Writers of Pro Football Prospectus 2008

Most Recent FO Features


» 2018 Free Agency Cost-Benefit Analysis

Is Kirk Cousins the best free-agent quarterback in recent memory? Should Trumaine Johnson or Malcolm Butler have gotten the larger contract? And what makes a free-agent contract good or bad, anyway?

25 Oct 2005

Week 8 DVOA Ratings

by Aaron Schatz

Issues discussed in this week's FOXSports.com commentary:

  • The one game that drops Pittsburgh out of the top spot
  • Should we dampen the power of the Washington blowout?
  • The unlucky fumble-squandering Saints of 2003 become the lucky fumble-recovering Saints of 2004 become the unlucky fumble-squandering Saints of 2005
  • An ancient Roman poem about a dog with two bones
  • Future Jets draft picks
  • Spanish vocabulary and conjugation

One issue not addressed in this week's commentary: Why we whiffed on our Atlanta projection. I mentioned in last night's open thread discussion that I would write about that, but I had problems this morning because Hurricane Wilma kept the stats people in Atlanta from sending a proper gamebook to NFL.com and I had to manually break down the Jets-Falcons play-by-play. I'll talk about the Falcons next week instead.

FOXSports.com commentary is right here. Individual pages for offense, defense, and special teams are now online. Player stats will be updated when the proper Jets-Falcons gamebook is ready.

* * * * *

These are the Football Outsiders team efficiency ratings for 2005, measured by our proprietary Defense-adjusted Value Over Average (DVOA) system that breaks down every single play and compares a team's performance to the league average based on situation in order to determine value over average. (Explained further here.)

OFFENSE and DEFENSE DVOA are adjusted based on strength of opponent as well as to consider all fumbles, kept or lost, as equal value. Opponent adjustments are currently set at 70% and will increase each week until they are full strength after Week 10. SPECIAL DVOA is adjusted for type of stadium (warm, cold, dome, Denver/Mexico City) and week of season. NON-ADJ TOTAL VOA does not include these adjustments.

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

1 IND 36.5% 2 54.6% 7-0 26.2% 2 -16.8% 6 -6.6% 29
2 CIN 33.9% 1 52.3% 5-2 24.2% 4 -9.3% 10 0.4% 18
3 JAC 33.1% 5 19.3% 4-2 -0.3% 15 -28.9% 1 4.5% 7
4 SD 32.1% 3 20.2% 3-4 24.6% 3 -4.2% 14 3.3% 11
5 SEA 27.3% 6 28.8% 5-2 28.8% 1 0.1% 17 -1.4% 22
6 PIT 24.7% 7 29.2% 4-2 9.0% 12 -14.9% 7 0.8% 16
7 WAS 24.4% 16 12.0% 4-2 19.5% 6 -7.9% 12 -3.0% 25
8 NYG 24.3% 4 29.4% 4-2 16.3% 8 7.8% 22 15.8% 1
9 DAL 21.1% 8 24.0% 4-3 -1.7% 16 -18.9% 5 3.9% 9
10 KC 19.4% 18 12.3% 4-2 17.8% 7 1.8% 19 3.3% 12
11 TB 17.5% 9 34.1% 5-1 -3.6% 17 -19.1% 4 1.9% 14
12 DEN 13.2% 13 14.4% 5-2 13.3% 10 -2.2% 16 -2.4% 24
13 CHI 12.3% 15 11.1% 3-3 -19.0% 27 -27.5% 2 3.8% 10
14 PHI 11.3% 11 14.3% 4-2 9.0% 13 -13.3% 8 -11.0% 32
15 OAK 6.2% 20 7.7% 2-4 21.1% 5 11.1% 25 -3.9% 28
16 MIA 3.8% 10 -5.2% 2-4 -8.4% 21 -8.1% 11 4.0% 8
17 CAR 2.2% 17 14.0% 4-2 -10.5% 23 -12.5% 9 0.3% 19
18 ATL 1.9% 12 20.2% 5-2 9.6% 11 10.2% 24 2.5% 13
19 NE -3.6% 21 -12.5% 3-3 15.3% 9 20.5% 29 1.6% 15
20 DET -8.0% 26 -12.6% 3-3 -26.7% 31 -22.5% 3 -3.7% 26
21 BUF -8.1% 14 4.0% 3-4 -17.7% 26 2.6% 20 12.3% 2
22 TEN -9.7% 23 -8.6% 2-5 -4.4% 19 12.0% 26 6.7% 3
23 CLE -11.4% 19 -24.5% 2-4 -5.9% 20 10.1% 23 4.6% 6
24 BAL -13.9% 24 -14.3% 2-4 -15.9% 24 -3.9% 15 -1.8% 23
25 GB -13.9% 22 -1.9% 1-5 6.8% 14 13.5% 28 -7.1% 30
26 MIN -22.4% 30 -39.5% 2-4 -16.6% 25 4.9% 21 -0.9% 20
27 ARI -24.5% 25 -9.7% 2-4 -23.7% 28 1.5% 18 0.7% 17
28 NYJ -25.2% 29 -27.6% 2-5 -25.8% 30 -4.6% 13 -3.9% 27
29 STL -26.6% 28 -21.6% 3-4 -4.3% 18 21.0% 30 -1.3% 21
30 NO -31.4% 27 -32.8% 2-5 -9.9% 22 13.2% 27 -8.4% 31
31 HOU -55.7% 31 -77.2% 0-6 -24.5% 29 36.8% 32 5.6% 4
32 SF -83.9% 32 -91.1% 1-5 -55.5% 32 33.2% 31 4.8% 5

  • FOX RANK represents the FOXSports.com Power Ratings which are 95% 2005 DVOA, 5% a special weighted DVOA for 2004 that includes the playoffs. This is the final week of a separate FOX rank. Beginning next week, FOX Power Rankings will be based on WEIGHTED DVOA.
  • ESTIMATED WINS uses a statistic known as "Forest Index" that emphasizes consistency as well as DVOA in the most important specific situations: red zone defense, first quarter offense, and performance in the second half when the score is close. It then projects a number of wins adjusted to a league-average schedule and a league-average rate of recovering fumbles.  Teams that have had their bye week are projected as if they had played one game per week.
  • PAST SCHEDULE lists average DVOA of opponents played this season, ranked from hardest schedule (#1, most positive) to easiest schedule (#32, most negative). It is not adjusted for which games are home or road.
  • FUTURE SCHEDULE lists average DVOA of opponents still left to play this season, ranked from hardest schedule (#1, most positive) to easiest schedule (#32, most negative). It is not adjusted for which games are home or road.
  • VARIANCE measures the statistical variance of the team's weekly DVOA performance.  Teams are ranked from least consistent (#1, highest variance) to most consistent (#32, smallest variance).

1 IND 36.5% 7-0 1 6.1 1 -24.0% 32 6.4% 8 3.9% 32
2 CIN 33.9% 5-2 2 5.0 5 -4.2% 23 1.3% 16 29.7% 10
3 JAC 33.1% 4-2 4 5.2 4 18.4% 2 -25.5% 32 29.3% 11
4 SD 32.1% 3-4 3 5.4 3 13.9% 4 9.9% 1 10.2% 30
5 SEA 27.3% 5-2 6 5.7 2 -3.8% 22 -18.9% 31 11.9% 26
6 PIT 24.7% 4-2 5 4.4 11 5.0% 12 -1.2% 19 32.2% 7
7 WAS 24.4% 4-2 7 4.5 10 1.6% 15 9.7% 2 25.3% 12
8 NYG 24.3% 4-2 8 4.7 6 -2.7% 20 3.9% 12 23.8% 14
9 DAL 21.1% 4-3 9 4.5 9 6.0% 10 4.0% 11 30.3% 8
10 KC 19.4% 4-2 10 4.7 7 5.6% 11 9.5% 3 12.1% 25
11 TB 17.5% 5-1 11 4.6 8 -12.3% 30 -10.5% 29 10.3% 29
12 DEN 13.2% 5-2 13 4.1 12 19.1% 1 5.5% 10 21.5% 16
13 CHI 12.3% 3-3 14 3.9 14 0.4% 16 -12.7% 30 49.1% 3
14 PHI 11.3% 4-2 12 3.9 13 -0.5% 17 9.4% 4 21.3% 17
15 OAK 6.2% 2-4 15 3.8 16 12.0% 6 8.4% 5 13.2% 24
16 MIA 3.8% 2-4 16 3.8 17 3.2% 14 -5.3% 27 20.2% 18
17 CAR 2.2% 4-2 17 3.9 15 -12.9% 31 -1.5% 21 10.9% 27
18 ATL 1.9% 5-2 18 3.6 19 -7.5% 25 0.2% 18 17.6% 21
19 NE -3.6% 3-3 19 3.7 18 13.4% 5 -1.7% 22 13.6% 22
20 DET -8.0% 3-3 21 3.3 21 -1.2% 19 -2.1% 25 57.7% 2
21 BUF -8.1% 3-4 20 2.8 24 -11.9% 29 8.0% 6 37.4% 5
22 TEN -9.7% 2-5 22 3.0 22 -3.7% 21 -1.2% 20 34.7% 6
23 CLE -11.4% 2-4 23 3.4 20 7.8% 7 2.5% 14 18.5% 20
24 BAL -13.9% 2-4 24 2.6 25 -0.9% 18 6.0% 9 13.5% 23
25 GB -13.9% 1-5 25 2.0 28 -8.9% 27 7.9% 7 30.2% 9
26 MIN -22.4% 2-4 26 2.9 23 3.4% 13 -1.8% 23 24.4% 13
27 ARI -24.5% 2-4 28 1.8 30 -11.0% 28 -2.1% 24 10.5% 28
28 NYJ -25.2% 2-5 27 2.0 29 7.7% 8 1.2% 17 9.8% 31
29 STL -26.6% 3-4 29 2.1 26 -8.8% 26 -7.7% 28 18.8% 19
30 NO -31.4% 2-5 30 2.1 27 -6.1% 24 2.0% 15 47.7% 4
31 HOU -55.7% 0-6 31 0.4 31 17.4% 3 -4.8% 26 21.9% 15
32 SF -83.9% 1-5 32 0.0 32 7.0% 9 2.5% 13 59.8% 1

Posted by: Aaron Schatz on 25 Oct 2005

173 comments, Last at 01 Nov 2005, 1:55pm by james


by MRH (not verified) :: Tue, 10/25/2005 - 5:43pm

Sometimes the numbers tell you things that make you scrath your head:

Was #7, lost to Den and KC
KC, #10, lost to Den and Phi
Den, #12, beat KC and Was
Phi, #14, beat KC, plays Den this week.

