Writers of Pro Football Prospectus 2008

Most Recent FO Features

SimBla15.jpg

» OFI: Don't Make Saban Angry

Notre Dame and Baylor entered the one-loss group in what is shaping up to be an extremely tight race for playoff consideration.

12 Sep 2006

Week 2 DVOA Ratings

by Aaron Schatz

You love them when your team is high! You hate them when your team is low! DVOA ratings are in effect for 2006!

This year we introduce a new feature of the DVOA ratings: An early-season rating that combines the preseason projection with the results of early games to give us a better prediction of how each team will rank at the end of the year. Can you guess how much of this rating is based on the first week's games?

Four percent.

Yes, that's it, and yes, I was pretty surprised too. In fact, when I tried to create a formula that combined preseason projection and Week 1 to get the most accurate prediction of DVOA for the final 15 games, the result was the same every time: the preseason projection, on its own, was the most accurate predictor no matter what the Week 1 result was. I simply could not build this "early-season weighted DVOA" in the same way I build the late-season weighted DVOA. So what we have here is a formula that predicts the total DVOA for all 16 games, not the DVOA for the remaining games. Those numbers are listed below as "PROJ + WEEK 1." I hope that makes it clear to everyone not to jump to conclusions based on the rating of a single game.

My goal was also to add home-field advantage adjustments to the DVOA ratings, but I didn't quite finish the mechanisms there. Hopefully that gets in by the end of the month. And of course, these are the improved DVOA ratings that are both more predictive, and more accurately correlated with winning. The changes in the method are detailed in Pro Football Prospectus 2006.

The plan at FOXSports.com this year was to run twin power rankings: one statistical, one conventional. To be honest, I think that's a little confusing, but it means I don't have to refer to these as "power rankings" anymore, and I'll feel less pressure to make any subjective changes or defend my numbers to the masses. The first commentary of the season is found here on FOXSports.com

You should also check out the new Football Outsiders FOX Blog. This is a new feature where I can put my random thoughts, mention the various little goofy stats I come across when I have no idea where else to mention them, and ask questions and just get data to see interesting things. Plus Mike Tanier will also be writing a little, and later I hope to use this to answer mailbag questions and post occasional scouting thoughts from the game charters. Enjoy the first couple posts, which have been completely taken the wrong way by nearly all commenters. In the proper FO tradition, I guess.

* * * * *

These are the Football Outsiders team efficiency ratings through one week of 2006, 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 DVOA is adjusted for type of stadium (warm, cold, dome, Denver/Mexico City) and week of season.

There are no opponent adjustments for the first three weeks of the season, which is why offense and defense are "VOA" and not "DVOA."

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

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


TEAM
TOTAL
VOA
W-L PROJ+
WEEK 1
RANK OFFENSE
VOA
OFF.
RANK
DEFENSE
VOA
DEF.
RANK
SPECIAL
DVOA
S.T.
RANK
1 BAL 74.3% 1-0 2.9% 15 1.9% 13 -72.1% 1 0.3% 15
2 CHI 67.5% 1-0 13.6% 8 7.8% 9 -31.1% 6 28.6% 1
3 SD 63.4% 1-0 23.0% 3 16.8% 5 -48.5% 2 -1.9% 20
4 ATL 57.7% 1-0 10.5% 10 16.1% 6 -43.1% 3 -1.5% 19
5 PHI 47.0% 1-0 13.6% 9 46.9% 2 2.6% 21 2.6% 13
6 STL 46.6% 1-0 -21.6% 27 -1.8% 15 -40.5% 4 8.0% 5
7 NYJ 36.9% 1-0 -24.0% 31 27.9% 3 -22.3% 10 -13.3% 28
8 CIN 36.0% 1-0 19.8% 4 -1.9% 16 -29.7% 7 8.1% 4
9 NO 30.3% 1-0 -31.3% 32 -4.2% 17 -35.0% 5 -0.5% 17
10 NYG 19.7% 0-1 2.7% 16 47.7% 1 13.6% 25 -14.4% 29
11 JAC 18.9% 1-0 4.8% 14 7.1% 10 -19.3% 11 -7.6% 26
12 PIT 13.5% 1-0 15.5% 7 5.4% 11 -26.1% 8 -18.1% 31
13 SF 11.2% 0-1 -24.0% 30 23.7% 4 15.8% 26 3.3% 11
14 SEA 9.5% 1-0 38.5% 1 -15.7% 21 -22.5% 9 2.7% 12
15 BUF 9.1% 0-1 -21.9% 28 -0.9% 14 -5.7% 14 4.3% 7
16 MIN 1.7% 1-0 -18.5% 26 -5.2% 18 -12.9% 13 -6.0% 23
TEAM
TOTAL
VOA
W-L PROJ+
WEEK 1
RANK OFFENSE
VOA
OFF.
RANK
DEFENSE
VOA
DEF.
RANK
SPECIAL
DVOA
S.T.
RANK
17 NE -0.9% 1-0 6.5% 12 -5.7% 19 -0.9% 19 3.9% 9
18 ARI -13.9% 1-0 -17.4% 25 15.8% 7 23.7% 29 -6.0% 22
19 WAS -14.3% 0-1 6.3% 13 -12.9% 20 -5.2% 15 -6.6% 25
20 MIA -20.2% 0-1 -1.3% 18 -26.1% 25 5.4% 22 11.4% 3
21 IND -20.4% 1-0 25.6% 2 13.6% 8 47.7% 32 13.7% 2
22 DET -21.9% 0-1 -9.3% 21 -22.5% 24 -15.7% 12 -15.1% 30
23 CLE -25.4% 0-1 -9.5% 23 -35.0% 28 -4.2% 16 5.4% 6
24 DAL -32.9% 0-1 -3.7% 20 -19.3% 22 7.1% 23 -6.5% 24
25 KC -36.7% 0-1 17.0% 5 -29.7% 26 -1.9% 17 -8.9% 27
26 DEN -38.8% 0-1 15.9% 6 -40.5% 29 -1.8% 18 -0.2% 16
27 HOU -45.0% 0-1 -22.0% 29 2.6% 12 46.9% 31 -0.6% 18
28 TEN -53.9% 0-1 -9.4% 22 -22.3% 23 27.9% 30 -3.7% 21
29 CAR -55.3% 0-1 8.2% 11 -43.1% 30 16.1% 27 3.9% 10
30 OAK -61.2% 0-1 -0.9% 17 -48.5% 31 16.8% 28 4.2% 8
31 GB -67.7% 0-1 -15.7% 24 -31.1% 27 7.8% 24 -28.8% 32
32 TB -73.0% 0-1 -3.1% 19 -72.1% 32 1.9% 20 1.0% 14

Posted by: Aaron Schatz on 12 Sep 2006

112 comments, Last at 21 Sep 2006, 7:39pm by Smartmonies

Comments

1
by The Mulgrew (not verified) :: Tue, 09/12/2006 - 10:48pm

The Eags are 5th! I knew they were going to have a good year. Of course, Houston wasn't a great test. After the Eags beat the Giants on Sunday, people will recognize.

2
by sicksock (not verified) :: Tue, 09/12/2006 - 10:56pm

I can't see Oakland reaching the PFP projection of 8 wins. It was worse than anything I saw from SF last year. Matt Leinart has to be happy they passed over him.

3
by David (not verified) :: Tue, 09/12/2006 - 10:58pm

Hey, unadjusted value over average! That's a blast from the past.

And seeing the Eagles rank 21st in defensive VOA after playing the Texans makes me scared of what the DVOA ratings are going to look like. Hopefully that's just skewed by the absurd number of shutouts this weekend.

