Writers of Pro Football Prospectus 2008

Most Recent FO Features


» Defense and Rest Time

Do defenses really wear out over the course of a game? Do defenses benefit from long drives that give them more time to rest on the sideline? Guest columnist Ben Baldwin investigates.

28 Sep 2010

Week 3 DVOA Ratings

by Aaron Schatz

Green Bay remains on top of the Football Outsiders ratings after three weeks, despite last night's close 20-17 loss to Chicago. The Packers' first two victories were big enough to keep a below-average game from knocking them out of the top slot. Before applying opponent adjustments -- those won't start until next week -- last night's game was the rare contest where both teams ended up with a rating below 0%. All of those False Start penalties result in a penalty for one team, but they don't boost the rating of the defense involved. The same goes for the two missed field goals (yes, that includes the one that was blocked, since blocking field goals isn't generally a predictive event).

That "double-negative" game is a big reason why Chicago is a surprising 14th this week despite starting 3-0. The Bears may have three wins, but all three have been very close. They certainly haven't been as dominating as a 2-1 team like Atlanta or Philadelphia. The flip-side equivalent to Chicago would be the Cleveland Browns, who rank 19th despite three losses. Those losses were all close, and Cleveland's single-game ratings are, in order -1.9%, -3.5%, and -24.6%. Another team that has been more impressive than its record is the Dallas Cowboys, who rank in the top ten even though they didn't win until this past week.

All of these ratings, of course, are still somewhat shaky because it is too early to include opponent adjustments. The 3-0 Kansas City Chiefs are a good example of this problem. When we predicted before the season that Kansas City would win the AFC West, it was more a prediction of San Diego's decline. We expected the Chiefs to sneak into the playoffs at 9-7. We did not expect them to start the season 3-0, including a triumphant 31-3 thrashing of San Francisco this week. According to our playoff odds report, the Chiefs now have the best chance to make the postseason of any team in the league: 93.9 percent of all simulations.

Now, it looks like Kansas City has built its good early record against bad teams, as its three opponents are a combined 1-8. But one of those opponents is San Diego, a team most observers expect to improve -- and the Chargers' two losses have been very close. Another of those opponents is Cleveland, which, as noted in the paragraph before this one, has played its opponents close so far. So has Kansas City played three bad teams? Or two average teams and one bad team? Or will the Chargers win eventually qualify as a win over a good team? It's just too early to tell.

San Diego's position in the rankings is even more questionable given how the Chargers managed to lose this week. The Chargers were almost entirely beaten by one man: Leon Washington. Seattle's performance this week qualifies as the best special teams game of the DVOA Era, and after looking at historical numbers, it's reasonable to state that Washington had the best kickoff return game in NFL history.

Back in Pro Football Prospectus 2006, we ran an essay in the New York Giants chapter discussing the best and worst special teams games in DVOA history. At the time, the best special teams game belonged to the 2002 New Orleans Saints, who were 16.3 estimated points worth of field position above average in a Week 6 win over Washington. (Click here for an explanation of how we turn field position on kicks and punts into an estimated point total.)

Right after we wrote that book, three of the best return men in history showed up in the NFL: Joshua Cribbs, Devin Hester, and Leon Washington. Cribbs has probably been the best of the three overall, but he's the one who has never had that one specific record-breaking superlative day. Hester was the main engine behind Chicago breaking the record in Week 12 of 2007, with 16.6 points worth of estimated field position. You may remember that as the "Why on earth is Todd Sauerbrun still kicking to Hester?" game, where Hester had two return touchdowns in the third quarter.

This week, the Seahawks broke that record by a very, very tiny margin. That Bears game was actually worth 16.59 points. This week, Seattle's special teams were worth 16.61 points. When we do our periodic upgrades to the statistical methods, adjusting the various baselines to make them more accurate, that Chicago game could move back in front. But for now, this Seattle game is the best special teams performance of the last two decades -- and again, it is almost entirely because of Washington. Olindo Mare hit two short field goals -- positive value, but almost meaningless. His kickoffs were very good, with two touchbacks and three other kicks returned to around the 20. On one of those, the Seahawks stripped the ball from Darren Sproles, which is definitely important. Golden Tate had a nice 31-yard punt return, but Jon Ryan's punts were pretty average.

Leon Washington, though... Washington averaged 63.3 yards on four kickoff returns. Besides the two touchdowns, he also brought a short 55-yard kickoff back 33 yards. (The average return on a 55-yard kickoff is only 18 yards.) Vince Verhei went back and looked at NFL history before the last two decades, and found that Washington had the highest single-game average of any return man with at least four kickoffs in a game. It's not even close -- Devin Hester previously held the record at 56.3 yards, and there are only two other players with games in the 50s. If we drop the minimum to three returns, then Washington ranks third with the highest average in 50 years; in 1960, Lenny Lyles of the 49ers had a game where he averaged 67.3 yards on three returns, while Ken Hall of the Oilers had one averaging 65.3 yards on three returns.

Here's a updated look at the best special teams games of the DVOA Era, going back to 1993:

Year Team Week Opponent Value
2010 SEA 3 SD 16.6 pts
2007 CHI 12 DEN 16.6 pts
2009 CHI 4 DET 16.4 pts
2002 NO 6 WAS 16.3 pts
2000 BAL 17 NYJ 16.0 pts
2002 CAR 14 CIN 15.5 pts
2000 TB 13 BUF 15.4 pts
1997 SD 10 CIN 15.3 pts
2003 OAK 17 SD 14.9 pts
1994 CLE1 1 CIN 14.8 pts
2005 NYG 1 ARI 14.8 pts

The good news for Chargers fans -- and bad news for Seahawks fans -- is that this kind of special teams performance is very unlikely to continue for the rest of the year. The Chargers may end up with the worst special teams in the league this year, but it won't be this bad.

* * * * *

All stat pages are now updated. The FO Premium database of DVOA splits will be updated later tonight. Also, please note that the KUBIAK midseason update will be available for download on Thursday afternoon of this week.

* * * * *

These are the Football Outsiders team efficiency ratings through three weeks of 2010, 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 VOA are adjusted to consider all fumbles, kept or lost, as equal value. SPECIAL TEAMS VOA is adjusted for type of stadium (warm, cold, dome, Denver) and week of season.

There are no opponent adjustments in VOA until the fourth week of the season, which is why it is VOA right now rather than DVOA. As always, positive numbers represent more points so DEFENSE is better when it is NEGATIVE.

DAVE is a formula which combines our preseason projection with current VOA to get a more accurate forecast of how a team will play the rest of the season. Right now, the preseason projection makes up 55 percent of DAVE.

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

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

1 GB 35.4% 1 24.4% 2 2-1 27.3% 5 -13.9% 5 -5.8% 29
2 KC 33.7% 11 14.8% 9 3-0 9.4% 13 -16.6% 4 7.8% 3
3 PIT 33.5% 9 29.8% 1 3-0 -8.4% 19 -33.7% 1 8.2% 2
4 NYJ 31.7% 4 21.1% 6 2-1 19.1% 7 -7.0% 11 5.7% 5
5 PHI 30.8% 15 21.8% 4 2-1 30.2% 3 -1.6% 17 -1.0% 18
6 NE 26.5% 7 23.7% 3 2-1 47.2% 1 20.0% 27 -0.8% 17
7 ATL 24.7% 2 21.7% 5 2-1 17.5% 8 -7.5% 8 -0.3% 16
8 DAL 20.2% 16 10.6% 11 1-2 22.4% 6 3.6% 22 1.4% 11
9 SEA 19.3% 6 5.0% 14 2-1 -9.3% 20 -7.3% 9 21.3% 1
10 IND 18.8% 12 19.7% 7 2-1 32.6% 2 9.6% 24 -4.2% 26
11 MIA 9.1% 5 11.2% 10 2-1 10.2% 12 -4.0% 15 -5.0% 27
12 NO 6.8% 14 8.3% 13 2-1 13.5% 9 10.7% 25 4.0% 7
13 TEN 4.0% 17 -1.9% 20 2-1 -19.3% 28 -22.8% 2 0.5% 13
14 CHI 3.8% 13 9.5% 12 3-0 1.7% 15 0.8% 18 2.9% 8
15 SD 3.6% 8 -1.4% 19 1-2 12.3% 10 -18.0% 3 -26.6% 32
16 CIN 2.5% 20 -0.3% 17 2-1 -5.2% 18 -5.9% 13 1.9% 10
17 TB -0.2% 3 -6.4% 21 2-1 -13.4% 22 -10.6% 7 2.6% 9
18 BAL -1.3% 22 16.7% 8 2-1 -12.3% 21 -12.0% 6 -1.0% 19
19 CLE -6.6% 18 -14.3% 25 0-3 -3.0% 16 1.3% 19 -2.3% 20
20 NYG -7.1% 19 1.6% 15 1-2 -4.1% 17 -6.5% 12 -9.5% 31
21 HOU -8.4% 10 -14.0% 24 2-1 29.4% 4 37.8% 32 -0.1% 15
22 DEN -12.5% 21 -9.2% 22 1-2 10.4% 11 20.1% 28 -2.8% 22
23 STL -12.7% 23 -17.9% 26 1-2 -15.2% 23 -5.3% 14 -2.9% 23
24 MIN -14.7% 27 0.8% 16 1-2 -18.9% 27 -3.5% 16 0.7% 12
25 WAS -16.6% 25 -0.9% 18 1-2 7.7% 14 20.3% 29 -4.0% 25
26 DET -22.0% 24 -28.2% 30 0-3 -18.7% 26 3.8% 23 0.5% 14
27 ARI -23.1% 30 -13.3% 23 2-1 -17.0% 25 3.0% 20 -3.2% 24
28 BUF -41.3% 32 -24.8% 27 0-3 -16.1% 24 32.1% 31 6.9% 4
29 OAK -43.8% 29 -36.5% 32 1-2 -33.0% 31 3.0% 21 -7.7% 30
30 JAC -44.7% 26 -26.4% 29 1-2 -22.6% 29 27.2% 30 5.1% 6
31 CAR -46.6% 31 -25.3% 28 0-3 -51.5% 32 -7.2% 10 -2.3% 21
32 SF -49.6% 28 -32.9% 31 0-3 -27.2% 30 17.2% 26 -5.1% 28

Posted by: Aaron Schatz on 28 Sep 2010

212 comments, Last at 06 Oct 2010, 10:31am by DeltaWhiskey


by MilkmanDanimal :: Tue, 09/28/2010 - 5:24pm

Wait, you mean Tampa isn't really the third-best team? Travesty!

by panthersnbraves :: Wed, 09/29/2010 - 9:23am

Just wait until those opponent adjustments come in! Why the Panthers game alone will..... oh, never mind.

by Bobman :: Tue, 09/28/2010 - 5:28pm

Every time I see someone ripping off good kick/punt returns regularly I die a little. Usually the last words on my lips are "they Colts coulda had him...." Washington was a risky move because of the injury, but a pretty inexpensive acquisition.

by Aaron Schatz :: Tue, 09/28/2010 - 5:34pm

I seem to remember someone around here suggesting the Colts draft Javier Arenas... who was that guy again?

by Bobman :: Tue, 09/28/2010 - 5:39pm

tall, dark stranger, mysterious glint in his eye, six-gun on his hip? No idea who he was. But he had a point.

Hey, at this pont I'll take Bethel Johnson--I see him in my nightmares anyway, why not on the field, too?

by Will Allen :: Tue, 09/28/2010 - 6:34pm

Once again, hearing woulda coulda from Mr. Fan of Seven Consecutive 12 Win Seasons, Recent Super Bowl Winner, is little hard on the ears of Mr. Fan of Four Consecutive Super Bowl Loser, followed by Five Consecutive Conference Championship Loser, expecting to be followed by Six Consecutive Divisional Round Loser, Seven Consecutive Wild Card Round Loser, and then Eight Straight Years of Missing the Playoffs!