If the DVOA ranks had been exactly the opposite, I wouldn't have noticed.

Noticed too that every NFC East and AFC West team is in the top 15 and of course they play each other this year. Is that confusing DVOA thru this part of the season? Or is DVOA just catching the things that W-L don't?

by tunesmith (not verified) :: Tue, 10/25/2005 - 5:56pm

MRH, it looks confusing at first, but in terms of actual wins and losses, all those teams were in beatpath loops with other teams.

WAS over SEA over ATL over PHI over KC over WAS
MIA over DEN over KC over MIA

So if you want to look at it in terms of wins and losses, the only victory worth paying attention to is Denver's win over Washington.

by pawnking (not verified) :: Tue, 10/25/2005 - 6:03pm

Excellent article on Fox to accompany these stats, guys.

Any ideas on an article on the predictive powers of DVOA and how they have evolved as DVOA has become "smarter?" I ask because DVOA is a champ at telling us what has happened, but the real test with any number system is how well it tells us what will happen. Not just in terms of wins a losses, but could DVOA have predicted Oakland's passing effectiveness this past week? How well has it done so? In retrospect, it's always obvious.

by Bill Moore (not verified) :: Tue, 10/25/2005 - 6:13pm

Im not a huge buyer of Jacksonville, but the fact that they are 4-2 after playing the second hardest schedule, and have 10 games to go with the easiest schedule is sure interesting. Indy is 6-0, but has played the easiest schedule, and has the 8th hardest going forward. Maybe JAC doesn't take the division, but it surely puts them in a good spot for a wildcard.

On the flip side, poor SD. Faces the 4th hardest schedule to date, but it gets no easier from here. Plus, they have played better than their 3-4 record should suggest. It will be a real interesting offseason if SD goes 8-8 (or worse) and management hasn't traded Rivers.

by Bowman (not verified) :: Tue, 10/25/2005 - 6:13pm

Why concentrate on Pit's ranking "except for" the Jax game? If you eliminate ALL team's worst game, would the rankings come out to be nearly the same? I'm sure that if you eliminate both the Jax-Pit and Jax-Den game, Jacksonville wouldn't move that much.

DVOA SHOULD account for injury, as a football team includes its backups, especially in the salary cap era. Not keeping a quality backup on the roster in order to pay for more quality elsewhere is a risk decided by the team. An injury to that specific player SHOULD be reflected in the DVOA, which measures ALL plays.

Why don't you blame the PIT front office for not having an adequate backup (assuming that Maddox's awefulness could be predicted)?

by Trogdor (not verified) :: Tue, 10/25/2005 - 6:18pm

Great job on getting the Browns helmet logo up there. Now, can you work on making them not suck as much?

by putnamp (not verified) :: Tue, 10/25/2005 - 6:21pm

I was thinking the same thing Bowman, except for this: if only one team gets to play the Maddox Steelers, that team looks really good compared to all the other teams that didn't. Yes, the Steelers' will be bumped down a notch for having a bad backup, but that doesn't affect the 16 teams they played.

Additionally, do they magically become a better team next year simply on account of not having to play Maddox at all (should they be so lucky)? If everything else stays the same except that one detail - they're the same team!

by Andrew (A.B.) (not verified) :: Tue, 10/25/2005 - 6:24pm

Right now, DVOA treats all plays equally, for the most part, but each game has a slightly different number of plays. This ends up giving one game more weight than another just because there happened to be a lot of plays. I wonder how the numbers would look if you weighted the plays in each game as if every game had theat team's average number of plays. (Or league-average number of plays)

In theory, any individual play is as likely to be predictive as any other. But the distribution of plays across a small number of plays is not likely to be uniform. For example, a team one day may have 15 plays in the last two minutes of the first half, and another day have 3. But the offensive and defensive performance in that situation is not very representative of the rest of the game.

This would seem to be doubly true when you look at separate ratings for offense and defense. If you've got a lot of offensive plays in a game, it's probably because you're behind and trying to catch up. So you're overweighting plays that occur in a specific game situation, on a day when the team is not doing too well to begin with.

by tunesmith (not verified) :: Tue, 10/25/2005 - 6:28pm

#3: Actually, I've been under the impression that DVOA sacrifices the need to accurately reflect the standings, in favor of being more predictive. What that means is that the DVOA power rankings might look a bit out of whack each week, but then they'll be vindicated over and over again as they turn out to be right.

Contrast that with other power ranking systems that try to always reflect how things are up to the moment, but then have huge changes week to week. They might seem more accurate each week, but as you go back over them at the end of the season, they'll be the ones that seem out of whack.

I love 'em both. DVOA is better for predictive purposes, while the other kinds are better for the horse race drama.

by Fnor (not verified) :: Tue, 10/25/2005 - 6:36pm

#5: Because we were lucky to have a backup as good as Maddox!

by James Gibson (not verified) :: Tue, 10/25/2005 - 6:46pm

I'm guessing the off-projections for Atlanta may have something to do with what was mentioned this week - blowouts. That 56-10 loss to Kansas City last year may have really raised KC's DOVA and dropped Atlanta's last year.

Of course, they're only 18th in DVOA this year so maybe I'm off.

by Michael David Smith :: Tue, 10/25/2005 - 6:49pm

James, KC wasn't Atlanta's only bad game last year. They lost to KC, Tampa Bay and Detroit -- not exactly a murderer's row -- by a combined score of 100-20.

by zlionsfan (not verified) :: Tue, 10/25/2005 - 6:50pm

Detroit and Chicago: second and third in variance, second and third in defensive DVOA. Will the offenses be mediocre or incredibly bad?

I think the over/under for Sunday's game is going to be 17. If they divided o/u into offense and defense + special teams, the offensive o/u (pun intended) might be 7.

by Matthew Furtek (not verified) :: Tue, 10/25/2005 - 6:52pm

I can understand why you'd want to dampen Washington's blow out, but every other team that has played San Francisco had the same opportunity to blow them out as well.

The Eagles, Cowboys, Cardinals and Colts all had the same oppurtunity to gain impressive victories, but only the Eagles were able to blow them out like the Redskins.

It's interesting to me how variable variance is...

And I will still contend Seattle didn't get unlucky against the Redskins.

by Vash (not verified) :: Tue, 10/25/2005 - 6:52pm

Wait... if Pittsburgh would move to #3 on offense without the Week 6 game, that means they would be above San Diego, which is at 24.6%.
Pittsburgh's total offensive DVOA is 9.0%.

That game was so bad that even as just 1/6 of the DVOA, it dropped the Steelers over 15%!?

by Jerry P. (not verified) :: Tue, 10/25/2005 - 7:11pm

"Not just in terms of wins a losses, but could DVOA have predicted Oakland’s passing effectiveness this past week?"

Yes, it did. As I said in the "Any Given Sunday" article, going into this past weekend's game Buffalo had faced zero passing offenses in the top half of the league. In light of that fact it came as no surprise that they failed to play up to the -40.7% DVOA they had. This weekend they will go to New England to face what is at this point the 9th ranked pass offense in the NFL. It won't be shocking if yet again they fail to post great pass defense numbers.

What I can't explain is why despite the fact Buffalo had played no good pass offenses their pass defense VOA was nearly the same as their pass defense DVOA. Now that they've played one good passing team their pass defense VOA is much higher than their pass defense DVOA. Shouldn't they have taken a huge hit in DVOA for playing bad passing offenses and a gain for facing a good one last Sunday?

by Richie (not verified) :: Tue, 10/25/2005 - 7:29pm

You mention Mike Martz' record when leading in the 4th quarter.

I think more attention needs to be paid to the bottom of that list. If I remember correctly, David Shula had a winning percentage somewhere around 52% in games in which he led at some point in the 4th quarter. So, half the time he had a 4th quarter lead, he lost. Amazing.

by Kachunk (not verified) :: Tue, 10/25/2005 - 7:37pm

I wonder if part of Maddox's awfulness stems from Cowher having too much confidence in him? When most teams have to play their backup they alter their gameplan so as to take the load off the QB's shoulders. Just ask him "not to lose" the game. They don't seem to have done that with Maddox though, possibly because he was thought to be more competent that his play showed. It might also be because Jax was stuffing the run so effectively.

by Brian (not verified) :: Tue, 10/25/2005 - 7:52pm

I think people overplay how bad Maddox played. Jacksonville stuffed their run, forcing them to throw WAY more than they typically do. While I think Roethlisberger would have definitely played better, I think there would have been a significant drop-off from his average. Their passing game typically excels because the opponent has to play run every down. More blame should be placed on the lack of a run game in that game than on Maddox.

by MDS (not verified) :: Tue, 10/25/2005 - 7:52pm

Pittsburgh tried to alter its gameplan with Maddox in, but Jacksonville shut down the running game. Stroud and Henderson were great that day.

by admin :: Tue, 10/25/2005 - 8:05pm

Yes. As I pointed out in Quick Reads last week, three of Maddox's turnovers came on third-and-11, third-and-14, and third-and-27. But there's no excuse for the one that lost the game, which came on second-and-10. And it isn't like he played well on the downs that weren't turnovers.

I will add that while DVOA is meant to be more predictive than win-loss record, it isn't psychic. This is the NFL and teams improve and decline all the time for reasons that can't be forecast by the numbers, not to mention that DVOA can't forecast injuries. We do our best.

by Andrew (not verified) :: Tue, 10/25/2005 - 8:15pm


How could the Eagles special teams performance on Sunday not have kicked them up from #32 (discounting the Field Goal block and return)? They must have been enormously bad on an individual game basis the last three weeks.

by Vash (not verified) :: Tue, 10/25/2005 - 8:22pm

Let's see... why did Jacksonville shut down the running game?
Because they knew the Steelers wouldn't pass well.
Maddox earns some of the blame for the ground game as well.

by Pat (not verified) :: Tue, 10/25/2005 - 8:25pm

Andrew: They were -13.0% last week. They're -11.0% this week. Assuming equal numbers of special teams plays (they're... not, but bear with me) that gives them a special teams DVOA of roughly 0% for this last week. Makes sense to me. They did look pretty mediocre. Better than before, obviously, but mediocre.