4
by Led (not verified) :: Tue, 09/12/2006 - 11:08pm

Wow. Indy's VOA was 40% lower than the Jints and they still won. That must be a record.

5
by DGL (not verified) :: Tue, 09/12/2006 - 11:20pm

After reading the comments on the FOFox Blog, all I can ask is, Why do you bother? Why expose yourself to such vitriol when you could put up the same content here on FO and get a lot more rational commentary?

6
by KnickerBlogger (not verified) :: Tue, 09/12/2006 - 11:22pm

I can understand that week 1 affects pre-season predictions by only 4% on average for the entire league, but is there a correlation between where this percentage is higher?

I would like to see the percentage for teams that finish either in the top/bottom 10 and begin the season on the other end of the spectrum. The hypothesis is that statistical probability doesn't account for offseason improvement like acquiring (or getting back from injury) one or more players that can change the fortunes of the team. Or having a coach that can turn the team around.

It would be cool if statistically we can account for changes like adding Bill Parcels to the Jets in 1997. The year prior they went 1-15, and in week 1 they demolished the Seahawks 41-3. They finished an impressive 9-7 with largely the same team.

7
by Derek (not verified) :: Tue, 09/12/2006 - 11:23pm

Wouldn't it make more sense to order it by projected rankings and not the week 1 rankings? Its not that its confusing, its that the 1st thing I would want to see is the overall ranking and then I would look for the rest.

8
by Duff Soviet Union (not verified) :: Tue, 09/12/2006 - 11:26pm

Holy crap. Two teams were actually worse than the Raiders this week? Neither of those teams has the excuse of playing San Diego either.

9
by Sunfish (not verified) :: Tue, 09/12/2006 - 11:32pm

Wow, Jets 7th... unbelievable...

10
by David (not verified) :: Tue, 09/12/2006 - 11:37pm

Remember this is pure value over average, which means the quality of the opponent isn't taken into consideration. A 15-yard pass on 1st-and-10 against the Raiders counts exactly as much as one against the Bears. In other words, this is no measure of how good a team is objectively, just how good their numbers looked this week.

I'm still shocked that the Bay teams were worse than Oakland, though.

11
by paytonrules (not verified) :: Tue, 09/12/2006 - 11:37pm

I hope you don't feel bad after reading comments like:

QUIT THE CEREBIAL ALREADY ..

So he want you to quit being Cerebral - or eating cereal? I'm really not sure.

It's probably just the East Coast bias of the numbers that go something like VOA + Coast + ManningIsGod.

When the Madden curse is the most reasonable response in the thread, it's bad.

12
by David (not verified) :: Tue, 09/12/2006 - 11:43pm

Oh, wait, I'm not shocked anymore. I forgot San Diego stopped trying to move the ball in the third quarter, thus making the Raiders look like they had a run defense (or at least a less horrible run defense). That makes sense, then.

13
by CA (not verified) :: Tue, 09/12/2006 - 11:46pm

Re: 10

A 15-yard pass on 1st-and-10 against the Raiders counts exactly as much as one against the Bears.

Well, the Raiders are projected to have the third best defense in the NFL, so that is about right.

14
by admin :: Tue, 09/12/2006 - 11:50pm

Update: Apparently, the reason Quick Reads never ran this week is that I never sent it in. So, I'm an idiot. Hopefully this is solved shortly.

As for comment 5, it's all about the kesef.

15
by Yaguar (not verified) :: Tue, 09/12/2006 - 11:55pm

The teams with real quality wins beyond what was expected are Baltimore, Atlanta, and St Louis. But without enough games to adjust VOA, it's hard to get much meaningful information out of it.

Jesus, check out this comment on the blog:

"After correctly blasting Mike Tanier last year regarding his idiotic suggestions about the Seahawks (and history shows he was wrong about everything, except Sea vs Car. in the NFC Champ game, BUT he gave the win to the Panthers [hey, Mike, GOOD CALL STOOPIT], allow me to continue my all out assault on the jackasses at F.O. Why dont you guys use this talent for analyzing numbers towards something useful. I agree with dadc43. The O-Line had a bad day, they have new guys (DONT blame this all on Hutch leaving either you nincompoops), and that accounted for BOTH Shauns bad day AND Matts bad day. Lets revisit this assessment after the game vs. the Cards. If Shaun has another bad game, well, you guys are still a*holes..."

Yikes. Teach them the angry troll format, please.

16
by Yaguar (not verified) :: Tue, 09/12/2006 - 11:57pm

12: What's sad is that SD still averaged 4.0 yards a rush despite continual 9-in-the-box efforts.

17
by PatsFanDan (not verified) :: Tue, 09/12/2006 - 11:57pm

Aaron, Given all the botched calls this weekend, you will have to develop a new measure - RVOA, or ref adjusted VOA.

Or you can determine what a ref's worth is to the team over a replacement ref named Billy Joe.

18
by steve (not verified) :: Wed, 09/13/2006 - 12:26am

How about some in season projections. Taking into account current record, remaining schedule and DVOA. Probability to win division, wild card, etc...

19
by navin (not verified) :: Wed, 09/13/2006 - 12:26am

San Fran outplayed Arizona? Was this because of the way the fumbles went? Arizona actually had 3 fumbles but only lost one. San Fran lost both fumbles.

20
by Pat (not verified) :: Wed, 09/13/2006 - 12:30am

The hypothesis is that statistical probability doesn’t account for offseason improvement like acquiring (or getting back from injury) one or more players that can change the fortunes of the team. Or having a coach that can turn the team around.

All of this would be true, if the DVOA projections didn't include offseason changes, coaching changes, quarterback changes, etc. But it does. So... yeah. Next?

21
by zip (not verified) :: Wed, 09/13/2006 - 1:04am

#14

Is that yiddish for nookie?

22
by CaffeineMan (not verified) :: Wed, 09/13/2006 - 1:06am

DVOA is here, excellent! Ask and ye shall receive. Ahhhh, the cool, refreshing sensation of actual data.

40% difference between NYG and IND? WTF happened in that game? I didn't see it, so I can't wait to see the commentary on that one.

Pats 19th O and D. Sounds about right for how they played. Oh, and your blog entry about Branch sounds about right, too, depressing as it was.

23
by centrifuge (not verified) :: Wed, 09/13/2006 - 1:21am

Man, I can't wait for DVOA to kick in. I tried to look at the "conventional" rankings on Fox, and I had to stop. I mean, I can be the biggest Steelers homer on Earth, but I can't accept them at #1 when I look at the game the Ravens had.

24
by centrifuge (not verified) :: Wed, 09/13/2006 - 1:26am

To be clear, I don't think the Ravens will end up being better than the Steelers, but they certainly were better this week. If that's what you're going to make your rankings on (how teams look right now), then do that.

25
by Pat (not verified) :: Wed, 09/13/2006 - 1:42am

The thing about early season DVOA (and VOA this week, of course) is that the early major outliers on defense almost always tend to be flukes due to small numbers of incredibly powerful plays - usually interceptions and fumbles. Last year it was Cincinnati's defense, which plummeted from Week 4's -61.9% to 1.0% by the end of the season. In 2004 it was Seattle, whose defense went from -48.3% in Week 3 to 6.3% by the end of the season, and in 2003 it was San Francisco and Seattle, who similarly went from -60% to 0.9% and 1.4% (you should be noticing a pattern of "eventual mediocrity" here). In 2005, it was Cincinnati's crazy wacky interception run. In 2004 it was Seattle causing 12 turnovers in 3 games. In 2003, Seattle caused 17 turnovers in 3 games, and San Francisco caused 14 turnovers in 3 games.

This too, shall pass.

The Ravens are a good defense, but they're not obscenely ultra-mega-powerful on defense. My money would be on them ending up around -10% defensive DVOA.

