this as far as we can with the numbers we have. I'm not planning at this point to go carving up more games to compare performances. Like you, I'd very interested to read an article on the subject.

23
by Anonymous Coward (not verified) :: Tue, 11/08/2011 - 5:10pm

It's a bit of a stretch to say the Pats dominated that game. The most the Jets fell behind by was 10 points, they were within 3 in the 3rd Quarter and within 7 in the fourth with 7 minutes to go. While the Pats held a lead the entire game, it wasn't exactly domination.

The story of that game was that the Jets played conservative on offense, and didn't attack the Patriots awful secondary. Instead they tried to limit their own turnovers, run the ball and win on defense...but their defense wasn't good enough especially after losing Bryan Thomas the week before.

DVOA is overrating the Jets, but the Jets weren't dominated by the Pats. In fact, both the Ravens and the Raiders did a better job against the Jets than the Pats did.

54
by RickD :: Tue, 11/08/2011 - 7:15pm

That depends on how strongly you read "dominated." When I have my math hat on, "dominated" can mean simply "had a greater value over several categories."

But in this case, let's see...

The Pats had 26 first downs, the Jets had 14 first downs.
The Pats earned 321 yards passing, the Jets got 166 yards passing.
The Pats had 152 yards rushing, the Jets had 97 yards rushing.
Mesko averaged 51.3 yards per punt over 4 punts. Conley averaged 45.6 yards per punt over 7 punts.
Wes Welker returned 4 punts for an average of 13.5 yards/return.
Jeremy Kerley returned 2 punts for an average of 10 yards/return.

The Jets did have a large advantage on kickoff returns (5 for 198 yards as opposed to 3 for 71 yards.)
Also, Brady threw one pick and the Pats had one fumble, which they recovered themselves. The Jets had zero turnovers.

Oh, and the Jets had 8 penalties for 89 yards while the Pats had 6 penalties for 50 yards.

All things considered, almost every statistic favors the Patriots. Oh, and they won the game 30-21 and the Jets were never closer than 3 points after the first Patriots TD.

Whether the Ravens or Raiders thumped the Jets more is immaterial to this discussion. We're trying to understand how a game seemingly dominated by the Pats in every statistical category ended up yielding a huge number of DVOA points to the Jets and seriously hurt the rating of the Pats at the same time. It is, at the very least, counter-intuitive.

56
by Aaron Schatz :: Tue, 11/08/2011 - 7:26pm

I don't think I've hidden from the fact that I'm not sure why DVOA favored the Jets in that Week 5 game. I've spent a lot of time going over it trying to figure out how to explain it without just literally going play-by-play through the whole thing.

However, the Patriots game is not the reason why the Jets are number one. Remove the Week 5 game, and the Jets' total DVOA would go UP to 35.5%. Offense would drop to 3.2% and special teams would drop slightly to 7.9%, but defense would improve to -24.5%.

58
by JetFanInMD (not verified) :: Tue, 11/08/2011 - 7:33pm

Perhaps the problem is that the adjustment the Jets D gets for playing the NE offense is too high?

61
by Milkman (not verified) :: Tue, 11/08/2011 - 7:41pm

Considering the Jets' Defensive DVOA and total DVOA both go down based on that game, I don't think that's the issue.

62
by RickD :: Tue, 11/08/2011 - 7:49pm

I suspect DVOA underrates Welker's 73-yard pass reception. That's really as good as 7 first downs from a dink-and-dunk offense.

129
by John Doe (not verified) :: Wed, 11/09/2011 - 3:43pm

Descriptive vs. Predictive. When "describing" the game a 73 yard pass reception is as good as a long, consistent drive. On the other hand long, consistent drives are more likely to be repeated than 40+ yard touchdown receptions.

74
by nat :: Wed, 11/09/2011 - 12:12am

The problem isn't that the Patriots game put the Jets on top. It's that the DVOA can't be explained, which makes us worry that the DVOA system has somehow become broken since last season.

You shouldn't need to explain each and every play. Split it up by down.

On first down, the Jets averaged 4.4 yards/play, and got a first down just twice in 21 plays. What's the VOA for that? What's league average?

On second down, the Jets averaged 5.6 yards/play, and got a first down nine times in 21 plays. What's the VOA for that?

On third down, the Jets averaged 3.6 yards/play, and got three first downs in 11 plays. What's the VOA for that?

Looking at league average yards/play and first down rates, it's obvious that the Jets were below or well below league average on first and third downs, although avoiding turnovers is worth something to be sure. They did this against an awful defense, so the DVOA for those plays should be below league average. Is it? If not, you had better check for bugs in the spreadsheet.

So the remaining question is: what was their VOA on second down? What was the DVOA after adjusting for a horrible defense? If the Jets got 29.7% on offense when including their poor first and third down performances and adjusting for a bad opponent, their second down pre-adjustment VOA must have been astronomical! Which is possible, but odd considering their yards/play were just about average.

So try it out, not with my rule-of-thumb numbers, but with actual VOA and DVOA. If nothing else, you should be able to confirm that it was only second down that the Jets offense was better than average.

93
by Exy (not verified) :: Wed, 11/09/2011 - 9:56am

Wouldn't red zone performance have a bigger impact with regards to both VOA and DVOA? Iirc the Jets were highly successful there and the Pats actually have/had a good red zone defense. Meanwhile Brady had that pick to Hernandez which was probably worth a lot.

97
by nat :: Wed, 11/09/2011 - 10:41am

It's the Jets offense DVOA that's the screwy one, so the Brady pick isn't in the story.

I don't think DVOA uses a different opponent adjustment in the red zone. It just weights the plays more in the overall total.

Yes, the red zone bonus is one suspect for the screwiness. It doesn't seem enough, since the Jet had just seven red zone plays out of fifty-three. But it's possible. Aaron could tell us what the VOA would have been without the bonus.