Glass half-empty guy? Who, me?

by Bobman :: Tue, 09/28/2010 - 8:43pm

As Manning says, there's always room for improvement. And verily, their field position annually ranks as slightly better than a case of botulism. In the last 13 years (Manning era) they have had consistent, great O, up and down D (but sometimes very good), but consistently putrid returns. I'd trade a competent DB for a semi-competent one who can average 5.0 more yards per KO return and 3.0 more per punt return.

A fair trade. It's not THAT much to ask for, really....

by Scott C :: Wed, 09/29/2010 - 2:17am

I'd wager that with a great offense, very good special teams returns and the better starting field position is not worth as much as a defense that produces more turnovers and therefore more possessions.

I know DVOA doesn't account for it, but it would be interesting research to see if the value of return teams is amplified or dampened by the biases of the offense and defense. Certainly, an offense that turns the ball over a lot ends up punting and kicking off less.

by QC Coach (not verified) :: Wed, 09/29/2010 - 12:45pm

Sure, but would you trade a third pass rusher (Jerry Hughes) for the "semi-competent" DB who excels at special teams?

by >implying implications (not verified) :: Tue, 09/28/2010 - 5:29pm

Worst 3-0 team: Chicago Bears
Best 0-3 team: Cleveland Browns
Worst 2-1 team: Arizona Cardinals
Best 1-2 team: Dallas Cowboys

by Chip :: Tue, 09/28/2010 - 5:51pm

I beg to differ on the Bears. When opponent adjustments kick-in, KC will take a tumble....

by kickerfan (not verified) :: Wed, 09/29/2010 - 11:04am

Chiefs will be tested in week 5 against Indy. Win or Lose, everyone will have an answer if they deserve the coveted "undefeated" title. Odds are, of coarse, on the proven Colts but the Chiefs are proving to be effective scrappers. Will hopefully be a good game.

by Jonadan :: Tue, 09/28/2010 - 9:29pm

As a longsuffering Lions fan... I seriously think the Lions are a better 0-3 team than the Browns. But I could be wrong, and I don't have even my own numbers to back it up (yet).

by tuluse :: Tue, 09/28/2010 - 9:31pm

I think so too, but I would believe they have played worse so far because Stafford got hurt. Shaun Hill sux.

by Jonadan :: Tue, 09/28/2010 - 11:25pm

Now that I've run the numbers for my own (experimental) personal rating system, I've got the Lions a spot above the Browns right now (#24 and #25 respectively). For what that's worth.

Caveat: I'm unwilling to say more than that I do have this system, as it's approximately 3 weeks old. (On two weeks of data, however, it went 8-7-1 on predictions last week: hardly stellar, but astronomically better than it did on Week 2, so with more data maybe it will be useful.)
Caveat II: I've only finished processing data for 9 of the 16 games, so ranking could change.

by Eric (not verified) :: Wed, 09/29/2010 - 1:17pm

My ranking system has Detroit above Cleveland too, although it's such a small sample size for ranking comparisons. Detroit is 22 and Cleveland is 23. Best undefeated team is the Chiefs in my system, best 0-3 team is the Lions. Worst 2-1 team is the Arizona Cardinals, and best 1-2 team is the Chargers. My system is based mostly off of PA and PF as well as "Strength of schedule" and Offensive yards/Defensive yards allowed per game.

by Viich (not verified) :: Wed, 09/29/2010 - 5:01pm

Why not run it through some historical data if you're trying to test its predictive accuracy? You don't have to wait for the results that way.

by Jonadan :: Wed, 09/29/2010 - 5:15pm

Because 1) I haven't had the time to collect the historical data and 2) I'm not entirely sure where to find it.

by chemical burn :: Tue, 09/28/2010 - 5:30pm

Hmm... Philly with the #3 offense. That Mike Bell and LeSean McCoy are really something. Also, their offensive line is clearly dominant. There's no other explanation for what's going on there.

Just kidding. Everyone knows Owen Schmitt is the man.

by t.d. :: Tue, 09/28/2010 - 5:59pm

Last year there was reluctance to acknowledge Favre's impact because the story was overplayed, too. I think it's at least partly sportstalk fatigue related. And, as is sure to be pointed out, opponent adjustments will probably knock a little of the bloom off the rose

by chemical burn :: Tue, 09/28/2010 - 6:06pm

Maybe - but so much of their offensive value is from the running game, where they are a 20% above the second ranked team and 40% above the 4th ranked team. Vick's passing value can drop significantly and their rushing DVOA is still giant.

Plus, they put up great numbers against GB, who is ranked highly on d. According to VOA, they've played a good defense, a mediocre defense and a bad defense - they truly haven't played 3 cream-puffs. I think they will drop to 8 or 9, but that ain't bad, considering they lost their Pro Bowl fulback and best offensive lineman in the first game of the season. And they haven't had their projected o-line starters make it through an entire game together (meaning, they've had o-line injuries in every single game.)

It's hard to think that Vick hasn't been the difference so far on offense...

by TruFalcons (not verified) :: Tue, 09/28/2010 - 7:49pm

Will be interesting to see what happens with opponent adjustments. They'd surely leapfrog Indy's offense in 2nd place who've faced the #28 and #32 defenses so far....

by chemical burn :: Tue, 09/28/2010 - 7:59pm

That's true - Peyton Manning is clearly nothing but the beneficiary of unwarranted hype. He's only beat up on chumps! I saw him fail to convert a third down, to boot!

by Kal :: Tue, 09/28/2010 - 8:38pm

Considering that's the same argument used to deflate Vick's performances, I think it's a fair point.

by TruFalcons (not verified) :: Tue, 09/28/2010 - 9:15pm

I think we need to wait until he faces a real defense. This might be the year he starts to decline. He hasn't even rushed for 100 yards once. And people are talking about him as a MVP contender? Please.

by RickD :: Tue, 09/28/2010 - 10:39pm

I'm confused - are you talking about Vick or Manning?

I'd like to see Manning run for 100 yards.


by TomC :: Tue, 09/28/2010 - 5:40pm

< Da Bears > is clearly ranked < too low > because < they are teh clutch and winnging close game >. < Peter King > is way better than this. < Cutty ruulz, Rogers sux. >

by Bobman :: Tue, 09/28/2010 - 5:40pm

Is it just me, or does it seem early in the season for the field of undefeateds to be winnowed down to just three?

by chemical burn :: Tue, 09/28/2010 - 5:45pm

It feels like this year in general there aren't any great teams, just some pretty good ones. Maybe Pittsburgh will fill that role when the only dude in the league I'd like to QB my team less than Mike Vick is back. A lot of dominant teams from the past decade look shakier than normal and none of the young teams that were supposed to have emerged really look impressive. Heck, some like SF and Dallas have flopped pretty hard. Dallas will be fine... but, yeesh, SF looks awful. Like 2006 SF awful.

by bravehoptoad :: Tue, 09/28/2010 - 7:16pm

Worse than 2006 awful, because we actually have a lot of talent on the roster this time around.

Have I said this before? WTF WTF WTF WTF WTF?

by jimbohead :: Wed, 09/29/2010 - 5:22pm

I'm pretty sure this is a secret plot to get Luck in next years draft. It has to be that, right guys? right? *sobs uncontrollably*

by Mostly Anonymous (not verified) :: Tue, 09/28/2010 - 5:42pm

#1 in Defense
#2 in Special Teams
#19 in Offense

#3 overall

Just wait until their quarterback comes back...

by chemical burn :: Tue, 09/28/2010 - 5:46pm

Then it truly will be the year of "Horrible Men Grab all the Pro Football headlines."

by Bobman :: Tue, 09/28/2010 - 5:54pm

Eeeeh, I suspect his D will be grabbing all the headlines unless Ben averages 300 YPG and 3 TDs. Assuming a more normal QB season, most pundits and fans will be talking about their dominant D. But I hear your point. I expect the other PA QB story to fade a bit with time as well.

Rightfully so, so the entire country can prepare for the annual Peytom Branning debate, which we'd all prefer.

by drobviousso :: Tue, 09/28/2010 - 8:45pm

You really need a new shtick

by chemical burn :: Tue, 09/28/2010 - 9:10pm


by drobviousso :: Tue, 09/28/2010 - 10:03pm

A bunch of your comments here and in a few other threads complaining about Roethlisburger and Vick as people were wearing a little old. It's probably just me. I notice you also have a lot of posts about other topics, so just ignore me.

I admit that I'm a huge hatter of the Jim Rome side of sports commentary, and I don't usually give a crap what other people think of third parties I (and usually they) don't know personally.

Basilicus - That's how wikipedia spelled it. I actually looked it up before I posted.

by Basilicus :: Wed, 09/29/2010 - 11:27am

I think it's fitting that for every article that hits on those players, a few of us hang around to hate them (the players, not the articles). After all, other people get to hate teams and players because they beat their team when they were young or were an annual thorn in the side for their beloved player, so why can't we be the ones who stick around to remind people that, despite whatever success Roethlisberger and Vick might otherwise see, they are people who don't deserve our respect for even bigger reasons. When I see Vick apologizing more than once a year for his prior actions, or NFL Network even once mention what he did as being awful rather than celebrating every mention of his name, then people like us will shut up.

As is, few seem to care, and the few who do care find that deeply aggravating.

by dbostedo :: Wed, 09/29/2010 - 1:41pm

Yeah...but you don't generally see a Redskins fan hang around just to say that "Dallas is garbage" every time they're mentioned in an article. (Or maybe I've gotten good at glossing over it?)

It's also slightly disconcerting to me that some people seem to have no forgiveness toward, benefit of the doubt for, or possibility of change in people they don't even know; And that makes it slightly more annoying.

by The Ninjalectual :: Wed, 09/29/2010 - 4:28pm

I don't recall Dallas torturing animals or being accused of raping women. That comparison is meaningless.

by dbostedo :: Wed, 09/29/2010 - 10:50pm

But I didn't bring up that comparison....it was being compared in the comment above :

"After all, other people get to hate teams and players because they beat their team when they were young or were an annual thorn in the side for their beloved player, so why can't we be the ones who stick around to remind people that, despite whatever success Roethlisberger and Vick might otherwise see, they are people who don't deserve our respect for even bigger reasons."

I figured as long as it was mentioned that way, I'd point out that commenters also don't seem to treat it the same way.

by TomC :: Wed, 09/29/2010 - 1:21pm

Yiddish is tricky, because the German gets transliterated into the Hebrew alphabet then back into Roman, but the standard is for the "sh" sound to be spelled with an "sh" (e.g., shlmiel, shpiel, shmate).

by DW94 :: Tue, 09/28/2010 - 10:04pm

Simpsons reference? Because I think it was "woozle-wuzzle" or however one would spell it.

by chemical burn :: Wed, 09/29/2010 - 12:46pm

It was a Simpsons reference. It was me trying out a new shtick after my whole "I didn't do" one wore out it's welcome.


by Basilicus :: Tue, 09/28/2010 - 9:43pm

You misspelled "schtick."

by panthersnbraves :: Wed, 09/29/2010 - 9:27am


by Bobman :: Tue, 09/28/2010 - 5:48pm

Yeah, that's pretty intimidating. As a Colts fan, I'm not easily worried. Even the 2007 Pats were a challenge I was pretty sure my team was up for (and for 50 minutes, they earned my faith). But this Steeler team is a wee bit scary....