But now they've got JOSE CORTEZ POWER!

by Andrew (not verified) :: Tue, 10/25/2005 - 8:26pm

I count 7 real teams this week (positive offense, negative defense).

Colts, Bengals, Chargers, Steelers, Redskins, Broncos, Eagles.

Jaguars, Cowboys, and Bucs are still searching for an offense.

Seahawks, Giants, and Chiefs are still searching for a defense.

Falcons, Raiders, Patriots, Packers are really missing a defense.

Bears, Panthers, Lions, Ravens, Jets are really missing an offense.

The remaining ten teams are upside down and just abysmal.

by Pat (not verified) :: Tue, 10/25/2005 - 8:31pm

I'm telling you, Andrew, it'll all end in tears when you find this year's Giants. :)

by Pat (not verified) :: Tue, 10/25/2005 - 8:39pm

Oh, and the other reason the Eagles didn't move up from #32 in special teams. Last week they were -13.0%. The nearest team to them is NO's -8.4%. That's a big, big gap to be fixed in one week.

by Tim L (not verified) :: Tue, 10/25/2005 - 9:13pm

Andrew, when you refer to "real" teams, you might consider the word "balanced" instead. By your definition, Super Bowl winning teams that had a great defense and a mediocre offense (2003 Patriots, 2000 Ravens) weren't "real" teams. It might come as a rude surprise to the 2000 Giants to discover that their 34-7 demolition wasn't administered by a team that was "real".

by Gorka Zubizarreta (not verified) :: Tue, 10/25/2005 - 9:17pm

The Pats ranked #19? Hey Einstein, why don't you take all your stats, turn them sideways and stick them directly up your ass. That way, maybe you can get a better look at them...seeing that your head is up your ass as well.

by Andrew (not verified) :: Tue, 10/25/2005 - 9:31pm


Okay, lets work with the equal game weight theory. Here's the progression starting with the rating after the Raiders game:

-8.7%, -12.7%, -13.0%, -11.0%

Not knowing the first two weeks, we can determine:

Chiefs game ~ -25% (What's that Coach Harbuagh? You mean we have to tackle Dante Hall? Man but that dawg is just so sweet to watch running down the field!)
Cowgirls game ~ -14% (Hey look coach, we didn't give up a kick-return TD this week!)
Chargers game ~ -1% (Hey, Coach, we can cover in the return game!)

Okay, that if they could just maintain being average the rest of the year they go to somewhere around 6th-9th overall in DVOA.

I still wonder what the Raiders game with Bartrum's squib kicks ranked as, and the 49ers game.

by Vince (not verified) :: Tue, 10/25/2005 - 9:32pm

Well, Aaron's deep-seeded anti-Pats bias has finally been exposed.

by Pat (not verified) :: Tue, 10/25/2005 - 9:45pm

psst, Andrew: you're forgetting about the jose cortez power!

No, seriously, their kicking game should improve markedly after this week. I think firing Cortez was one of the stupidest things Parcells has done in a long time (starting the Other Drew last year is close). You can't expect a kicker to be consistent without a consistent long snapper. Thankfully, the Eagles have one.

And I doubt the Eagles will have Akers in the game without Cortez active for at least a week or two. And Akers is expected to play either next week or the week after.

France has been - well, better than Mike Bartrum - but that's about it. I will appreciate getting touchbacks again.

by RCH (not verified) :: Tue, 10/25/2005 - 9:51pm

I must admit that when Toomer caught that pass the first thing I thought about was how many Denver posts we'd see in this thread.

by calig23 (not verified) :: Tue, 10/25/2005 - 10:04pm


Yay for random convoluted insults!

by admin :: Tue, 10/25/2005 - 10:11pm

Of all the hate mail I receive, and all the angry message board posts I read, by far my favorites are the ones accusing me of anti-Patriots bias. I mean, you have to hunt around this site for what, 30 seconds maybe, before you read something where I talk about being a Patriots fan. I mean, last week I was inventing new swear words because I ran out of things to scream at my television whenever Duane Starks blew yet another play.

Here's the deal on Cortez: very good on kickoffs, very bad on field goals. The man was born to be a kickoff specialist. So he may be half an upgrade on Todd France, in fact less than half because he's probably a small downgrade on the field goals.

On the other hand, XFL!

by Pat (not verified) :: Tue, 10/25/2005 - 10:21pm

Simple solution. Use Akers for the field goals for a few weeks, use Cortez for the kickoffs. Less strain on Akers, and in my opinion, less costly in case something goes wrong.

by Parker (not verified) :: Tue, 10/25/2005 - 10:22pm

If SF's Total Non-Adjusted VOA drops to -100%, do they cease to exist?

by MRH (not verified) :: Tue, 10/25/2005 - 10:26pm

Re #35: Good points. I'm guessing that Reid thinks Akers will soon be able to handle FG/EP duty but wants to avoid strain on his leg and so grabbed a kickoff specialist. Normally Akers is good enough to do both.

Looking at the raw numbers, it looks to me like Cortez' will be a 10 yd net gain on kickoffs over France - a huge advantage, and worth the roster spot especially if it helps Akers heal:

France's kick-offs: 60.6 yds, no touchbacks (16 kicks), and 25.1 yd ave return.

Cortez was 65.5 yds, 5 touchbacks (34 kicks), and 21.7 average return.

Yeah, the return yardage is partly a function of the coverage but when Akers was kicking hurt his numbers were 64.0, 1 touchback (9 kicks), and 21.0 ave return, so I think it was more France than the coverage.

by MRH (not verified) :: Tue, 10/25/2005 - 10:27pm

#36 Pat beat me to the point.

by Tim L (not verified) :: Tue, 10/25/2005 - 10:32pm

I can only hope Cortez's poor field goal kicking continues, at least through the Monday night game with Dallas. The Eekgirls were beyond lucky to win Sunday; they are overdrawn on their karmic account.

by Tim L (not verified) :: Tue, 10/25/2005 - 10:45pm

While watching how DVOA stacks up, I can't resist the temptation of occasionally comparing it with the Sagarin ratings. Ironically, the roles of Denver and Cincinnati are completely reversed by the two systems. Denver ranks 12 overall in DVOA, but is the second-best team under Sagarin; while Cincinnati ranks a second and 11th under DVOA and Sagarin, respectively.

I realize that how these teams rank now may have little to do with where they rank at the end of the season--I have a hard time believing that either will be anwhere near second-best towards the end of the year. But it is fun for me to identify the contrasts between the two systems and then see how the season plays out.

by Tim L (not verified) :: Tue, 10/25/2005 - 10:55pm

Okay, I don't get the Carolina comment. Did it refer to Belichick? I thought he was a Snoop Dogg kinda guy.

by admin :: Tue, 10/25/2005 - 10:57pm

I would be curious to see the correlation of the Sagarin ratings a) between the first half of the season and the second half, and b) between one year and the next, and then compare to DVOA. Someone tell me if I'm wrong, but I think Sagarin uses the exact same system for every sport, right? Of course, since his system is in large part based on wins and losses, it naturally should correlate better with wins and losses than DVOA. And, of course, unlike DVOA it cannot be broken down into various splits in order to deeper analyze each team's strengths and weaknesses.

The Cincinnati-Denver thing, I'm guessing, has to do with the importance of wins compared to plays. Sagarin sees Cincinnati as beating five cupcakes and losing to two good teams, while DVOA sees Cincinnati as WHALLOPING five cupcakes and losing to two good teams. Sagarin sees Denver as narrowly beating a number of good teams, while DVOA sees Denver as being outplayed by a number of good teams because it doesn't care about the win, but rather the issue of what the play-by-play says about the team's inherent quality.

To get philosophical for a minute, DVOA is an attempt to figure out each team's quality in terms of a Platonic ideal. Wins and losses are the external reality, but they are an imperfect reflection of a team's true abilities.

Belichick is personal friends with Bon Jovi.

by Becephalus (not verified) :: Tue, 10/25/2005 - 11:00pm

I like looking at divisional strength based on the how the teams do in your power rankings. I have seen I did about 2 mins of math and came up with the following (where 16.5 is average). It would probably be more accurate to do it with sraight DVOA...but I need to go to bed about 20 min ago :(

AE 20.5
AN 13.5
AS 14.5
AW 10.25

NE 9
NN 22.75
NS 19
NW 23.75

by James (not verified) :: Tue, 10/25/2005 - 11:10pm

any heads up on where to find the the defense against positions?

by admin :: Tue, 10/25/2005 - 11:13pm

That's coming soon.

by Tim L (not verified) :: Tue, 10/25/2005 - 11:20pm

Yes, Sagarin uses the same system for all sports. I thought the system correlates principally to point differential, with a dampening effect for blowout wins and losses. It measures the penultimate product, points (wins obviously being the ultimate product), but unlike DVOA does not examine the underlying behavior to achieve it.

What fascinates me about the difference between the two systems is: Is Cincinnati's DVOA peformance in their wins representative of how they will perform later in the year, or does their Sagarin results imply they are not a top shelf team?

I've thought about charting the dynamics of Sagarin to see how it acts as a predictor (they even have an ELO predictor metric), but lack time and ultimately interest have intervened.

by Todd S. (not verified) :: Tue, 10/25/2005 - 11:42pm

#31 It's clear to me that Gorka is complaining that the Patriots are ranked too HIGH. Thus the anti-Outsiders vitriol...

by jim (not verified) :: Tue, 10/25/2005 - 11:47pm

re: #14
what you fail to relize is that S.F played Dallas, at home and wash.and philly on the road,west coast to east coast is a tuff trip not to mention starting a rookie qb but I think Indy beating them by 25 would also be considered a blow out and that was in S.F.,and washington still gave up 17 points to an awful injured team that has just been turned upside down(trades).Yes they should have scored 52 both times they have gone to the east coast they have lost by 35 plus,don't put to much into whipping up on a bad team especially now the team has given up on this year after dumping rattay you telling your team we are done

by BlueStarDude (not verified) :: Tue, 10/25/2005 - 11:52pm

Yeah, there's not much data to go on, but SF at home w/ Rattay seems like it was a decent team (beat st. Louis, made some nice plays against Dallas).

