Writers of Pro Football Prospectus 2008

Most Recent FO Features


» Weight and Injuries

NFL football is a violent game, and traumatic injuries are unfortunate but unavoidable. But are bigger players more likely to be hurt than their smaller peers?

08 Jan 2008

2008 Postseason DVOA I

by Aaron Schatz

Hey kids, time for postseason DVOA ratings. Like last year, based on reader requests, we've ranked all 32 teams, whether they are in the playoffs or not. Teams which did not play in the wild card round are treated as if they had a bye week. (That includes both the 20 non-playoff teams and the four teams with byes.)

All numbers are weighted DVOA. That means that Weeks 1-4 are not included, while Weeks 5-10 are somewhat discounted. If you want to see the actual regular season final ratings, click here.

There will be the usual comments over on AOL sometime on Wednesday, and they will be linked from the "FO Goes Mainstream" page. For those wondering, yes, we are updating the playoff odds report; it just didn't get updated until late last week due to some New Year's confusion. (It is updated now through the wild card games.) Loser League results are finally fully posted as well, and you'll find the matchup pages available for all four second-round games at the DVOA Premium database.

* * * * *

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

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

I'm not going to bother to run the whole DVOA explanation; if you are new to the website, you can read about it here. As always, positive numbers represent more points so DEFENSE is better when it is NEGATIVE.

1 NE 40.4% 1 16-0 39.5% 1 0.9% 18 1.8% 11
2 JAC 32.0% 2 12-5 24.7% 2 -8.7% 3 -1.4% 18
3 SD 27.9% 3 12-5 8.2% 11 -14.5% 1 5.3% 4
4 IND 21.2% 4 13-3 20.0% 3 -7.7% 6 -6.5% 31
5 GB 17.5% 5 13-3 17.7% 4 2.6% 20 2.4% 8
6 CLE 15.4% 7 10-6 7.7% 14 -0.8% 17 6.9% 2
7 DAL 14.5% 6 13-3 10.9% 6 -4.4% 11 -0.9% 16
8 SEA 14.2% 8 11-6 8.6% 10 -7.7% 5 -2.1% 21
9 CHI 10.2% 12 7-9 -12.8% 25 -12.0% 2 11.0% 1
10 WAS 9.3% 10 9-8 4.0% 15 -6.1% 8 -0.9% 15
11 PHI 8.2% 11 8-8 10.0% 8 -5.2% 10 -7.0% 32
12 MIN 7.8% 13 8-8 7.7% 13 1.2% 19 1.3% 12
13 TB 7.6% 9 9-8 1.9% 16 -6.9% 7 -1.3% 17
14 PIT 5.8% 14 10-7 1.3% 17 -8.3% 4 -3.7% 27
15 NO 4.6% 15 7-9 16.3% 5 9.2% 26 -2.6% 22
16 HOU 2.6% 18 8-8 10.9% 7 13.6% 30 5.3% 3
17 CIN 2.5% 19 7-9 9.6% 9 10.4% 28 3.3% 6
18 TEN 2.2% 16 10-7 -4.1% 19 -5.3% 9 1.0% 13
19 DEN 1.7% 17 7-9 7.8% 12 4.2% 21 -1.9% 20
20 BUF 1.3% 20 7-9 -4.5% 20 -3.9% 12 1.9% 10
21 NYG 0.7% 21 11-6 -5.6% 22 -3.6% 13 2.7% 7
22 NYJ -11.7% 23 4-12 -13.6% 26 -1.9% 15 -0.1% 14
23 ARI -12.0% 22 8-8 -1.4% 18 5.9% 22 -4.6% 29
24 BAL -14.4% 24 5-11 -12.0% 23 -1.0% 16 -3.3% 25
25 CAR -21.3% 25 7-9 -19.3% 31 -2.2% 14 -4.2% 28
26 ATL -25.0% 26 4-12 -12.3% 24 14.7% 31 2.0% 9
27 OAK -26.1% 29 4-12 -16.2% 28 6.2% 23 -3.7% 26
28 DET -26.3% 28 7-9 -4.6% 21 19.0% 32 -2.8% 23
29 KC -27.5% 27 4-12 -19.0% 30 6.6% 24 -1.8% 19
30 SF -30.2% 32 5-11 -25.8% 32 9.6% 27 5.2% 5
31 MIA -30.4% 31 1-15 -15.8% 27 11.8% 29 -2.9% 24
32 STL -31.5% 30 3-13 -17.3% 29 8.2% 25 -6.0% 30

Yes, the Giants are that low. It's pretty simple: New York has played very well the past two weeks, but really struggled for two months before that. From Week 7 through Week 16, the Giants played five games with single-game DVOA below 0%. The highest game during this period was the Week 11 victory over Detroit, and that game had a rating of just 26.8%. New York is trending up now, and Dallas is trending down, which makes Sunday's game very interesting -- but it still doesn't make the Giants a good team according to the weighted DVOA formula.

Here are the one-game DVOA ratings for the first round of the playoffs. Remember that these include opponent adjustments.

SEA 18% 1% -16% 0%
WAS 3% -8% -14% -4%
JAC 34% -2% -29% 7%
PIT 10% -15% -36% -11%
NYG 55% 43% -5% 7%
TB -37% 5% 36% -6%
SD 38% 27% -8% 3%
TEN -8% -4% 5% 1%

Posted by: Aaron Schatz on 08 Jan 2008

117 comments, Last at 13 Jan 2008, 9:28pm by Matt Saracen - QB1 - Dillon Panthers


by The Original Sam (not verified) :: Tue, 01/08/2008 - 4:50pm

Can't WAIT for the whining about Jacksonville at #2.

Although, even I am surprised by their ratings for the Pittsburgh game.

by turbohappy (not verified) :: Tue, 01/08/2008 - 4:54pm

Interesting. By these numbers, NE-JAC should be close and IND is actually the underdog to SD. Makes some sense only looking at the second half. All the games this weekend look pretty close except DAL-NYG and that one might be closer based on recent trends.

by Chris Heinonen (not verified) :: Tue, 01/08/2008 - 4:55pm

Since we're already weighing early weeks less, is it possible to weigh week 16 (and possibly 17) less for teams (like the Colts) that really didn't give a full effort during that time? I imagine Indy would be higher if they didn't lose to the Titans that last game.

by Al H (not verified) :: Tue, 01/08/2008 - 4:56pm

#1: The main reason they're still rated high is that the defense had an exception DVOA as did special teams. While the offense struggled much of the game, the rest of the team didn't.

by turbohappy (not verified) :: Tue, 01/08/2008 - 4:56pm

If Indy could muster average (0%) special teams, the SD and IND ratings would be equal.

by navin (not verified) :: Tue, 01/08/2008 - 4:58pm

San Francisco is not the worst.

by White Rose Duelist (not verified) :: Tue, 01/08/2008 - 5:03pm

Do I understand WDVOA wrong, or will eliminated teams will get shuffled around for no other reason than the number of games getting weight and exactly how much. I wonder if it would be more reasonable to freeze DVOA after a team is eliminated (except for changes to their opponent adjustments). It seems weird to think of how a team like the 49ers or Rams is playing now, because, well, they aren't.

by Fourth (not verified) :: Tue, 01/08/2008 - 5:06pm

Very interesting lines this week. All home teams favored by more than a touchdown...my first instinct would be to take all the road teams. Sea +8, Jax +13, SD +8.5, and NYG +7.5. Of course, one of the dogs will get blown out, but which? I'm thinking San Diego, despite that handy chart up there...is Gates playing?

by JasonK (not verified) :: Tue, 01/08/2008 - 5:07pm

The Giants' outperforming their DVOA ratings down the stretch is, I think, yet another argument that the next major revision of the formulae should include some weather adjustments. A meaningful portion of their offensive decline towards the end of the season (certainly not all of it, but a decent percentage) was heavily influenced by the playing conditions against Miami, Washington, and Buffalo. If Aaron can find some way for DVOA to take this into account, their rating would probably be more in-line with their recent performance.

by B (not verified) :: Tue, 01/08/2008 - 5:08pm

According to WDVOA, the Giants are the road team least likely to win and the Chargers a the team most likely to win. I feel like that's backwards.

by Jim (not verified) :: Tue, 01/08/2008 - 5:10pm

Wow, that's a pretty monstrous gap after the Giants at 21. The Giants are closer to 9 (Chicago? Really?) than to 22.

by Scott (not verified) :: Tue, 01/08/2008 - 5:15pm

Are guys like Robert Mathis, Raheem Brock, Marvin Harrison, Anthony Gonzalez, Ryan Diem and Ben Utecht going to play Sunday? If so I'd say SD gets crushed this week.

GB-SEA will be close, DAL-NYG will be close especially if TO is out, and I don't know what to expect from JAX-NE. Jags could give them a tough game like 4 other teams have this year, but then again, they could lose by 30. Either way, I expect the Patriots to win the game.

by admin :: Tue, 01/08/2008 - 5:17pm

Three notes:

First, as far as weather adjustments, remember that any weather adjustments which improved the rating for the New York offense would equally depress the rating for the New York defense.

Second, before you go putting your mortgage down on San Diego because they are listed above Indianapolis, remember this thing called home-field advantage. Also remember that DVOA does not include any adjustments for the effects of injuries, and that the Colts will not be starting any players this week named either "Pitcock" or "Craphonso."