I've called Parcells to see if he has any annointing oil handy, but he said he was all out. Guess we'll have to wait.

by BJR :: Wed, 09/29/2010 - 8:25am

FO's preseason analysis of Pittsburgh is looking remarkably prescient at this moment. Whilst the mainstream media caught up forecasting impending doom, they noted that last season's drop-off was down to key injuries on defence and historically inept kick/punt coverage performance. Get the starters back on defence, sort the special teams out with a new coach and some outstanding athletes in the draft and they should be able to withstand 4 games without their starting QB. Bingo.

by galactic_dev :: Wed, 09/29/2010 - 12:55pm

I'm ready to crown Polamalu's ass.

by SirKev (not verified) :: Tue, 09/28/2010 - 7:59pm

The most shocking number in that list is the #2 for special teams. Watching the coverage teams give up long returns for the last several years, how on earth are they #2? Is that soley due to 8 FG's and the return TD from last week or have the coverage units actually improved?

If that is the case, then this team could truely be scary. I can't wait for this weekend. This is the first time in along time that I have not been nervous for a Baltimore game. I expect to be no better than 2-2 in the first four so I feel like they are playing with house money. Based on last season, I'm more concerned about coming out of this game with all the star players as healthy as possible.

by drobviousso :: Tue, 09/28/2010 - 8:46pm

Jeff Reid has a new approach on kickoffs that is apparently more uncomfortable on his knee, but produces much better results. He's got, I think, 3 touchbacks and at least 1 more kick off into the end zone this year. It'll either be his best year, or the year his knee explodes.

Also, Stevenson Sylvester, Worlids, and a few other rookies are playing good coverage.

by Mystyc :: Tue, 09/28/2010 - 9:07pm

The coverage is indeed much better. Only one kickoff has been returned past the 30 so far. (I'm pretty sure.) I didn't know about Reed changing his technique - I recall him booting it pretty far early in the season before, only to flag off to the 4-yard line by December...

Returns are still pretty middling except for the TD. As the year wears on, I expect this ranking to taper off a bit, but even staying in the top half of the league would be such an improvement it's ridiculous.

by drobviousso :: Tue, 09/28/2010 - 10:07pm


Ed Bouchette mentions it here. Can't find the article that talks about it more, but the article wasn't too enlightening. No real info on the technique, just a few quotes.

by Geo B :: Wed, 09/29/2010 - 9:09am

Totally lost faith in the defense after watching all the 4th quarter blown leads last year. Starting to get that faith back. The difference in getting Big Ben back is not that he's guaranteed 300 YPG or 3 TD's but that he has more of a comeback capability in case they do get behind.

I think Baltimore got caught in a look ahead situation not taking Cleveland seriously, they won't have that issue this week. It will be a real test as always for all aspects of the Steeler's game. Can the O-line handle Baltimore's D this time? Can the D slow down the Baltimore running game that killed them last year? This is why they play the games ;-)

Props to FO's preseason work for predicting a better Steelers team this year while everyone else was down on them. I need this "prayer" one more time:

Oh Football Gods, we beseech thee for successful results and no injuries this weekend. Blessed be those in Black and Gold, and Blue with White, as they do glorious battle and seek to bring you glory with the pigskin. May preposterous punts be avoided and glorious line play be hailed. In the names of Noll, Lambert and Ham we pray, amen.

Steeler fan trapped in Houston!
Six Time SB Champs! ;-)

by AnonymousOne (not verified) :: Wed, 09/29/2010 - 10:18am

Personally, I think the difference in getting Ben back wont be that he is guaranteed 300ypg or 3 TD's, but that the Steelers would all of a sudden be a threat to get 300ypg or 3TD's every game.

The running game will benefit from Ben's return just as much as the passing game IMO. Assuming Arians doesnt decide to give up on it.

(and of course, the comeback capability you mentioned)

by MidnightAngler (not verified) :: Wed, 09/29/2010 - 2:08pm

I'm still a little skeptical about the steelers offense. Their O-Line is still a wreck. It might even be the worst line they've had. I'd like to see how the steelers ranked in 3-and-outs last year, even with Big Ben. They were 21st in 3rd down conversions - not exactly great. A four week suspension is not insignificant either. Its hard to say how rusty and out-of-sync Roethlisberger will be when he returns. He'll be running for his life as usual if the steelers go back to throwing first...

Everybody went overboard thinking the steelers season was shot when the suspension was announced. The defense was far more responsible for their recent super bowl wins compared to Roethlisberger's contribution. It would be nice if people learned about going overboard... its still a bit early to proclaim them SB champions.

by AnonymousOne (not verified) :: Thu, 09/30/2010 - 3:53pm

Actually, I thought the OL looked improved. Charlie Batch wasnt sacked a single time against TB, though maybe that wasnt a great test.

As for Ben being rusty, could be, but he rejoins the team Monday. Then he has two weeks of practices, due to the Steelers bye week, followed by a Cleveland game. You couldnt really ask for better timing as far as that goes.

by AnonymousOne (not verified) :: Thu, 09/30/2010 - 3:55pm

"The defense was far more responsible for their recent super bowl wins compared to Roethlisberger's contribution"

I thought it smelled stupid in here.

by Jacob Stevens (not verified) :: Tue, 09/28/2010 - 5:48pm

Credit for that Seattle special teams performance should go to the rarely-credited special teams coach, Brian Schneider.

Although after the 1st wave of blockers, it was all Leon. But Pete Carroll talked on Monday about the special teams plan against SD, that they are very aggressive up front but lack discipline behind. Schneider used a different blocking scheme to open up that first hole and leave the rest to the returner. Carroll said before Leon was even through that first hole Schneider was shouting, "we got em! We got em!"

by Aaron Schatz :: Tue, 09/28/2010 - 5:50pm

Schneider was the special teams coach in Oakland when they had that huge leap in special teams rating between 2007 and 2008. He's a good one.

by The Powers That Be :: Tue, 09/28/2010 - 5:49pm

"There are no opponent adjustments in VOA until the fourth week of the season, which is why it is VOA right now rather than DVOA"

The chart says DVOA, but I assume that's an editing error?

by Led :: Tue, 09/28/2010 - 6:04pm

They don't change the chart template. They just remind us every week that the "D" hasn't kicked in yet. I'm cool with that.

by The Powers That Be :: Tue, 09/28/2010 - 6:34pm

The chart headers have been different each week so far this season, so I'm not sure how you can say that.

by Led :: Tue, 09/28/2010 - 6:54pm

Really? I never noticed the difference, but I'll take your word for it.

by Vonnie Rotten :: Tue, 09/28/2010 - 5:53pm

As a 49ers fan, it is strangely comforting to be back at the bottom of the rankings. Like coming home again.

by chemical burn :: Tue, 09/28/2010 - 5:59pm

I feel genuinely bad. I've never seen a fan-base have their hopes crushed so quickly and decisively.

by greybeard :: Tue, 09/28/2010 - 6:40pm

One of my co-workers send me a link to Australian Rules Football on wikipedia after hearing that I am a 49ers fan. He thinks I should start looking for joy somewhere else.

by Thok :: Tue, 09/28/2010 - 11:38pm

To be fair, I'd be angry or disappointed about the 49ers, except that I've used up all my anger on the Cal Bears. That 10-9 loss to Arizona is just gah grerg;n;kdgnlfgmhlkfnmkfhn

by Tundrapaddy (not verified) :: Wed, 09/29/2010 - 5:24am

If you look back far enough, I think you'll find some harder-suffering fan bases.


by dmstorm22 :: Wed, 09/29/2010 - 3:18am

Just imagine if DVOA could tell you how good the '84, '87 or '89 teams were.

by kbukie :: Tue, 09/28/2010 - 6:05pm

Specific question: Just how bad was the Bears' pass defense VOA last night?

by tuluse :: Tue, 09/28/2010 - 6:21pm

I can't answer pass specifically, but their overall defense went from -13.1% to 0.8%, which I believe means a single game VOA of 28.6%

However, the Packer's offense was worse than it was the first 2 games.

by kbukie :: Tue, 09/28/2010 - 7:50pm

By my calculations, the pass D was 3.2% after week 2 and 17.3% now, so if each game counts equally that would put them at 45.5% for the game. Granted, the Packers passing offense is really good, but that will have to improve if they are going to seriously contend for anything.

I'd also say the rushing offense has to improve, but it won't.

by Jake (not verified) :: Tue, 09/28/2010 - 6:06pm

Nice to see the Pack back where they belong ... in the special teams rankings.

by n8 (not verified) :: Tue, 09/28/2010 - 7:12pm

exactly what I was going to write... Crosby kicks out of bounds (NOT to Hester) but Mathsay kicks IN bounds (to Hester)? WTF?

by justanothersteve :: Tue, 09/28/2010 - 7:15pm

That's Shawn Slocum for you. Always keeps the other team guessing - just what stupid thing the Pack ST will do next.

by bigtencrazy (not verified) :: Tue, 09/28/2010 - 6:12pm

For the first time ever Mark Tauscher looked washed up. Yes Julius Peppers is a great player but last night was just brutal as Tauscher was outright abused. He's never been pretty but MT has always been effective. Against the Bears he looked slow and overwhelmed.

by justanothersteve :: Tue, 09/28/2010 - 7:16pm

I noticed that too. I almost thought Barbre was playing again.

by Will Allen :: Tue, 09/28/2010 - 7:47pm

Even when Mckinnie was above average he always looked like an Arena League player against Peppers, so don't give up on Tauscher just yet. I expect Mckinnie to look like he is playing for the Ohio version of Miami when he faces Peppers this year.

by InTheBoilerRoom :: Tue, 09/28/2010 - 10:19pm

Speaking of Peppers, last night's game was a great example of how uninformative box score stats can be for defensive players. Anyone who watched that game knows just how significantly Peppers' pressure impacted that game, but if you look at the box score he only had 2 tackles and 1 QB hit.

As a Panther's fan, I can say I have frequently seen him dominate similarly without actually registering a sack, but instead, forcing multiple holding and multiple false start penalties. And four weeks from now "pundits" will have forgotten how important his play was since there aren't really any box score stats next to his name to "validate" it. It'll all just be about how he only had one sack through three games. I know I'm preaching to the choir on these boards regarding traditional stats failing to tell the story of a player's value, but I just need to vent, considering how terrible the Panthers have been so far. I miss Peppers already.

by RickD :: Tue, 09/28/2010 - 10:42pm

Yes, it was a very nice game for Peppers. With him on the team and Urlacher back, the Bears again have an elite defense.

Shame about the QB.

by Bad Doctor :: Wed, 09/29/2010 - 10:34am

Totally agree with this point. All of the talk about this game has been how the Packers played circles around the Bears, but they shot themselves in the foot. That's because most football fans think about the game through the window of the box score and, to a lesser extent, fantasy stats. The Packers outgained the Bears by more than 100 yards, and their penalties and turnovers were just "sloppy" or "undisciplined" play. And some of the penalties were undisciplined -- an illegal formation, an offsides on a kickoff, the German suplex on Forte in the 4th quarter. But there were 3 holding penalties on passing plays, plus the intentional grounding, that are a result of the Bears' D, often Peppers, beating the Packers' O line. (Perhaps the false starts too, if you believe Gruden, though he did himself no favors making this point on a false start on the opposite side of the line from where Peppers was lined up.) If the Packers play "clean" and "disciplined" on these plays, they probably give up a couple of sacks (maybe even a strip sack) or hurried incompletions (maybe even a pick) that show up in the box score and fantasy stats. There were 2 DPIs (plus a third that was declined) that were the result of the Packers' secondary being beaten by the Bears' receivers. If they play "clean" and "disciplined" on these plays, they probably give up a big completion or two that show up in the box score and fantasy stats. The "clean" and "disciplined" play would negatively affect the Packers' NFL yardage ranks, and sack totals, and then the talk would be about the problems the team has on the O line and the secondary. Instead, the prevailing chatter just says that the Packers looked great, they just have to get some discipline and stop playing so chippily.