As a Cowboys fan I was really bummed when SF decided to go with Smith because you knew they were going to completely tank, thus making Dallas's effort against them look even worse than it was (and it was bad, but there was some bad luck going on as well).

by Tim L (not verified) :: Wed, 10/26/2005 - 12:07am

I think San Fran's performance against Dallas said as much about how unsettled the latter's defense was as how not-so-awful SF looked then. When people like Brandon Lloyd torch a defense, it's not playing very well.

Recent finishes notwithstanding, Dallas defense is playing as well now as it has in at least a decade.

by bowman (not verified) :: Wed, 10/26/2005 - 12:15am

Any ideas why Den has the strongest strength of schedule, but has a lower DVOA than VOA, while JAX has the 2nd strongest schedule, and has a DVOA 14 points higher than VOA?

I know there are adjustments for fumbles, and I haven't watched many Denver games to know thier fumbling status.

by Andrew (not verified) :: Wed, 10/26/2005 - 12:27am

Tim L #28:

A balanced team is a real team, because it is not one dimensional. One dimensional teams have been one shot wonders (2000 Ravens, 1998 Broncos) or they improve into balanced teams (Patriots). Perennial playoff teams of late (00-04 Eagles, 99-03 Rams, 99-03 Titans, 00-02 Raiders, 01-03 Packers) have usually been balanced teams year after year.

by Jerry P. (not verified) :: Wed, 10/26/2005 - 12:40am

Isn't it funny that out of those 21 teams they have one Super Bowl. And it came matched up against one of the other listed teams so one had to win. And the fake Ravens have one Super Bowl with Trent Dilfer. Seriously, I hope Trent Dilfer is proud of himself.

by Jay B. (not verified) :: Wed, 10/26/2005 - 12:47am

What's the difference between "Aaron" and "The Outsiders?" in the comments?

God, I hope Cortez will only be kicking off. I also hope Akers is really healthy enough for FGs.

by charles (not verified) :: Wed, 10/26/2005 - 12:51am

Not only is DVOA showing washington love, traditional nfl yards stats have the redskins with the second best offense in the league and the fourth best defense in the league.

by Jacko (not verified) :: Wed, 10/26/2005 - 1:00am

Tim, re:53, I would argue that the Ravens would have been a perennial playoff team had they not messed with the QB position, Dilfer was getting it done. Had they gone the draft route for the QB, had him learning, they may have that QB starting now, rather than the sorry situation they find themselves in.

Broncos the same if their QB was not vacated for uncontrollable reasons, with the unfulfillable expectations placed on Grise.

But... I say this as a Bears homer. Our D is awesome, good depth at RB, QB is getting more steady each week, with Grossman to return in a month.

The only time where I can't see a 1 sided team being successful is when the O is carrying the team.

With a good D, the D can not only stop points, but set them up through good field position. The O just needs to control the ball and hit the limited opportunities. The pressure is not on to score 30 points.

When the D is awful, they are not stopping the flow of points, but the field position for the O is generally poor as well. The O is forced to take chances and more turnovers occur, etc.

Maybe this needs more investigation from someone that does not live in Australia, and brought up on Rugby.

But, maybe it doesn't...

by Rollo (not verified) :: Wed, 10/26/2005 - 1:04am

Weighing in on the Maddox Question (full disclosure, I'm a Jax homer) - I think all the people who blame him and solely him for the loss miss the point that he played a thus far elite pass defense (#2 according to DVOA beforehand iirc) without any semblance of the running attack around which the Pittsburgh offense is based.

When Maddox got the ball back after the Jax punt in OT, Pittsburgh had gotten off two running plays thus far, for -1 yards. Pittsburgh's only touchdown came off of 50 or so Jax penalty yards that were imo stupid and not necessary. On first down of Pitt's second OT possession, Maddox throws to his left quickly and is almost picked off.

The second down call was a good one - take advantage of the aggressive Jax defense, focus the attention on the left side of the offense, then come back right and fire it quickly to Morgan. Pittsburgh is tied, and needs passing yards to move the ball - playing for a tie at home against an AFC rival is not really an option. The way the play runs out is that Maddox scans right, then comes quickly back to Morgan and fires and is intercepted.

Personally, I credit Rashean Mathis for the int and don't blame Maddox. Mathis jumps the route (based on recognition), and still has to reach his hands around the receiver to catch the ball (pretty impressive grab actually). If Mathis doesn't gamble or is slightly less athletic, its an incompletion at worse.

by admin :: Wed, 10/26/2005 - 1:05am

The difference between "Aaron" and "The Outsiders" in comments: Because of the new re-code I have to go into edit to change my name in comments back to my name and sometimes I'm just too damn lazy. Will be fixed soon. :)

by Born a Bronco Fan/Die a Bronco Fan (not verified) :: Wed, 10/26/2005 - 2:48am

Yes, I am back this week even though the Broncos took it on the chin. They probably deserve things to go a little against them because they have had their share of good luck.

Regarding Wash: it seems that Wash plays according to DVOA. The strategy Gibbs uses focuses on what it takes to be good at what DVOA measures. (that is good) The two close losses against good teams were covered up by a blowout against the worst team. Broncos and SHanahan tend to focus on strentghs and weaknesses more--maybe that is why the end of the season isn't good for my team.

Aaron, I liked the comments about Saragin rankings. Your way definitely takes a lot more time and gives us all a lot to think about. SD and JAX are still overrated!

by Purds (not verified) :: Wed, 10/26/2005 - 3:05am

Re: #53

Andrew, what then do you make of the Colts? Playoffs 99-04 (minus '00), but not what has been described as balanced offense and defense.

by Israel (not verified) :: Wed, 10/26/2005 - 3:30am

You write about the Bears On 13 possessions, they've got nine touchdowns, two field goals and zero turnovers.

What about the other two? If they were failed fourth downs, then they are the same as fumbles (near the goal line). If they are missed field goals, then that should count against the special teams, not the offense. Maybe the clock ran out?

by Fnor (not verified) :: Wed, 10/26/2005 - 3:38am

I should really know this, but does anyone recall JAX's pass DVOA before the Tommy Maddox experience? I'm wondering if that game had a large part in prodding them up so high.

by 2 cents (not verified) :: Wed, 10/26/2005 - 4:32am

How can Cincy remain at #2 in Total DVOA after the drubbing? Now it has lost both games against winning teams. 1 might be a fluke. But 2 out of 2 means that Cincy cannot be ranked that high.

Also, Phylli's #8 Defense DVOA seems a little low. Especially, Detroit at #3 doesnt make any sense! How could that be?

by Becephalus (not verified) :: Wed, 10/26/2005 - 6:55am

@ #8 Yeah Andrew I have been thiking the same thing since I saw this system. Make each play weighted to the proportion it was of its games action.

by James Gibson (not verified) :: Wed, 10/26/2005 - 9:17am

Two comments:

Re #43: Yes, Sagarin uses the same system for all sports; however if you want to compare DVOA to his system, compare it to PURE POINTS. That one doesn't care if you won or lost, just what the score was. The ELO Chess system only uses wins and losses, and his overall rating is some combination of the two. Only the PURE POINTS is meant to be predictive. Right now, San Diego is the #1 team by PURE POINTS, whereas the Chargers are only #13 in ELO-Chess and #7 overall. I think he started breaking them up on the USA Today website after the BCS banned all their computers from using scoring margin. ELO-Chess is the one Sagarin has to submit there. I would also note that the top 9 teams by PURE POINTS (in order: San Diego, Pittsburgh, Cincinnati, Denver, Indianapolis, Kansas City, Jacksonville, Washington, New England) are all ranked within one field goal of each other, meaning according to his system, they would all be favored at home against each other.

Re #53: I don't see how you can call the 1998 Denver Broncos one hit wonders. In 1996, the Broncos went 13-3, but were upset by Jacksonville in the early rounds of the playoffs. In 1997, the Broncos were a wild card, but they were 12-4, and they had the highest pythagorean winning percentage in the whole NFL, and only Green Bay and Kansas City had better winning percentages. (Click my name to see this and the list of turnover adjusted drive ratings from that year, according to my personal webpage that I ran from 1996-2000). Then, the Broncos won their second of back-to-back Super Bowls the next year. One hit wonders? I don't think so. They fell off the next year after Elway retired and was replaced by Brian Griese and Terrell Davis blew out his knee 4 games into the season.

by James Gibson (not verified) :: Wed, 10/26/2005 - 9:24am

Let me follow up op on my last post: Three teams had better records than Denver in 1997: San Francisco (who I missed), Green Bay, and Kansas City. However, Denver did have the best pythagorean percentage of that year. If you go to my page, I know I'm not using the exact pyth% calculation. Instead of exponent 2.37, I was using 1/2 exponent 2 and 1/2 exponent 3, but the order will still be correct.

by Andrew (not verified) :: Wed, 10/26/2005 - 11:16am

Jerry P.

Yeah it is funny. But I think it comes from the Patriots playing to their offensive strengths in the playoffs versus the way they played in the regular season - in the 2001 and 2003 playofs it was give Brady the ball and have him throw it 2/3 of the time. The Patriots played as a balanced offensive/defensive team in the playoffs by ignoring their weakness at running offensively prior to Dillon.

The 2002 Bucs were a balanced team as well, but that was the only year their offense had a pulse (and amusingly, it was the passing game). That slight heartbeat in passing was all they needed to get over the hump with their defense.

I think it will be found that the 97 Broncos and 96 Packers were both balanced, 95 Cowboys, 94 49ers, 93 and 92 Cowboys, and 91 Redskins were all balanced teams. That takes us back to the 1990 Giants as a probable fluke unablanced team like the 2000 Ravens. I wouldn't count on winning a championship that way very often.

by Andrew (not verified) :: Wed, 10/26/2005 - 11:42am

Purds #61:

And how many games did Mr. Manning win 1999-2004 in the playoffs?