26
by Bobman (not verified) :: Wed, 09/13/2006 - 1:43am

Indy's ST Rank #2! Whoo-hoo, read it and weep! Rob Morris is a tackling machine! (Shame he never was one when he started at MLB for five years, but y'know, maybe he's a late bloomer....)

What's that, their D is ranked, uh, a bit lower? Must be a glitch in the system.

But Number 2 ST! Whoo-hoo! Now all we have to do is get that D ranking up to, about, oh, 30 or so.

27
by Bobman (not verified) :: Wed, 09/13/2006 - 1:52am

Aaron, cool Fox blog. It hurts me to praise anything with the Fox name on it, Simpsons excepted. Especially after Arrested Development was cancelled.

28
by turbohappy (not verified) :: Wed, 09/13/2006 - 1:53am

Re: Indy vs. NY

Do penalties count in VOA? I can't remember from last year and didn't find anything about it in the explanation. If not, this explains a lot about how Indy was able to win. Other than the penalties, the Giants were definitely the better team. That the gap was THAT big is surprising though.

29
by turbohappy (not verified) :: Wed, 09/13/2006 - 1:57am

Re: 26

He really has come to embrace his ST role since no other team wanted him when he tested the waters, it seems. I don't know if this will hold up at all, but it could make up (somewhat) for their probable decline on both offense and defense.

30
by Dave Brude (not verified) :: Wed, 09/13/2006 - 1:57am

RE the SF-ARI and IND-NYG games.

SF did enough to at least play to a tie ball game on a play by play perspective but had a couple bad lost fumbles. 1 by Vernon Davis and another by Bryant I think. Vernon's was in the Niners own zone which set up an easy TD.

Arizona got back two of Warners 3 fumbles. Also SF failed to punch it in from close range once and Nedney shanked an easy FG.

Indy definitely looked bad. NYG shot themselves in the foot with penalties, turnovers and bad timing. Should of had an easy victory.

31
by Yaguar (not verified) :: Wed, 09/13/2006 - 1:59am

It's partly the penalties, which IIRC don't count at all. Also, DVOA loves run defense. The Colts last Sunday appeared to have studied at the prestigious Kansas City Chiefs School of Not Tackling Tiki Barber. Hence their absurdly bad defensive VOA.

32
by jonnyblazin (not verified) :: Wed, 09/13/2006 - 2:09am

Re: 26
Uh, yeah, I don't think anyone will think that the Ravens are a -70% DVOA defense, but I do think the fact that Air McNair marched them down the field and scored a TD (on the Bucs D no less) on the first drive was instrumental in generating the turnovers the Bucs later committed. If the offense is able to actually score TDs on a semi-regular basis, the opposing offense will be forced to play more aggressive, and thats where Reed, McCalister, Rolle, Suggs, Thomas, etc, will be able to make big plays. Last year opposing offenses didn't take chances, they waited around until Boller (or Wright) f*cked up (it usually didn't take long) and then took a chance in good field position. They'll get more than 11 INTs this year, that's for sure.

33
by Michael (not verified) :: Wed, 09/13/2006 - 2:16am

the future looks bright in san fran!!!!

34
by K. Derek7 (not verified) :: Wed, 09/13/2006 - 2:31am

Falcon fan here... can someone explain in simple terms how the DVOA works?

35
by Richie (not verified) :: Wed, 09/13/2006 - 2:32am

Miami is clearly ranked too low because the NFL has a collective man-crush on the Pittsburgh Steelers and will do anything to make sure they win games. Miller's touchdown should have been called back and then it should have been reviewed on Saban's flag, but the incompetent senior citizens in funk striped shirts are biased against the fabulous team from Miami. My ranking system is way better than this because I factor in the hart of Dante Culpepper and the inherited genious of Nick Saben. Did you know that he once shook hands with Bill Bellicheck? Bellicheck taut Saben everything, so the Dolfins will be in the Super Bowl!!!!!!! F.O. sux!

36
by David (not verified) :: Wed, 09/13/2006 - 2:44am

This is a bare-bones, 1:30 AM summary. For detail, see the official explanation, linked in my name.

Value over average looks at the conditions of each play a team participates in (down, distance and field position...I think it also includes score, but I'm not sure), and compares their success on that play to other teams in similar situations. The play's VOA is how much success it had relative to the league average in similar conditions. The formula for "success" is complicated and the site addresses it in greater depth on the "our new stats explained" page. Suffice to say it's not just how many yards were gained. Average that out through an entire game, and you have team VOA for the week.

Defense-adjusted VOA does the same thing, but then adjusts for the overall quality of the offense or defense the team was facing. For example, if you're playing the 2005 49ers, picking up three yards rushing on first and ten would be below average. Those three yards against the Bears or Steelers would get you a higher score - maybe even a positive DVOA.

Right now, since each team has played only one other team, it's impossible to know whether their performance is due to their own quality or their opponent's. For example, we have no way to know, working from pure stats, whether the Giants' rushing attack is really good, Indy's rush defense is really bad, or if it's somewhere between the two. Defense adjustments will only enter into the picture once there's a broader array of data to analyze.

Scoring: For offense, positive DVOA is good. For defense, negative is good - it represents the amount by which the other team's offensive production was decreased compared to normal. Total DVOA is, of course, a combination of the two, plus special teams (where positive is also good).

Hope that helped.

37
by Elan (not verified) :: Wed, 09/13/2006 - 3:30am

Re: #21

It's Hebrew for money.

38
by Todd (not verified) :: Wed, 09/13/2006 - 4:24am

Come on Schatz. Make us Jews proud. You can figure out the intro.

Interestingly, the conventional rankings which were recently posted on FOXSports.com was done by another one of us...haha! Mr. Peter Schrager.

http://msn.foxsports.com/nfl/powerRankings

Unfortunately I dont think Carucci (NFL), C.Robinson (Yahoo), Prisco (CBS), Alder (About.com), etc. are Jewish.

You are part of a rare breed.

One last thing:
Check out my blog, Aaron.
I track your power rankings against all the other rankings at my blog in a nice big excel sheet. I also use all the ranking in a conventional way...higher ranked = pick for each weekly game. I suggest you check out how your rankings are doing in my picks excel sheet...

All at my blog, which I set as my website when posting this comment.

39
by Mr Shush (not verified) :: Wed, 09/13/2006 - 4:46am

David (3) - Honestly, I think that the Eagles' mediocre defensive VOA is the result of a pretty good defense going against a pretty good offense on the road. My pre-season expectation was that the Texans would have an offense in the 8-16 range (by DVOA) but suck on defense once again and finish with a 5-11 or 6-10 record. I expected the Eagles to be pretty good on both sides of the ball but amazing on neither, making them a genuine contender in the NFC and a likely division winner, but probably not good enough to go to the Superbowl and definitely not good enough to win it. Nothing I saw on Sunday made me think I was wrong about any of this.

40
by ammek (not verified) :: Wed, 09/13/2006 - 6:02am

I’m surprised at the four percent figure. I guess it shows your confidence in your preseason projections. I have no data to contradict you, so I’ll let myself be carried along by the DVOA breeze.

I agree it’s dangerous to jump to conclusions after week one. However, using the simplest calculation possible - far less “fastidious� than DVOA - it appears that week one games are congruous with the rest of the season 75% of the time. Hence my surprise at the four percent figure.

(The calculation: take the week one results, fast forward to the end-of-season record, adjust for home field advantage (roughly .600), and see which of the teams would be favorite. Eg, last season, 12-4 Jacksonville beat 13-3 Seattle at home: a predictable result. In 2004, the 9-7 Jags beat 9-7 Buffalo on the road: an upset.)