Now throw in the Hester TD, which robbed the Bears' offense of a possession, and you get the Packers running rings around the Bears on the stat sheet, but not so much on the field. When you break it down per play, the Bears had 4.3 yards per rush (propped up by some good Cutler scrambles); the Pack had 4.2 (propped up by some good Rodgers scrambles). The Bears had 6.6 yards per pass play; the Pack had 7.0 (not counting the penalties discussed above). Each team honked a makeable FG and had a turnover that affected the scoreboard. I'd say the Packers probably played a bit better play by play, mostly because Cutler was a bit erratic and the Pack's turnover seemed more fluky, but the difference was maybe that of a 10 or 11 win team versus an 8 or 9 win team, nothing more.

by GnomeChumpsky (not verified) :: Wed, 09/29/2010 - 11:07am

Now here's a man who doesn't succumb to outcome bias or post hoc rationalization. Excellent points, sir.

by Q (not verified) :: Tue, 09/28/2010 - 6:14pm

I am curious to see how PITT's Defense does against top Passing Offenses that can spread the field and go attack the secondary. So far, Pitt has gone against two teams that rely heavily on running (ATL and TENN) and a team lacking any playmakers on Offense (TAMPA) which makes all 3 opponents great matchups for PITT's great Run Defense.

by AnonymousOne (not verified) :: Wed, 09/29/2010 - 10:42am

I dunno, Tenn and Tampa werent that good but I think Atlanta has a very good passing game.

Ryan dropped 225 yds/3tds on Arizona, and 228 yds/2tds on the Saints. They do have a great run game, but they also have the passing game to punish teams that can defend it.

by AnonymousOne (not verified) :: Wed, 09/29/2010 - 10:44am

CANT defend it, that is.

by the cat in the box is dead (not verified) :: Tue, 09/28/2010 - 6:29pm

Looking at that, I'd think Pitt are going to be terrifying when they get their QB back.

by Mort (not verified) :: Tue, 09/28/2010 - 6:33pm

I have to say it's pretty concerning as a fan to see the Ravens at 8 in DAVE but 18 in VOA. I realize opponent adjustments aren't in yet but I can't imagine that would change things all that much.

Flacco looks to have lost some of his mojo and I don't quite know what to make of it.

Regardless, poor 49er fans. At least the Bills, Browns, and Rams weren't saddled with high expectations.

by The Ninjalectual :: Tue, 09/28/2010 - 8:06pm

Early theory: the Ravens hired Jim Zorn as their QB coach, and he's been busy fixing things that weren't broken.

by JimZipCode :: Wed, 09/29/2010 - 3:14pm

Well, Zorn is still pretty respected as a QB coach. His rep isn't all Hasselbeck either: Zorn helped cajole one of the best rookie-QB seasons in history from Charlie Batch, plus he had 9 seasons as a college coach. If Zorn is making some fundamental changes in Flacco's game, then it would be legit to see Flacco take a step backwards before progressing. Consolidation phase or whatever.

The thing I wondered about Zorn was, his pro coaching background was primarily with the "West Coast Offense" or Bill Walsh offense. That's not what the Ravens run. Cam Cameron is more of a Coryell/Zampese/Suanders/Norv kind of coach. Is it possible there is some kind of incompatibility between the QB skills taught in the WCO, vs the things required in that other system? Footwork or something?

Eh, I don't know a thing about coaching QBs, not one thing, so I'm just wondering aloud. Anyway, Zorn played QB for a dozen years in the NFL, mostly before there even was a "WCO". He probably knows how to coach the position, either for that offense or for some other.

by The Ninjalectual :: Wed, 09/29/2010 - 4:27pm

I don't know a thing about coaching QBs either. But just because a coach has stuck around and has a long history of employment does not mean he's above incompetent. Look at hoe Eric Mangini continues to find work... Maybe Zorn knows how to be successful. And maybe he forgot after his embarrassing stint in Washington.

by Skeeter McGee (not verified) :: Tue, 09/28/2010 - 10:44pm

Eh, there were some expectations. Anybody remember the FO preseason DVOA prediction that the Bills would have the #3 defense?

Man, that prediction came out of *nowhere.* Injury bounce-back must have been way, way overestimated. They got back some good players, not Jesus "Ballhawk" Christ and his Turnover Apostles. If you looked closer, you'd see a 3-4 defense with three lineman ill-suited for their positions and a damn poor linebacking corps that depends upon a very injury-prone guy who's never proven himself to be much better than average. (He is, of course, already injured.)

So terrible. So very, very terrible.

by jmaron :: Tue, 09/28/2010 - 6:53pm

Atlanta is giving up 1.2 yards per play more than they gain. I find that a strange stat given their 24.7% VOA.

SD is gaining 2 yds per play more than they've given up. I suspect we will see them make their usual 2nd half run.

by johnnyxel :: Tue, 09/28/2010 - 7:23pm

Since DVOA is a per play stat, the three huge plays that account for 20% of Atlanta's yards against are probably downplayed in DVOA (since there is little difference in value for such long plays). The other 161 plays yielded 5.03 y/p. Atlanta's defense has given up the fourth-fewest first downs in the NFL, and stopped opponents on 25/33 third and fourth downs, and has collected 6 ints. Not saying that is sustainable, but it sure explains why the defense is better than just looking at a 6.2 y/p.
On offense, Atlanta's y/p is low (5.0), but they are tied with the Texans for the most first downs through three games, and leading the league in plays from scrimmage. So #1 in first downs on offense, #4 in first downs on defense.

by Rich Arpin (not verified) :: Tue, 09/28/2010 - 7:05pm

The Pats Defense is terrible. I wonder when Belichick and all the rookies and 2nd years will gel, figure things out, and start gellin'. At least Chung is the real deal. Too bad Meriweather isn't seeing the field much.

by Nate :: Tue, 09/28/2010 - 8:29pm

Their defensive performances in the first three weeks are going to look (even more) funny bad once opponent adjustments are factored in.

by RickD :: Tue, 09/28/2010 - 10:47pm

Yes, the Pats (pass) defense is godawful. Oddly, they've been getting a little more pass rush than last year, and the run defense is reasonable. But the coverage is just awful.

It should be a good test of BB's abilities. According to his reputation, he should be able to whip these guys into shape.

by RichC (not verified) :: Wed, 09/29/2010 - 9:05am

Merriweather played 40 of 56 defensive snaps.

This whole "Merriweather is riding the pine" thing is overdone.

by muteant :: Tue, 09/28/2010 - 7:25pm

A bit surprised to see the Vikings defense only ranked 16th—they've certainly played better than that according to my human eyes.

"We're the worst thing since sliced bread" - Steve Francis

by outsidersisajoke (not verified) :: Tue, 09/28/2010 - 7:25pm

This is where your rankings are a joke I am sorry the raiders in yardage have the 10 ranked offense 4th ranked defense and somehow they are 29th?

KC has been running a offense on gimmick plays etc... (how many passes have they completed to WRs and their one td to a WR was a trick play) this look like this years Denver.

Problem with outsiders is it does not incorporate common sense. Like the raven were the best team in the league last year at 9-7 right!

As Parcells would say you are what you are. And Cleveland is 0-3 not the 19th best team in the league.

by JIPanick :: Tue, 09/28/2010 - 9:17pm

Improper formatting. Allow me.

The Raiders are clearly ranked too low because total yardage is an infallible measure of football prowess. Common sense is way better than this. Chefs suk cuz their offense is only gimmiks!

by RickD :: Tue, 09/28/2010 - 10:52pm

You want "common sense"? Go read something that isn't based on numbers. Or do you typically spend your time whining about statistics?

I am sorry the raiders in yardage have the 10 ranked offense 4th ranked defense and somehow they are 29th?

It's about getting points on the board, not moving the ball. Having a kicker who horks (sp?) three figgies in one game is a good way to lose.

by RaidersAreAJoke (not verified) :: Wed, 09/29/2010 - 4:41pm

Seriously dude, the Raiders are an awful football team, and they have no idea what they're doing. Try to hide your homerism a bit next time. It's blatantly obvious. You just sound like you have no clue what you're talking about (which is probably true).

by Basilicus :: Wed, 09/29/2010 - 5:46pm

I want this dude and Raiderjoe to have an epic throwdown.

by JasonG (not verified) :: Tue, 09/28/2010 - 7:26pm

I'm very much behind FO and what they are about, but I've been paying attention to DVOA for years now and I just have to say I'm getting farther from "believing in it" than closer. I understand the idea of 50% of 1st down yardage, 100% of 3rd, etc. as definitions of successful plays, but DVOA's limitation are becoming increasingly glaring.

Sure, it's frustrating watching the Bears D give up CONSISTENT 5-7 yard passes. BUT, that type of defense is intentional. I'm not saying it's wrong or right, but philosophically Lovie Smith is playing a strategy on both O and D, and last night at least it worked.

For all Rodgers "domination", the Bears FORCED Rodgers to gain 70 yards in 10 plays instead of allowing him to do so in 1 (Btw, he average 7.0 yards per pass. Not only is that NOT dominating, it was worse than Cutler's 8.2 average). So he had 75% (or whatever the number was) successful plays. Smith is taking the chance that over 10 plays (or 60 in the game), his opponent can't make enough successful short plays to win. You need to string a lot of successful plays together to beat the Bears D.

On the other side, "mediocre" Cutler didn't string a number of successful drives together so he's rated poorly. Yes, he was 3 and out some, but on other drives all he had to do was get 1 or 2 decent chunks to ultimately lead to points.

Yes, this is the typical boom or bust argument, but the Bears are quite the one-team self-contained example of it. Their O doesn't put together consistent 5-7 yard (deemed successful) plays, but the ones they do hit are huge (30-50 yards). The D "allows" successfully plays, but only some singles, no home runs.

So, is DVOA rating both of these teams in this game properly? For the record, I don't care what any one ranking says, 1st or 32nd, about the Bears. I'm talking about systematically understanding what plays lead to winning. One 50-yard play vs 10 5-yard plays. Which is better? Personally, I'm in Lovie's (and Martz') corner here. I'd rather force the other team to string together 10 successes. You put the burden on the other team and give yourself 10 chances to do something disruptive yourself. On offense, if you can gain 50 in one play, of course that's better than having to take 10 plays to do it.

Sure, it felt like Rodgers was more in control than Cutler (and that's commentary on more than just those two players, like the O lines, the opposing Ds and the WRs: all advantage Pack btw), but does that mean he was more successful? Because for all that "control," it really just felt Rodgers was more "contained." Ultimately, (besides the ST domination, a welcome sight) letting Rodgers have the underneath stuff won out. If the Bears D had played differently (like the Pack D did) and been frenetic, maybe Rodgers hits on one big play (like in the 4th quarter of last year's Week 1 game) that blows the entire game plan and wins it.

Btw, this all says nothing of the fact that nine yards on 3rd and 10 can be successful. It could get your punter out of the end zone. It could get him closer to pinning the opponent inside the 10. It could get you into FG range. It could get you in position to go for it on fourth down. All these scenarios are deemed "unsuccessful" by FO stats (as I understand it) and basically I have a fundamental problem with that.

I'm not arguing the Bears as a team are better than the Packers as a team, but is it really 13th vs 1st? I have a very hard time buying into that.

by tuluse :: Tue, 09/28/2010 - 7:36pm

I would say Rodgers thoroughly outplayed Cutler.

1) The Packers offense scored 17 points, the Bears scored 13, with Hester adding a return TD.

2) They both had a missed field goal, but the Packer's attempt was from a much closer range.

If the Bears offense and special teams had played worse, and the Packers were starting their drives 10-15 yards closer on average, would the sit back and watch them make a mistake strategy work?