But a better explanation - 13-3 in 1999 with a powderpuff schedule. 10-6 the next year with a tougher one AND a better defense (24th vs. 29th, and 9.0% DVOA vs. 13.7%) 6-10 the following year as they still had a tough schedule, but the defense regresses back to 29th and Special Teams collapses from average to among the bottom 3. 2002, new creampuff schedule, defense improves to 22nd, special teams comes back to average, but the offense regresses relative to the league, even though it improved compared to itself. 2003 the offense comes back to life, and Peyton finally has a semblance of an NFL defense (2.2% DVOA) - hey look, he won a game! Ditto for 2004.

Now, how were they in the playoffs almost every year? Because except for 2000, the AFC was horrible from 1999 to 2002, with only a 7 really good teams (12+ wins) in 4 years, meaning a 13-3 fluke like the 99 Colts, or a 10-6 wild card wonder like the 2000 and 2002 Colts would get a playoff berth, even though they had no idea what to do with it. In no year until 2004 were they anything more than a 10.5 pythagorean win team. 10-6 teams don't go far in the playoffs, and few teams are consistently 10-6 or 11-5 year in and year out - either you get better or get worse in a hurry.

by Andrew (not verified) :: Wed, 10/26/2005 - 11:55am

James Gibson #66:

98 Broncos were the tail end of a successful three year run, and I was only looking at 1998-2004. They had an offense, but the defense was somewhat suspect. So when Elway and Dvais took away the offense, they were left with nothing in 1999. In 2000, they rediscvered an offense, but then lost it again 2001, wasting another good defensive effort, then they lost their defense in 2002. Finally, they had a moderately balanced team in 2003, but an offense only slightly above nothing. The showing after 1998 makes that team a one hit wonder when looking at the period of 1998-2004 - they weren't able to keep up a consistent effort over that period - same goes for the 98 Packers, who came off of a long streak that year and collapsed into obscurity for a couple of years, as their offense, which had collapsed in 1998, vanished in 1999.

by Chris (not verified) :: Wed, 10/26/2005 - 12:07pm

I wish we could take the '76 Buccaneers DVOA and compare it with the '05 49ers. It might be the first time the Niners are favored in anything all year.

Right now their offense is making a strong case for just punting on second down and eliminating 66% of their offensive mistakes in one shot.

by James Gibson (not verified) :: Wed, 10/26/2005 - 12:08pm

1990 N.Y. Giants look pretty balanced to me - 3rd in turnover adjusted yards/drive on offense, 4th in turnover adjusted yards/drive on defense.

I still don't think it's fair to call a team a one-hit wonder when they had good years before the years you are looking at. That Super Bowl champion was clearly part of a longer sustained success, and not just a one year fluke, which you were talking about above. The Packers that won the Super Bowl were tops on both offense and defense. So what if 1998 was the last of a successful 4 year or so run? I don't see how that fits into your point at all. As players age, teams will rise and fall, but if you're going to call teams one-hit wonders, you need to extend your examination back to the point where that team's run existed.

by Chris (not verified) :: Wed, 10/26/2005 - 12:12pm

I found this link while searching for some of the worst teams in NFL history - '52 Texans. It's a pretty interesting read

by TomC (not verified) :: Wed, 10/26/2005 - 12:50pm

You write about the Bears: On 13 possessions, they’ve got nine touchdowns, two field goals and zero turnovers.

What about the other two? If they were failed fourth downs, then they are the same as fumbles (near the goal line). If they are missed field goals, then that should count against the special teams, not the offense. Maybe the clock ran out?

Israel: One missed short field goal (vs. Detroit) and one 4th-down non-conversion in garbage time vs. the Vikings.

by Tom Kelso (not verified) :: Wed, 10/26/2005 - 12:53pm

That same season -- the last time before this year's Saints that a team played virtually all its games on the road -- was the source of Artie Donovan's now-legendary "elephant crap" story on Letterman.

Interestingly enough, when the team was relocated to Baltimore, they found the band already there and waiting. The Colt Band had existed when the original AAFC franchise joined the NFL in 1950 (with the 49ers and Browns, ironically enough), and had stayed together when that franchise collapsed in 1951. They incorporated as a separate organization.

This move prevented a threatened lawsuit by the Irsays for continuing to use the name "Baltimore Colts" when the band stayed together again after the "midnight run". George Young and Wellington Mara made a point of having the Colt Band play at Giant games during those years. Some people do have a proper respect for tradition and history.

by Tom Kelso (not verified) :: Wed, 10/26/2005 - 12:56pm

Oh, btw, what's up with the Loser League page? I keep getting a "fatal error" from the link -- but it's not crashing anything; it simply doesn't go anywhere.

by Proudhon (not verified) :: Wed, 10/26/2005 - 1:25pm

Are you going to post the pass defense numbers vs. #1, #2, #3 receivers, etc., this week? If you already have, can you tell me where to find it?

by ChicagoScott (not verified) :: Wed, 10/26/2005 - 1:51pm

Loved the comments about the Colts-Pats hype machine & the Iraqis outlawing Manning-Brady debates. Very clever.

With 2 weeks to go, I'm already sick of all of the talk about the game...coming from the voices in my head.


by admin :: Wed, 10/26/2005 - 1:52pm

Sorry about that. The fact that there is still no official Jets-Falcons gamebook in HTML format is screwing me up. Plus, we're working on automating the defense vs. receivers thing. You'll now find it on the team defense page (click my name).

by Freddie (not verified) :: Wed, 10/26/2005 - 2:10pm

Let's see here.
Falcons win and drop 6 spots to 18th.
Am I the only one here that thinks this
has to be too low. Obviously, the system needs some fine tuning. A 5-2 team worse than 2-4 team. And one of those loses was without Vick. This system is just not a good predictor of how good a team is.

by JonL (not verified) :: Wed, 10/26/2005 - 2:12pm

Nice job this week. As someone who hasn't yet bought PFP05, I found the fumble discussion interesting.

Looking at these ratings, I'm relatively surprised how unimportant special teams seems to be for a team's overall rank. I also noticed that there seems to be a preponderance of good punters and lousy kick/punt returners. I suppose the two could be related, but does anyone else have a better handle on this?

by B (not verified) :: Wed, 10/26/2005 - 2:22pm

"Falcons win and drop 6 spots to 18th.
Am I the only one here that thinks this
has to be too low."
I'm going to go out on a limb here and guess that yes, you are the only person who thinks the Falcons are underrated by DVOA.

by admin :: Wed, 10/26/2005 - 2:39pm

Yeah, punters have been crazy good this year. I may need to re-jigger the punt baselines. Now go buy the book, will ya?

I would just like to point out that the person in comment #80 wants us to move the Falcons up in the ratings because of the game without Vick. He ignores:

a) the fact that Schaub was better in that game than Vick has been in any other game

b) the fact that there is a huge section of the commentary that describes in detail why we don't move teams up and down based on games missed to injury

c) the fact that the section below that one discusses in detail the luck involved in Atlanta's turnover margin this season

I would have to say that 80% of the hate mail and 95% of the message board posts attacking me come from people who clearly haven't bothered to actually read the commentary. I want to answer critical and negative e-mail when it is thoughtful, but I'm going to have to stop responding to anyone who clearly didn't give me the courtesy of actually reading what I wrote.

One line has been in every single commentary and clearly nobody can be bothered to read it: "Remember, of course, that any statistical formula is not a replacement for your own judgment, just a tool to use in analyzing performance."

by Fiver (not verified) :: Wed, 10/26/2005 - 3:22pm

Aaron, just some totally unsolicited advice: There are many, many, many, many people out there reading your stuff now, and not all of them are going to be thoughtful people asking intelligent questions that advance the debate. Lots of them just want to blow off steam with some hatemail or get into an argument because it entertains them. Assuming that every message board post or email is worth reading, reflecting upon, and responding to, is a sure road to unhappiness.

Your work is making you more of a public figure, which means you need to filter out more of the public if you want to continue your work.

by Mshray (not verified) :: Wed, 10/26/2005 - 3:32pm

"...but I’m going to have to stop responding to anyone who clearly didn’t give me the courtesy of actually reading what I wrote".

As well you should. I've been here since the 1st time TMQ gave FO a callout & you, Aaron, as well as all the rest of your writers, have an uncommonly excellent mix of intellectual analysis and it's-still-only-a-game irreverant humor. Unfortunately there will always be people (and far too many of them, imho) who get an unhealthy does of their self-worth from their sports affiliations. You really should not worry about those who haven't gotten beyond stage 2 pre-conventional Kohlbergian moral development, and just keep doing what you do best. And anyway, getting more email/posts than you can respond to is an indication of success. Which you & all the Oustiders richly deserve.

by JonL (not verified) :: Wed, 10/26/2005 - 3:43pm

Because I'm still thinking about this, I noticed that teams are doing fairly well on kickoffs as well (only 12 teams have a neg. DVOA, and that includes four teams that lie between -0.5 and 0). Could the success of kickoff and punt teams have to do with the lack of good return men? Why don't more teams use their good WRs/RBs? I know they are afraid of injuries, but how many returners get injured every year?

by Ferg (not verified) :: Wed, 10/26/2005 - 3:54pm

Re 86: Perhaps a factor is that kicking tends to get worse as the season goes on (due to the weather getting colder). So the everybody-above-average effect may fade out by the end of the year.

Unless the special teams data already adjust for this.

by Ferg (not verified) :: Wed, 10/26/2005 - 3:57pm

Of course, if I had bothered to look at the special teams page *before* posting, I would have noticed that the data *are* adjusted for week as well as location. So, please ignore what I just said.

by B (not verified) :: Wed, 10/26/2005 - 4:02pm

Speaking of weather, a common NFL cliche is "It's harder to throw the ball as the weather gets colder." And, conversely, running is more important in cold weather. I've suspect this is should be files right next to "defense wins championships." But it would be an interesting study. For purposes of the study, we can define cold as "kickoff temperatures at or below 32 degrees fahrenheit" And we should look at the passing performances of both the home and away teams, and compare those to the same team's performances in warmer games.

by TomC (not verified) :: Wed, 10/26/2005 - 4:41pm

re: 89-

Anecdotally, I would say temperature alone has almost no effect on the relative importance of passing vs. running (my most painful data point on this is the 1988 NFC championship game, known around Chicago as the deathknell of the concept of "Bear Weather".) However, I would guess that other weather conditions that tend to obtain in the late autumn / early winter --- high winds, cold rain, sleet, snow, etc. --- do in fact influence the run/pass importance hierarchy.

by Trogdor (not verified) :: Wed, 10/26/2005 - 5:01pm

Shouldn't the "cold" temperature be 34 degrees, or whatever the cherrypicked stat is for the Bucs never winning a game below that temp? Or was it Favre never losing a game? I forget, the dumb stats all roll together.