Third, remember that weighted DVOA was never really meant to be continued into the playoffs so take all the ratings for non-playoff teams with a nice big chunk of salt. If we are going to freeze the ratings for those 20 teams at the end of the regular season, there's no reason to list them here. This is what people asked for two years ago.

by Catfish (not verified) :: Tue, 01/08/2008 - 5:17pm

re: 7

I was gonna suggest the same thing. It makes sense to drop the importance of earlier games when we have the evidence of more recent games to contradict them, but not when the teams are done. We have learned nothing more about Cleveland or any other nonplayoff team over the past week, so there is no reason for their rating to change, besides opponent adjustments.

by JasonK (not verified) :: Tue, 01/08/2008 - 5:23pm


Wouldn't a weather adjustment apply differently to passing than it would to rushing, though? Wouldn't that open up possibilities where an offense was more adversely affected by poor weather (e.g., because it fell behind and had to pass a lot) than the defense was (e.g., because the opponent was ahead and was running the ball)?

Or am I misunderstanding something?

by Purds (not verified) :: Tue, 01/08/2008 - 5:26pm

#5: "If Indy could muster average (0%) special teams, the SD and IND ratings would be equal."

Yeah, and if I could run a 4.2 40, gain 40 pounds of muscle mass, and fell no pain, I could be the first 40+ year-old DB to play for Indy.

(PS: I am a Colts fan, just not an optimistic one when it comes to ST coverage)

by TGT (not verified) :: Tue, 01/08/2008 - 5:28pm

#3: That has been tried, it doesn't help predict success.

by jonnyblazin (not verified) :: Tue, 01/08/2008 - 5:29pm

re: 3

Along those lines, I wonder what the Jacksonville WDVOA would look like if you remove their -50% DVOA week 17 vs. HOU. (If that game hasn't been removed already, that is...)

by Charles the Philly Homer (not verified) :: Tue, 01/08/2008 - 5:33pm


Anecdotally, there were a lot more "truly bad" teams this year.

by Doug Farrar :: Tue, 01/08/2008 - 5:54pm

Gadzookies. That's quite the cliff-dive by Seattle's special teams in the last two months (furrows brow).

by Are-Tee (not verified) :: Tue, 01/08/2008 - 5:55pm

Wow. The Jets' defense finished in the top half of the league. I know they made some changes after their bye week (week nine)and had one of the top defenses based on conventional stats since then. I would really love to see how their DVOA breaks down between the first and second halves of the season.

by Jacob Stevens (not verified) :: Tue, 01/08/2008 - 6:06pm

I agree, Doug. I guess that's why Aaron keeps saying prolonged special teams excellence is unsustainable.

0% on the Washington game, but it seemed a bit more than decent to me. Plackemeier's punts were OK for the first time in forever. Not much hang time, though, but good placement, and so much of it was from mid-field already, he did a fine job of pinning -- I'm pretty sure there were no touchbacks. I guess that weird kickoff thing, where the wind was carrying the ball and then suddenly dropped it straight down towards the right side of the 3rd row of blockers is keeping that from looking better.

Anyhow, if that kind of special teams performance is repeated on Saturday, rather than the dive ST took over the past 2 months, then GB-SEA is even more even. 20th ranked GB defense? What in the world happened to them? Everyone is saying Atari Bigby has become much less of a liability than he previously was...

by Bobman (not verified) :: Tue, 01/08/2008 - 6:06pm

No Gates for SD but that isn't huge--Indy's decent against TEs and the WRs blossomed last week for them. I suspect the Indy blowout will happen, mainly because the DVOA from the last 10 weeks measures a different Indy team.

Over that time they missed about 15 DL starts (not even counting Freeney, who won't obviously play), 15 OL starts, 3-5 LB starts, 13 WR starts, and 2-3 TE starts. ALL of those guys who missed significant time are expected back. A lot of them were held out the last few weeks to protect their health and the Colts still won before letting Week 17 go, so I'd venture to guess that their playoff DVOA will be a big step up from their second half DVOA, since their dinged-up starters should be both healthy and rested.

#5 Turbohappy, in our dreams! I sometimes fantasize about the Colts coverage teams running down the field in a line with half-inch stainless steel cable strung from guy to guy to pull down opposing KRs. Of course then the returners will start hurdling the Maginot Line and I'd be fired as ST coach. Maybe Darrell Reid's emergence will somehow Bob Sandersify the coverage units.

Obviously the game could go either way, but I don't even think it will be close. I know SD has improved since then, but in that Week 10 game, won by SD ST returns and 5 meaningful Manning picks, Indy's O was without 2 OTs and 2 WRs and the TE had a concussion, plus the D was without 2 LBs. And they still almost pulled it out on the road. (No idea about SD's health from then til now, but I THINK Castillo was playing hurt back then. Still? Merriman seems slowed down currently, at least to Dr. Z.) I can see the Colts playing with a Pats-like vengeance at home this week.

by Bobman (not verified) :: Tue, 01/08/2008 - 6:11pm

#16 Purds, one word: Pharmacology.

Or did you want to do this legally?

by MJK (not verified) :: Tue, 01/08/2008 - 6:22pm

Something I still don't understand about weighted DVOA is whether it is "iterative".

I get that it counts early games less. However, does that just mean that it calculates single game DVOA's normally, and then does a simple weighted average, with the early games discounted? Or does it actually take into account varying levels of play over the season when figuring the current "D" for the adjustment against a team?

For example, suppose some team had an abysmal defense in Weeks 1-10, and then made some magic switch to personnell or scheme (or get a weather benefit), and their defense started playing really well.

Now suppose some team in Week 16 or 17 scores very few points against them.

Overall, the team's defensive DVOA over the whole season will be bad, but their weighted defensive DVOA by Week 16 or 17 will be good. So does the team that scored very little on them then get slammed in its single game offensive DVOA, and hence in weighted DVOA, for scoring few points against a bad overall defense, or is "weighted DVOA" smart enough to figure out that it should get the understanding that it was playing a team with a good defense that was trending upwards?

I.e. do you take temporal trends into account when calculating the opponent adjustment in weighted DVOA? Does my question make sense?

(I image doing so would require some kind of iterative technique...)

by MJK (not verified) :: Tue, 01/08/2008 - 6:25pm


On the Jets defensive improvment--some of that may be weather related. Remember, they play in an open stadium in the Northeast, and at least two of their divsion opponents do as well. I bet a bit of their defensive DVOA boost towards the end came from "holding" New England to 20 points in Foxboro, but remember that the conditions had a lot to do with that.

by Tundrapat (not verified) :: Tue, 01/08/2008 - 6:26pm

#19 -

You nailed it on the 'really bad' teams. The fact that the 0.0% line falls just below the 21st-ranked team (NYG) tells the story. 7 teams had a DVOA of -25% or worse, while only 3 teams had a DVOA of +25% or better.

Doug - I admit to some confusion on the our 'Hawks ST play. I thought that they pretty much beat Washington based on special teams and field position (until desperation set in, and the 2 INTS were returned etc.). Plackemeier had 4 or 5 punts, and only one started at or past the 20-yard line for Washington. The coffin corner was beautiful, and Washington had 0 return yards from punts. Zero! Does that only affect Washington's ST DVOA, or does '0 opponent punt return yards' also improve Seattle's ST DVOA?

And of course, Josh Brown nails figgies from 50+, while Suisham chokes from 30 yards (big momentum shift there). So, I'm baffled about the almost negligible disparity (0% to -4%) in ST DVOA.

The bigger story in overall DVOA, to me, is New England's slow, steady decline in defensive DVOA. Looks like the age is starting to show later in the season. And it's not like they sat their starters, or weren't playing 'meaningful' games towards the end.

by Jon (not verified) :: Tue, 01/08/2008 - 6:27pm

Really interesting stat for the Giants for me is the S/T ranking. Remember folks, especially if you're praising Tom Coughlin now: he thought it was a good idea to bench Ahmad Bradshaw for Reuben Droughns. We were screaming, begging, pleading for Bradshaw for months while Droughns relentlessly killed drives, and returned kicks about as well as Shaun Rogers could. It's astounding how the S/T instantly improved with Bradshaw taking over.

Also, David Harris taking over for Vilma was really big for the Jets from what I saw.

by Toast Patterson (not verified) :: Tue, 01/08/2008 - 6:30pm

How is Chicago's D # 2? They were 28th in the league in yards allowed and 16th in scoring defense. I know DVOA looks at more than overly simplistic metrics like how many yards and points a defense allows, but 28th and 16th looks closer to the mediocre to sucky scale rather than #2 overall.

by Are-Tee (not verified) :: Tue, 01/08/2008 - 6:33pm


I watched that game, and the effect of the weather conditions was way overblown (pun only partially intended). Pennington threw for more yards than Brady, without his #1 receiver in the lineup. The Jets ran some very good schemes, and Revis did a great job covering Randy Moss.

by Tundrapat (not verified) :: Tue, 01/08/2008 - 6:33pm

Toast -

It's probably because the Chicago defense is good at preventing yards on a play-by-play basis, but they're on the field for 45 minutes due to 'less than ept' offensive play. So the opponent yards and points come, but only at a clip of a few at a time .

by BadgerT1000 (not verified) :: Tue, 01/08/2008 - 6:36pm


I have been trying to explain for about a month that the GB defense is not what the media claim.

Ryan Pickett has missed the last several games due to a sore knee and he is the focal point of the run defense.