Over the past two seasons, there have been 8 upsets in 32 opening-week games. (The proportion is similar through five years, too.) Five of them have been road wins by equal or marginally superior teams (eg, 11-5 Tampa at 9-7 Minnesota in 2005); two were home wins by significantly inferior teams (ie, .200 or more below the opponent), namely 9-7 Miami over 13-3 Denver in 2005 and 4-9 Cleveland over the 9-7 Ravens in 2004; and the other was the extraordinary Saints’ road win in Carolina last season.

Oh, and from a purely subjective point of view, I trust the DVOA Week 1 ranking (32nd by miles) rather than the preseason projection (24th) regarding the Packers’ special teams. Watching them vainly grope after Hester on the Bears’ punt return reminded me of the good old days of botched snaps in the snow and 4-12.

41
by ADF (not verified) :: Wed, 09/13/2006 - 7:58am

Four percent is one twenty-fifth. Speaking very broadly, this implies that last season's DVOA is worth about 1.5 season's worth of games. That seems unlikely, as it contains only one season's worth of games.

The only reason I can even fathom why this might be so is this is a result of the "unconnected graph" problem. That is, as you mention, at the moment you can't tell whether Week 1 winners were good, or Week 2 winners are bad. At some point, the network graph of teams becomes completely connected (no islands). Will the weighting factor then rapidly adjust upwards? After all, you might expect it to be at least 6%, 1/16th of all games...

42
by Moe (not verified) :: Wed, 09/13/2006 - 9:58am

I'm in Korea on business so watched the monday night games on my computer through a Slingbox. Pretty cool technology. Even cooler that the Vikings and old man Opie came through in Washington.

And oh by the way its true that Hines Ward is in a TON of TV commercials over here.

43
by KnickerBlogger (not verified) :: Wed, 09/13/2006 - 10:12am

#20. Then my question still stands. What is the correlation between teams that have preseason DVOA which rank them in the top/bottom X%, but then have a week 1 that is in the opposite end of the spectrum?

44
by Andrew (not verified) :: Wed, 09/13/2006 - 10:38am

steve #18:

How about some in season projections. Taking into account current record, remaining schedule and DVOA. Probability to win division, wild card, etc…

If you have the patience to wait until next week, you can effectively write off the post- season hopes of the probably 8+/- teams that will end up 0-2. The chances that a team that starts 0-2 pulls it together are slim to none.

2005 - 8 teams start 0-2, all no playoffs.
2004 -7 teams start 0-2, all no playoffs.
2003 - 8 teams start 0-2, only Eagles make playoffs (and do so as a 1st seed!).
2002 - 8 teams start 0-2, only Falcons and Steelets make playoffs (Steelers 3rd seed, Falcons 6th seed)
2001 - 9 teams start 0-2, only Patriots make playoffs (as a 2nd seed)
2000 - 10 teams start 0-2, none make playoffs
1999 - 8 teams start 0-2, none make playoffs
1998 - 11 teams start 0-2, 3 make playoffs (Jets 2nd seed, Bills 5th seed, Cardinals 6th seed)
1997 - 6 teams start 0-2, none make playoffs
1996 - 10 teams start 0-2, 1 makes playoffs (Patriots 2nd seed)

That's 7 teams of 85 over 10 years, or 8% of teams starting 0-2.

So for 8 or 9 teams, next Sunday, the playoff dream for 2006 essentially will already be over.

45
by throughthelookingglass (not verified) :: Wed, 09/13/2006 - 11:35am

checking the fo blog...ow, that's painful to read. and how funny is it that the deion branch trade generated one response. (titan troll!)

46
by Wanker79 (not verified) :: Wed, 09/13/2006 - 11:44am

Just in case anyone else was wondering about strength of victory, here are the Total VOA differentials in each of the games.

Bal/TB: 147.30
Chi/GB: 135.20
SD/Oak: 124.60
Atl/Car: 113.00
Phi/Hou: 92.00
NYJ/Ten: 90.80
StL/Den: 85.40
Cin/KC: 72.70
NO/Cle: 55.70
Jac/Dal: 51.80
Pit/Mia: 33.70
Sea/Det: 31.40
Min/Was: 16.00
NE/Buf: -10.00
Ari/SF: -25.10
Ind/NYG: -40.10

47
by WookieeBH (not verified) :: Wed, 09/13/2006 - 12:03pm

44 - So essentially Buffalo-Miami, Kansas City-Denver, and Washington-Dallas are effectively playoff games for both teams? That's especially interesting since each of those matchups are divisional.

48
by Jason (not verified) :: Wed, 09/13/2006 - 12:56pm

#47:

That's an excellent point, though if each of the underdogs wins those games, I would have a hard time writing off the loser/favorites for the rest of the season, historical percentages notwithstanding.

49
by weinsteinium (not verified) :: Wed, 09/13/2006 - 1:13pm

And on sunday the Ravens end the Raiders postseason hopes.

BTW #12, I was at the Raiders game on monday and I don't think that the Chargers were actually doing anything differently in the 3rd quarter. I think that the Raiders finally realized that the Chargers were just going to run out the clock and the Raiders starting playing more men in the box. And calling run blitzes, at one point Huff (SS) had two staight tackles at or behind the line of scrimage.

50
by Scott de B. (not verified) :: Wed, 09/13/2006 - 1:40pm

Four percent is one twenty-fifth. Speaking very broadly, this implies that last season’s DVOA is worth about 1.5 season’s worth of games. That seems unlikely, as it contains only one season’s worth of games.

Preseason DVOA is not simply last season's DVOA.

51
by Ferg (not verified) :: Wed, 09/13/2006 - 1:54pm

44: Interesting. Do you think the record in weeks 1 and 2 is any more significant than any other two weeks? For example, I have a hunch that if you looked at teams that went 0-2 in, say, weeks 4 and 9, you'd again find about the same small fraction making the playoffs.

52
by Nate (not verified) :: Wed, 09/13/2006 - 2:04pm

Aaron, you really have to disable comments on that blog. Ouch.

53
by Bobman (not verified) :: Wed, 09/13/2006 - 2:07pm

Ferg,
Interesting idea and it probably holds some water, though I wouldn't look much past week 12 when dinged-up starters sometimes get rested "for bigger upcoming games" and some teams clearly know where their postseason fates will lie.

The reason your suggestion might not be as accurate is that when a team that thought it was a bubble team going in to the season has their old vet QB/RB start and stink things up going 0-2 or 0-3, by game 3 or 4 they might put in the rookie to get him experience and shut the fans up. This results in a few more losses than they'd otherwise have, but good experience, and definitely no trip to the postseason that year. (Think, potentially Ariz this year or NYG when Eli was a rookie)

But if that team wins one of the first two, say they're 2-2 after four games, they may stick with the program longer and try to eke out a winning record, rather than season the youngsters. They may even make the playoffs.

No idea how often an 0-2 record causes coaches to start an all-rookie roster (I suspect an 0-6 start is more likely to do that, and by then you're pretty much guaranteed to miss the postseason anyway, unless you're playing in the NHL), but it's a theory.

54
by B (not verified) :: Wed, 09/13/2006 - 2:18pm

49: The Raiders postseason hopes were ended when they signed Aaron Brooks.

55
by Tom V (not verified) :: Wed, 09/13/2006 - 2:18pm

The effort to create a formula to rank football teams is futile, particularly early in the season. The reason? It's all based on stats, and stats lie. While a useful tool in ranking the teams, the human element and common sense should override the raw data. Ever see the 4 computer models for a hurricane track? 4 different directions? Same with equation based systems. Use this system to help make a "power ranking" but don't rank the Jets at #7 or the 49ers at #13 while placing the Colts at #23.