Aaron, I'd be more than happy to answer your hate mail for you. Most of them would get an automated response full of terms involving their mouths and my lower abdominal region, but some of them could get personalized ones as well.

by Pat (not verified) :: Wed, 10/26/2005 - 5:13pm

Trogdor: Favre is 38-0 at home below 34 degrees when not playing an African-American quarterback with the number 7.

by shonk (not verified) :: Wed, 10/26/2005 - 5:22pm

Re weather conditions:

Anecdotally (though I seem to recall reading some analysis to this effect a couple years ago), I'd be inclined to say that crappy weather tends to actually be good for the passing game. Snow, especially, seems to be conducive to passing (the Raiders-Broncos game from last year comes to mind). DBs have to react to WRs, so whenever the footing is unsure, they tend to be much more likely than the WRs to slip/fall. Of course, high winds (QB has no idea where the ball will end up) or heavy rain/sleet (can't get a good grip) probably tend to mitigate the DBs-falling-all-over-themselves factor, so I guess I have no general point.

by MJK (not verified) :: Wed, 10/26/2005 - 6:12pm

Weighing in on the weather issue,

Tom Brady was once quoted as saying he perferred running the passing game when there was bad footing (e.g. muddy or snow covered field). As Shonk mentioned, WR's have the advantage over DB's in bad footing because they KNOW in advance when and where they will make their cuts. And, possibly more importantly, Brady talks about how bad conditions slow down the pass rush and give him more time in the pocket to make his reads.

On the other hand, I would expect bad footing to slow down a running game. In the running game, defenders are just running towards the guy with the ball, and he must rely on speed or agility (e.g. making cuts) or power to avoid tackles and gain yards. The runner loses all three in sloppy conditions.

Remember the Snow Bowl? The Pats miraculous drive to tie the game occurred after the conditions got really bad, and it was almost entirely passes.

The slippery ball issue is solved by most QB's by wearing gloves. Yes, it affects their throwing ability a little, but if they have practiced with them, it doesn't seem too bad. Of course, bad visibility and high winds certainly would slow down the passing game.

by Purds (not verified) :: Wed, 10/26/2005 - 6:25pm


You wrote, in typical Brady-Manning fashion instead of looking at the Colts as a team, "And how many games did Mr. Manning win 1999-2004 in the playoffs?" That Manning-centric comment aside, you need to adjust your earlier statement to include playoff success as the mark of balanced teams. In your first post, you did not. You simply stated as a universal truth that perennial playoff teams were balanced, and I am suggesting that you're wrong.

Now, I am glad to see you have an excuse for every season of Colts playoff appearance, noting that only "powderpuff schedules" allowed them to get there, but is that really the only explanation you can find? Isn't it possible that one-dimensional teams CAN be consistent playoff teams, even if they won't get much past the opening round?

Clearly, it's better to be a balanced team. But, I think it's short-sighted to lump together that connection of balance and playoffs. And, I don't care how powderpuff a schedule, you say the 13-3 season wasa "fluke." I'd suggest to you that the real fluke for the Colts was going 24-8 over last two years and not getting a single playoff bye -- it's NEVER happened before, and clearly did not benefit Indy, though they won 3 of 5 playoff games in that stretch.

by Pat (not verified) :: Wed, 10/26/2005 - 6:47pm


Actually, the balanced/playoff success connection was looked into by Aaron here. It's a little limited by statistics, but it does look like balanced teams fare better in playoffs.

by jebmak (not verified) :: Wed, 10/26/2005 - 7:14pm

Other lucky defenses include ... the 49ers (5 of 5).


That doesn't bode well for things to come.

by Fnor (not verified) :: Wed, 10/26/2005 - 7:43pm

re: Cold weather offense
I think the idea is that passing is great if you can get a really quick drive going, but if you get bogged down it's better to slog it out with the run, just because the QB's hands will get friggin' cold. I base this, of course, on complete personal experience, and also the memory of trying to throw a ball in the pouring cold rain this past sunday. Granted, I'm not an NFL quarterback, but they get worse conditions for longer and actually get GLOVES.

by thad (not verified) :: Wed, 10/26/2005 - 7:45pm

esoteric email alert.

re james gibson,
I read your website, it was very good. However as i recall one of the reasons Bud Goode picked the packers was that he felt that Denver would stop running the ball. One of his indicators is rush attempts and I think he said that Denver would probably not achieve their season average. I think thats what he said, it was a while ago

by Purds (not verified) :: Wed, 10/26/2005 - 7:45pm


I completely agree with you and the article. What I don't agree with is the earlier assertion that only balanced teams will regularly MAKE the playoffs.

by Kibbles (not verified) :: Wed, 10/26/2005 - 8:27pm

Re #33: Aaron mentioned the absolute bull the refs were playing with the phantom defensive pass interferences and the blatant offensive pass interferences, so I'm content. That's all I wanted.

I swear, I saw at least two instances of Plaxico Burress putting both hands on the defenders shoulders, used his position to leverage himself into the air and hold the defender down, and the DEFENSIVE player was called for interference. I guess it's now illegal to allow the offensive player to use your body as a platform to go up for the ball. Of course, it's also illegal to try and stop him.

I had 2 thoughts when watching that final Giants scoring drive (which I decided was absolutely inevitable after the first first-down). The first thought: I hate having a 6 point lead. I'd much rather have a 3 point lead, because then I feel like the other team will settle for a field goal and go to overtime, but with a 6 point lead, the entire field becomes 4-down territory and a score means a loss. The second thought: I can't believe Denver started playing softer defense when NY got the ball down 6 with 3:30 to go. I understand when they played softer defense with NY down 13, but if NY's down 6 with 3:30 to go, I'm calling ridiculous blitzes. I'm sending a corner, both corners, 6 men, 7 men. I'm bringing the house. Why? Because the blitz had been effective in getting you the 13 point lead, in the first place, AND... because if you blitz and they beat it, then they're going to score in a hurry, giving you the ball back down by 1 point with several timeouts and probably at least 2 minutes to work with. I like the chances. Meanwhile, by playing soft, you're running down the clock, but they have timeouts and the 2-minute warning, so they aren't concerned with clock. Really, you're just running the clock out on your chance to score again if they score.

One final note, in the "insult to injury" category... not only did NY score with 5 seconds left, they left Denver with no recourse but the miracle-lateral return. They kick off, and Denver obligingly attempts to lateral like mad to get the TD. NY predictably recovers one of the laterals, ending not only the game, but Denver's bid at an NFL record. They had gone over 5 complete games without a turnover, which would have been the NFL record, but the NFL only tracks COMPLETE games, and the fumble with 00:00 showing on the clock ended their streak at "4".

One other thought that I always have. Why do teams attempt the miracle return? I always felt like their chances would be better if they just fair-caught the ball and attempted to run an offensive play.

by James Gibson (not verified) :: Wed, 10/26/2005 - 8:42pm

Thad - thanks for the comments about my website. If you follow the links through my main page (linked in #67 above) to my essay about the Broncos-Packers Super Bowl, in it, you can find linked Bud Goode's article. However, I've linked his article in my signature on this post so you don't have to go through all that.

Bud Goode did adjust the Packers numbers due to the thought that they would run more and I can see that. The paragraph that really bugged me in his article was this one and the following line:

"John Elway has not been an "efficient" qb in previous Super Bowl games. He
relied on his arm too much. Payoff? Mucho sacks and INTS at a cost of 3 points
per sack and 5 points per INT. What was the score in the Redskin Super B.?
Something like 42-10.

So give Elway one INT and one Sack more than Favre."

He gives Elway blame for the Broncos debacle against Washington and assumes something similar will happen.

I don't mean to be a crazed Broncos fan like the ones that were sending Aaron hatemail, but I'm pretty passionate about the '97 Broncos as a team - I don't think they're given enough credit - and the '96 through '98 Broncos as a collection.

by Will Allen (not verified) :: Wed, 10/26/2005 - 8:48pm

Wind is what kills the passing game, and, I suppose, rain, if it is torrential enough, as in tropical depression-like conditions. A very slippery field, and a typical rainfall or snow, however, really harms the pass rush, perhaps even more than db performance.

I have no idea whether northern cities are much more likely to have very windy days in December and January than in September and October, although that the memory from my youth.

by James Gibson (not verified) :: Wed, 10/26/2005 - 8:48pm

Kibbles - I'm not sure I agree. The Giants were making short yardage against the Broncos soft defense, but they only made the long play to Shockey when Darrent Williams blitzed and left Ian Gold one-on-one with Jeremy Shockey. Granted, they only rushed 3 D-lineman on that play so they only had a 4-man rush, but dang, the Broncos kept not completing passes because they had Will Allen guarding Putzier and Tatum Bell. The Broncos seriously needed a CB on Shockey on that play. On the last play of the game, Lynch came in on the delayed blitz as Denver had been using all game, and Eli bought himself time by backpedaling and then Toomer was open in the endzone - I'm not sure Toomer ends up open on that play if Lynch stays back in his safety role.