Kampmann clearly wore down as the season progressed due to the loss of three backups causing the regulars to stay on the field even more than usual

KGB lost his explosive first step in the Detroit game and it has yet to return making him all but useless as a designated pass rusher.

Corey Williams is looking to land a big contract and pretty much played patty cake with opposing O-linemen to avoid injury. It was disgusting. He was playing at a very high level the first 10 games and then just packed it in the rest of the year.

These are all D-linemen and the GB defense is predicated on this unit controlling the action.

Because Bigby and Collins stink in coverage.

Al Harris has lost a step and can be exposed by quicker receivers

Woodson has been in and out of the lineup

The defensive secondary backups are erratic in coverage to say the least.

The linebackers (Poppinga/Barnett/Hawk) are solid but cannot compensate for all the leaks that have sprung up the last month.

If the weeks off have refreshed Pickett/Kampmann/KGB then then the defense should be fairly functional. If not expect the offense to have to score a lot, and I mean a LOT, of points on Saturday. Holmgren is too good in gameplanning for Seattle not to exploit issues......

by patriotsgirl (not verified) :: Tue, 01/08/2008 - 6:39pm

5, 16, 23: If I recall correctly, Indy ST had a solid game against Oakland - sure, it's Oakland, but you have to take your victories where you can find them.

23: I think no Gates would be fairly huge, because I suspect that the Colts will do a much better job stopping the big play by Chambers, Jackson et al than Ten did. If Rivers gets frustrated, no Gates will have a big impact (and allow more focus on that Tomlinson guy). Of course, no Gates may mean it's less likely that Rivers will throw to the TE with 5 Colts in the area like last time.

And I think Bobman's right about the Colts - someone mentioned in the Audibles thread that people seemed scared of the Chargers, but the Colts frighten the bejesus out of me (assuming NE can beat Jax). They've been way too quiet/under the radar for a defending SB champion for my liking...

by Tundrapat (not verified) :: Tue, 01/08/2008 - 6:43pm

BadgerT1000 -

I hadn't paid close attention to recent GB games, but that would explain their DT DVOA drop (much like New England's).

As a 'Hawks fan, I can find little to complain about with that. I'm pinning my hopes on the Seattle front 7 hitting Favre even half as often as they got to Collins. That should bring 'bad eveel Farvie' out of the musty old trunk, just in time to throw 3 interceptions.

One can hope...

by Joe in Seattle (not verified) :: Tue, 01/08/2008 - 6:46pm

I thought the Hawks ST DVOA might be higher for the game on Saturday, but they did have that phantom kickoff which the Redskins recovered, that probably hurt the rating. Plus Rock (or is it Rod?) Cartwright was getting pretty solid returns all game.

by MJK (not verified) :: Tue, 01/08/2008 - 6:48pm

I wonder how much of the Pats defensive decline is due to two injuries--Sammy Morris and Rosie Colvin.

Morris was a good runner to get 4-5 ypc, especially when teams backed a safety off to help with Moss, and so the Pats could run longer, sustained drives with Morris than without him, letting their defense rest more. Without Morris, they have Faulk and Maroney, who seem to be more "boom or bust" runners who get either 10 yards, or none, and they have been more forced to rely on their passing game. This in turn leads to faster drives, either to score or to go 3-and-out, which means the defense gets less rest than they did for the first half of the season. I haven't looked up the statistics to check this out, but I would guess that time of possession dropped a bit for the Pats after Morris got hurt, or possibley number of defensive snaps per game increased.

Then, when Rosie went out, that left the Pats with essentially four and a half LB's, instead of five and a half, to man four LB positions. Obviously, you're going to see a decline there.

Also, I think the Pats defense was rated artifically high early in the season--I think it was benefitting from teams trying to play in "shootout" mode instead of trying to run clock-sustaining drives.

by Kitione (not verified) :: Tue, 01/08/2008 - 6:51pm

What is the NY Giants DVOA counting only away games?

by BadgerT1000 (not verified) :: Tue, 01/08/2008 - 6:55pm

Post 34:

Well, the GB offensive line does a pretty good job in pass protection. Clifton and Tauscher are vets who know how to "use their hands" if you know what I mean. And Scott Wells the center has done a good job of calling protection changes at the LOS.

Like I wrote in the TB/NY thread, I continue to be surprised by folks getting all giddy about D-linemen who are light in the pants. Having watched KGB his entire career I know firsthand the downside of guys who are dependent on speed. Looks great if you can keep them on the field for 25 minutes give or take but like we saw on Saturday a d-line can start spinning its wheels pretty quick.

I expect Seattle to get some hits early but if Grant and Co. can keep them honest over time the pounding will take the starch out of their shorts. And the more a Holmgren offense stands on the sidelines the more impatient Mikey gets when he finally has a chance to do something.

By the way, forecast right now is for 25 degrees with maybe some snow showers.....

by panthersnbraves (not verified) :: Tue, 01/08/2008 - 6:55pm

#7 As long as the Panthers keep rocketing up (especially ST's!), they can keep adjusting through Pro Bowl Weekend - when they should have made it to average.

by Joe in Seattle (not verified) :: Tue, 01/08/2008 - 7:00pm


25 degrees??? Why so hot? I was hoping for something with a wind chill of 0-5 degrees.

by BadgerT1000 (not verified) :: Tue, 01/08/2008 - 7:03pm

Post 40:

So it limits Mike Holmgren's "Airing of Grievances" to less than 50 items.

by Tundrapat (not verified) :: Tue, 01/08/2008 - 7:07pm


Based on the Wild Card games, 'using their hands' probably won't be penalized. Still, since the Seattle defense is solid, and not just speed-based, I like their chances of getting to Favre. If Walter Jones (aging) can pin down the GB LE, KGB, (can I string more acronyms together?), then Hass should have enough time to make passing decisions.

Whether he makes the correct decisions is a valid question.

Split cheering for the weekend; doesn't Green Bay have a 2nd-year lineman, Colledge? He's one of 2 Alaskans in the NFL (hard not to cheer for them).

by Chris M (not verified) :: Tue, 01/08/2008 - 7:07pm

25: It's not iterative. I asked Aaron about this, and he said it's currently too complicated to do (although it's one of the future projects).

by turbohappy (not verified) :: Tue, 01/08/2008 - 7:09pm

Re: 23

Maybe Darrell Reid’s emergence will somehow Bob Sandersify the coverage units.

If a DT is the star of your coverage unit, that seems like more of a problem than a solution.

by BadgerT1000 (not verified) :: Tue, 01/08/2008 - 7:11pm

Post 42:

KGB only comes in on obvious passing downs. Cullen Jenkins, who is now healthy and rushing the passer much better, handles more than 50% of the snaps at the end position. Jenkins can both bull rush as well as do an inside dip of the shoulder or spin move to get to the qb.

by Karl Cuba (not verified) :: Tue, 01/08/2008 - 7:15pm

Apparently the niners have hired Mike Martz as offensive coordinator. I'm really not sure he's the best fit (our best receiver is a TE and he mighht ignore Gore) but he will probably at least get the niners offense moving.

by MattB (not verified) :: Tue, 01/08/2008 - 7:18pm

Interesting that the Chargers now have the best defense in the league, and the 5th best special teams (both by far the best of the teams left), yet their offense is the worse than any other team left in the playoffs. Who could have pictured this after what they did last year.

by John (not verified) :: Tue, 01/08/2008 - 7:20pm

One thing about KGB or Cullen Jenkins...Walter Jones is the LT. Remember that. He seems to be revived late in the year.

by Bobman (not verified) :: Tue, 01/08/2008 - 7:28pm

#36 MJK, I think you are very right about Morris and Colvin. Especially Colvin; I suspect the O can compensate most of the time for Morris's absense (duh? Ya think so, Bob?), but there is less wiggle room on D. Of course, last time I checked, it's not like it's a glaring weakness.

#44 But he's a crowd favorite with deceptive speed! Okay, maybe I am trying too hard to make lemonade out of lemons.

The crowd just can't chant "Luuuuke!" when he comes out, and "Daaaa-Relll!" makes it sound too much like a Mets/Yankees game, where you can't tell if they're cheering on or ridiculing Strawberry. Not sure how a 288 Lber gets to be a ST star, but he's also a Pro Bowl alternate at that slot, so it's not totally a fluke.

by BadgerT1000 (not verified) :: Tue, 01/08/2008 - 7:35pm


I am well aware of Walter Jones as well as the bulk of the Seattle roster. I follow Mikey's team with great interest.

I was merely pointing out the DE rotation.

by Purds (not verified) :: Tue, 01/08/2008 - 7:39pm

MJK (#36):

I think the loss of Colvin is important to NE's defensive WDVOA decline, but I think it's interesting to note that in 2004, when Indy's offense was going lights out, their defense dropped to something like #18. Now, the Colts have been hit or miss with defense the past decade, with no real trend, so it might mean nothing.

On the other hand, though, the loss of Colvin is just one guy, and while he's good, there have to be other reasons. Simply the age factor could be playing a role. Here is the NE defensive starters' age:

D line: 26, 26, 28
LB: 32, 38, 34, 25 (Alexander the 4th LB?)
DB's: 24, 26, 35, 24/27 (Sanders/Wilson)

That's certainly not an old defense across the board, but it's not a young group, either. I know the Colts are very young (too young), but here they are for comparison:

D line: 24, 29, 26, 26
LB: 27, 26, 25
DB's: 23, 26, 24, 24

In other words, the Colts have Brock (29) and Brackett (27), and then everyone else is 26 or under, and no one 30.