63
by Anonymous Coward (not verified) :: Tue, 11/08/2011 - 8:59pm

It might seem counter intuitive at first glance that the Pats went down in DVOA and the Jets went up after losing the game, but it also seems like a great predictor. Since that game the Jets are 3-0, the Pats are 1-2.

Regarding the Ravens and the Raiders, I was simply using them as a basis of comparison. We were discussing what "dominated" means in the context of a football game, so yes it's material.

98
by jonnyblazin :: Wed, 11/09/2011 - 11:28am

Why would total yardage indicate dominance? Pats ran 72 plays, Jets 53, and the last 5 Jet plays were "garbage time" (down 2 scores w/ 1 minute left).

Going over the Jets play by play, I counted the following (probably missed 1 or 2):

22 good plays
8 meh plays
16 bad plays

The Jets only had 1 play for negative yards until the last play of the game, and that was on 3rd down so it was just as successful as any other failed 3rd down conversion.

So if the Pats threw an INT and fumbled, maybe we should subtract ~75 yards from their total? Just a guess. That would bring their yards per play down to about 5.1, which is close to the Jets 4.8. Then you can go through NE's plays to see how many good plays vs. bad plays they made.

6
by Bobby Wommack (not verified) :: Tue, 11/08/2011 - 4:17pm

Worst year in existence for DVOA continues...

134
by Vicious Chicken Of Bristol (not verified) :: Wed, 11/09/2011 - 5:14pm

DVOA appears to have jumped the shark.

7
by Travis :: Tue, 11/08/2011 - 4:20pm

The 1993 Jets went on to blow 21-0 and 17-0 leads in the two weeks after Week 4. Highlights included this Eric Allen game-winning 94-yard interception return; 38-year-old Vince Evans, who hadn't started a game in 6 years, leading the Raiders to a comeback; and and the clock operator in L.A. improperly stopping the clock long enough to give the Raiders enough time to score on the last play of the game.

At 21-0 of the Eagle game, they genuinely looked like the best team in the league.

8
by tunesmith :: Tue, 11/08/2011 - 4:20pm

Denver climbs and their variance took a dive, which both combine to mean they really did improve as a team this week.

9
by Uncle Luke (not verified) :: Tue, 11/08/2011 - 4:23pm

D-V-O-A DVOA DVOA DVOA!

11
by Dales :: Tue, 11/08/2011 - 4:28pm

This is one year that I think DVOA has the Giants rated too high-- they've had the easiest schedule so far and have needed fourth quarter comebacks to win five of the six wins they have.

Yet, I look at who is above them, and I think they'd beat at least two, if not three, of them more times than they'd lose. We'll get to see one of them this weekend.

Strange season in the NFL.

12
by ChicagoRaider :: Tue, 11/08/2011 - 4:28pm

At least the Raiders are consistently inconsistent. They were really high in variance last year too.

78
by Eggwasp (not verified) :: Wed, 11/09/2011 - 4:27am

I think thats what happens when you have some decent talent but absolutely NO depth, and a tendency to give up 2+ enormous plays (usually runs) for TDs every other game. Never mind Campbell for Palmer, I think the McClain, Sebass, Shaughnessy, McFadden and secondary injuries are way more important right now and explain much of the recent decline. Annoying this happened when we had our first contending season in a decade but them's the breaks (literally) and no-depth is what happens when you spend too much money on Tommy Kelly, Kamerion Wimbley and your kickers.

24
by JIPanick :: Tue, 11/08/2011 - 5:22pm

You know, for all the talk about how high the Jets are, the Texans at 3 also looks really weird... I doubt they are really that good.

I'll buy Jets at #1 a lot faster than Texans in the top 3.

30
by nat :: Tue, 11/08/2011 - 5:39pm

Look at the net drive stats - still only up to last week. Houston is much, much better than even the Jets in those. And their special team stats are quite good, too.

Actually, lots of teams are better than the Jets in net drive stats. Which is why people are worried that the Jets DVOA is screwed up.

36
by bravehoptoad :: Tue, 11/08/2011 - 5:59pm

Won't great special teams have a tendency to make net drive stats look worse for the offense? Better starting field position for offense = lower yards per drive because maximum drive length is capped. SF, CHI, and NYJ are all mediocre in net drive stats, and all have great special teams. When you look at points/drive or tds/drive, SF and NYJ climb into the top third.

DVOA is nice in that it counts special teams -- some advanced stats sites don't at all.

48
by Arkaein :: Tue, 11/08/2011 - 6:48pm

I'd think that great defense might drag down drive stats even further than great special teams, since the Jets probably force a lot of punts from deep in opponent territory.

113
by bravehoptoad :: Wed, 11/09/2011 - 1:16pm

Yes, but defense is directly counted in net drive stats. Special teams aren't.

57
by nat :: Tue, 11/08/2011 - 7:26pm

Net means Offense minus Defense.

Field position created by good special teams will not affect DSR at all, should help most of your net stats by lengthening the field for the defense and shortening it for the offense, and could have a (small, I suspect) effect on your net yards/drive for the same reason.

105
by bravehoptoad :: Wed, 11/09/2011 - 12:43pm

I know what "net" means -- you'll notice special teams aren't present in the equation "offense minus defense."

I had been thinking about yards per drive, which is how teams are ranked in drive stats (on FO's page, in any case). Consider two hypothetical teams that are exactly equal, except that team A always starts drives on the 40 yard line, and team B always starts drives on the 10. Team A will have fewer yards per drive than team B, but score more points.

As for DSR, if you have a shorter field, you can have fewer successful drives and still score more points than a team with a higher DSR but poorer starting field position. So special teams may not affect DSR, but it will effect Pts/Dr, which seems like a better stat because it includes all aspects of team play -- offense, defense, and special teams, while DSR considers only offense.

132
by nat :: Wed, 11/09/2011 - 5:08pm

Yes, we aren't talking about special teams.