Also, remember opponent adjustments haven't kicked in. By my calculations, the single game DVOA for Bears and Packers was respectively -8% and -4.8%, which is to say this was a very close game.

I do think the Bears might not be getting their due credit for "forcing" the Packers to commit penalties, but I doubt they're going to force another team to commit 18 penalties, so who knows if they should get credit for that.

Also, DVOA does take into account field position. Getting into field goal range is considered a good thing, and you get credit for that. Also, DVOA isn't really about making plays that are considered "successes," but about making plays that are more successful than the average team in the same situation. If the average team gets 5 yards on 3rd and 10, and you get 9, you will have a positive DVOA.

by JasonG (not verified) :: Tue, 09/28/2010 - 7:58pm

"If the offense and ST played worse"

You could play that game all day. How about "If the refs had called all the holdings that actually occurred maybe this wasn't even a close game." "If Gould makes his FG, if Desmond Clark holds on to the TD, f the Bears didn't play for a FG at the end of the game once inside the 5." Tack another 7-10 points on right there. What if Hester doesn't go 62 yards and instead the O goes 62 yards and takes 4 minutes to do it? Then the Pack loses before Jones even gets a chance to fumble.

As for "would the sit back and watch them make a mistake strategy work?" that's the game you play. No, one strategy is going to win every time. The sit back strategy doesn't have to be in stone. Against a Rodgers who can burn you, you play it. Against a weak O, you might play more aggressively. These aren't the questions I'm asking or arguments I'm trying to start.

I want to know how DVOA can consider one of these teams league average and the other #1. I can concede, despite the Bears victory, the Packers are the "better" team, the one more likely to have more success over the rest of the season. However, I can't concede such a gap between the two and so I am questioning DVOA, in particular how it rates consistent 5-yard plays versus semi-consistent much bigger plays.

by TomC :: Tue, 09/28/2010 - 8:24pm

The short answer is that through three games the Packers have done more of the things that FO has determined correlate long-term with winning than the Bears have. A lot more. It could certainly be true that Lovie's bend-don't-break strategy is fooling DVOA into thinking that the Bears actually can't stop anybody, and that Martz's high-risk, high-reward offense is fooling DVOA into thinking the Bears offense is just lucky --- I don't think those things are true, but I can't prove they aren't.

Personally, I feel that the Bears' current rating is about right, and that we're more than a bit lucky to be 3-0. But I'm very optimisitic for the rest of the season, because 1) the offense is still finding its way, and 2) we've played the best two passing offenses we're going to see until very late in the year.

by tuluse :: Tue, 09/28/2010 - 8:37pm

I think you missed my point. I wasn't saying "if the Bears had played worse, they would have lost." I'm saying if the Packers had shorter fields, would the defensive strategy work? It worked because the Packers had to go 70 yards, and they had to run 10 plays. If they only had to 50 yards and run 7 plays, things could have been a lot different. That's what DVOA sees. An offense putting together a lot of successful plays, so if the offense had shorter fields, they would score more points.

Also, you are putting too much emphasis on this one game. DVOA saw the teams as very close in this game. However, it saw the Packers dominate their other opponents better than the Bears did theirs.

The Bears passing offense is actually rated pretty well. 32% is very good, and it's 13th in the league so far.

Finally, opponent adjustments aren't in yet, I could see them looking a lot closer after that.

by JasonS (not verified) :: Tue, 09/28/2010 - 9:24pm

I was about to write this same post (though no doubt less eloquently). I find it very interesting that the Bears defense and offense are predicated on doing things that DVOA doesn't seem to value. In the case of the defense, the Cover 2 gives up plenty of first downs, but at least in theory fewer scores. In the case of the offense, it seems to suffer from more of its share of 3 and outs, but produces lots of big plays.

It could just be that the Bears offensive and defensive philosophy is completely wrong. But it seems like the Bears are trying to accomplish something entirely different that DVOA is attempting to measure.

by tuluse :: Tue, 09/28/2010 - 9:27pm

The Bears only had 1 3 and out against the Packers, and the passing DVOA is 32%. DVOA like the Bears passing attack a lot. It hates the running game though, which drags the total offensive DVOA down.

by dmstorm22 :: Wed, 09/29/2010 - 3:26am

I would agree that DVOA seems to not like Tampa-2 (or even base cover-2) teams, because they basically give their opponent a five yard pass each time, limit YAC, and force their opponent to have flawless execution.

However, either that is not true, or the 2002 Bucs were easily, by far, unequivocally the best defense ever, if they could put up those numbers playing a true Tampa-2.

BTW, I think you are right, but teams are better playing against cover-2 teams now than they were in 2002, but not necessarily better in scoring against those teams.

If a team like the Packers converts four different 3rd downs, but then doesn't convert the fifth which leads to a field goal, I think DVOA would think this was a nice drive (four 3rd down conversions and the like), but it is a win for the defense.

By the way, this isn't about FO or DVOA, but I love how so many people in the general media all say that the Packers "outplayed" the Bears, mainly on the evidence that they outgained them by 100 yards (again, this is not talking about FO but the msm). Of course, they all say the Giants suck because they lost when outgaining the Titans by 225, or that it is the same old Raiders, because they outgained the Cards by 150. Double standard much, msm? (end of rant)

by DisplacedPackerFan :: Wed, 09/29/2010 - 11:56am

The Packers did outplay the Bears. I will grant that several of the offensive penalties the Packers had were created by the Bears defense, but having watched the game the Bears got a lot out of the Packers defensive penalties (5 of their 18 first downs, and I believe it was over 80 yards). The one on the 9 yard line when I was watching it I was waiting for the offensive pass interference call against Bennett as he started the contact and was pushing against Burnett. To me even on the replays that looked like the receiver was trying to push the defender out of the way and the defender simply got tangled. Burnett didn't even have his hands on Bennett his arms were tangled with him and so were their legs but Bennett had the hand contact and was even holding Burnett's Jersey. So I saw the flag after the interception and thought "well the Pack will decline that offensive pass interference call" but nope.

The unnecessary roughness call against Woodson was dodgy as well because the ball carrier wasn't down yet. Forward progress was stopped so yeah he picked the guy up and dropped him to the ground, it was a tackle to stop a guy.

There could have been other holding calls against the Packers too, but I actually think those were called consistently for both teams. What wasn't called as a hold for the Packers wasn't called as a hold for the Bears either as both lines were holding.

The roughing the passer on Zombo was the right call as it was helmet to helmet, but again that was one of those where it was more incidental. You have a rookie getting pressure there and hitting hard, if his head was 2 inches lower that changes the complex of the game a lot too.

It wasn't a poorly called game, but there were some borderline calls and plays, that made a huge difference.

The eyeball test to me looked like the Packers generally outplayed the Bears. I would have expected them to win 7 of 10 games if both teams played like that. It was a close fought football game. But really the Packers had more "unforced errors" if they played cleaner, more like they did in the first 2 games, I think they win a close game. I didn't expect a blow out but my eyes say the Packers should win about 70% of the time vs what I saw from the Bears. On Monday they made more mistakes they lost like they should have, I think they cracked under pressure honestly.

This could be one of those "wake-up call" games for the Packers, because the team does look good and has in all 3 games. They have flaws, but they have playmakers on both sides of the ball, they have solid game plans usually. They were executing fairly well though they didn't on Monday night.

The Bears also have play makers in all 3 phases of the game. They execute pretty well, they look to have good game plans as well. They are a better team than I thought going into the year, but I see them as a 10 win team, I still see the Pack as a 12 win team.

by Jonadan :: Wed, 09/29/2010 - 12:13pm

This is possibly a bias due to my work as a soccer referee, but one of the things that irritates me most in football analysis is the tendency to downplay penalties. I don't remember offhand how DVOA/DYAR treats penalties, so this may not apply so much to FO, but most places just sort of ignore them, unless they were egregious...

...And Green Bay was beyond sloppy. So even the regular outlets ought to be mentioning that. If you're outplaying when you're not committing penalties, but you're getting penalized a lot, that's just as "inconsistent" as a regular boom-or-bust offense (like Reid's or Martz's).

by DisplacedPackerFan :: Wed, 09/29/2010 - 2:37pm

Oh I agree Green Bay deserved to lose this game, but the term outplayed tends to refer to what the opponent did to you not what you did to yourself. That's the hang-up I think. Even saying that the Bears caused 10 of the Packers 18 penalties I still think the Packers played better offense and defense. The Bears played way better special teams. Penalties matter, but as mentioned not all penalties are created equal, but even if all the questionable ones are removed that Packers still committed at least 13.

The other factor is that I think penalties should be easier to correct than other problems which is why I value them lower when evaluating the strength of the team. If the Packers don't clean them up then, yes that will make them a very bad team playing like that all the time. But I don't expect that to be the case. I still think they are the better overall team, they have more potential, and I feel they did more to beat themselves than the Bears did to beat them.

by dmstorm22 :: Wed, 09/29/2010 - 1:00pm

I don't necessarily disagree with this, or the statement that the Packers are a 12 win team and the Bears are more of a 10 win team. However, to hear the media talk about this game, one would think the Bears did nothing but get giftwrapped TDs and had no business being in that game, and are a 6 win team at best, which is totally untrue.

by RickD :: Tue, 09/28/2010 - 10:59pm

For years some of us have wondered if the Pats' defense has been underrated because they often seem to be a "bend but don't break" approach. But the best answer is that a defense that doesn't bend is considerably better than one that tries to "bend but not break."

I don't think the DVOA system is optimal in the sense that it is the best possible statistic to compare the levels of different teams. But it is considerably better than most of the statistical systems out there.

For example, only one website liked the Chiefs before the season started. That was FO, and that's because their system picked the Chiefs to win the division. Right now that's looking like a great prediction.

by Pat (filler) (not verified) :: Sat, 10/02/2010 - 5:14pm

Yes, this is the typical boom or bust argument,

Great, because that allows a simple response.

On defense, giving up consistent couple-yard gains is bad even if the other team doesn't score. You're giving up field position. Your defense might not seem bad - you might not give up a ton of points, for instance - but it'll hurt your team anyway, because your offense will be in consistently poor starting field position, so you'll score fewer points.

Equivalent argument for inconsistent offenses (3-and-out versus a couple-yard gain, etc.). In that case you hurt your defense. In both cases, it might seem like the boom-or-bust unit isn't that bad, just based on points scored/allowed (depending on offense/defense). But the lost field position hurts you on the other side of the coin.

So with a boom-or-bust offense, even if it scores as much on average as a more 'consistent' offense, you'd find that the opponent scores more than you would expect. With a bend-but-don't-break defense, you might prevent scoring as much as a 'good' defense, but you'd fine that your own offense scores less than you would expect.

2006 Bears vs. 2006 Ravens are a good example.

by DeltaWhiskey :: Tue, 10/05/2010 - 1:47pm

Nice comment...makes sense. Any sort of data that would support these descriptions?

by Pat (filler) (not verified) :: Tue, 10/05/2010 - 10:47pm

Sure. Just look at the drive statistics and locate teams that have a disproportionate points-per-drive compared to their yards-per-drive. Then look at the starting field position of the opposing unit, and it will tend to be way worse than average, and the points-per-drive of the opposing unit will tend to be worse than you would expect for yards-per-drive as well.

I really should compile these into a quick regression to maybe explain it, but there's no real question whether this sort of thing is true, so I just don't have a ton of drive to do it. To do it really, really well, you'd have to fold in the special teams ability of the teams as well, which would take a fair bit of effort. But I have to say, I don't know why I thought Chicago in 2006 was a good example; it's not. My brain must be mixing up teams.