That all said, I don't think I'd worry about the defense's age if I were a Pats fan. That bye week is just the ticket, and they have the experience to cope with playoffs, etc. But, I do think age explains the regular season DVOA drop, but it's just not that big of a deal.

by Karl Cuba (not verified) :: Tue, 01/08/2008 - 7:49pm

Purds, who do you have at Lb for the Pats? I thought their four starters were Thomas, Vrabel, Bruschi and Seau.

by hwc (not verified) :: Tue, 01/08/2008 - 7:57pm


I don't think fans around the country truly appreciate that almost every game the Pats have played since the half-way mark has been their opponent's "Superbowl". Four of the last seven games were nationally televised night games that were just insane at places like Buffalo and Baltimore and New York.

In every one of these games, the opponents threw everything but the kitchen sink at the Patriots, basically figuring they had nothing to lose -- very aggressive defensive and offensive playcalling. Running teams came out throwing. Defenses blitzed the Pats 66 times during a single game. And, so forth.

Not that there is anything wrong with that, but the Pats have, in effect, been in playoff intensity games for two months now. An odd situation where one team is pumped and jacked like a playoff team and the other team is still in regular season, "just another week", "next game on the schedule" mode. A very unusual dynamic. And, no surprise that the Pats got the best game of the season from almost all of these opponents.

by patriotsgirl (not verified) :: Tue, 01/08/2008 - 7:58pm

52: From the ages, it looks like Purds included Vrabel, Seau, Bruschi, and Alexander, and didn't include the 30-year old Thomas.

by Toast Patterson (not verified) :: Tue, 01/08/2008 - 8:05pm

Toast -

It’s probably because the Chicago defense is good at preventing yards on a play-by-play basis, but they’re on the field for 45 minutes due to ‘less than ept’ offensive play. So the opponent yards and points come, but only at a clip of a few at a time .

:: Tundrapat — 1/8/2008 @ 4:33 pm

I agree that having a ridiculously bad offense keeps the D on the field longer, thus more yards and points but that still can't make up the difference between 28th in yards and 2nd in DVOA.

Also, they were 25th in yds/play giving up avg. of 5.5. Right between the Rams at 24th and the Bengals at 26th. Just wondering what they did so well that DVOA had them at #2?

by hwc (not verified) :: Tue, 01/08/2008 - 8:13pm

Purds, who do you have at Lb for the Pats? I thought their four starters were Thomas, Vrabel, Bruschi and Seau.

Those are the nominal starters, but the Pats have really mixed up their personnel groups a lot this year. They've play a bunch of 4-3, maybe even more than they have played 3-4. They've also played a lot of nickle with Rodney Harrison playing a hybrid safety/LB role.

They've also really tilted toward the relatively passive "bend but don't break" mode this season. Given the Pats offense, allowing the other team to make long, slow clock-eating drives is basically letting their opponent bleed themselves to death because they can't match the Pats points production.

It's just not smart football for the Pats to be aggressive on defense. Averaging 38 points a game, they don't really need to be. Stopping a running back after a four yard gain is almost as effective (in determining the outcome of the game) as taking more risks, shooting the gaps with blitzers, and trying for big negative plays. It's really, really hard for an opponent to put more than 38 points on the board, four yards at a crack.

I do think the Pats are missing one component of a really great defense: an athletic explosive play-maker in the middle of the field -- either at inside LB or an Ed Reed type safety. Rodney Harrison has played well this season, but he's no longer that guy...as he was the first few years in New England.

From a bigger picture standpoint, I think we may be seeing then demise of "shutdown" defenses. The current rules of the game (outlawing pass defense, permitting offensive holding, not being allowed to touch a QB) change the equation for a defense. The rules force defenses into softer zone coverages because no defensive coordinator is going to make a living trying to man cover athletic WRs and TEs with defenders that can't touch them downfield. That's a recipe for getting torched. Plus, more and more teams are taking advantage of the favorable rules to chuck the ball downfield (even if the QB heaves it with his eyes closed).

by Bobman (not verified) :: Tue, 01/08/2008 - 8:24pm

#53, While that's true, it's usually the treatment the SB champs get every year as well. The schedule-makers did neither team many favors in terms of prime time games (they each had around 6, including Indy's two Thursday nighters) and the late-season shifted Pats games.

What is amazing to me is that they both are this healthy right now. Those "other guys' SB" games have high potential for injury. Colts probably benefitted from (and I use that term loosely) their mid-season injury spate, as a lot of those guys sat out stretches in the season's 2nd half and are now considered to be healthy (knock wood). The Pats managed to avoid the injury bug most of the year (certainly compared to years past), probably because of a pact with the devil. (AHA! I had to step off the reasonableness train for a moment there.)

Related to Purds's point above, I think the theory on Indy's 2004 performance was that the O scored too fast and the D was back on the field too quickly. Doesn't have to happen all the time, but just a few 40 yarders a game can take away maybe 10-15 minutes of scattered rest time for the D players. (that's real time, not game time, and I suspect it adds up over the course of a game). While it's nice to score big and fast on the scoreboard, that kind of activity can stress a D. At least that was the thinking in Indy, and since 2005, they seemed to throttle down the O to make it more diverse, make drives last just a minute or two longer (game time, meaning about 5+ minutes real time), and the D has responded to that in 05 and 07. Clearly other factors play into this, such as opponents, personnel, scheme, etc, but an explosive O might expose some flaws in an otherwise good D, especially if a few guys are over 32 or so.
Sorry, not very concise.

by JV (not verified) :: Tue, 01/08/2008 - 8:24pm

An important point to remember about the Pats D is that it allows less points at home than on the road, by an 8.5 point margin (12.9 to 21.4). Indy, by contrast is almost evenly split (16.75 to 16), while Jacksonville allows almost 10 more points per game on the road than at home (around 24 to 13). This bodes well for this weekend's game and less well should the Colts come to Foxborough.

by John (not verified) :: Tue, 01/08/2008 - 8:25pm

I figured! More of a general comment that I think Blue Men Group can win that battle. Still, I'm a newcomer to the website, very interesting indeed! I read KC Joyner's column on ESPN.com, but am new to the 'sabermetric' approach to football. Very good stuff.

by MJK (not verified) :: Tue, 01/08/2008 - 8:32pm


I agree totally with the pace of the offense impacting the defense. That was my argument for years about why Manning and Harrison and Wayne were actually making their defense's life more difficult, and why I started to really get scared of the Colts once they discovered the 6-yard pass last year.

The long bombs to Randy Moss are pretty and fun for Pats fans, but I bet Bruschi and Seau, every now an then, internally are saying "Awww, come one, guys! We were just out there! Why'd you have to score a TD in 3 plays!"

by Tundrapat (not verified) :: Tue, 01/08/2008 - 8:36pm


Nice discussion about the league rules trend. And I think we're starting to see the results of that trend; for example, more money is being spent on athletic pass-rushers, since corners can only 'shut down' a wideout for so long. And I wonder if it's actually going to be detrimental to QBs (and therefore offenses) because the defense has to get so much faster to take down the QB.

To put it another way - I'd rather be dragged down to the ground by a 300-lb non-athletic tackle than get blindsided by a 270-lb DE or LB at 30 mph. One you get up from, the other you don't.

Anyhow - current trends (which I don't like) aside, the Pats are starting to look vulnerable at exactly the wrong time. Just ask those very dominant Bills teams. I think that, assuming the Pats survive what will probably be an extremely physical game against Jacksonville, they're going to face a fully healthy Colts squad that will beat them down.

by Alex (not verified) :: Tue, 01/08/2008 - 9:14pm

Without Morris, they have Faulk and Maroney, who seem to be more “boom or bust” runners who get either 10 yards, or none, and they have been more forced to rely on their passing game.

Actually, Maroney is 2nd in the NFL in success rate, so definitely not 'boom or bust'.

by Optimistic Packer Fan (not verified) :: Tue, 01/08/2008 - 9:20pm

40: It's been close to 50 for the last few days, so I'm just happy that something resembling normal weather is returning.

by randplaty (not verified) :: Tue, 01/08/2008 - 9:38pm

#23 Dr.Z doesn't know anything about the Chargers. I don't know if he even looked at a boxscore much less watched that game. Merriman had a great game with a key forced fumble in the redzone and also a key sack late in the game when Young was trying to bring his team back.

by BiteTheDust (not verified) :: Tue, 01/08/2008 - 9:45pm

While SD has better WDVOA with good margin,playoff odds shows IND 61.1%.
Considering HFA which is around 8%,it does not look right.And it is said that playoff odds was caliculated by using WDVOA and HFA.

by Pat (not verified) :: Tue, 01/08/2008 - 9:58pm

From a bigger picture standpoint, I think we may be seeing then demise of “shutdown” defenses. The current rules of the game (outlawing pass defense, permitting offensive holding, not being allowed to touch a QB) change the equation for a defense.

Most of those rules were there in the 80s as well - they've come and gone over time, especially the "reemphasis" which occurred in the late 1990s too. Really, only the change to roughing the passer is real. Remember, the NFL's scoring average is still well below what it was in the mid-1980s.