Yes, in theory, good field position could affect your net drive stats. It will help things like net TOs/drive and net points/drive, and hurt net yards/drive. The effect won't be large, because an average team scores a TD on less than 20% of their drives, and moves the chains about 66% of the time, meaning a TD will curtail a drive's length about 13% of the time. (66% of 20%)

DSR isn't about complete drives. It's about sets of downs, and what portion of them result in a first down (or TD). So except inside the 10 yard line, field position isn't a limiting factor. Very few teams start drives just 10 yards from a TD.

Points/drive is a less good way to judge an offense, and net points/drive a less good way to judge an offense/defense combination precisely because they can be dependent on special teams play or play of the other unit.

My suggestion in using drive stats to judge offenses and defenses is to focus on DSR first (move the chains), yards/drive second (improve field position), and TOs/drive third (it's better to punt than throw an interception or fumble). Points/drive should be a result of the other three. When it's not, you look to the other units and how they affect the LOS for the drives, and then to other factors, such as anomalous red zone play.

The relative importance of DSR, yds/dr and TOs/dr is an interesting question which I can't answer. DSR correlates best with DVOA, but perhaps DVOA undervalues turnovers or extra yardage.

Net DSR considers both offense and defense. That's what Net means. But you knew that.

135
by bravehoptoad :: Wed, 11/09/2011 - 6:08pm

Yes, we are talking about special teams. This is the comment of yours I was responding to:

Actually, lots of teams are better than the Jets in net drive stats. Which is why people are worried that the Jets DVOA is screwed up.

Jets DVOA includes special teams. The net drive stats you're concerned with do not. I was suggesting a reason that DVOA would value the Jets higher than net drive stats, and why net drive stats are not a reason for "worry" when you factor in good special teams.

Points/drive is a less good way to judge an offense, and net points/drive a less good way to judge an offense/defense combination precisely because they can be dependent on special teams play or play of the other unit.

Yes, which is precisely why they're a better way to judge a team, which is what we're talking about.

True about DSR...I was misunderstanding the definition.

136
by nat :: Wed, 11/09/2011 - 10:11pm

Ah, then you read the first sentence of that very same post, which talked about net drive stats and set the topic for this branch of the thread.

No, we are talking about net drive stats, and what they say about DVOAs. Certainly they don't include special teams performance and therefore say anything about special teams DVOA. But no one is asking about those.

On that subject, points/drive is a terrible way to judge special teams, even though special teams have a theoretical effect on them. Points per drive completely misses TD punt and kick returns. No, net drive stats are best for judging the two main units together. If you want to judge special teams, drive stats are completely off the mark. Even field goals per drive would be useless for judging field goal kickers, and that would be the best of them.

Please don't misjudge my concerns. This isn't about disrespecting the Jets, or even judging them. It's about finding out why DVOA came to such a strange conclusion about the Jets offense. I am hoping Aaron can find us some insights into both DVOA and the game of football by examining such an unexpected result.

150
by bravehoptoad :: Thu, 11/10/2011 - 1:27pm

Ah. So what you meant to say was: "Which is why people are worried that the Jets offensive DVOA is screwed up."

Or some modifier like that.

38
by SFC B (not verified) :: Tue, 11/08/2011 - 6:03pm

The Texans are the only team that is Top 10 in Offense, Defense, and ST. I'm a Texans fan so I'm biased, but GB is the only team in the NFL I'd pick who would clearly beat the Texans in a hypothetical best of 3 this season. I think the Texans could outscore most of the teams who could torch their defense, and I think their defense can stop the teams whose defense can slow down the Houston offense. Between being above average in all phases of the game, and a very weak remaining schedule, I'm really hopeful to get to see my first Texans playoff game.

49
by Chappy (not verified) :: Tue, 11/08/2011 - 6:50pm

Dude, what makes you think they would beat all non-GB teams when they've already lost to New Orleans, Baltimore and Oakland? I think you have a lot of reasons to be hopeful for a playoff game, but I wouldn't hold your breath on Houston as a dominant team. I think the problem this is that there aren't really any dominant teams, so it all comes down to match-ups.

84
by Mr Shush :: Wed, 11/09/2011 - 7:23am

He didn't say "would beat". He said "wouldn't clearly lose a best of 3 against". The Texans had a bizarre shoot-themselves-in-the-ass meltdown against Oakland (and still had an opportunity to win that game at the death) and lost by one score in a shootout in New Orleans which they led for most of the game and could easily have one. They were thoroughly beaten in Baltimore, but the Ravens are for my money the second best team in football and are also a horrible matchup for the Texans. As for the wins, they bossed the Steelers more comfortably than the scoreline would suggest, and have dominated the various middling or worse teams on the schedule. If you believe in the value of stomps, the Texans have been doing some stomping.

For my money, Green Bay are in a class of their own. Significantly behind them, but just ahead of the rest of the pack, are the Ravens (and conceivably the Saints). There's then not too much to choose between the Texans, Giants, Jets, and maybe a few others. I don't buy the 49ers, because 1. they play in the NFC West and 2. too much of their ranking is down to a level of special teams performance I doubt they can sustain. They'll still make the playoffs, of course.

I wouldn't quibble with a ranking for the Texans anywhere between 3 and 7. With a healthy Andre Johnson, they could even conceivably be #2. But they are most definitely not an elite team - there is only one of those this year.

101
by InTheBoilerRoom :: Wed, 11/09/2011 - 12:06pm

Not a 49ers fan, but I did want to point out to you that the 49ers have only played one divisional game so far, a 33-17 win over the Seahawks in Week 1. None of their six other wins to date have come against the NFC West. They have wins against Cincinnati, Detroit, Tampa Bay, Philadelphia, Washington, Cleveland, and Dallas. Are any of those super impressive? Maybe not. I'd consider the wins @Detroit and @Philly to be solid. The stomping of Tampa Bay was solid. Cincinnati has put up a good record against bad competition, so that win is not overly impressive. But, their current record is not a reflection of crappy NFC West competition, because they have not had the opportunity to face their division rivals yet.