I'm sorry, let me use the template:(this is INTENTIONAL, I'm following the template that Football Outsiders provided.. it's intended to be funny)

is(template error, should be 'are') clearly ranked because . is way better than this.

or

is(template error, should be 'are') clearly ranked because . is way better than this.

Enjoy...

56
by Bobman (not verified) :: Wed, 09/13/2006 - 2:21pm

Crap, Aaron, Nate's right. If this is typical of the FOX blog comments: "Do the world a favor and go drink some anti-freeze. Moron." Turn off the comment box.

It DOES make me appreciate the civility and mental acuity of FOers. And the fact that trolls tend to get stomped by the regular inhabitants. Even the Peytom Branning thread didn't feature this kind of mindless bile. (Even though I am pretty sure I was hyperventilating a few times as I contributed to it.)

57
by Andrew (not verified) :: Wed, 09/13/2006 - 2:29pm

WookieBH and Jason:

44 - So essentially Buffalo-Miami, Kansas City-Denver, and Washington-Dallas are effectively playoff games for both teams? That’s especially interesting since each of those matchups are divisional.

Yes, especially if Philadelphia, San Diego, and New England or the Jets win. 0-2 is just a really big hole to climb out of when you are chasing a team at 2-0 that already has a leg up in divisional match-ups, and other teams at 2-0 that are ahead of you for the probable wildcard. 2 or even 3 games thanks to tie-breakers, is a lot of ground to make up in just 14 games.

The vast majority of the time, teams that start 0-2 end up at 8-8 or less, and most end up 6-10 or less. I noted that 7 of 85 teams starting 0-2 made the playoffs. Only another 4 teams made a winning record and missed the playoffs after starting 0-2 (2000 Steelers @ 9-7, 2000 Packers @ 9-7, 2004 Bills @ 9-7, 2005 Chargers @ 9-7). An 11 in 85 chance of a winning record is just 13%, meaning that of the likely 8 teams that will be 0-2 next Monday, just one of them will pull it together enough to win 9 of their remaining 14. And of that 1 team in 8 that might pull it together to have a winning season, just 36% of the time (4 chances in 11) will they pull it together enough to be a real contender, like the 1996 Patriots, 1998 Jets, 2001 Patriots, and 2003 Eagles.

Especially for teams that some thought had legitimate playoff and even Super Bowl chances - such as Dallas, Washington, Giants, Miami, Kansas City, and Denver, an 0-2 start is almost certainly the kiss of death to those hopes because of the divisional tie-breakers being set in week 2. Ask the 2005 Chargers about that. In the NFC East and AFC West especially, at least one and possibly two teams people thought had a chance to be good are going to be in serious trouble come Monday. Tampa Bay-Atlanta is suddenly a huge game too. The Falcons have a chance to drive a dagger into the heart of the Bucs season hopes. Lastly, Minnesota has a chance to knock out Carolina at the Metrodome, thanks to the hollowness of the Panthers without Steve Smith. While the Eagles have a chance to put a season long hurting on the Giants at the Linc.

I'd say the supposedly good teams in the most trouble if they start 0-2 will be the Bucs, Panthers, Giants, Broncos, because the Bucs and Panthers must then play each other in week 3 (leaving the potential for one of them to be 0-3 and thus officially done for the year), and the Giants must visit the Seahawks, while the Broncos must visit the Patriots, both probably losses for the visitors.

And I don't know why you'd have a hard time writing off Dallas, for example, if they choke at home to Washington again. The big choke last year is what put them permanently behind Washington in chasing the wild-card spot, leaving them in the end not just 1 game behind Washington, but 2 games behind, because Washington held the tie-breakers over them. If they start out 0-2 this year, their season is even more effectively over than last year's 1-1 start left them. Dallas is a better team than Washington talent-wise. They should definitely win at home. To not win at home to a lesser team is a bad sign.

58
by coltshomer (not verified) :: Wed, 09/13/2006 - 2:48pm

The Indianapolis Colts are clearly ranked too low because Peyton Manning Rocks!!!. Throwing darts while reading tea leaves is way better than this. FO needs 2 C that real life is more than #s. COLTS ROX FO SUX

59
by DavidH (not verified) :: Wed, 09/13/2006 - 2:56pm

Re #51, 53, 44
It's not too difficult to do a quick look at the other weeks, if you don't mind including EVERY SINGLE COMBINATION OF WEEKS, and not throwing out the late weeks. Since team with X number of losses has a specific number of loss pairs (i.e. every 4 loss team will have 6 pairs of losses), all you have to do is count up the teams with each number of losses, and the number of playoff teams with each number, and do a little math. I just did that for 1996 through 2005, and here are the short and simple results.

ALL LOSS PAIRS: 10115 pairs, 1053 by playoff teams, 10.4%
Andrew's wk1 stats: 85 pairs, 7 by playoff teams, 8.2%
(Total - wk1 stats): 10030 pairs, 1046 by playoff teams, 10.4%

So on first glance it looks like weeks 1 and 2 are more important, but I would guess that Bobman is right, that taking out the last 4 or 5 weeks might make a difference. Unfortunately that is WAAAY harder. (At least I haven't thought of an easy way to do it. Scratch that, if there is a source with archived standings after week 12 for every season, I can do it the same way. But I don't know of such a source.)

I also tend to agree with him about teams giving up on the season after they start 0-2. It becomes a kind of self-fulfilling prophecy.

60
by Crushinator (not verified) :: Wed, 09/13/2006 - 3:03pm

44

I don't really believe in that stat. I mean, I believe its true, but its deceptively set up. What was the average record at the end of the season of those 0-2 teams? Many of those teams that did lose their first two games were probably just bad teams, rather than some sort of indicator that "after 2, your season is screwed".

61
by Aaron Boden (not verified) :: Wed, 09/13/2006 - 3:10pm

What were the reasons that Atlanta's defense ended up ranked higher than the Bears? I would think blanking a team would be quite a big achievement, especially if this is not taking into account oppontent adjustment.

62
by Scott C. (not verified) :: Wed, 09/13/2006 - 3:12pm

#55, Tom V

Uh, you do realize the rankings above say the 49ers are 30th, not 13th, right? And the colts are ranked 2nd.

The sorted order and rank in the first column reflects the quality of play (unadjusted by opponent strength) for this week ONLY. The 49ers played better against Arizona than the Colts did against the Giants. Having watched both games myself I'd have to agree with the numbers. Though I'm sure if the Colts played Arizona and the 49ers played the Giants the situation would be reversed.

Blanket statements about common sense being superior to statistics are downright silly. Have you taken any Psychology classes before? How about Economics? Or Physics? Common sense says the world is flat....

"Stats lie" is also hillarious. Human interpretation of the meaning of stats can lie and mislead, stats themselves cannot lie at all.

For example, you interpreted the rankings above as a statement that the 49ers were a better team than the Colts (ranked higher). Its your mistake in interpreting the numbers, not the numbers themselves.

63
by JSR (not verified) :: Wed, 09/13/2006 - 3:12pm

I know everyone's sick of the Mannings but it seems like a -40 VOA differential would have made for an interesting article about the Colts/Giants game. How can a team that apparently was beaten so badly on the field still end up winning?

Oh, and after reading the comments in your blog on Fox I'm ashamed to be associated with those Seahawks fans. Idiots like that aren't exactly going to help us overcome the post-SB reputation for whining, but we're not all that bad.

64
by B (not verified) :: Wed, 09/13/2006 - 3:22pm

55 & 58: You realize the Colts are ranked #2, right? The first colum ranks the teams purely on thier week one performance, the actual ranking is based on week 1 & the preseason projection, which is a combination of last season, off-season transactions and the age/experience of returning starters.