I'm not really sure I believe that an increase in zone coverage is extending from those changes - I think it's more money related. It's just cheaper to spend less on moderately fast defensive backs, and spend more on an elite pass rusher.

by Pat (not verified) :: Tue, 01/08/2008 - 9:59pm

#65: HFA is 17%, not 8%. It's +8.5% for the home team, -8.5% for the away team.

by Yinka Double Dare (not verified) :: Tue, 01/08/2008 - 10:23pm

Toast 55: Since it's WDVOA, they lose a few early bad games out of the calculations, and I think they had some of their better defensive performances of the season in the last few games on the schedule.

Their overall defensive DVOA ranking was still top 10 though, and one of the reasons is this stat that we heard trotted out repeatedly in broadcasts of Bears games -- the Bears led the league in forcing 3-and-outs. That would definitely be a contributing factor to their good DVOA.

by Nicky P (not verified) :: Tue, 01/08/2008 - 10:45pm

Bobman - 57 - "#53, While that’s true, it’s usually the treatment the SB champs get every year as well. The schedule-makers did neither team many favors in terms of prime time games (they each had around 6, including Indy’s two Thursday nighters) and the late-season shifted Pats games."

As a lifelong Pats fan, I have now seen what you speak of three different times. However, while it's true that the SB Champs definitely have a bullseye on their backs, I've never seen anything like what this year's Patriots went through in terms of opponents' intensity levels.

It's been a pretty wild ride.

by Jimmy (not verified) :: Tue, 01/08/2008 - 11:45pm

#55, Toast

If you take a look at the drive stats for the Bears they had 210 defensive drives the most in the league, by quite a way. The only other team over 190 defensive drives was Seattle with 204. The blame for this can probably be blamed on the offensse failing to sustain drives.

Also the weighted DVOA places heavier emphasis on recent games in which players like Urlacher, Briggs and Harris were more healthy and they were getting better play form Manning and McGowan at safety.

by Herm? (not verified) :: Wed, 01/09/2008 - 12:14am

Saturday, 8PM! Superbowl 41 3/4! Be There!
Then Sunday Sunday Sunday! 1PM! Superbowl 41 4/5¶²

CBS = no robots. I hatest the robots.

by Purds (not verified) :: Wed, 01/09/2008 - 1:39am

Sorry about the ages of the Pats LBers. I was thinking of a top LB being out with injury, and I was thinking Thomas was out, not Colvin. I took the lineups mostly from the NFL site, and they still list guys on IR in the depth charts (for example, they have Rob Morris as a starting LB for the Colts, and Morris has been out since the first quarter of the season).

So, the LB would be 30, 32, 34, 38, even older.

That said, I think the bye will solve any of age problems.

by Pat (not verified) :: Wed, 01/09/2008 - 2:44am

The blame for this can probably be blamed on the offensse failing to sustain drives.

Partly, but it's twofold. Strong defenses make games longer, and so they face more drives as well.

Really, the answer to #55 is simple: third down. Chicago is 2nd in the league at getting teams off the field on third down, behind only Kansas City. Kansas City's D is the "anti-Bears D", though - they started off strong, and collapsed - their Weighted DVOA is now 22nd.

Anyway, they're giving up yards each play, but their yards/drive are good (10th in the league) and their drive success rate is good (4th in the league).

(Hint: not a good sign for Chicago next year.)

by MdM (not verified) :: Wed, 01/09/2008 - 3:11am

"They’ve also really tilted toward the relatively passive “bend but don’t break” mode this season. Given the Pats offense, allowing the other team to make long, slow clock-eating drives is basically letting their opponent bleed themselves to death because they can’t match the Pats points production."

Making a team win by throwing a boatload of 4 yard passes may not be a bad idea, but can you prove that it is a lower risk strategy than playing "normal" sound defense? If you mean to concede long drives to the other team because there's always a chance they could get lucky and hit on a long pass, that would reduce the amount of offensive possessions for the Patriots and thus reduce their ability to get double digit leads.

What you suggest is the exact opposite of what a high powered offense would want to do. It's like saying the Phoenix Suns should play ball control, because "after all they're better shooters anyway and who can keep up with them". It doesn't compute.

by hwc (not verified) :: Wed, 01/09/2008 - 4:19am

RE: #74

I'm saying that there is a tradeoff in playing defense. Generally speaking, if you really, really want to stop an offense, you've got to make an aggressive play such as sending the blitz on 3rd and long. There's a cost associated with that, namely, leaving your secondary exposed. For the most part, the Pats don't like that risk. They preach that the DBs should keep everything in front of them. If they make a mistake, let the mistake be a 15 yard out pattern because the safety was too deep, rather than a ball that sails over the safety's head for a touchdown.

The same thing applies to defending the running game. Any good defense can stop a running game if they really want to. Just put eight in the box. But, again, that puts the secondary at risk.

You can see the mode shift in close Pats games where they go three quarters playing softer defense and allowing the other team to move the ball (and maybe score). Then, all of the sudden at crunch time in the fourth quarter, the Pats get aggressive, start blitzing, etc.

This is nothing new. It's the Belichick approach to defense. He's been doing it since he's been in New England. Especially with an offense that can score points, why take the risk until the game is on the line? Let 'em have a 100 yard game rushing the ball. It takes more than 100 yards rushing to match 38 points.

by Bobman (not verified) :: Wed, 01/09/2008 - 4:24am

#74 MdM You BOTH have a good point but what I think you may be missing is that the D does not want to give up 80 yard TD drives 3.3 yards at a time. They want 3 and outs or 5 and outs, etc. Worst case scenario, if their plan works out, is a 14-play TD drive for the other guys.

This is the same for Indy's D this year and they have an explosive O as well. Their Pass D is ranked in the top 5, but is near last in terms of completion % against, yet #2 in average yards per pass against. Think about that--roughly speaking, they allow the most completions but the fewest yards. Is that the essence of bend-don't-break? Basically, you will not scorch us (unless you are named Moss) and you will earn it IF you get it is their message. Part of that might be their opponents taking what they can get--maybe if the other guys challenged more on deep balls there would be more big gainers agianst Indy. Or INTs. Don't know.

They certainly don't give the yardage up and are 2nd in the league in turnovers, so they're pretty aggressive. But they keep the play in front of them and don't risk diving for the INT or making the highlight reel blow-up tackle only to give up a long gainer when they whiff.

by thestar5 (not verified) :: Wed, 01/09/2008 - 4:46am


Yeah, and it also takes more than a long pass or two to score 38 points. You're not really making much sense. Its in the best interest of the Pats to stop the other team, not let them have long, clock killing drives. Afterall, most every other team is trying to shorten the game versus the Pats, which is exactly what clock killing drives does.

And I've really said this enough already but the reason I think Dallas was trending down is because they had many of their injuries near the end of the season, as opposed to sat the Colts. As long as everyone is healthy by this Sunday I would really be shocked if they weren't back to normal. Of course, if Romo, TO, and co. are still hurt than all bets are off.

by hwc (not verified) :: Wed, 01/09/2008 - 5:43am

RE: #77

The part of the equation you are missing is that it is very difficult for most NFL offenses to execute sustained long drives. Think how many times you've seen an offense moving the ball, only to stop themselves with a false start, or a holding call, or a dropped pass. Generally speaking, executing a long, sustained drive involves converting third downs. It's tough. Most NFL offenses make mistakes.

I'm not making this stuff up. Belichick has talked about it on many occasions. His mantra to his defense is "live to play another down".

He has also talked about the concept of "defending the entire field". When the opposing offense is on their own 30 yard line, the defense has to defend the area from the line of scrimmage to roughly 50 yards downfield. As the offense nears the red zone, that territory is reduced by the goal line and the defense automatically gets more aggressive (i.e. moves closer to the line of scrimmage) because it no longer has to defend 50 yards.


Absolutely right. Even though the Pats and Colts play completely different defensive schemes with totally different personnel, the underlying philosophy is identical. Keep the play in front of the defense. Live to play another down. Make the offense earn everthing they get.

BTW, on the flip side, Belichick WANTS Brady to throw the ball downfield. He wants him to take shots to Moss. Why? Because he knows that it is frickin' hard for his offense to score, four yards at a crack. Ultimately, putting up points in the NFL requires getting the ball downfield in some occasional big chunks.

by TracingError (not verified) :: Wed, 01/09/2008 - 10:28am

Aaron, please explain this to me:

You say that overall DVOA correlates better with playoff success than weighted, yet you present weighted come playoff time. Why not just continue as usual with weighted and overall DVOA?

I suspect the reason for that result, by the way, is that the best teams, who have the highest playoff success, often have abnormal results at the end of the season as they rest starters, keep wrinkles in the game plan under wraps for the playoffs, and so forth. So why not let us see the overall DVOA?