I don't buy the 49ers as being a strong Super Bowl contender right now, but, to be fair, you can't hold it against them that they play in the NFC West when discussing their performance to date this season.

116
by t.d. :: Wed, 11/09/2011 - 1:38pm

When did they beat Dallas?

121
by InTheBoilerRoom :: Wed, 11/09/2011 - 2:46pm

Yep, my mistake. My intention was to list the other six wins, but for some reason I ended up listing all of their additional opponents. Thanks for pointing out my brain fart.

123
by t.d. :: Wed, 11/09/2011 - 2:53pm

Well Dallas fans need to take credit for whatever good moments they can from this mildly disappointing season.

117
by justme_cd :: Wed, 11/09/2011 - 1:42pm

Ah hem.

"They have wins against Cincinnati, Detroit, Tampa Bay, Philadelphia, Washington, Cleveland, and Dallas."

Not Dallas. Don't take that win away from us. However, as a Dallas fan I'd say the 49ers played clearly better against the Cowboys than other teams we've lost to. (Detroit, Jets) but that's only looking at my impression of the teams in the Cowboy games

143
by Mr Shush :: Thu, 11/10/2011 - 9:20am

That was more a lame attempt at humour than an actual justification for suspicion of the 49ers' quality. As in, "they are an NFC West team --> they suck", as opposed to "they are an NFC West team --> they have had a soft schedule --> they aren't as good as their results".

Truth be told, it probably has more to do with how absolutely bloody awful they were against the Texans in pre-season Week 3. I know it's only pre-season, but the Texans ran all over them and they couldn't move the ball on offense for toffee (not to mention the turnovers). In fact, it looked very much like the regular season games the Texans have played against inferior opposition. It was probably just a fluke, but the (admittedly Gore-less) 49ers 1s were awful.

163
by Karl Cuba :: Fri, 11/11/2011 - 3:33pm

The niners went against the grain and only played their starters for a couple of series in the third preseason game, where teams normally use starting players for two or three quarters. Davis, Willis etc left pretty soon and Gore didn't play. Harbaugh wanted to see how the back ups performed against first team players in order to evaluate them properly. I do remember Alex Smith looking a bit ropey though.

After about 5 plays the defensive line was Demarcus Dobbs, Ian Williams and Will Tukuafu while the middle linebackers were Blake Constanzo and Larry Grant. I remember the Texans running amock through the middle of the niners defense and being really worried at how Willis and Bowman were playing, it was a relief when I realised that 52 and 53 weren't on the field.

If anyone has ever wondered how team of reserves would fare against a starting squad, it would probably be as ugly as that.

164
by tuluse :: Fri, 11/11/2011 - 4:20pm

I figured Saints vs Colts gave a pretty good approximation.

168
by Mr Shush :: Sat, 11/12/2011 - 9:12am

Ah - that does make a certain amount of sense. Factor in that the Texans defense seems to be much better than I thought it was at that point, and it all no longer seems so wierd.

109
by bravehoptoad :: Wed, 11/09/2011 - 1:03pm

Yes, I keep reading all around the internets about how SF's record can be discounted because of the division they play in. It's an ignorant statement, since SF has only played one division game. When (and if) they're 14-1, then you can discount them for the division they play in, because likely 5 of those wins will have come against division opponents.

Why wouldn't special teams be sustainable? They have a smaller variance than either offense or defense. Good coaching produces reliable improvements (Brad Seely). Kickers and punters are cheap and reliable.

I'm not buying SF as a serious threat, either, but because I think:

1) their secondary will be exposed by good quarterbacks. The defense works by stopping the run and making the opposing offense be one dimensional. They get good pressure without blitzing, and all that plus creative scheme protects the secondary. But they get burned by big plays because they're not actually all that talented. I'd think they'd also have fits with a really accurate quarterback -- Green Bay, for instance, is already one dimensional, and doesn't care if you force them to pass.

2)their offense just relies on too few tricks to keep moving. They have Gore doing things up the middle, breaking occasional big runs because he's Gore, and Alex Smith being super good against blitzes and okay on short to intermediate routes. Their offense would add 5 points of DVOA if Gore or Davis were better catchers, but they both drop a lot of balls. They'd add another 5 points if Smith could hit deep routes -- their receivers can get open on them -- but he can't hit them. If they ever meet a defense that can shut down the run and apply pressure without blitzing -- in other words, if they ever meet their own defense's mirror image -- SF is toast.

IMHO. But it really is a humble opinion, because I'm not the most educated viewer.

118
by dmstorm22 :: Wed, 11/09/2011 - 2:06pm

They are just that team that does enough to get a bye, but everyone thinks and knows that it probably won't do any good. A lot like the 2001 Bears who were QBed mainly by Jim Miller and went 13-3. It wasn't a huge surprise to anyone when they lost to Philly. Another example is probably the 2005 Bears, who went 11-5 with Kyle Orton for much of it.

They'll get a bye (unless the Giants or Saints go on a huge run), but I doubt they'll be large favorites in the divisional round.

140
by zenbitz :: Wed, 11/09/2011 - 11:58pm

49er fan here.

I think there is some truth in what you are saying. However, they have played 3 good-but-not-great pass offenses (Detroit, Philly, Dallas) and while they gave up yards - all these games were close. They are not the Ravens defense - or even the Jets - but seem to be solid enough to hold Green Bay to 30-ish. Even the Packers run it 40-45% of the time.

On offense... well, they are a mystery. They run the ball a lot... but they are not very efficient at it. However, it sets up their play-action game nicely because their two good receiving TEs lets them pass out of running formations. They are terrible on 3rd down when they are forced to pass.

The question on the niners boards is... can the Niners keep up in a shoot out? Eh, unlikely right? But they did blowout the Bucs -- a team with a terrible pass defense -- essentially riding Alex Smiths' arm in the first half.