65
by KnickerBlogger (not verified) :: Wed, 09/13/2006 - 3:52pm

#60 - that's exactly what I'm trying to get at with my question (see #43, #6). Aaron's point at the top was that the first week wasn't highly significant when compared to his predicition DVOA. However this is a blanket statement, taken over the average of the entire league. I'm curious if there exist conditions where week 1 DVOA is more predictive of the season.

For example the 0-2 data presented above is probably weighted down by teams that were already expected to miss the playoffs. But I wouldn't take those stats and apply them to a team like Denver or Carolina if they start off 0-2 and say "Oh well they only have a 1 in 10 chance of making the playoffs". This is because the 0-2 data is noisy with teams that aren't expected to do well.

To make this data relevant to the teams like Denver/Carolina, I'd like to see that 0-2 study looking only at teams that were .500 or greater the year before.

66
by Johnnyel (not verified) :: Wed, 09/13/2006 - 4:47pm

@61 (Chicago and Atlanta defense)

I just spent a little bit of time looking at both play-by-plays, and it was ugly.

Carolina ran 59 plays from scrimmage, 29 of which were for 2 or fewer yards, were sacked four times, fumbled three times (lost one), and threw one interception. They ran the ball 16 times, and were stuffed 5 times. The Panthers had 4 rushing and 10 passing first downs. They never reached the red zone, but they made it into Atlanta territory 3 times in twelve possessions.

Green Bay ran 56 plays, 28 of which were for 2 or fewer yards, were sacked 3 times, and threw 2 interceptions (their fumble was on ST). They ran the ball 23 times and were stuffed 10 times. The Packers had 6 rushing and 8 passing first downs. They never reached the red zone, but they made it into Chicago territory 5 times in 11 possessions.

Neither team crossed their opponent's 30-yard line. I don't know how I could create a system to determine which of those offenses were more inept (or which defense more dominant depending on your POV).

It does look like Carolina could have easily been shutout. Their FGs were from 47 and 54 yards, and their scoring drives were 9 plays and 42 yards combined, which has more to do with starting field position than anything the offense did.

67
by zlionsfan (not verified) :: Wed, 09/13/2006 - 5:00pm

I used pro-football-reference to check each of the 84 teams that started 0-2 in the last 10 years (I believe only 7 teams started 0-2 last season). Of these, 23 had winning records the previous season, (so .500 teams were excluded, but the '98 Redskins, who finished 8-7-1 the previous year, are included), and 3 of them made the playoffs.

So the percentage does jump (to 13%), but not by much. It still suggests that if you start out 0-2, you've got an uphill road to the playoffs, which makes sense when you think about it: you've already lost one-eighth of your games, and you'll usually have played one or two teams in your conference, or even in your division, which means you're also starting off at a disadvantage with respect to tiebreakers.

68
by KnickerBlogger (not verified) :: Wed, 09/13/2006 - 5:24pm

Thanks zlionsfan! Nice research. But you've only piqued my curiosity. ;-) What about sub-.500 teams that go 2-0?

BTW is there an easier way to get this info other than going to each individual team & looking at the record?

69
by Wanker79 (not verified) :: Wed, 09/13/2006 - 5:38pm

I thought Aaron's stat in the Tampa Bay commentary...

"From 1993-2005, 40 teams lost in Week 1 by 20 points or more. Only six of those teams went on to post a winning record."

...was pretty surprising. I guess I never really mentally separated Week 1 blowout from Week 1 loses. I've just always taken every Week 1 result with a heafty helping of salt.

70
by Catfish (not verified) :: Wed, 09/13/2006 - 5:42pm

re: 55

You make a comparison to computer hurricane predictions. You're right, none of them are perfect or even close, but that's not the issue. They are still a heck of a lot better than guessing based on whether grampa's knee aches today. Remember the offical FO motto: "The best is the enemy of the better."

71
by KnickerBlogger (not verified) :: Wed, 09/13/2006 - 5:44pm

#69 - that would be Tampa Bay, Green Bay, and Oakland.

72
by navin (not verified) :: Wed, 09/13/2006 - 5:49pm

Another thing about San Fran. They were 0-9 on 3rd down (but 2-2 on 4th). I'm impressed they put up 27 points with that bad of a 3rd down conversion rate.

73
by underthebus (not verified) :: Wed, 09/13/2006 - 5:52pm

Aaron, regarding Alexander being pushed out of bounds, shouldn't you also account for td's scored untouched? Shaun had 28 td's in 2005 and 20 in 2004. I know from watching the games last year a good percentage of those were 'untouched' td's. Oh i also posted this on the FO Fox blog, but after yesterday's comments I wouldn't blame you if you didn't read them.

74
by Andrew (not verified) :: Wed, 09/13/2006 - 6:19pm

KnickerBlogger #65:

For example the 0-2 data presented above is probably weighted down by teams that were already expected to miss the playoffs. But I wouldn’t take those stats and apply them to a team like Denver or Carolina if they start off 0-2 and say “Oh well they only have a 1 in 10 chance of making the playoffs�. This is because the 0-2 data is noisy with teams that aren’t expected to do well.

To make this data relevant to the teams like Denver/Carolina, I’d like to see that 0-2 study looking only at teams that were .500 or greater the year before.

37 teams were .500 or better one year, and opened 0-2 the next. Only 3 of these teams ended up in the playoffs - the 98 Jets, the 2002 Steelers, and the 2003 Eagles - 8.1% of the total.

14 teams had 10 wins or more and then opened 0-2 the next year. Only 2 of these teams ended up in the playoffs - the 2002 Steelers, and the 2003 Eagles - 14% of the total (but the sample is smaller).

21 teams had 6 or 7 wins and then opened 0-2 the next year. 3 ended up in the playoffs - the 96 Patriots, the 98 Bills, and the 2002 Falcons - 14% of the total again.

22 teams had 2 to 5 wins and then opened 0-2 the next year. 2 ended up in the playoffs - the 1998 Cardinals and the 2001 Patriots - 9% of the total.

The numbers really don't vary a whole lot between classes of teams. You start 0-2, and you have around a 10% chance of making the playoffs. And given that 6-10 teams fall into this every year, in 5 of 10 years, none of the teams opening 0-2 have gone to the playoffs.

This year, 6 0-1 teams play each other in week 2, while the other 10 teams play a winner. So there are going to be anywhere from 3 to 13 0-2 teams. 8 of the 10 teams playing a winner are going on the road, and 5 are almost certain weaklings (Browns, Lions, Texans, Raiders, Titans) playing much better teams, while the 2 home teams are also likely weakinglings (Packers and 49ers) playing weaklings who seem just a bit stronger than them (Saints and Rams). Its doubtful that all three of the Panthers, Giants, and Bucs are going to win on the road against other strong teams. This says to me, playing the probabilities, only 2 or 3 of the 10 are likely to come home with a victory, leaving us with a cornucopia of possibly 10 or 11 0-2 teams this year. Can a playoff team or two emerge from this mess? Anything is possible, and 1 out of 10 or 11 is a likelihood, but its just as likely to be someone unexpected, such as the Lions or Bills or Texans as it is to be the Broncos, Chiefs, Cowboys, Panthers or Giants if they lose and fall to 0-2.

Every year there are 4, 5, and 6 win teams which reach the playoffs the next year, and there are 12+ win teams that collapse. Its going to happen to someone, and Week 2 is going to make a strong push for helping to determine it.

75
by Kal (not verified) :: Wed, 09/13/2006 - 6:39pm

49: The Raiders postseason hopes were ended when they signed Aaron Brooks.

Corrected.

76
by The Mulgrew (not verified) :: Wed, 09/13/2006 - 6:40pm

re: 14

Kesef? They gave you a Polish pastry to post there? I don't know, I guess that's good if it comes from a fine bakery.