Also, I for one would like to see VOA for playoff games it order to ascertain who should have won, rather than, or in addition to, DVOA.


by Jake (not verified) :: Wed, 01/09/2008 - 11:37am

#78 and earlier:
I've been arguing the same thing on this site for years. Offensive DVOA correlates well with points per game. Defensive DVOA correlates much more strongly with yards/game. For the most part thats fine, but with some teams a conscious strategic philosophy means that defensive DVOA underrates that team.
I've made the argument too many times to repeat it, but look at the historical numbers. This regime's Patriots team has a 'bend-but-don't-break' mindset. This leads to more yards given up but fewer points. The counter-argument is that this approach is inherently bad and DVOA correctly punishes the Patriots and other like minded teams for it. The 2003 Patriots were the #1 scoring D by a wide margin but #2 DVOA defense to the #6 scoring D Baltimore who allowed almost 3 more points a game. That year the Patriots also led in interceptions, defensive scores and had more FF than Baltimore. Yet the Ravens were the #1 D by 6%, adjusted or not and two other teams were within 3% of New England. A similar but less blatant situation arose in 04 (#2 scoring, #6 DVOA), in 06 (#2 scoring, #8 DVOA) and even in 01 (#6 and #12). Its a fairly consistent thing that can't be explained by lack of turnovers, or defensive scores.
However, scoring defense still correlates more strongly to success than (defensive) DVOA. It seems a philosophical mantra is creating a resistance to the idea that the bend-but-don't-break defense not only exists but can be an effective philosophy and that DVOA should be adjusted for this.

by V4Velvetta (not verified) :: Wed, 01/09/2008 - 12:04pm

I bow to the always insightful hwc (#75). The Pats do have the ability to shift into reverse Miller Time...meaning turning it up a notch at crunch time. And the playoffs are nothing but crunch time. The casual fan may believe that the Pats always come into the playoffs with vaunted defenses, but those of us who know better know that they have always come into the playoffs with a serious weakness. (I believe they faced the high-flying Manning machine of 2004 with a defensive backfield of Obama, McCain, Huckabee and Hillary). But they use the by-week to make the necessary adjustments. This year I think the adjustments will hinge on two things. The health of Seymour, which appears to be back and will be significant. The use of A. Thomas. I've just finished watching replays of the entire season and I swear Thomas arrives at 80% of the plays a half-step too late. Either he has been grossly overrated, or the Pats never figured out how best to use him. I'm guessing it's the latter and they'll have it fixed by Saturday.

by Vern (not verified) :: Wed, 01/09/2008 - 12:25pm

Re: 79

Simpler than that. More data points are better than fewer. Total DVOA is a truer measure of what a team really has than less, no matter how you filter more down to less (late season games, etc.)

Re: bend but don't break and similar

DVOA is really a measure of a team's "innate" abilities - how likely they are to succeed in all the situations that MAY be required to win the game. While it takes into account down-and-distance, it really doesn't take into account game strategy, primarily score-and-clock situations(except in extreme and obvious situations like FG set-ups or end-game downs). A team with a game plan to allow the other guys to run all day from the first snap, for example, would not be accounted for at all. We all know football play selection is hugely based on things like leading, trailing, close, big lead, etc. Field position is another category that is weakly if at all covered by DVOA (getting 4 yards on 3rd and 5 when you're in 4 down territory is a GOOD thing).

Having a higher DVOA is thus like having more or better pieces left on your side of the chess board than the other guy. But the object of chess is not to have the most pieces but checkmate. The greater "innate" strength - more pieces - can easily be overcome by the strategy of the game, and may often be a product of that strategy.

by MattB (not verified) :: Wed, 01/09/2008 - 1:25pm

On the Pat's defense. I think that the "bend don't break" folks are ignoring the fact that too often recently, the Pat's defense actually has broken. Look at the numbers from recent games:

Week 17: Defense gave up 35 points to the Giants (19th ranked -2.5% DVOA offense) in a close game
Week 16: Defense gave up 7 points to Miami (23rd ranked -8.8% DVOA) while Cleo Lemon was playing
Week 15: Defense gave up 10 points to the Jets (25th ranked -10.8% DVOA) in bad weather
Week 14: In the first half vs Pittsburg the Pats gave up 13 points and Pittsburgh (12th ranked 6.8% DVOA) was driving the ball rather well and the score at the half was 17 to 13. In the second half the defense stepped it up and Pittsburg didn't score again.
Week 13: Defense gave up 24 points to Baltimore (26th ranked -12.8% DVOA offense) in a close game
Week 12: Defense gave up 28 points to Philly (8th ranked 10.2% DVOA offense, but playing with backup QB) in a close game

So in the last 6 weeks the Pats have faced the 19th, 23rd, 25th, 12th, 26th and 8th ranked offenses and given up an average of 19.5 points per game. The only teams they have really stopped are 23rd ranked Miami playing a backup, the 25th ranked Jets and 12th ranked Pittsburg for the second half of that game.

Now, granted all of the points in the above games weren't put up against the Pats offense. Some were probably put up on turnovers or vs special teams, but still the main point here is that the WDVOA for the Pat's defense looks correct, even if you look at the numbers through the "bend don't break" glasses.

by B (not verified) :: Wed, 01/09/2008 - 1:28pm

80: The reason the Pats were better in scoring defense than the Ravens despite being worse by DVOA is because they had a better offense, so their defense wasn't left in bad situations.
Say, for example, McNair turns the ball over on his own 20 yard line, the Ravens defense allows no yards, and the opposing offense kicks a field goal. By DVOA and yards, that's a great job by the defense, but by scoring defense, it's not. The Patriots (Colts) have a great defense when tied to their great offense, but if you took the offense away, their defense wouldn't look as good. But we know they have a great offense already, and there's no need to give the defense credit for being associated with a great offense.

by MdM (not verified) :: Wed, 01/09/2008 - 1:56pm

RE: Pat's D

I try, but can't imagine that the bend-not-break D can be something the Pats and Colts willingly choose to do. If they have shifted their philosophy in that direction, why? Is it to cover for an extreme weakness in some part of their D?

Again, the penalties to this style of play are huge. Not only are you aiding other teams in shortening the game, an axiomatic way of playing against superior opponents, you are exposing your D to fatigue. Also, there is the fact that every extra snap is another chance for the other team to get lucky and breat a long one.

I can see that perhaps the O might want to slow down a little to give their D a chance to rest, but I have a hard time seeing why the D would willingly prolong their time on the field--or does Randy Moss get too tired running post patterns!

by B (not verified) :: Wed, 01/09/2008 - 2:07pm

85: The answer is pretty simple, Salary Cap. Both have a substantial investment in offense, and are in the position where adding players to offense has a better cost/benefit ratio than adding players to defense. I.E, a receiver will have the benefit of a top-teir QB, a running back won't face 8 man fronts. Also, recent rule changes/emphasis have placed a higher value on offense than defense. It's not that they won't pay for top-teir talent on defense, just they'll only pay for it in select positions (A. Thomas/Seymour on the Patriots, Sanders and Feeney on the Colts).

by Brian G. (not verified) :: Wed, 01/09/2008 - 2:45pm

Actually, I think the real reason for the drop-off in the Colts defense in 2004 is because anytime the Colts get any sort of substantial lead, Dungy will put 'em in the prevent, which allows teams to get big chunks of yardage up the middle, but eats the clock and shortens the game. I do agree that the Pats D declined primarily due to injuries and the fact that everyone played the like it was the Super Bowl.
What I would be worried about if I were a Pats fan is that, since the Bills game(the last huge blowout), they struggled against generally lesser opponents, playing only 2 playoff teams, one of which was clearly a fraud (PIT). As noted earlier, yes they were facing the 'kitchen sink' from everyone, but does anyone think JAX/IND/SD won't also throw their own kitchen sink? Every game from here on out IS the Super Bowl. I'm not saying NE will lose, or that they can't handle it b/c so far they have, but they are about to face far better 'kitchen sinks' than they did in the 2nd half of the season and if I were a Pats fan, I'd be concerned.
(As a Colts fan, my concern lies with the lack of pressure from the D-Line post-Freeney, which I think may be our downfall).

by Nicky P (not verified) :: Wed, 01/09/2008 - 3:13pm


You make some good points, but what you fail to mention is the Patriots' own kitchen sink. I think it's safe to say that, especially defensively, the Pats have played relatively conservatively. I'm pretty confident that we have not seen everything the Pats can bring to the table.

Also, if my memory serves me correctly, nobody was calling Pittsburgh a fraud prior to that beatdown they received in Foxboro (aside from Pats fans, of course). Most fans and critics kept pointing to the Steelers game as the one the Pats would lose.

And one question: How many playoff teams did the Colts play that weren't frauds? They played Jax and Tenn twice, TB, the Pats and SD. If we're considering one and done in the playoffs to = "frauds," that would mean the only legit playoff teams the Colts have played are Jax, NE and SD.

by Jake (not verified) :: Wed, 01/09/2008 - 3:16pm

84: False. The stats are available here at FO. The Ravens were in the better field position per defensive drive in 04, 03 and 06.
85: Its about discipline and philosophy for BB defenses. It reduces variance for one in that you're less likely to give up a big gain. The longer you're on the field, the most likely you are to make a mistake. Few offenses can consistently put together 15 play drives. Add to that a focus on whatever the other team does best - with a bias towards focusing on the run - and the other offense will have to use its weaker component consistently in order to keep a drive alive long enough to score. They're not going to blitz or get real risky with a long field. Instead, they'll frustrate you by taking away an explosiveness you have.
The offense has always reflected this philosophy too. Its not a big play offense (until this year). Its a death by a thousand cuts offense. By playing defense the way they do in combination with a highly efficient offense, they shorten the game in terms of numbers of drives. Shortening the game isn't less legitimate than lengthening it. Now the other team has less chances to get lucky.

by Nathan (not verified) :: Wed, 01/09/2008 - 3:32pm

Bob Sanders doesn't read offenses. He just stares them down until they give up the information he wants.

by Nathan (not verified) :: Wed, 01/09/2008 - 3:33pm

Indy's Defense is called the Cover 2 because the team is only responsible for covering 2 people...Bob Sanders covers the rest.

by nat (not verified) :: Wed, 01/09/2008 - 3:55pm

87: Will JAX, IND, or SD throw their kitchen sinks at the Patriots (Assuming the Patriots make it past JAX)?