The short yardage seemed to be doing a pretty good job of eating the clock, and it might have run out were it not for that Shockey play. Granted, Tim Carter did almost catch a TD on the play before that.

by jim (not verified) :: Wed, 10/26/2005 - 9:35pm

NON-ADJ TOTAL VOA is a head scratcher for me,the skins are 12 but adj 24 while most teams NON-ADJ TOTAL VOA is higher than adj ,the skins are doubled no other team is doubled in the way of positive and jacksonville has a 12 point gain but is not doubled.so does this mean playing in 1 dome game, 1 game in Denver in the rain and a game at K.C mean that much.Maybe I have misunderstood what NON-ADJ TOTAL VOA as described at the top of rankings means.Or maybe it is Denver fans are that scary

by Andrew (A.B.) (not verified) :: Wed, 10/26/2005 - 10:39pm

RE: jim 105

Please read Aaron's commentary on foxsports.com, which can answer your question.

by Born a Bronco Fan/Die a Bronco Fan (not verified) :: Wed, 10/26/2005 - 10:50pm

Kibbles and J Gibson
Like both the viewpoints expressed about the NY/Den game. Interesting how the same game has two different takes on it.

I agree with both of you a little. Offensive pass interference is the least called and most occuring penalty by #1 WR--watch Terrell Owens, Randy Moss, Michael Irvin, etc...-tall and strong means they can use their hands more than the average WR.

Denver is still trying to figure out how to finish a game defensively. The game plan sucks right now. Agressive for 3 quarters and then afraid to get into or stay with something other than hoping the other team gives up. If you want to win just hang close until the final 10 minutes. It continuously shows up, Wash, NE and now NY. This one was on the road and they didn't hang on.

Anyone buy into the fact that NY had to win that game with both owners facing their deathbed? Ref's were pretty bad (for NY)
It is still all about the Money!!

by Fnor (not verified) :: Thu, 10/27/2005 - 1:21am

#107: The skill is to do it while your hands are going towards the ball. So long as they are, then you're making an effort for the ball, and the defender cannot impede your path. You see a few defensive penalties drawn when the back pushes back. Bad back! Bad!
#104: They were able to effectively move the short game, though, because they knew that the Broncos were "playing the clock" as opposed to "playing the opposing team." They set up their prevent defense like they always do and it bit them in the arse like it always does. While your point is valid concerning the one blitz in the series, I would call that more of a coverage error than a byproduct of the blitz. In any case, any time you tell your opponent "sure, we'll give you the short ball, but not the long one" with over a minute left, you're shooting yourself in the foot. You need to stay agressive or a disciplined QB will dink you to death, with an ocassional big play when they see favorable coverage.

by 2 cents (not verified) :: Thu, 10/27/2005 - 3:10am

I say Colts finally get the Pats monkey off their backs. Everyone then prop Colts to win the SB this year. Until Pitts kicks their azz in a monday nighter. But remember dont jump onto Pitts bangwagon when that happens. Because Cowher and Marty can never win a big one.

by Brian (not verified) :: Thu, 10/27/2005 - 4:37am

I wish I could have gotten tickets to Colts-Steelers. Shoulda waited in line outside instead of trying to buy them the next day :( Should be a good one!

by Another in the District (not verified) :: Thu, 10/27/2005 - 8:39am

It would be interesting for us to discuss what someone from my hometown in another thread.

Has FO "jumped the shark" on us? Some of us are coming in here to read what our favorite critics say not what FO says. It is almost as if since they went to Fox they have become media insidery on us.

I do not think they are as funny or as insightful. They have as much shtick as Easterbrook.

I have been reading a lot of Salon.com to compensate. Should I have to do that?

This post will be deleted I am sure but I thought we should be intellectually honest and discuss it.

by P. Ryan Wilson :: Thu, 10/27/2005 - 10:06am

Just for the record, I jumped the shark a while back, when I got my own talk show, moved out of Lanford, and won the lottery. Oh wait, that was Roseanne. And if an "insider" means that you have to pay face value for Steelers tickets, sit in the nosebleeds, and have absolutely no access to players or coaches, then I'm "in."

by JonL (not verified) :: Thu, 10/27/2005 - 10:07am

I'm not sure why I'm so hung up on the special teams issue, but could it be that kickoffs and punt returns for touchdowns rely on chance, much as fumbles do? The chance would come in mainly on blocking, where the returner decides to run, etc., and in that sens I suppose every play relies on some level of chance, but I wonder if this could explain the fluctuating levels of ST touchdowns year-to-year.

by Fnor (not verified) :: Thu, 10/27/2005 - 10:28am

#112: Oh yeah? I am the shark!

by thad (not verified) :: Thu, 10/27/2005 - 10:52am

Yes Denver is not given enough credit. But i think there are several reasons why.
1. the Bulls and Jordan, the Broncos were great, the Bulls won 6 in 8 years.
2. The Yankees, who also won more.
3. The loss to the Jaguars. It was a really fun game to watch, and going in I totally thought Denver would win, but if you lose to a second year expansion team at home, you are going to be punished in the court of public opinion.
4. Elway. OK there are many reasons why Denver got crushed in three Super Bowls, but Elway sucked. I think its fair to say that most people are going to blame a loss on one person or one play or one bad call rather than look at the whole game or come up with new and better drive stats, seriously, what does Aaron get more of, goofy emails about the Bronco's dvoa ranking or emails about coming up with new stats using the matrix function on excel.
4. Time. As time goes by the 90's are quickly forgotten, especially since there is so little info on the internet regarding the 97 NFL season, the gamebooks are gone, the espn stories are gone, the sortable cbs sportsline stats are gone.
5. The perception has clearly changed, but at the time the idea that an AFC team could compete, not win, just play a close game against the packers, was considered idiotic. I remember saying things like, Gee I dont know, maybe Denver can win, and people would start going on and on about Farve and the nfc and talking to me like i was a small dull child.
6. While the 98 team was clearly fantastic, the 97 team won one blowout and two close games before the Super Bowl. I would guess that most people would sort of rank them with the 2001 Pats as a "good but lucky team" rather then the 2003 Pats which seemed pretty unstoppable.
btw, it kind of bugs me that the 2001 pats are not as aclaimed as i think they should be, but whatever, I think Lance Armstong is probably the most underrated athelete in my lifetime, it doesn't keep me up.

by Rick "32_Footsteps" Healey (not verified) :: Thu, 10/27/2005 - 11:03am

I don't know what surprises me more - that people complain about biases in forcing every team into the same equation, or that I read two different DEVO references in this week's discussion.

by admin :: Thu, 10/27/2005 - 11:06am

I just want to make sure what you understand what you mean when you say FO has jumped the shark.

FOX is not editing our content in any way. We're still doing the same thing we did before. They give us total freedom. I still live in the same house using the same spreadsheets on the same computer and writing the same style articles with the same sense of humor.

What you are saying is that you don't want Mike, Mike, Al, Vivek, Benjy, etc. to get paid. Because that's the difference between FO now and FO before the FOX deal. Before, I scraped by and barely made a living. Now, my wife and I can provide for our daughter and there's money left over to pay everyone else.

If that's a problem for you, you are welcome to stop reading. If you don't have a problem with FO making enough money that I can actually pay the other guys, but you think our content has somehow changed since the agreement was signed with FOX, the CONTACT FORM is up in the top menu bar, and we welcome your e-mail.

by Ray (not verified) :: Thu, 10/27/2005 - 11:13am

If FO has jumped the shark, why do I sit here and keep refreshing the page all day long waiting for the new FO articles to be posted? You know guys, you CAN post articles before 4pm on the day they're due. It's okay to not post all three Wednesday articles within an hour of each other. I mean, even Easterbrook usually gets his million-word article up by 1pm! ;^)

by Ray (not verified) :: Thu, 10/27/2005 - 11:22am

Psst, Aaron, leave the feeding and/or repulsion of the trolls to us. People who post crap like "This post will be deleted I am sure but I thought we should be intellectually honest and discuss it." obviously haven't been here long enough to know what it was like before the deal with Fox. He's not worth your time, which would be better spent writing the articles I crave. ;^)

by admin :: Thu, 10/27/2005 - 11:22am

Speaking of how I spend my time ... My wife's new job requires a long commute, which means that I need to spend a lot more time as a daddy this season. Therefore we've assigned the task of "assistant editor" to Tim, which means that our posting schedule is often dependent on the academic schedule of the University of Texas Law School. If you ever get frustrated about late postings, just think of the smiling face of little Mirinae, who turned two yesterday.

(Also, trying to write a comment about all 32 teams for FOX along with my usual long-winded statistical analysis is really hard, and I seem unable to get the Tuesday DVOA article done before 2pm. That's my fault, and I'm working on it.)

Never let it be said that the staff of FO is not completely open about the reasons why things do or do not get done around here ... it's my executive policy that nobody at FO should sacrifice their family time to write about football, and that even includes me. And now, Ray, I shall let you fight the grumpy old trolls that live under the bridge.

by Arthur Fonzarelli AKA Fonzie (not verified) :: Thu, 10/27/2005 - 11:28am


by Tony D (not verified) :: Thu, 10/27/2005 - 11:34am

Any ETA on when the stats for QB/RB/WR/TE will be updated (i.e. when you expect the Jets-Falcons play record)?

by Tony D (not verified) :: Thu, 10/27/2005 - 11:47am

And anyone dogging these guys for accepting the deal with Fox needs to have their head checked. To be able to play stay-at-home dad while getting paid to do what you love to do is nothing short of spectacular and I'm sure a dream come true for many people. If you're going to say anything about the Fox deal, say congratulations.