77
by Andrew (not verified) :: Wed, 09/13/2006 - 6:42pm

Crushinator #60:

I don’t really believe in that stat. I mean, I believe its true, but its deceptively set up. What was the average record at the end of the season of those 0-2 teams?

Counting ties as 1/2 a win, the 83 teams averaged 7.18 wins the prior year, and when we add the 1999 expansion Browns who had no prior record, the 84 teams in total (thanks for the correction to zlionsfan) averaged 5.92 wins the year they opened 0-2.

The 36 teams with prior winning records averaged 9.82 wins the previous year, and 6.40 wins the year they opened 0-2.

Really, taking the bigger picture view, only 4 of these 84 teams really put themselves in a position to win a Super Bowl after opening 0-2 by subsequently winning at least 11 games and getting a #1 or #2 playoff seeding - the 96 and 2001 Patriots, the 98 Jets, and the 2003 Eagles. This shows you just how far behind the 8-ball losing your first two games puts you.

The 98 Bills and Cardinals and the 2002 Steelers and Falcons were 9 and 10 win teams that were lucky to make the postseason and were obviously not going very far, while the other 9 win teams - the 2000 Steelers and Packers, 2004 Bills, and 2005 Chargers and Vikings clearly cost themselves their shot at the playoffs by their poor opening play.

While its nice to daydream about you 11+ win team last year that opens 0-2 coming roaring back and winning 11+ again, when you look over the history of this, only the 2003 Eagles ever did it. The more normal result for a previously very good team (11+ wins) opening 0-2 is the epic collapses of the 99 Jets, Broncos, and Falcons, the 2001 Titans and Vikings, the 2002 Steelers and Rams, the 2004 Chiefs and the 2005 Chargers. Once again with yet another look at a subset of these teams, only 1 team of 10 winning 11+ the previous season opened 0-2 and managed to win 11+ again.

So if the Broncos, Giants, Bucs, or Panthers end up on the wrong side of the game next Sunday, its most likely that these teams are spiraling towards a season of just 5 to 8 wins, not that they are going to turn it around and end up in the #1 or #2 seed again.

Sorry to be the bearer of bad news, but to turn it around and win 11+ of 14 is extremely difficult. If you were good enough to win 11+ of 14, typically you wouldn't have started 0-2.

78
by Richie (not verified) :: Wed, 09/13/2006 - 6:42pm

On the Jacksonville comment: "Reggie "Dancing Machine" Williams looks like he's running crisper routes this year."

Are you actually able to watch a WR route-running?

Good route-running seems to be this year's catch phrase du jour when talking about receivers.

79
by Kal (not verified) :: Wed, 09/13/2006 - 6:43pm

on #75 - I guess the strikethrough tag doesn't work. Ah well - funny fails at the hands of the FOMB.

80
by NewsToTom (not verified) :: Wed, 09/13/2006 - 7:02pm

Kal winsloses!

81
by The Mulgrew (not verified) :: Wed, 09/13/2006 - 8:34pm

My fiend Phil says kesef is not a Polish pastry and he would know since he's Polish. He doesn't know what exactly kesef is. Can somebody please list what it is?

82
by B (not verified) :: Wed, 09/13/2006 - 8:57pm

81: It's Hebrew for money.

83
by The Mulgrew (not verified) :: Wed, 09/13/2006 - 8:59pm

Thank you, B. Much obliged.

84
by Trogdor (not verified) :: Wed, 09/13/2006 - 10:23pm

"That's why these ratings may seem a little strange, with teams that won this week showing up lower than teams that got blown out like Oakland and Tennessee."

Tennessee got blown out? From what I saw (granted, on a small TV off to the side of Cle/NO, during commercials), they got pretty well dominated, but were tied late in the 4th due to the Jets sucking. The Packers at 24 would be a better choice, I think.

Have to agree on the Cleveland comment. Winslow's a beast, but due to total offensive ineptitude he couldn't do a whole lot. It's amazing, how much money they spent on the O-line this offseason (granted, much of it's on IR), how many returning starters for 'stability', yet they got absolutely destroyed. Either it was a sign that New Orleans' highly-drafted D-line is finally coming around, or it's going to be a looooong season. Or both, actually.

I wouldn't be too hard on Rivers for only throwing twice to Gates. He only threw 11 times total, after all. I'm sure he'll look to Gates more when they play an NFL team.

85
by KnickerBlogger (not verified) :: Thu, 09/14/2006 - 1:03am

#77 Thanks for the hard work.

BTW if either you or zlionsfan are interested in hoops, I'm always looking for guys that are handy and thorough with numbers that can follow through on a hypothesis. :-)

86
by Fnor (not verified) :: Thu, 09/14/2006 - 1:22am

Trogdor: Losing your center is huge. As is losing your backup center. And the 3 other guys you brought it to be backup backup center, backup backup backup center, and backup backup backup backup center.

87
by Travis (not verified) :: Thu, 09/14/2006 - 1:50am

Re: 77

While its nice to daydream about you 11+ win team last year that opens 0-2 coming roaring back and winning 11+ again, when you look over the history of this, only the 2003 Eagles ever did it. The more normal result for a previously very good team (11+ wins) opening 0-2 is the epic collapses of the 99 Jets, Broncos, and Falcons, the 2001 Titans and Vikings, the 2002 Steelers and Rams, the 2004 Chiefs and the 2005 Chargers. Once again with yet another look at a subset of these teams, only 1 team of 10 winning 11+ the previous season opened 0-2 and managed to win 11+ again.

Half of these teams (though not the more recent ones) suffered early injuries to or retirements of key players, so it was clear they were not going to post the same record, regardless of how they started out:

1999 Jets - Vinny Testaverde injured in Week 1, out for rest of the season.

1999 Broncos - John Elway retired in the off-season.

1999 Falcons - Chris Chandler injured in Week 1 (returned in Week 3, then didn't play until Week 6); Jamal Anderson (1,846 yards rushing) injured in Week 2, out for rest of the season.

2001 Titans - Steve McNair injured in Week 1, didn't return until Week 4.

2001 Vikings - Robert Smith (1,521 yards rushing) retired in the off-season.

88
by Trogdor (not verified) :: Thu, 09/14/2006 - 7:39am

Yes, losing all those centers can't be good. But, they traded for a highly-regarded backup (Fraley) to fill the hole - supposedly he would've been an improvement over Faine straight-up. And it wasn't just the center - the entire line was just getting whooped. Part of it may be having a new center, since they make the line adjustments, so maybe some of it was being in the wrong blocking scheme. But they were certainly just getting beaten physically - on runs they were shoved into the backfield, on passes Frye had people in his face before his drop was finished, and it's not like NO was doing any crazy LeBeau-style blitzing. Obviously, the Saints have some talent there - they have spent like seven first-rounders on DL. But it looked like they were going against a college team on Sunday, just getting the best of every individual and collective battle. It was U-G-L-Y, they ain't got no alibi.

89
by Sam! (not verified) :: Thu, 09/14/2006 - 10:33am

RE: 60

I think what the stat is saying is that, regardless of what the fans think or expectations are... if a team starts 0-2, they are probably not a good team. Of course there are exceptions, but teams start out 0-2 despite being picked by many to make the playoffs or contend for a title every year. Just because they are expected to do well does not make their 0-2 start a fluke. 0-1, maybe a little more flukey. 0-2 is starting to show a trend.

90
by Mr Shush (not verified) :: Thu, 09/14/2006 - 10:49am

"If you were good enough to win 11+ of 14, typically you wouldn’t have started 0-2."