Just as a quick list, though:
Dallas, 2009: Defense 11th in yards per drive, 3rd in points per drive: Offense 5th in yards per drive, and 13th in points per drive (and 29th in LOS/drive).
New England, 2009: Defense 17th in yards per drive, 7th in points per drive: Offense 1st in yards per drive, and 5th in points/drive.
Minnesota, 2007: Defense 12th in yards/drive, 7th in points per drive: Offense 15th in yards/drive, 20th in points/drive.

There are counterexamples, but they tend to have really good special teams (hence the 'fold in special teams ability'). The teams listed above all had average special teams. Also, NE 2007/IND 2007 kindof are counterexamples, but when you've got an offense that ludicrously good, field position doesn't matter that much (and your defenses tend to start with great field position).

Again, the logic here isn't complicated - giving up yards is bad. Period. The end. If Chicago has to choose between giving up yards and giving up a touchdown on defense, they're not as good a defense as one that *doesn't give up the yards in the first place*. The choice to not give up the touchdown might be a good one, but that doesn't change the fact that giving up the yards isn't a good thing, period.

And as an aside, my personal opinion is that 'bend but don't break' defenses occur when the defense has one of the three main components - defensive line, linebackers, and secondary - which is *much* worse than the other two. In the middle of the field, the offense can exploit the weak unit, but when the field compresses, that goes away. For Philly, I used to say "it's not a bend-but-don't break defense, it's a 'God our linebackers suck' defense."

by Rocco :: Tue, 09/28/2010 - 7:27pm

So does GB maintaining their #1 spot despite the loss make them this year's Philly, that is the team that DVOA loves no matter what they do?

by chemical burn :: Tue, 09/28/2010 - 7:38pm

A) Wait for opponent adjustments B) Pick an example where the team didn't beat themselves.

by Rocco :: Tue, 09/28/2010 - 8:13pm

I wanted to get ahead of the curve on complaining that DVOA is in love with a team for no reason.

by chemical burn :: Tue, 09/28/2010 - 8:16pm

Fair enough. My feeling is that after opponent adjustments, it will be Atlanta.

by tuluse :: Tue, 09/28/2010 - 8:39pm

If we're making predictions, I'm going with the Steelers.

by Basilicus :: Tue, 09/28/2010 - 9:47pm

Eagles. Why break with tradition?

by chemical burn :: Tue, 09/28/2010 - 10:05pm

Yeah, truthfully, I was shocked to see the Eagles as high as they are. If opponent adjustments don't torpedo them, there's no reason for their DVOA darling streak to break!

by Kal :: Tue, 09/28/2010 - 8:44pm

I thought last year GB was in the same boat - they 'should have' won a bunch but lost and had a huge differential.

I wonder when things like sloppy play and penalties become systemic enough in an organization that they should be predictive? GB was horrible on penalties last year as well, IIRC, and it did cost them.

I also wonder how much of DVOA nowadays is somewhat tailored incorrectly. IIRC, DVOA was a response to show how effective Brady was despite his subpar traditional stats, and one way they adjusted DVOA was to see what worked (ie, playoff and regular season game success) and tweaked it so that those teams that had success had higher DVOA on average. Has the game changed enough to where that's not reasonable?

by tuluse :: Tue, 09/28/2010 - 8:57pm

Even if penalties were systemic for the Packers, if it's not affecting the whole league, it would add predictive capabilities.

If one team does something vastly different from the rest of the league, a computer model isn't going to be able to capture it.

by chemical burn :: Tue, 09/28/2010 - 9:07pm

You actually got DVOA's creation exactly backwards - Aaron thought the media was wildly over-rating the 2001 Pats and Brady in particular and came with DVOA as a way to measure the value of what was really happening on the field. He was promptly proven correct when the 2002 Pats underperformed mainstream expectations and missed the playoffs completely.

by nat :: Tue, 09/28/2010 - 10:01pm

The Establishment Clause (2003) was the first article in FO's history. It addressed the myth that teams need to establish the run to win. It was Aaron's first use of play-by-play data in spreadsheet form. So DVOA was not about proving or disproving the Patriots worthiness prior to the 2002 season. The timing is wrong.

Other early articles discussed Pythagorean Wins in football, an early attempt at pre-season predictions with DVOA, the introduction of DPAR (vastly superior to DYAR, in my opinion), and the value of turnovers and the "fluidity" of field position. So, other than the focus on debunking the Run-to-Win idea current in Boston during the 2002 and 2003 seasons, FO wasn't really about the Pats at all.

by tuluse :: Tue, 09/28/2010 - 10:03pm

DPAR and DYAR are the same stat, just on a different scale. I fail to see how one can be vastly superior to the other.

by Lou :: Tue, 09/28/2010 - 11:25pm

because the scale matters. theres a reason VORP is described in runs and not hits.

by The Ninjalectual :: Wed, 09/29/2010 - 4:39pm

DYAR literally claims that a "replacement level back" would have rushed for 1,700 yards in Chris Johnson's place last year. DYAR literally claims that a "replacement level" quarterback would have thrown for 1,900 yards in Tom Brady's place last year. To believe this is to be stupid. DPAR is better because it's easier to take it for what it's worth--an estimate of relative cumulative value. DYAR is too literal.

by tuluse :: Thu, 09/30/2010 - 6:03am

I think it says a replacement level QB would throw for 2277 yard in Brady's place last year. Which sounds about right to me. Rookie Kyle Orton threw for 1869 yards with Moose and a bunch of crap at receiver and Ron Turner as his coach. I think he could have got 400 more yards with Moss and Welker, and one more game started.

by chemical burn :: Tue, 09/28/2010 - 10:04pm

Aaron - we need a ruling on this. He's definitely given the version of FO history that I laid out in interviews. It's possible he was working on DVOA and hadn't completed it when he started publishing articles...

by Aaron Schatz :: Tue, 09/28/2010 - 10:04pm

Yes, Nat is correct. DVOA was developed totally separate from any question about the Patriots. The only Patriots connection was in the inspiration for the first piece I wrote, on establishing the run, which I actually first wrote on my own in November of 2002. I started creating DVOA in December, but that really had nothing to do with the Pats.

Actually, the most important player in the development of DVOA may be Mike Alstott of Tampa Bay, because he was the guy who made me realize you had to create a baseline and compare every play to the average performance in the same situation -- otherwise, Alstott came out as the best back in the league in 2002 because he converted so many third downs.

by chemical burn :: Tue, 09/28/2010 - 10:07pm

I respectfully accept that I am wrong.

Did you talk about the 2002 Pats early on in terms of DVOA though? Am I creating a memory wholecloth?

by Will Allen :: Tue, 09/28/2010 - 11:04pm

Well, then I need to express my profound gratitude for your efforts once again, Aaron. Mike Alstott was my Zombie King when he was in his prime; it drove me effin' crazy the way people used to overrate him, even smart people like Dr. Z. I was convinced that half his reputation was built, and his obvious weaknesses (like fumbling a lot) were overlooked, in good part due to Chris Berman's sound effects when narrating Bucs highlights.

by nat :: Wed, 09/29/2010 - 6:41am

I used the excellent "Analysis: Complete Archives" link to seek out the earliest articles, and found the reference to The Establishment Clause in "About: FO Basics". The early articles are a good re-read.

by The Ninjalectual :: Wed, 09/29/2010 - 4:34pm

I would like to briefly interrupt this thread to agree with nat that DPAR was and in far superior to DYAR. I know, they are iterations of the same thing, but conceptually, DYAR is too easy to disprove. Bring back DPAR!!!!!!!!

by Led :: Tue, 09/28/2010 - 7:44pm

Go Jets! It's amazing how clever and accurate DVOA is when it says your team is good and how stupid it is when it doesn't.

by Bobman :: Tue, 09/28/2010 - 8:47pm

Truer words were never spoken, except, of course, when the Colts came out on top.

(Actually, after a few years of Indy in the 1/2 slots but not winning it all, I started to covet the 5/6 spots.)

by chemical burn :: Tue, 09/28/2010 - 9:09pm

Well, I'm actually in a different boat - the numbers are saying my team's offense is good and FO is insisting that our QB stinks. DVOA actually seems a very clever and accurate measure... that their writers really want to undermine for some reason...

by Led :: Tue, 09/28/2010 - 9:26pm

I've learned to forgive a certain amount of pointless contrarianism at FO because it arises out of the same instinct that drove Aaron to develop DVOA in the first place.

by chemical burn :: Tue, 09/28/2010 - 9:44pm

Yeah, that's definitely true. And no one can ever accuse them of failing to go out on a limb. I really think DVOA does a really good job, just I am a little frustrated with the Vick writing because there's nuance there that the mainstream media is getting wrong (he's probably not a Top 10 passer and certainly not an MVP candidate) and that knee-jerk "he was over-rated then and he's over-rated now" gets the story just as wrong. It seems to me so far Vick's performance indicates the intriguing possibility that Andy Reid didn't have fluke success with McNabb and maybe really knows how to develop a running QB into a capable passer. Sure, he hasn't been tested yet by fierce opposition, but he is so far playing VERY differently than he ever did in Atlanta.

Yeah, I hate his guts as a guy, but so what? Really, let me know what I should do other than stop watching the Eagles, which is not going to happen. Write bad analysis for my football website seems like the worst option and the one FO is pursuing.

by tuluse :: Tue, 09/28/2010 - 9:50pm

It doesn't have to be anything that specific. Reid could just be much better at coaching quarterbacks than anyone in Atlanta. It seems quite reasonable to me.

Also, his receivers are quite a bit better than his Atlanta days.

by drobviousso :: Tue, 09/28/2010 - 10:12pm

Good point.

Is there anywhere else you'd like to be, if you where Vick? Good QB coach, good receivers, good game planning, and hey, it's Philly. You don't have to take the boos, headlines, and occasional battery toss personal.

I think "is Vick good?" is the wrong question. I think "is he a good fit for the system?" is the question, and the answer is yes. His strengths are magnified, and his weakness are covered up. Enjoy it while it lasts, because it's both rare, and a long time coming.

by Basilicus :: Tue, 09/28/2010 - 9:55pm

I don't know if they're really getting that down on Vick. They haven't said he's terrible, just that his success has been exaggerated. Everything I've read here has pointed to, "If Vick plays a full season, his rating will be closer to 80 than 100." I think that's probably pretty accurate, and a lot of people are making a lot of inferences that FO is saying Vick is terrible when they're just saying, "Vick's good but not as great as these two weeks."

(And I hate Vick and WILL refuse to watch the Eagles - it's my choice, I don't want to hear it - unless it's in the playoffs, since then I'll have no other choice.)

by chemical burn :: Tue, 09/28/2010 - 10:01pm

Listen, I understand anyone who literally can't stomach to watch Vick play, it;s just not at that level for me. Think if I were a Pittsburgh fan, I would have to stop watching when Roethlisberger comes back, just because what he did hits me personally a lot harder and I couldn't even look at him. But, yeah, I don't begrudge anyone who says "No way, can't do it."

As for FO's tone, it just feels like too much of a simple continuation of their take on him from Atlanta. In Atlanta, he really did suck, with passing DVOA normally down at around 37th in the league. And he would make one read before taking off or launching it randomly at whatever reciever he felt like throwing it towards regardless of coverage. Purely in terms of passing, according to DVOA he did SUCK in Atlanta and he hasn't sucked thus far - there's a crucial difference there.

by Aaron Schatz :: Tue, 09/28/2010 - 10:06pm

Hey. I just want to make sure: Do you mean FO's tone, or do you mean the tone of a specific (or multiple specific) FO writers? I don't remember writing anything negative about Vick this year, and Doug Farrar wrote some very positive things about him over at Yahoo.