As to who is their worse nightmare... well the Jets are like the same team -- great special teams - but with worse offense and better pass defense. But that just makes a 6-3 game either way. Similar with Baltimore - although that game could be 0-0 except for one deep Flacco bomb. It seems a team like Detroit or Atlanta (balanced offense and defense) would be trouble, but they beat the Lions in Detroit already - not any guarantee or anything but I think a rematch would be close.

157
by battlered90 (not verified) :: Thu, 11/10/2011 - 5:56pm

Ravens game was a lot closer than the final score indicated. Texans were within one possession until 4 min left in fourth quarter if I remember correctly. Also, the Ravens have looked a lot more beatable than they did against the Texans.

That said, if I had to guess I would say Texans high DVOA is a result of beating up on bottom dwellers for the past three weeks straight. Even if they finish third in the rankings, I don't expect them to make it deep into the playoffs.

39
by tuluse :: Tue, 11/08/2011 - 6:11pm

If the Texans were healthy I could see them as a top 3 team. The problem of course is that they're not going to be healthy this year.

68
by dmstorm22 :: Tue, 11/08/2011 - 9:54pm

The Texans are a team that might quietly be really good. They have a great running game. Very good o-line, and a good QB. They've been good even though they have missed Andre Johnson since week 4. This is conventional stats, but they are #1 in defensive yards allowed (again, conventional, they've had an easy schedule). Their defense is solid. Their offense is good and should get better with Andre. Missing Mario hurts, but their pass rush is still there.

They also have an easy schedule the rest of the way. They might be favored in every game left on the schedule and a 12-4 finish is very forseeable, and running the table isn't out of the question.

27
by 0tarin :: Tue, 11/08/2011 - 5:33pm

Interesting to see Cincy pulling ahead of GB in Variance. And I never would've guessed I'd be so pleased to see Baltimore ranked #30 there. Glad to see them finally putting together some semblance of almost-consistency sometimes maybe.

28
by Schrute Farmer (not verified) :: Tue, 11/08/2011 - 5:35pm

The Packers' passing DVOA is 82.2%. In a year where 'average' is higher than it has been before. The Packers' as a team aren't a juggernaut but, given THAT offense with THOSE receivers and THAT No. 12 playing with them, they are as hard to beat as a juggernaut. So I'm okay with the Jets being ahead of the Packers, because at least the DVOA nails down just how good the Packers' passing attack has been.

Though I am curious how their offensive DVOA is only 39.3%. Has their running game been THAT much of a drag?

Instead of Aaron trying to explain the Jets' ranking, I'd have prefered Aaron explaining that, even though Green Bay isn't the top dog, they have an aspect of their game that makes them the most difficult team to beat. Not the most complete team, but the most difficult team to beat.

33
by QQ (not verified) :: Tue, 11/08/2011 - 5:45pm

I'd love to see what GB's Drive stats are for Quarter's 1-3. It seems like every 4th Quarter GB shifts to clock kill mode which of course kills their Offense as they run the ball play after play then have to keep punting to their opponent. Exhibits A and B would be the 4th Quarter of the Saints Game and GB's final possessions vs SD

40
by Schrute Farmer (not verified) :: Tue, 11/08/2011 - 6:12pm

From ESPN's NFC North Blog:

Rodgers didn't have time to set his feet and simply flung the ball about 55 yards in the air. Second, that pass helped elevate Rodgers' completion percentage on throws of 20 or more yards downfield to 65.2 percent this season. That's a better completion percentage than all but three NFL quarterbacks have on all of their throws. Amazing.

65.2% on throws that travel 20 yards? Are you serious!?!?

If any team was ever better equipped to throw the ball every damn play, I'd like to see them.

72
by McDaniken :: Tue, 11/08/2011 - 10:50pm

ummm... the 2009 Arizona Cardinals?

90
by Aaron Brooks Go... :: Wed, 11/09/2011 - 9:51am

Warner-era Rams?

They were basically GB, but with a RB who demanded 8 in the box and was a better receiver than any of GB's RBs or TEs.

And the Rams did it in an era where DBs could still mug WRs, and hitting a WR going across the middle was still legal.

95
by QQ (not verified) :: Wed, 11/09/2011 - 10:04am

I think GB's Offense is better than those Rams teams due to the depth of its skill position talent. Yes the combination of Faulk, Holt, Bruce was insane but their #3 WR (Hakim) would not be able to make GB's team this year. GB's team this year simply has more options and mismatches that it can exploit than did any of those Rams teams.

102
by Aaron Brooks Go... :: Wed, 11/09/2011 - 12:31pm

I think you're severely underestimating the value of Faulk. Faulk was a HOF-caliber RB who was also a premier receiver. Jaman Grarks is neither of those things, and defenses play GB accordingly. The mere presence of Faulk forced defenses to honor the run, and prevented them from playing wholesale dime defenses, which allowed StL to pass all day, despite a lesser 3rd receiver. That Rams offense was more versatile than this year's Packers.

110
by Independent George :: Wed, 11/09/2011 - 1:11pm

I'd also mention the presence of Orlando Pace at his prime completely eclipses anyone on the Packers' line. I'd give Rodgers an edge over Warner for his mobility, the GB receiving corps an edge over STL, and Faulk & Pace a huge edge over GB.

In conclusion, I would characterize both offenses ridiculous to the point of obscenity.

119
by dmstorm22 :: Wed, 11/09/2011 - 2:09pm

The one difference is that the Rams turned it over. A lot. Of course, it didn't matter a lot since the defense, except for 2000, was good.

This is where I think DVOA cannot guage just how scary the Rams were. It really was like the only way to stop them was to get turnovers. They weren't really stoppable in the conventional sense of forcing punts. To me, they were the most unstoppable offense I've ever seen because the only way to really stop them was for them to make mistakes. That was the real story of Super Bowl XXXVI. People laud BB, but really, if Ricky Proehl doesn't fumble before halftime and if Warner doesn't throw a terrible pass in the 3rd quarter (not even counting the pick-6), the Rams probably win that game easily.