I'm not so sure they will. The "blue print" from the Eagles etc was basically a desperate one: make the game as unpredictable as you can, and hope the breaks/officiating fall your way. But each of JAX, IND, and SD have reasons to believe they are not too far below the Patriots. I think pride will keep them from kicking onsides multiple times, or abandoning the play-mix that got them here for high-risk strategies.

Each of them will prefer to force the Patriots to beat them at their 'A' game, rather than to lose through unorthodox play calling.

In a way, it's too bad. I think teams should kick onsides, go for fourth down plays, run trick plays, etc, far more than they do.

by erik fast (not verified) :: Wed, 01/09/2008 - 4:23pm

RE: PATS D & Running Game.

People who don't watch all Patriots games seem to be under the mis-impression that their D is weak. Yes it is weaker than their offense,but that is a comparative thing. Likewise, their running game is weaker than their passing game, but it is also a comparative thing.

1) Their defense is #4 in points allowed. Yes the "bend but don't break", but who cares. They always seem to step up with the big stop (or turnover) when the game is hanging in the balance. This is due to not taking big risks to may give up the big play until they need to. If you could do a DVOA for specific scenarios (like 4th Qtr defense when the game is +/- 2 scores) I have a feeling the PATS D would be near the top of the list.

2) They don't run often because they don't generally need to. But, when they need to, they are generally successful. They average 4.1 yds/rush. while that is only #14 in the NFL, it is more than enough to to keep the chains moving. By the way, Jacksonville has the highest yds/rush left in the playoffs at 4.6 yds/rush - only .5 more..so the difference isn't much.

by Wanker79 (not verified) :: Wed, 01/09/2008 - 4:52pm

NOTICE: The following post isn't meant to point out any deficiencies in the Patriots or Colts. It is only meant to refute the premise that DVOA misses something about the quality of a "bend-but-don't-break" defense.

The goal of every defense is to get the ball back to the offense as quickly as possible while surrendering as few yards and points as possible. Every yard or every point a defense gives up is another yard or point their offense has to gain (not quite literally, but you know what I mean). If NE or Indy had defenses capable of allowing the opposing offense even fewer sustained drives resulting in fewer yards and fewer points, they'd be better teams. Given a high-powered offense, it may be the optimal scheme to disguise a relative weakness, but "bend-but-don't-break" is not an optimal defensive scheme regardless of whether or not you have an offense good enough to rationalize it.

I think what clouds this discussion is the return-on-investment that someone brought up. If NE or Indy spend more on defense, they'd have less to spend on offense. So while they could potentially improve the defense, overall it could make leave them worse off since they don't require incredibly good defense. But just because "bend-but-don't-break" may be the best strategy to get the most out of your talent, it doesn't make them a better defense.

by Andrew (not verified) :: Wed, 01/09/2008 - 4:57pm

The huge jump in Chicago D ranking is due to using weighted DVOA instead of regular DVOA.

Week 17 Chicago D in regular DVOA is -6.9% and rated 7th.
Week 16 is -5.6% and rated 10th.
Week 15 is -2.3% and rated 13th.
Week 14 is -2.0% and rated 14th.

Falling off the chart on the other end during this period in weighted DVOA were stinkers like the 34-10 loss to Dallas, the 37-27 loss to Detroit and the 34-31 loss to Minnesota, while gaining on the chart were the Green Bay and Saints games, where the D did well against good offenses.

The Chicago D being good is predicated on being good against #1 WR's and the run. Sucking against everything else was not much of a formula for success though.

by Purds (not verified) :: Wed, 01/09/2008 - 5:10pm

NickyP: "Most fans and critics kept pointing to the Steelers game as the one the Pats would lose."

I have no problem with much of what you've offered in your post, except the above statement. Critics pointed to the Steelers game as the one most likely to be a loss, not the one the Pats would lose. Big difference. The other teams down the stretch for NE? Baltimore, Jets, Miami, NYG. It wasn't that people expected NE to lose to Pitt, but that if they were going to lose any game, it was most likely against Pitt.

by Yinka Double Dare (not verified) :: Wed, 01/09/2008 - 5:28pm

95: It would have really helped them a few years ago though when they were beaten in the playoffs almost entirely by a #1 receiver...

by Nicky P (not verified) :: Wed, 01/09/2008 - 5:37pm

Purds - agreed. It was the game people said they would most likely lose.

I guess my point was that there weren't many people out there calling Pitt "frauds" before that Pats game.

In fact, I "guarantee" it.

by Travis (not verified) :: Wed, 01/09/2008 - 6:13pm

Re: 98

In view of this performance, there have to be grave doubts about the Steelers the rest of the way. The rosy prediction, here and elsewhere, about the team, look grossly unfounded today. It was one thing to lose eight days earlier to the one-win New York Jets. Those kinds of upsets, particularly on the road, are fairly standard in the NFL. It's quite another to play so poorly at home against a team that has been wrecked by injuries and without a win.

Pittsburgh Post Gazette, November 27, following the Steelers' ugly (albeit rain-soaked) 3-0 win over Miami. They did beat the 4-7 Bengals by two touchdowns the next week, however.

by MJK (not verified) :: Wed, 01/09/2008 - 6:15pm


Not that I disagree with you, but your numbers need to be corrected slightly, if we're just talking about the Pats defense:

Week 17: The Pats D only gave up 28 points (the other TD was a runback after a stupid penalty moved the kickoff)
Week 16: Correct, they gave up a TD to Miami (when leading 28-0 late in the 3rd quarter, already playing defensive backups, if I recall correctly).
Week 15: They gave up only 3 points to the Jets (the other TD was a blocked punt recovered by the Jets in the endzone).
Week 14: Correct, they gave up 13.
Week 13: Correct, they gave up 24.
Week 12: Correct, they gave up 28.

So, neglecting points given up through special teams, and omitting the essentially garbage time TD against Miami, the Pats DEFENSE was giving up 16 points per game on average, not 19.5 And if you arbitrarily cut it at Week 11, instead of Week 12 (i.e. you include all the Pats games since their bye), when they gave up 10 points to Buffalo, then they averaged giving up only about 15 ppg.

Yes, that's not elite, and yes, they were playing some horrid offenses. But it's not quite as bleak as you were painting.

by MJK (not verified) :: Wed, 01/09/2008 - 6:21pm


From my last post, I just noticed that the Patriots defense played fantastic against teams in the AFC East (giving up a garbage time 7, 3, and 10 points to Miami, the Jets, and Buffalo), good to mediocre to AFC North teams (13 to Pittsburg, 24 to Baltimore), and badly against NFC East teams (28 to Philly and the Giants).

In other words, they played well defensively against teams that they play a lot, average against teams that they play a moderate amount, and poorly against teams that they hardly ever play.

This lends credence to the idea that Belichick is very good at designing a scheme to beat someone if he knows them well, and not as good if he doesn't. (Not that that's news).

Granted, this is a small sample size, but following this logic out that would put the Jags in the 13-24 point range. I wonder if the Pats offense can score 13-24 points, plus or minus whatever special teams points affect things, this weekend...

by Scott (not verified) :: Wed, 01/09/2008 - 6:52pm

Interesting pattern I've picked up on...

In the last 4 meetings between SD and IND, the Chargers have blown (or nearly blown) leads of 9, 15, 16 and 23 points.

1999 - Chargers led 19-10 in the 3rd and 19-13 to start the 4th. Manning ran for a TD and threw another in the 4th quarter to win the game 27-19. Marvin Harrison had 13 catches for 196 yards and a TD. Manning went 29/54 for 404 yards, 2 TDs, 1 INT, 1 rush TD.

2004 - Everyone probably remembers this game for the record breaking TD pass. The Chargers did lead 31-16 to start the 4th quarter, but choked up another game as Manning converted a 4th & 4 to Reggie Wayne to keep the drive going and later hit Stokley for his 49th TD of the season. Edge ran in the 2pt-conversion and the Colts added the FG in OT for a 34-31 win. Manning went 27/44 for 383 yards, 2 TDs, 1 INT. Stokley and Harrison both over 100 yards.

2005 - And everyone will remember this game for Michael Turner busting an 83 yard TD in the 4th quarter to make the Colts fall to 13-1. But even this game was crazy, since the Chargers led 16-0 in the 2nd half before quickly giving up 17 pts and the lead to Indy. Manning went 26/45 for 336 yards, 1 TD, 2 INTs. Harrison had 135 yards, and it was one of the most epic regular season games in the past decade.

2007 - Probably the craziest one of them all. 6 Manning interceptions, 2 return TDs by Sproles, 2 missed FG's by Vinatieri, and a 23-0 lead that nearly didn't hold. Manning was 34/56 for 328 yards, 2 TDs, 6 INTs. Reggie Wayne had 140 yards.

I don't know if we'll get a 5th straight crazy game or not, but that's pretty interesting. 4 straight games where Manning had to throw 44+ passes, over 328 yards every game, a receiver over 100 every game, and a multiple score deficit overcome or damn near overcome (thanks, Vinatieri).