Again, not a hive mind.

by chemical burn :: Tue, 09/28/2010 - 10:10pm

I've singled out Farrar a couple times, but failed to do so just now. I'm never sure if Barnwell writes all of the little Quick reads blurbs or just the commentary, so I suppose I only mean everything Tanier and Barnwell have written. But dammit, can't you exercise some editorial control and reign in Barnwell every now and then?

by Will Allen :: Tue, 09/28/2010 - 11:13pm

Yeah, I'd say everybody would be better off waiting at least for a few hundred attempts before making broad generalizations. For the anti-Vick sentiment (as in performance, not value as a human being) is the fact of how rare it is for a guy to become a much different player after so many years out of college. For the pro-Vick sentiment is the fact that it isn't completely unprecedented, especially when a guy has been inconsistently coached, and there is a chance that guy who lost a hundred million dollars and a couple years of his life will seize on a second chance to make a pile like, well, I guess I'll forgo the canine metaphors.

My penetrating analysis is......wait for it......we'll see.

by Basilicus :: Wed, 09/29/2010 - 11:41am

Haven't Quick Reads always risked very specific analysis in the commentaries that never fails to piss someone off? Speaking from a purely statistical aspect, Barnwell's criticisms are spot on: remove the Jackson TD and Vick opened up 2-for-8 for 6 yards and a sack.

Quick Reads have, more than any other weekly statistics report here, always taken sides, but those commentaries are one of the things that make me look forward to it. I like that the writers comment on the statistics and sometimes disagree with them. I don't always agree with the opinions, but I'm fine with writers expressing opinions. There are always at least five or ten comments with which I genuinely disagree, but when you're writing up forty-to-fifty performances, I don't see how that can be avoided.

by ohearn :: Wed, 09/29/2010 - 10:09am

Spawn more overlords.

by marcusm (not verified) :: Tue, 09/28/2010 - 9:18pm

Even if their DVOA doesn't match the understood level of quality that is supposed to accompany a 3-0 team, these numbers still bode pretty well for the Bears' chances to build up a really good record over the next several weeks. Of Chicago's next six games, only one is against a team who currently has a higher DVOA (and that's when Seattle comes to town), and most are against teams with pretty significantly lower ratings. I don't think the Bears have played especially well so far, but I think it's fairly likely they'll improve as their offense learns how to execute Martz's offense better. Just based on these rankings, Chicago has a really strong chance of being 8-1 before they hit the tougher portion of their schedule -- which still includes games against Detroit and Minnesota. The Bears just might not need to be a much better team to end the season with at least 10 wins. Which, IMHO, would rule.

by tuluse :: Tue, 09/28/2010 - 9:22pm

The playoff odds page agrees with you. http://www.footballoutsiders.com/stats/playoffodds

Mean wins: 10
Chance of making the playoffs: 72%

by Eddo :: Tue, 09/28/2010 - 11:43pm

This is a really good post. People need to remember that being 3-0 and having DVOA not like your team is better than being 0-3 and having DVOA like them.

You also articulate how I view this Bears team quite well. The schedule is set up nicely for them to capitalize on a fortunate 3-0 start.

by MainerRaider (not verified) :: Tue, 09/28/2010 - 9:50pm

I know it's getting close to my old-man-esque bedtime, but how can five teams have a 0.0% chance of making the playoffs as a top seed? I can understand 0.1%, especially with 0-3 teams, but there remains a possibility, albeit not a plausible one, that the Lions, say, go 13-3 and get the 2 seed.

by JIPanick :: Tue, 09/28/2010 - 9:55pm

It means that in 10,000 simulations that scenario happened less than 5 times (assuming FO rounds off the way I would expect).

I always found DAVE kinda sketchy, I wouldn't put any faith in the playoff odds until Thanksgiving or later.

by SteveNC (not verified) :: Tue, 09/28/2010 - 10:10pm

If they simulate the season 10000 times, and the chance of such a thing is < 1/10000, it could easily happen that such a team didn't get the top seed in any of the 10000 simulations, so the estimated probability is 0.0%.

by justanothersteve :: Tue, 09/28/2010 - 9:59pm

FWIW, while Washington may have had the best game for returns overall, back in 1967 Travis Williams had a game with only 2 returns. Both were TDs. He had 4 TDs total on only 18 returns for a 41.1 yd average. Doubt any of the current return specialists can match that.

by mm (not verified) :: Tue, 09/28/2010 - 10:01pm

Opponent adjustments will certainly reduce it, but the hole in the Saints' passing defense is really quite clear from the numbers:

vs. #1 receivers = -16.1%
vs. #2 receivers = -22.6%
vs. other receivers = 12.1%
vs. RBs = -29.0%
vs. TEs = 49.9%

I don't know that they'll improve their tight end coverage much this season. However, I do expect the rushing offense to improve from their current disaster of -31.3%, which should leave them good to make the playoffs this year.

by GnomeChumpsky (not verified) :: Wed, 09/29/2010 - 11:34am

I think that this comports with my intuitive perception, but last year, I recall the Saints being excellent against tight ends (except for the elite ones). I think that at the end of the year, the Saints might be better against tight ends than we thought, especially considering that the TEs they've faced so far this year are: Shiancoe (excellent, usually one of the highest DVOAs), Vernon Davis (exceptional, potential freak of nature despite a low DVOA), and Tony Gonzalez (Future Hall of Famer, in the discussion for best of all time). Yeah, adding a linebacker with better cover skills might help, but those three tight ends are pretty good.

by Treima6 (not verified) :: Tue, 09/28/2010 - 10:12pm

Chiefs with a 0.4% chance at 16-0.

I think my head would explode.

by drobviousso :: Tue, 09/28/2010 - 10:22pm

Ok, I feel a little vindicated with the Steeler's defensive numbers this week. After the Atlanta game, a Steelers fan blog linked to a Falcon's fan blog, and the comments read like the fans were giving up all hope. A few Steelers fans tried to point out that the Steelers D is *really really good* and you should wait a few weeks before jumping out the window.

Then the Titans game came. CJ goes from boasting about 2500 yards to shut down and the lowest stakes QB controversy outside of Cuyahoga County reared it's head again. Meanwhile, the Falcons are lighting up the Cards.

And I said "Hah! This defense is *really really good*" And now these numbers back that up. (So do these)

by marcusm (not verified) :: Tue, 09/28/2010 - 11:05pm

Wait, are you trying to tell me a bunch of Steelers fans went to a Falcons fan blog after the Steelers beat the Falcons and then tried to encourage Falcons fans to feel better about their team? Because... Either that is the most hilariously passive-aggressive form of taunting I've ever heard of... Or Steelers fans are weird...

by Spielman :: Wed, 09/29/2010 - 8:03am

Can't it be both?

by Intropy (not verified) :: Wed, 09/29/2010 - 2:54pm

I do not think it's meant to be taunting. I didn't post comments on the Falcons blog, but I did on the Steelers blog that pointer to the Falcons blog.

Here's my thinking. I don't want anyone to be down on their team. Fans of other teams should be justly excited by their successes and dissapointed by their losses. But in general I want other football fans deriving enjoyment and satisfaction from the sport.* I do not like fair-weather fansmanship. Being more interested when your team is playing well is fine and natural. But picking some random team to root for because they play well, no.

Another side of it is that you want your team to beat other good teams. I don't want to say "The Steelers beat you, haha you suck." I want to say "That was a solid matchup against another talented team, but the Steelers were good enough to win."

*Exceptions are Cowboys (American's team, WTF?) fans, Ravens fans (but only when they beat the Steelers), and any team I feel has been unsporting (for example, by videotaping other teams' playcalls) or disrespectful recently.

by qsi :: Thu, 09/30/2010 - 3:51am

Which reminds me: I saw a Steeler fan post some encouraging and friendly comments on a Bucs blog after last week's game. Strange, but I'm not complaining.

by The Powers That Be :: Tue, 09/28/2010 - 10:45pm

I wrote this in the preseason DVOA thread:

"Well, speaking as a Cowboys fan, the beginning of the season will be an interesting test of the more unconventional rankings here. DAL (ranked much lower here than by conventional wisdom) starts out with WAS and CHI, both ranked much higher here than by conventional wisdom. Then they get HOU, who many expect to improve but look terrible by DVOA. Conveniently, one of the reasons that DAL is ranked low by DVOA is the expectation of injuries on the OL, and lo and behold, they're likely to have two backups on the OL for at least the first game."

Well, Dallas lost the games to WAS and CHI, and beat HOU handily. And the OL injuries played a huge part in the week one loss. So, big points to FO and DVOA for that. Of course, the Cowboys are ranked higher than any of those teams now (by VOA and DAVE), so that's weird.

In that same thread, we got these words of wisdom from DrewBrees4MVP:

"If the Saints beat the Vikings handily tonight, as I think they will, then let's see if you show up..."
"If [the Saints] don't score 30 pts a game again this year, I'll eat my shorts."

I wonder what happened to that guy.

by JIPanick :: Tue, 09/28/2010 - 11:01pm

They may have lost to Washington, but they killed them for 60 minutes and just somehow found a way to lose. Even DVOA agrees that was the case (Cowboys #5 after week 1), so I'd hardly give the preseason formula credit for that game. The season long prediction's merits remain to be seen.

CHI is looking like a good call, but Washington's now 1-2 and the 1 was a fluke (although the Houston loss was close...dunno who outplayed who). Never mind the fact they just got clobbered by East Missouri State.

Out course, we're still not out of jump to conclusions month yet, so things could change a lot.

by Treima6 (not verified) :: Wed, 09/29/2010 - 12:45am

Don't forget this pearl of wisdom:

"Saints 7th in offense? HAHAHAHAHAHAHA"

After three games and a very mortal 20 points per game, I'm sure the Saints wish they could get to 7th in offense.

by Eric (not verified) :: Tue, 09/28/2010 - 11:20pm

San Francisco is clearly ranked too high because they are worse than the worst team in football. Placing them in hell would be way better than this. You are a slave to the limitations of the number of slots available in a ranking for a 32-team league.

by Aaron Schatz :: Tue, 09/28/2010 - 11:39pm

Unverified Eric Internet WIN.

by Eddo :: Tue, 09/28/2010 - 11:44pm

This is probably my favorite zlionsfan template usage yet. The second sentence is particularly great.

by Shattenjager :: Wed, 09/29/2010 - 1:13am

That is truly inspired.

by BigCheese :: Wed, 09/29/2010 - 3:08am


- Alvaro

by the cat in the box is dead (not verified) :: Wed, 09/29/2010 - 3:36am


On a related note, there's a guy in my office who's my Football pal- we've typically spent a few hours of each working week talking NFL.

Then he put his whole annual bonus (about £600) on the 49ers to win the West. At the time, I thought that was a fairly safe bet, QB situation notwithstanding. Now, whenever I see him he just kind of mumbles into his shirt, and walks away.

Just to add insult to injury, he checked his Fantasy football team yesterday, and remembered that for three weeks, he's been starting Kevin Kolb.

by dmstorm22 :: Wed, 09/29/2010 - 3:42am

Just be lucky he didn't put his mortgage on it.

by bravehoptoad :: Wed, 09/29/2010 - 1:26pm

I think zlionsfan template will become a new form of poetry, like sonnets or haiku. 100 years from now universities will be teaching from the Norton Anthology of zlionsfan template.

by RichC (not verified) :: Wed, 09/29/2010 - 9:09am

Has there ever been any thought of counting penalties (besides PI) as plays?

by Arkaein :: Wed, 09/29/2010 - 12:19pm

Aaron has said he's tried incorporating it into DVOA, but in general it doesn't improve the predictive power in any way (except for False Starts which I think are already incorporated). If it doesn't improve the model in a quantifiable way, it doesn't get incorporated.