65
by Turin :: Tue, 11/08/2011 - 9:10pm

I'm almost starting to wonder if McCarthy and Capers aren't trying to protect the more interesting chapters of the GB playbook for later in the season.

34
by Mr. X (not verified) :: Tue, 11/08/2011 - 5:50pm

The question is this: if the Jets are tops of the DVOA, how have they lost 37.5% of their games?

82
by ammek :: Wed, 11/09/2011 - 6:03am

Precisely. I mean, who ever won a championship by going 10-6 in the regular season?

91
by Aaron Brooks Go... :: Wed, 11/09/2011 - 9:52am

Who said a single-elimination tournament was a robust means of determining the best team?

(Does anyone legitimately argue Villanova was better than Georgetown, for instance?)

108
by FrontRunningPhinsFan :: Wed, 11/09/2011 - 12:56pm

Certainly not the BCS.

130
by panthersnbraves :: Wed, 11/09/2011 - 3:48pm

NCSU 1983 would defend that to the death!

35
by Hank (not verified) :: Tue, 11/08/2011 - 5:55pm

A question about what goes on under the hood-
Does TOP impact defense dvoa? As in a team with an offense that consistently stays on the field an additional 5 minutes, means the defensive unit has 5 less minutes and similarly so for offenses that are 3-outs.
Much like an int, are defenses unfairly impacted by what the offense leaves them with?

69
by Jerry :: Tue, 11/08/2011 - 10:05pm

DVOA doesn't look at time of possession at all. It's just the total of the values for individual plays. Those individual play values reflect field position and some other game circumstances, so "what the offense leaves them with" isn't a direct effect.

79
by Eggwasp (not verified) :: Wed, 11/09/2011 - 4:35am

which is why it doesn't work. I suspect TOP will be overrated for offense, as it seems to prefer a string of 7-yard plays to an 80 yard TD catch, which takes longer.

On a related point, I also suspect it undervalues the 80 yard TD catch because a string of successful plays still leaves the chance that one will be very unsuccessful before any points are scored (a fumble, a tipped pick etc), whereas the long, scoring strike reduces this chance to zero.

80
by Eggwasp (not verified) :: Wed, 11/09/2011 - 4:36am

*it is viewed by some as to not work

111
by bravehoptoad :: Wed, 11/09/2011 - 1:12pm

... it seems to prefer a string of 7-yard plays to an 80 yard TD catch, which takes longer.

As has been said time and again, this is simply untrue.

DVOA prefers 11 successful 7-yard plays to an 80-yard catch and 10 unsuccessful plays. Who wouldn't? 10 unsuccessful plays is going to give the other team the ball at least 3 times.

DVOA doesn't count TOP because it's a play-based system. It's only evaluating individual plays.

You're also confusing things...if DVOA is evaluating 11 successful 7-yard plays, it's because they've already all been successful. If they had been unsuccessful, DVOA would have evaluated them differently.

37
by phillyangst :: Tue, 11/08/2011 - 6:01pm

"DVOA loves Philadelphia!"

'nuf said.

71
by B :: Tue, 11/08/2011 - 10:18pm

DVOA is always sunny on Philadelphia.

42
by Kb (not verified) :: Tue, 11/08/2011 - 6:31pm

Im not a packer fan but have made it a point to watch everygame this year. Their running game isnt great but in no way would i call it below average. Their number 1 back is averaging 4.6 ypc and grant is above 4the ypc.. also trying to claim gbs wrs havent been the best in the league is completely false. There has been two games where their wrs dropped at least 10 passes but besides those games they have been lights out. Their defense is a different story but if i remember right this d isnt much worse than it was at the beginning of last year then they averaged wat. 10 points a game the last half. I know there best corner has been hurt all year. Burnett played really good the first few games then he broke his hand and has played wit a club.. imo there biggest loss has been nick collins who was the leader in the secondary. I have to think this team will be fairly complete on the defensive side of the ball once the playoffs come around... ok my friend just mentioned they will be getting their second round pick from last year back in michael neal. They desperately need him because they need a inside pass rusher to help bj raji(eho has played way to many snaps) and complement clay matthews. Anyway it is obvious the packers are much better than the jets and i dont believe that game would be within 10.. sorry for this rant

44
by MilkmanDanimal :: Tue, 11/08/2011 - 6:41pm

San Francisco is really disappointing me by their apparent plan to utterly screw up the NFC West's fine tradition of sucking completely and utterly. Three teams in the bottom six, and SF has to come along and make a mess of things.

45
by Jonadan :: Tue, 11/08/2011 - 6:44pm

On the other hand, the Jets' raw DVOA splits don't look that ridiculous, all things considered and taken by themselves. Given how I think DVOA works, the beat-downs they laid on Miami and Jacksonville I would expect to be the major factor at the moment, and their losses are "good".

But that still doesn't add up to #1 - The record seems to agree with the majority: they've beaten exactly two probable playoff teams (counting Bills, Chargers, and Raiders as a total of two). Taking the numbers by themselves, my diagnosis is that DVOA is over-rating the effect of the Jets defense right now, but there is also the Pats game question people keep bringing up.

---
"When you absolutely don't know what to do any more, then it's time to panic." - Johann van der Wiel

55
by RickD :: Tue, 11/08/2011 - 7:25pm

The Bills are a probable playoff team? I'd count them as a contender, but they have 3 losses already and still haven't played any road games in their division. They also have a road game in Dallas against a team that would seem to match up well against them. And they have a road game in San Diego.

I expect they'll finish with 6 or 7 losses, which will put them behind the Jets, Pats, Steelers, and Ravens.

The home loss to the Jets really hurts them, since they have nothing other than a home win against the Pats to balance that.

Of the three teams, the Jets clearly have the advantage, with the road win and two home games left against the Pats and Bills. And both the Jets and Pats have far easier schedules down the stretch than the Bills do.

46
by Raiderjoe :: Tue, 11/08/2011 - 6:44pm

Raiders at 21th is outrageuos. The Raiders beat the Jets head to head action and only have one less losss than Jrts. Id dvoa computer broken? Did Aaron spill cofee on it?