The Chargers must be doing something right when Indy is on the schedule.

I think I'd still go Indy by a score of 35-17 though.

by B (not verified) :: Wed, 01/09/2008 - 7:06pm

98: For the record, after the Steelers lost to the Jets, Pittsburgh was being called "frauds" by many sources, not just Patriots fans.

by PatsFan (not verified) :: Wed, 01/09/2008 - 8:35pm

Speaking of frauds...

Matt "Turnstile" Light in the Pro Bowl - fraud.

Matt "Turnstile" Light on the All-Pro Team - unforgivable fraud.

by Nicky P (not verified) :: Wed, 01/09/2008 - 10:01pm

103 - so losing in OT, on the road, to a bad team - who's coming off a bye - makes for instant fraud status?

As a Pats fan, I watched them lose in 2003, on the road, to a soon-to-be 5-11 Redskins team, then rip off 15 in a row to win the SB. The following year, I sat in horror as the Pats threw one away in Miami, only to finish the season by winning five in a row and putting the finishing touches on their back-to-back SB run.

Any Given Sunday right?

Losing to a bad team can happen to any team, any time in today's NFL. It doesn't necessarily turn the team into frauds. Pittsburgh's loss to the Jets may have knocked them down a notch leading up to that Pats game, but they were not viewed as frauds - not by fans, "experts," or DVOA (I believe Pitt was #1 in Defensive DVOA prior to that game - I could be wrong on that).

by hwc (not verified) :: Wed, 01/09/2008 - 11:05pm

a) Have no fear. No Pats fans this year are under the illusion that the Pats defense will be pitching any shutouts in the postseason.

b) On this bend, but not break thing. Of course, Belichick would rather have a three and out everytime the opponents get the ball. The question is: "At what cost?". For example, if you really wanted to maximize three and outs, you would have your CBs jump every route and go for the INT rather than sitting back with a little cushion. But, that kind of aggressive play carries a risk -- a seven point risk to be exact. An eight yard reception, even a twenty yard reception, does not put points on the board. The defense lives to play another down. Getting beat over the top puts points on the board. Belichick hammers his DBs that, if they are every confused or having ANY doubt about their assignment on a play, make the mistake of playing too deep. The odds say that, if you limit your offensive turnovers and don't give up big plays on defense or special teams, you've got good chance of winning the game.

by Pat (not verified) :: Thu, 01/10/2008 - 12:22am

b) On this bend, but not break thing. Of course, Belichick would rather have a three and out everytime the opponents get the ball.

Then can we please stop calling it "bend but don't break?" It's a silly name. The coordinator doesn't say "oh, FINE, we'll give up 50 yards and then be forced to stop them ten yards away from the end zone." 'Bend but don't break' happens when two out of the three components (D-line, linebackers, secondary) of a defense are solid, and the third *!^(!ing sucks.

When the field compresses in the red zone, the other two components make up for the third, and poof! 'Bend but don't break.' It's not a "choice." It's a facet of a defense with a liability.

For example, if you really wanted to maximize three and outs, you would have your CBs jump every route and go for the INT rather than sitting back with a little cushion.

What?!? What kind of wacko defense are you running? That's a recipe for disaster, not a recipe for three and out. You want constant three and outs, you have cornerbacks who can stick with a receiver and not get fooled by double moves.

Cornerbacks jumping routes are situations where the head coach says "great play, but I sure as hell hope you knew what route he was taking."

Now, what you might be saying is, "if you don't have corners who can stick with a receiver easily, it's better to have the receiver stay in front of them," you're dead on right. But that's an acknowledgement that your corners *!^(!ing suck. I'm exaggerating, of course. But you get my point.

by Yaguar (not verified) :: Thu, 01/10/2008 - 10:40am

104: Light is the third-worst player on the Patriots offense behind Kaczur and Stallworth. I don't get it either.

by Starshatterer (not verified) :: Thu, 01/10/2008 - 11:11am

Yaguar (#108 ) --

To be fair, I'd put Light ahead of the Patriots' right guard platoon, since he's actually on the field (ahead of Neal) and at least identifies and tries to block the defender running past him (ahead of Hochstein). He's probably at least even with the collection of missed games and dropped passes the Patriots have at tight end, too. So that puts him behind Brady, Moss, Welker, Koppen, Mankins, and Maroney/Faulk (assuming the Patriots are in their bread-and-butter 3WR/TE/RB set). It's no great shame to be worse at your position than those guys.

Now, I don't think he's an All-Pro tackle, either, but he's probably average at worst, and durable compared to a lot of other guys who are usually rated better.

by PatsFan (not verified) :: Thu, 01/10/2008 - 11:32am

Re: #107

I don't think it's that the other components make up for the third, I think it's that on the compressed field, the third component can function better because the compressed field hides more of the third component's flaws.

by Andrew (not verified) :: Thu, 01/10/2008 - 11:59am

MJK #101:

Extend your thinking a bit.

NFC East - 28, 28, 27, 7 (cowardly Redskins)
AFC Central + SOS - 13, 24, 20, 17, 13, 14
AFC East - 7, 10, 10, 28, 7, 14

Essentially, ignoring outliers like the Redskins 7 and the Dolphins 28 (with its Rusher McFumbles participation) its:

NFC East - ~4 touchdowns
AFC Central + SOS - ~2-3 touchdowns
AFC East - ~1-2 touchdowns

From this perspective, the Patriots are at a small risk of losing if something bad like a return TD against them happens in an AFC game because other AFC teams haven't shown they can score over 20-24 points against them. They appear to be at a much bigger risk in the Super Bowl, given that they relied upon a return TD to beat the Eagles and surrendered return TD's against the Giants and Cowboys, while all three teams scored 27-28 points on offense and gave the Patriots a very close game.

by Jake (not verified) :: Thu, 01/10/2008 - 12:07pm

#107: We've had this argument before. You've never been willing to give an explanation on why the offensive DVOA correlates well with PPG, but defensive DVOA doesn't for certainly styles of defense.
Points allowed is a better measure of past success and predictor for future defensive success than defensive DVOA. The FO staff don't deny this, Pythagorean wins still work better for an overall team than DVOA. Defensive DVOA is the inaccurate one in this case (DVOA has many advantages to traditional stats of course, especially the ability to break it down to components).
Your next response is generally that a defense that allows no points but does so slowly hurts the offense, but this is insufficient to explain the discrepancy.
Taking New England (and Indy which I don't think fits the lack of correlation most years at all) out of the equation lets look at the top defensive DVOA this year.
Tenn - #1 D DVOA, 5th y/game, 12th pts/drive, 11th TO/drive, 8th points, 4th in 1st/game, 5th y/p, 24th 3rd down %
In other words, a top 5-ish team in terms of yards, top third in terms of points but certainly not the best defense in the league.
It maybe in 3rd down % that the issue arises but I'm not confident in my memory of explanations of the FO outlook on 3rd down % (IIRC it reduces its importance greatly as semi-random?) to say more on that yet.

by Nick (not verified) :: Thu, 01/10/2008 - 1:30pm

#109: Um, Stephen Neal isn't part of a platoon. He's been injured for much of the year. When healthy he's their second or third best on the O line.

by Starshatterer (not verified) :: Thu, 01/10/2008 - 3:56pm

Nick (#113 )--

You've just said why he's in a platoon, rather than refuted it. Any player actually on the field is performing better than the guy who's inactive half the season due to injury, which is why Light is better than Neal.

by Pat (not verified) :: Thu, 01/10/2008 - 6:41pm

You’ve never been willing to give an explanation on why the offensive DVOA correlates well with PPG, but defensive DVOA doesn’t for certainly styles of defense.

Because defensive points allowed aren't a good measure of a defense. A defense's job is not to stop points. A defense's job is to stop drives.

It's just semantics, really. If you want to measure a team's defense as "points per game," fine, but you're not really measuring the 11 guys on the defense, as offensive yards correlate inversely to defensive points per game.

Your next response is generally that a defense that allows no points but does so slowly hurts the offense, but this is insufficient to explain the discrepancy.

Huh? Adjusting defensive points allowed by average starting defensive field position, opponent, and correcting for the negative effect on average offensive starting field position basically looks like DVOA, with a few exceptions that all looked like teams with crappy special teams.

In other words, a top 5-ish team in terms of yards, top third in terms of points but certainly not the best defense in the league.

Tennessee VOA: -7.2%, behind PIT, SEA, TB, IND, SD. It's opponent adjustments.

by Pat (not verified) :: Thu, 01/10/2008 - 8:37pm

Whoops, I forgot to mention: adjusting points per drive, not points per game. Points per game is biased by the offense's pace. Crappy offense = longer game = more points per game.

Easy enough to see, just compare the average points per game for 10 drive games versus the average points per game for 15 drive games. It's almost 50% higher, which suggests (gasp) that the fundamental pace-free scoring stat in the NFL is the drive, not the game.

by Matt Saracen - QB1 - Dillon Panthers (not verified) :: Sun, 01/13/2008 - 9:28pm

Sorry for the late post everyone, but I had to type this. If Light, Kaczur and RG of the day all suck, how the hell does Tom Brady seem to always have a good 4-5 seconds of pocket time?? Sure teams often rush only 3-4 against the Pats and put everyone in coverage, but still you need linemen of some competency to keep the pocket he has enjoyed this year. I've never seen a guy have so much time...ever.