Some of the reasoning behind this is that penalties of aggression (offsides, personal fouls, DPI, holding) often come with the territory of playing right up to the limit of the rules, and that even teams which get a lot of penalties are not really hurt much overall because it allows them to play aggressively in all downs, most of which are not penalized.

For non-aggression penalties other than false starts, I'm guessing that the penalties are too infrequent to contribute much to the data. Getting called for 12 men on the field a few times a season doesn't really say much about a team, for example.

by RichC (not verified) :: Wed, 09/29/2010 - 6:05pm

I was thinking holding specificly, which generally gets called on a play when a tackle/guard gets beat.

by Jetspete :: Wed, 09/29/2010 - 9:49am

thank you for ruining my morning by referencing the debacle that was week 17 between the Jets and Ravens in 2000. Vinny shredded the vaunted Ravens D that day to the tune of like 480 yards, but was pick-sixed and jermaine lewis had two PRTD's to knock the one time 6-1 jets from the playoffs.

by Led :: Wed, 09/29/2010 - 10:03am

I had to go my in-laws house that day and their TV was broken. So I BOUGHT THEM A TV just so we could watch that game. Money well spent. I can still see Vinny T throwing the pick six inside the ten yard line shortly before half-time, which totally changed the game. Not good times.

by TomC :: Wed, 09/29/2010 - 1:30pm

That story made my day, thank you.

by Sean McCormick :: Wed, 09/29/2010 - 9:55am

I always like to compare the DVOA ratings with the drive stats to see where there are big discrepancies. As it happens, there are some really big ones right now, and it's interesting to see why those numbers are so different.

According to DVOA the Steelers are playing much better ball than the Ravens heading into their matchup, but when you look at their net yards per drive, things look just the opposite--Baltimore is generating 9.43 net yards per drive, while Pittsburgh is down at 3.31 per drive. The big differences are that the Steelers are doing very well in the turnover department, while the Ravens are the worst team in the league in terms of turnover differential per drive, and as a result, the Ravens are generally playing from significantly worse field position.

San Diego is a middling team according to DVOA, but they are absolutely killing everyone in net yards per drive, topping the league at 12.51. Unfortunately, they've had farther to go, as they've generally had the worst field position of any team in the league, and clearly their special teams are putting them in big holes.

The Jets look like the team everyone expected them to be when you look at their DVOA, but their net yards per drive is bottom five in the league. If you look at the bottom five, you've got the usual suspects like Detroit, Jacksonville and Buffalo, but you also have the Jets and the Saints. It hasn't shown fully because the Jets are best in the league in turnover differential and because the Saints and Jets have benefited from consistently good starting field position.

by Led :: Wed, 09/29/2010 - 10:38am

Isn't that a little bit circular? All other things being equal, teams with bad field position will tend to gain more yards per drive because there are more potential yards to be gained and give up fewer because there are fewer potential yards to give up. That effect may not fully explain the discrepancies you identified, but it's a partial explanation. Moreover, the net yards per drive approach basically shows which teams are better if you ignore turnovers and special teams. For that to be more accurate than DVOA you'd have to posit some fluky, non-predictable results in turnovers and the kicking game that DVOA fails to filter out. That's certainly possible with a small sample size. Sanchez, for example, isn't going to throw zero interceptions the rest of the year and San Diego's special teams won't continue to be THAT bad. But DVOA factors those performances in because the general trends are likely to hold. Do you think that's not the case with these teams? (I'm open to the argument, but you haven't made it.)

by Sean McCormick :: Wed, 09/29/2010 - 11:01am

All things being equal, field position won't impact how many yards per drive you manage, but it will impact how many points per drive you put up. Anyway, I'm not positing that the yards per drive are more accurate than DVOA--they don't take opponent into account, for starters--but that they might give some clues about which teams are likely to rise or fall in the DVOA rankings if there is some outlier in their performance that will not hold up over the full season. If San Diego is gaining lots of yards on offense and not giving up very many on defense, it's likely to outweigh bad special teams over the course of a full season, even if their special teams deficiencies have had a big impact in their overall performance through a small number of games. The Jets defense has been pretty bad so far, but it's been disguised because they are generating more turnovers and because the offense has cut way back on their turnovers, to the point where the defense is getting an extra five yards worth of field to defend every drive vis-a-vis last year. If Baltimore gets their turnover situation under control, they look like they could make a big jump, because their offense and defense are winning the yardage battles drive in and drive out. It doesn't mean those things are going to happen, but I do find the numbers interesting.

by Dan :: Wed, 09/29/2010 - 1:10pm

Field position does impact yards per drive. Starting a drive on your opponent's 1 yard line will reduce your yards per drive, since you can't get more than 1 yard. Starting at midfield rather than your own 30 will also tend to reduce yards per drive, since some drives from your own 30 will go more than 50 yards. Having good field position puts a cap on the number of yards that you can get on that drive, and it also means that you'll get to the red zone sooner, where yards are harder to gain.

But I do like drive stats, especially DSR early in the season as a measure of which teams are best at moving the ball consistently (and preventing the other team from doing so). Net DSR puts San Diego, Atlanta, and Baltimore out in front of the rest of the league.

by Sean McCormick :: Wed, 09/29/2010 - 3:57pm

In the short run, absolutely. Over the course of a full season, the impact of those very short-field drives gets washed out some.

by JimZipCode :: Wed, 09/29/2010 - 3:01pm

Baltimore gets their turnover situation under control = Flacco stops having 4 INT games

by Julio (not verified) :: Wed, 09/29/2010 - 12:31pm

Why is the Pats defense ranked so high? Shouldn't they be 42nd?

by nat :: Wed, 09/29/2010 - 12:48pm

Please use the correct format for these:
[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]

For example:

The Pats defense is clearly ranked too high because they truly rank badly in more than one meaning of the word. Comparing teams by watching highlights only is way better than this. Brady rulz, but da Pats D drulz, so u all r fulz! 42! Woot!

by Treima6 (not verified) :: Wed, 09/29/2010 - 8:37pm

I had no idea the NFL incorporated the UFL while I wasn't looking.

by jebmak :: Thu, 09/30/2010 - 2:37am

Wait for opponent adjustments to kick in.

by Jonadan :: Wed, 09/29/2010 - 2:11pm

Bizarre thing I found out that I don't know where else to mention: in three weeks, no Giants opponent has attempted to return a kick or punt from the endzone.

by Marko :: Wed, 09/29/2010 - 6:20pm

What is bizarre about the kickoff return part of this? Maybe none of their kickoffs reached the end zone. Or if they did reach the endzone, maybe they were mishandled or were too deep to return. I have no idea, since I haven't watched much of the Giants or paid attention to their kickoffs.

As for the punt return part of this, not only is this NOT bizarre, but it is the same for every team. No team has attempted to return a punt from the end zone this year, nor will they do so. In fact, no team has attempted this for several years, since the rule change that made any punt that reached the end zone a touchback and a dead ball. So you can't return a punt from the end zone anymore, even if you want to. Which means that Robert Bailey's record 103 yard punt return TD for the Rams will forever remain the record for longest punt return, barring a future rule change.

by Jonadan :: Wed, 09/29/2010 - 11:39pm

On further inspection, yeah, it's not as weird as I thought, especially since I didn't know about the punt rule. For instance, the Panthers haven't attempted a single return from the endzone either.

by riccja :: Thu, 09/30/2010 - 11:06am

I know a lot of people of some doubts about the predictive ability of these numbers. I have found that they are very predictive but only if you incorporate them into a handicapping routine along with other significant variables. I have been doing that successfully for some time now.

If you use these numbers alone however you wont have much success.

Twitter: rickjsportplays

by doofman (not verified) :: Thu, 09/30/2010 - 1:11pm

Can someone explain why Dallas has a running offense ranked 11th, while the Bears are 30th, despite all the conventional stats showing little separation? Is it entirely situational running, or is there something else weird going on here?

by nat :: Thu, 09/30/2010 - 1:32pm

Looking at conventional stats, I notice that Dallas has more than twice as many first downs rushing than Chicago does. That's probably either successful long gains on passing downs, or successful runs in situations where defenses are expecting a run. Either one scores well in DVOA.

Chicago must be getting a lot of its yards either while falling short of a first down, or in a smallish number of long runs. DVOA discounts very long runs as nonpredictive, and doesn't much reward medium length runs in passing situations.

Nothing weird. These are exactly the distinctions DVOA was designed to make.

by doofman (not verified) :: Thu, 09/30/2010 - 1:42pm

That makes sense, I see the first down numbers now - 20 rushing first downs in only 69 rushing attempts versus only 9 for the Bears would make a big difference. It seems unsustainable in the long run, either they'll have to run more to keep getting those first downs, or they'll start converting at a more "normal" rate and the rankings will even out, but anything is possible. Thanks for the explanation!

by tuluse :: Thu, 09/30/2010 - 1:46pm

Well Chicago isn't getting a lot of long runs, I'll tell you that.

I think it's two things. One the first downs that you mentioned, 2 yards on 3rd and 2 is a lot better than 2 yards on 1st and 10. The other is goalline situations. The Bears have been stuffed on about what 7 goalline attempts now?

by nat :: Thu, 09/30/2010 - 3:11pm

Good point about goalline situations. Those are plays where it's impossible to get an "average" result. You either get in or you don't. So success or failure gets amplified or at least emphasized in DVOA. Of course, it really does matter whether you can run the ball in from the two.

So there's hope for Chicago. If their goalline failures were due to lucky defensive guesses, we can expect a "regression" towards better luck, more TDs, and a higher rushing DVOA. Or they could just really be bad in short yardage runs. Time will tell.

by doofman (not verified) :: Thu, 09/30/2010 - 1:51pm

Yeah, I thought about the long run issue but neither team has had a rush longer than 18 yards, so I discounted that idea, but maybe Dallas has significantly more 10-15 yard runs, which would also explain it (although given their equal yards per rush it would mean they were also getting stuffed more which seems unlikely, especially given the Bears goal line rushing problems that you mentioned)

by ammek :: Thu, 09/30/2010 - 4:50pm

Adjusted Line Yards (one of FO's stats) says that the Cowboys have gained fractionally fewer yards 10+ yards past the line of scrimmage.

Chicago's problems are power running (worst in the league) and getting stuffed (next-to-worst).

Also, Dallas is averaging 0.8 yards more than Chicago per running back carry. The closeness in conventional stats may be in part due to Jay Cutler's nine rushes for 66 yards.

by doofman (not verified) :: Thu, 09/30/2010 - 5:11pm

I forgot those numbers were on the O-line page, thanks for the tip!

by StatMan (not verified) :: Mon, 10/04/2010 - 10:42pm

I was a little disappointed to find that DAVE provided no good predictions of this week. All I did was subtract the away team's DAVE from the away team's, and found that there's no linear relationship between point differential and the DAVE differential. 8 home teams won these week: only 4 were predicted to win.
Now, I probably could have taken a look at things like comparing pass offense vs pass defense, but I would've liked to have seen predictive measures here.

by dbostedo :: Wed, 10/06/2010 - 8:50am

There is no way that sports teams are anywhere near consistent enough - especially not in football where sample sizes are small - to show a clear linear relationship between any rating and their points scored/allowed on a weekly basis. Maybe if every team played exactly the same from week to week, it would be more predictive - but if it was perfectly predictive, why play the games?

Add to that the fact that DVOA is based on defenses that have already been faced, and you'd need some way to factor in the upcoming opponent (which you could probably make a good guess at from the current values).

What I'd guess is that it's somewhat predictive on average - i.e. some things will be way off, and others right on, but overall it will be better than random. How much, I don't know.

by DeltaWhiskey :: Wed, 10/06/2010 - 10:31am

Week 8 DVOA predicts end of season record as well as Week 8 win percentage.