Raiders grtetting back into first plave thursdsy night. Goimg to teach chatgers a lesson. We will see abiut Jets. Let's see Jets rtry to win at Oakland in playoffs. Cuz thst going to happen. Raiders getting home playoff ham

59
by RickD :: Tue, 11/08/2011 - 7:36pm

The Raiders need to show that they can beat a good team without Jason Campbell at QB. They've lost two consecutive home games to divisional rivals. Haven't looked good in either game.

Oh, and their schedule isn't very kind, either.
Hmm...every team in the AFC West has negative net points. The Raiders could pull this out.

75
by commissionerleaf :: Wed, 11/09/2011 - 12:58am

The Raiders need McFadden back. Without him their offense is a still-learning Carson Palmer throwing to the U.S. Olympic track team.

81
by Eggwasp (not verified) :: Wed, 11/09/2011 - 4:39am

More important would be actually making some tackles on defense, on say, every play instead of taking a few off each week.
By all accounts they have talent on the DL and even some at LB, its maddening that they just can't fix this thing. I'm fed up with Seymour and Kelly talking about this after every game.

It also would help if they didn't give opposing offenses a do-over every fifth play.

104
by WesOMG (not verified) :: Wed, 11/09/2011 - 12:34pm

Home playoff ham is delicious.

53
by Johnny Socko (not verified) :: Tue, 11/08/2011 - 7:03pm

The Colts are currently 32nd on defense, 31st on special teams, and 25th on offense. Is is possible they could finish the year with a perfect tri-fecta, meaning last in all three categories? Has that ever happened?

64
by Tom Gower :: Tue, 11/08/2011 - 9:07pm

Just checked: the 2009 Lions came closest, as they finished 31 O/32 D/31 ST, and not far ahead of the Rams on offense (go ahead and cry about your special teams that year, Packers fans). 2003 Cardinals were 32nd on offense and defense, but only 25th on special teams.

92
by Aaron Brooks Go... :: Wed, 11/09/2011 - 9:56am

2009 was the season Hanson was hurt. He can usually single-handedly elevate the Lions STs to mediocre.

60
by kbukie :: Tue, 11/08/2011 - 7:40pm

Looking at the % of offensive plays that lose yardage, does that include offensive penalties as well?

107
by Aaron Schatz :: Wed, 11/09/2011 - 12:47pm

No.

128
by kbukie :: Wed, 11/09/2011 - 3:26pm

That doesn't reflect well on the Bears, then, who are failing that frequently on offensive plays on top of all the false starts on their offensive line.

Makes me wonder how good Chicago's offense would be if the line performed at an average level.

133
by tuluse :: Wed, 11/09/2011 - 5:12pm

Probably about how they looked Monday night.

66
by TomKelso :: Tue, 11/08/2011 - 9:49pm

Just wondering if DVOA picked the Ravens to lose to the Jets as well (or Houston, for that matter).

Heck, I'M still trying to figure out how they lost to the Titans or Jax -- why would I expect a system designed by people to do any better?

76
by jonnyblazin :: Wed, 11/09/2011 - 1:25am

They run the same kind of defense, Flacco always struggles when rushing 4 D-lineman slowly collapse the pocket on him. The DBs muscle up on the WRs and crowd the box, and Flacco pump fakes, wanders around, gets strip sacked, etc. I'm interested to see how he plays the Bengals this year, they've had his number in the past.

70
by tuluse :: Tue, 11/08/2011 - 10:13pm

Looking at the Bears DVOA in week 9 last year compared to this year is fun. They had the same record, but last year were ranked 26th. The defense was about the same, the special teams is even better this year 8.8% vs 5.8%. However, the real difference is the offense, which was an absurdly bad -29.2% last year and is a bad but respectable -0.7% this year.

73
by kbukie :: Tue, 11/08/2011 - 11:28pm

I believe that most of that can be chalked up to what I like to call "The Todd Collins Experience".

77
by tuluse :: Wed, 11/09/2011 - 2:01am

There was also the first half of the Giants game, still that those combined would still only have been 3/16th of the total offensive DVOA. The offense is much better this year.

94
by Aaron Brooks Go... :: Wed, 11/09/2011 - 9:59am

Cutler seems miles better. He seems to have accepted that Martz contains a lot of mad mixed in with the genius, and occasionally just Favres a play when the call doesn't make any sense, then just wanders back to the huddle with that wry smile.

Basically, Cutler was the only reason the Bears were even in that Lions game, despite the O-line actively trying to kill him all night. He stayed patient and contained, and made some killer throws under duress. Last year, he would have either self-destructed or been ground into a fine paste by the end of the first half.

88
by jesse.hoff (not verified) :: Wed, 11/09/2011 - 9:41am

Focusing on the mystery of why the jets are getting rated so highly by DVOA is one thing, but it may be much easier to explain why they are ranking so high.

Their d is just very good at times when most d's have been terrible. compare them to the steelers, who rank around 10 better in offense and 10 worse on d and st. the dvoa gap at these levels is much larger on d then it is on offense (20-10). The Steelers have a positive defensive dvoa, as do the pats, packers and bills. If the packers had a merely adequate defense, with a 0 rating, they'd be a better team than the jets. Now, maybe the garbage time rating's have unfairly affected the packers (dvoa holds defenses fully accountable for garbage time, but not offenses).
The jets offense is capable of generating success, particularly against mediocre to bad defenses, while most of their opponents defenses have been incapable on average of such a feat.
The ravens offense has looked much worse than the jets, except in the jets game, and their special teams are definitely not as good.

99
by Sean McCormick :: Wed, 11/09/2011 - 11:58am

And the Ravens offense was terrible in the Jets game. Just not otherworldly awful like the Jets offense.

Nick Mangold for MVP!