Typically, no. But if the Panthers do, there will have been a pretty obvious reason why. The no-Smith Panthers are a losing team, the with-Smith Panthers are a Superbowl contender potentially capable of winning 11 of 14. Especially if the Bucs really do fall apart.

91
by jebmak (not verified) :: Thu, 09/14/2006 - 12:59pm

The blog is a great idea! It keeps the comments out of here.

92
by zlionsfan (not verified) :: Thu, 09/14/2006 - 1:08pm

Re 68, 85: Oh yeah, you're the KnickerBlogger. :) I've seen your site and like your work, although I'm probably, at best, a casual hoops fan.

Yes, a good way to produce these numbers would be to dump the week-by-week results into a (insert favorite database app) table so that you could write queries to do the rest. I, however, took the long way around ... at least this time. Either Andrew did it that way or put in a lot of manual labor. For his sake, I hope it's the former - nice work, by the way.

I already have a table with franchises and years, so that I can say, for example, that since the introduction of the 16-game schedule in 1978, the third-worst records in a five-year span that does not include strike-affected seasons (basically, starting in '88 instead of '78) belong to the 1991-95 Bengals and the 2001-05 Lions. (The worst records belong to the 1998-2002 Bengals and the 1989-93 Patriots.) They're not quite to the level of the wartime Cardinals, thankfully ...

So, I guess I ought to drop in the games as well ... that way, in a couple of weeks, I'll be able to help answer similar questions ...

93
by Jake (not verified) :: Thu, 09/14/2006 - 1:53pm

Does this mean Reuben Droughons was a bad choice? Eep.

94
by Bencoder (not verified) :: Thu, 09/14/2006 - 2:18pm

I will second those who have said an analysis of why the DVOA domination by the Giants did not produce a victory. It would be interesting insight into an extremely significant statistical outlier.

95
by Pat (not verified) :: Thu, 09/14/2006 - 5:20pm

But, they traded for a highly-regarded backup (Fraley) to fill the hole - supposedly he would’ve been an improvement over Faine straight-up.

Fraley is smarter than he is athletic: he'll probably get better as the year goes on as he figures out exactly how the line performs. But if there's just nothing that can be done with the line, he'll be a liability.

96
by Wanker79 (not verified) :: Thu, 09/14/2006 - 6:30pm

Re: The Fox Blog

Holy Jesus. I just read through some of the comments. Check out this little gem of wisdom from the same guy that said "QUIT THE CEREBIAL ALREADY":

"I HAVE A HIGH REGARD FOR SEAHAWK FANS .. BUT THE POWERS TO BE WILL NEVER LET THEM IN ANOTHER SUPER BOWL.. BRANCH THIS BRANCH THAT.. BELICHEK KNOWS TEAM AND WINNING .. HE WILL NURTURE A COMPETANT REPLACEMENT.. MY VIKES HAD MARONEY IN OUR BACKYARD AND DID NOT TAKE HIM.. PATS WILL PLAY WITH WHAT THEY GOT.. RUN THE BALL AND STOP THE RUN AND WILL BE IN THE HUNT COME PLAYOFF TIME...MANNING AND MANNIG DONT = A BRADY.. CHECK MY BLOG OUT"

What is it with the correlation between stupidity and the caps-lock key?

97
by Richie (not verified) :: Thu, 09/14/2006 - 9:30pm

What is it with the correlation between stupidity and the caps-lock key?

1.0

98
by Smartmonies (not verified) :: Fri, 09/15/2006 - 12:29am

TAKE THE 49ERS OVER THE RAMS SUNDAY.

BEST OF LUCK TO ALL!

99
by Independent George (not verified) :: Fri, 09/15/2006 - 1:40am

Sweet merciful crap... Until reading the comments on the Fox FO blog, I didn't realize how spoiled I've been by the FO message boards.

Aaron, if that's in any way representative of what you get in your mailbox every morning, you have my deepest condolences.

100
by Plummerception (not verified) :: Fri, 09/15/2006 - 3:09pm

Denver is clearly ranked too low because we all know that they mailed it in the first game of the season, AGAIN. I could poop on the floor in the shape of a horse head and it would be way better than this. If there rankings are going to be so unfair I'll go too a real foot ball web sight like NFL.cmo. L4M3

101
by Wanker79 (not verified) :: Fri, 09/15/2006 - 3:21pm

Plummerception, I'm having a real had time distinguishing whether that was actual anger or sarcasm. It followed the template a little too close for me to take completely serious, but the spelling was too good to take completely unserious.

102
by B (not verified) :: Fri, 09/15/2006 - 3:36pm

Re 98: Hmmm, I was thinking of taking SF over Saint Louis, but seeing your post in all caps, I know you must be foolish. On the other hand, even a broken clock is right twice a day. Can a troll be as useful as a broken clock?
Oh, by the way, I know plummerception is serious because nfl.cmo.l4m3 is the best football site on the internets.

103
by Aaron Brooks (not verified) :: Fri, 09/15/2006 - 5:54pm

my back hurts. does anyone have any motrin?

104
by Timdog (not verified) :: Fri, 09/15/2006 - 8:07pm

Why aren't the unadjusted VOA's symmetrical? Or, more specifically, why does BUF have 9.1% and NE -0.9%? Are there things that are kinda bad for the defense but super-good for the offense in VOA terms? Is there some sort of adjustment for swagger?

105
by Chris (not verified) :: Fri, 09/15/2006 - 8:52pm

This is the most pointless stuff in the world when it comes to football. All these projections and predictions and ratings mean absolutely nothing. Who cares about this junk.

106
by Pat (not verified) :: Fri, 09/15/2006 - 9:43pm

Why aren’t the unadjusted VOA’s symmetrical? Or, more specifically, why does BUF have 9.1% and NE -0.9%?

Psst! The word is symmetric, not symmetrical. :)

Special teams. Note that BUF's defensive VOA is NE's offensive VOA, and vice versa, but the special teams are semi-independent - see special teams VOA, "w/HIDDEN". Teams don't get penalized for opponent kickoff length, for instance: a touchback doesn't hurt the other team's kickoff return VOA, and it boosts the kicking team's VOA.

107
by Crushinator (not verified) :: Fri, 09/15/2006 - 11:57pm

105 -

If nobody cared, the site would have caved years ago and Fox Sports wouldn't be hiring them to do the power ratings.

108
by Spielman (not verified) :: Sat, 09/16/2006 - 2:14pm

I'm relatively new to this, so my apologies if this is a particularly stupid question.

I'm curious after glancing through the numbers for the first week... in most cases, the Total VOA numbers for the teams that played each other balance pretty well. For example, Philly beats Houston, and Philly is rated as +47, Houston as -45. Most of the Total VOA numbers for week one opponents are within a few points of each other, but there are a few outliers where the two numbers don't really come very close to balancing, like the Jets and Tennessee. What's causing this effect?

109
by Spielman (not verified) :: Sat, 09/16/2006 - 2:16pm

Bleh. Never mind. It appears 105 answered my question, and I just didn't see it.

110
by gleebergloben (not verified) :: Sat, 09/16/2006 - 9:18pm

wow, i'm a football fanatic and i have to say 'what a piece of absolute fecal matter.' so, if i carry the one and divide the integer by the square root of 4, then i'll come up with the expected turnover ratio for the texans when the wind is blowing from the southeast in the second quarter and blah blah blah.

111
by smartmonies (not verified) :: Thu, 09/21/2006 - 4:44pm

re102How do you like me now?
This week Take Mich +4, Packers -6.5, Auburn +42

112
by Smartmonies (not verified) :: Thu, 09/21/2006 - 7:39pm

URGENT TYPO

Take Mich+4 Lions -6.5 & Buffalo +42 (not packers & auburn)

Sorry