100
by Jesus (not verified) :: Wed, 11/09/2011 - 12:03pm

The Jets' defense is good enough to keep them in the game with anyone. Even the beloved Texans . . . and they continue to improve a bit each week.
Their offensive line is becoming a force to be reckoned with. The Jets are very well rounded - since they have weapons at receiver, tight end, & running back. Their special teams are impressive -take a look at their kick return game. They are able to handicap the opposing passing game with 3 very competent defensive backs, and if you watched Fred Jackson try to run the ball on the Jets the way he did to every other team he faced, well, their run defense is becoming a thing of defensive beauty. (Running into the middle of the Jets run defense is pretty much a failed effort nowadays).
They have an all-pro center, and a QB that has won more play-off games on the road than -well, most any other QBs in the league. Tell me what other QB besides big Ben has taken their team to 2 consecutive championship games in a row... in their first 2 seasons in the NFL.
Hard to knock the Jets. And it is amusing the way their defense makes a great offense look rather pedestrian. As the defense continues to improve, and as the running game continues to become more of a factor - seems to me they are going to be in good shape come the rest of November and December.

120
by Gnopaine :: Wed, 11/09/2011 - 2:18pm

The Vikings have the least negative DVOA, without being positive.

Why does this strike me as the perfect analogy for the Leslie Frazier era?

126
by Independent George :: Wed, 11/09/2011 - 3:05pm

Brilliant!

122
by t.d. :: Wed, 11/09/2011 - 2:50pm

I don't really see the AFC East champ getting a bye, whoever wins the division. Looks like Balt and Houston have the inside track. Given the Pats makeup, if they win the division, that probably helps them

124
by Moridin :: Wed, 11/09/2011 - 2:58pm

I think DVOA could be improved if it incorporated a way of measuring drives in comparison to other drives. aka, essentially treating a drive as a 'play' and being able to compare other drive 'plays' in the same way that normal plays are compared to others (by down and distance, position on the field, game situation, etc). This would allow extra comparisons by TOP, # of plays run, field position vs drive outcome, TOP vs previous drive (3&outing or otherwise quickly giving it back right after a long possession is tiring out the D), etc.

Now, I think that this addition would be more descriptive that predictive, and that some of the things in DVOA currently attempt to make up for this (3rd down offense would probably overlap somewhat with drive 'plays', since the drives die with bad 3rd down performance), but I think being able to incorporate strategic decisions and a 'forest and the trees', instead of just the 'trees' approach would expand the number of factors being able to incorporate. I'm not sure how you would blend them together, though maybe just like O/D/S it would be a 9/10s of the O is play DVOA, and 1/10 would be drive DVOA to create the offense DVOA.

However, I realize that drive 'plays' are so much less available for comparisons that normal plays, because a 12 play drive is still just a single drive 'play', and that gives FO much less to work with for balancing and figuring out useful pieces to incorporate.

I just feel that there are a lot of things that can be done with this inclusion that 'sound' good and useful, but I don't know if they are statistically or scientifically useful.

137
by JoeDog (not verified) :: Wed, 11/09/2011 - 10:42pm

I'm a Jets fan but mostly I'm impressed by the author's use of "Internets"

139
by nat :: Wed, 11/09/2011 - 11:38pm

I'm revising my suggestions about how to use drive stats, at least for the offense. Yards/Drive and DSR are .97 correlated, so there's no point in looking at both of them. Take your pick: they mean the same thing. Either one is correlated with DVOA at .84, which is pretty good considering that DVOA is opponent adjusted, while drive stats aren't.

TOs/Drive and DVOA correlate at -0.15. The sign is right: fewer turnovers mean higher DVOA.

So it's correct to think of DVOA as measuring the ability to sustain drives, with an adjustment for turnover rates. There doesn't seem to be much room for long yardage plays as a separate skill. What puts Green Bay above New England on offense is a lack of turnovers, not the long ball.

153
by Tigera (not verified) :: Thu, 11/10/2011 - 2:38pm

I think that there are some areas of both defense & offense that cannot
be measured quantitatively. Either that, or their is no attempt to do so.
What is really an enigma is the fact that GB, as bad as their defense has been,
is leading the league in interceptions. Last week, the defense scored 2 TDs
on interception returns, and just missed a 3rd. They give up a lot, but they also take away quite a bit. The defense of the Packers is really hard to weigh. They
are definitely missing Collins, out for the year with injuries.

165
by Richard Grijalva (not verified) :: Fri, 11/11/2011 - 9:18pm

Is DVOA meant to be a more accurate measurement of past performance than conventional statistics or is it meant to be predictive? If it is meant to be predictive, is there any way of adjusting for a team like the Oakland Raiders who essentially got stuffed with a 2.5 game span of playing preseason football in the middle of their season? If we want to predict the future with DVOA, do we want to give the 2nd half of the Browns game and the Chiefs/Broncos game the same weight as any other football games?

Similar situations are pretty rare as I can't think of a lot of starting QB's that went down for an extended period of time without a competent backup in recent years, and QB is truly the only position that can sink a team hard if you go from top-10 performer in the league (Campbell) to a guy who shouldn't be in the league (Boller) or a guy who simply isn't ready for NFL speed yet (Palmer).

It would be fun if you guys built a what-if tool where curious fans could isolate DVOA of specific games (i.e. lookup what the Raiders DVOA would be excluding certain games, or with McFadden injured, or look at the Chiefs with healthy Matt Cassell instead of the broken rib version).

167
by Jerry :: Fri, 11/11/2011 - 11:36pm

For what you're asking, DVOA is being descriptive. When you evaluate this year's Raiders, you need to include Boller and Bad Carson. (The couple of similar examples that come to mind are 2005, when Charlie Batch was a competent replacement for an injured Roethlisberger, but Tommy Maddox wasn't, and the 2008 Cowboys, where Brad Johnson filled in terribly when Romo went down.) It's actually possible to separate offensive DVOA by QB, since those changes are mentioned in the play-by-play, but you can't do it for any other position. It's even harder to meaningfully separate Good Carson from Bad Carson.