And Aaron, I work in Excel all day for the military, if you need a hand with crunching numbers or anything, gimme a holler. I eat this stuff up, and maybe I wouldn't be able to support a family on it but I could help out if you need it. :)

by Johnnyel (not verified) :: Thu, 10/27/2005 - 11:55am

Re: 83
In PFP '05, I read something about fumbles being significantly more likely on a sack than other plays from scrimmage. Is fumble recovery different on sacks as well, or is it still 50/50?

by B (not verified) :: Thu, 10/27/2005 - 12:24pm

RE #124: Aaron mentioned this before, and I forget the details, but it's my understanding that the defense's likelyhood of recovering a fumble depends on where the fumble happened, and a fumble behind the line of scrimmage (IE, a sack) is most likely to be recovered by the defense.

by dryheat (not verified) :: Thu, 10/27/2005 - 12:26pm

I'm sure fumbles born out of sacks are going to favor the defense, just because most of the offensive players aren't going to know there's a fumble, whereas there's probably seven defenders who see the ball come out and get a several-step headstart. And the one guy who's in the best position to recover for the offense is typically lying under 275 lbs worth of linebacker.

by dryheat (not verified) :: Thu, 10/27/2005 - 12:32pm

In re: 115

I agree regarding the 2001 Patriots. There is no doubt in my mind that when late December rolled around they were the best team in the NFL. I think if they played the Rams again the next night, they would have beaten them again. Let's save the amazing-underdogs-win-against-all-odds stories for the 1980 US Olympic Hockey team.

by Tony D (not verified) :: Thu, 10/27/2005 - 12:34pm

That leads me to another thought I had watching football this past week. Do teams yell to one another for things like that? You'd think that communication would be valuable -- O-linemen misses his block, so he yells "FREE!" to the QB; or the snap is mishandled the ball is loose, the QB yells "LOOSE!" or something (in this case the defense already has the advantage and you need to let the O-line know to jump on the ball).

by dryheat (not verified) :: Thu, 10/27/2005 - 12:57pm

Tony D, I thought of that exactly. I know the defense usually yells "Ball!" to let them know the ball's loose, or "Ice!" on an interception to let the other defenders know to start blocking. So why doesn't the offense? Then I thought, how funny it would be if a Defensive End, to use the terminology from your post, yelled "Loose!" and an offensive lineman came off his block to look for a fumble that wasn't there, leading the QB to get killed.

My guess is that in a fumble situation, the linemen should just keep blocking their defensive counterparts and hopefully the QB or RB can recover the ball. While not easy to do, if they can keep the man they're blocking from getting to the fumble, it's just as valuable as chasing after it.

by Fnor (not verified) :: Thu, 10/27/2005 - 1:13pm

The problem with fumbles in the backfield isn't that no one knows the ball is loose, but that the offense is busy running a play downfield that no longer exists, while half the defense is running to where the now-loose ball is.

by dryheat (not verified) :: Thu, 10/27/2005 - 1:54pm

Fnor -- I agree in cases on a botched snap or handoff, but in the case of a sack-fumble, aside from the receivers, the entire offense is relatively close to the ball...they're just looking in the other direction. I'm sure they quickly realize what happened, just not as quickly as the defense.

by Ray (not verified) :: Thu, 10/27/2005 - 1:57pm

Part of the issue with too much verbal communication during the play would be that it wouldn't take long for the opposing team to figure it out, and then they could start sending false signals. For instance if the RDE yelled "FREE!" while he rushed, he could cause the QB to panic and leave the pocket even though there was no danger.

by Pat (not verified) :: Thu, 10/27/2005 - 2:08pm

I think the other problem might be that for the visiting team at least, I doubt you'd be able to hear it over the noise.

by Andrew (not verified) :: Thu, 10/27/2005 - 2:22pm

thad #115:

It was good for Denver to lose in 1996. They were not yet ready to defeat Green Bay, like they would be in 1997. Had they gone to the SB in 1996, it would have been Elway's 4th loss. The magic of 1997 and 1998 might never have happened.

by Leon Lett (not verified) :: Thu, 10/27/2005 - 3:04pm

I just wish that there was more communication after we pick up the fumble, like" there's a really fast little guy behind you trying to knock the ball out even though they are down by 30 points!"

by James (not verified) :: Thu, 10/27/2005 - 3:10pm

The Denver, GB superbowl was the most played championship game I ever saw. Everyone I've ever spoken with counts Denver as oneof the great NFL teams.

The 98 team had the misfurtune of the VIkings losing. If two 15-1 teams had met in the superbowl and Denver had beaten the Vikings(juggernaut like team IMO) then they would have talked about a little more.

Finally, that Elway is now considered the best qb ever in many eyes as a result of the two titles that team won. IMO, that is a credit to how good those teams were. Before they came along Elway was maybe a step below Marino in most eyes outside of Denver. After he became even Joe Montana.

That team gets alot of credit.

by dryheat (not verified) :: Thu, 10/27/2005 - 3:17pm

Weren't the Broncos 14-2 that year, same as the Falcons?

by dbt (not verified) :: Thu, 10/27/2005 - 4:54pm

Look, the 2001 Pats get panned because of the tuck rule play and because nobody realized they were going to elevate last second wins on field goals after Ricky Proehl scores a potential game winning TD to an art form.

Speaking of which, I still say Belicheck and Proehl is the closest you'll ever get to seeing Lucy and Charlie Brown in the modern NFL.

"You won the game!" "Yay!" *yoink*

by Doc (not verified) :: Thu, 10/27/2005 - 6:26pm

Have you looked into converting DVOA differential into point format? That is to say, if the Colts have a DVOA 15% higher than Dallas, that means they should be a 7 point favorite (on a neutral field), if the difference was 30%, they would be 14point favorites, etc???.... And then you could add a premium or discount for home field advantage.. speaking of home field advantage, while I know you've said there is a 17% premium that should be added for home teams... dont certain home teams have greater home advantages (not just in terms of weather, but a team like Rams is much more competitive at home than on road)--could there be a home DVOA and a road DVOA?

by B (not verified) :: Thu, 10/27/2005 - 6:50pm

Doc: Where were you 3 months ago? Aaron briefly tried the DVOA as points thing, it proved to be less popular than DVOA as percentages. And yes, different teams have different home field advantages. Denver is first, then the NFC west teams, then dome teams, then cold weather teams in winter.

by Tony D (not verified) :: Thu, 10/27/2005 - 8:02pm

Did the points thing not work out or something? What was the conversion rate?

And is there really a 17% bump for home field? Wow.

by james (not verified) :: Thu, 10/27/2005 - 9:05pm

I was doing more research into estimated wins.

The team to win the superbowl has been 1st or 2nd in estimated wins in every year except 2001. The pats won 11 games but had 7.5 estimated wins.

Interestingly enough Carolina in 2003 had similar totals. 11 wins 7.5 estimated wins approximately.

It seems that a team who has 11 wins though they "should" be a .500 team is very dangerous.

by NF (not verified) :: Thu, 10/27/2005 - 10:53pm


About the first two points on the mis-rating of teams:

1) I think the Steeler-Jacksonville game will likely diminish in effect as more games are played, but remain. I think the effect of it at the end of the season will be meaningful, as it shows that with the exception of Charlie Batch, an unknown, the depth at QB for the Steelers is awful. Granted, this is not that likely to matter for the rest of the season, but based on the play of the two qbs who have started, there is no safety net for the Steelers if Roethlisberger is injured for an extended period of time.

2) I think the Washinton-SanFran game will not result in an obvious abnormality in Washington's DVOA at the end of the season. Right now, opponent-adjustments are at less than 3/4 strength, and the Washington blowout will go from representing around 1/6 of the total DVOA for the Redskins to only 1/16.

by Pat (not verified) :: Thu, 10/27/2005 - 11:01pm


More or less, a 0% offensive DVOA versus a 0% defensive DVOA scores 21.5 points per game. A 100% offensive DVOA versus a 0% defensive DVOA would score 43 points. It's not quite that simple, but that's probably close enough.

by Dave (not verified) :: Fri, 10/28/2005 - 12:00am

I'm stunned that no one has corrected the Spanish conjugation, that Aaron even referenced in the teaser. "Plaxico" is Spanish, not for "pushing off", but for "I push off."

by james (not verified) :: Fri, 10/28/2005 - 2:34am

So, the 17% home field advantage represents about 3 points of the 21 ppg of an average offense against an average defense. That makes alot of sense since this is what is used by odds makers.

My next question is how accurate/elaborate where the point projections?

by Freddie (not verified) :: Fri, 10/28/2005 - 10:29am

Regarding comment #80 and your response.
I did read about the system. All I am saying is that something has to be wrong. Maybe where putting too much weight on certain factors and too little on others. Atlanta IS a better team than 18. Finding out why can help you make your system better. I can't beleive that they made it to the NFC champion game last year or are 5-2 this year because of luck. I do love the approach.
Keep up the good work.

by Pat (not verified) :: Fri, 10/28/2005 - 10:41am

I can’t beleive that they made it to the NFC champion game last year or are 5-2 this year because of luck.

I can definitely make an argument for them making it to the NFC championship last year because they were lucky. They faced the NFC West, the worst division in football, and then managed to face one of the worst playoff teams ever in St. Louis. Those two things aren't indicative of a strong team. They're luck.

by Tony D (not verified) :: Fri, 10/28/2005 - 1:08pm

Ok, so for offense it's 1%DVOA=0.215points, what's the conversion rate for defense?

by james (not verified) :: Fri, 10/28/2005 - 1:21pm

I've been looking back at the old numbers and it seems a -30% defense is a good indicator of a immovable defense. Anyone input on that number?

I also experimented with capping a game by starting with pass offense, coming up with an expected score from that and either subtracting 4, adding 4, or adding 0.

what i came up with:
Strugglers(10 pts may be tough)

Easy Days(30 pts might be easy)
san diego

add in my coming off a loss corollary and you get
cinc, NE, SD, Dallas should all win by 10 points.

Just a little try at predicting with DVOA. Hopefully I haven't found fools gold.

by James Gibson (not verified) :: Fri, 10/28/2005 - 2:45pm

Thad (back in #115) - I'm not really arguing in this post, just expanding on my thoughts. It's really only the '97 Broncos that I feel are disrespected a little bit. The '98 Broncos, not at all. Yes, they did have the long undefeated streak and that's why. I will note that the Broncos '97 that had narrow victories in the playoffs had to go on the road for two of them. They had a worse record than Kansas City, but a better record than Pittsburgh, although the Steelers won the division. I guess the thing that tends to bother most now is that when the national media puts up polls of what the best Super Bowl was, that one is nowhere to be found, although at the time it was declared the best ever by several different sources. I still have articles from the Sporting News and Sports Illustrated claiming as much. I'm also surprised that given that's the Super Bowl that ended the NFC's streak it's not talked about more.

I'm really interested in seeing the '97 DVOA ratings as well. Denver had a higher pythagorean winning % than anybody else, but I do agree that Green Bay (and likely San Francisco) had harder schedules, so I wouldn't be surprised to see Denver drop a little in opponent adjustment, but I'd still bet the anlalysis there wouldn't put Green Bay as a 14 point favorite.

And I guess the real reason I feel as strongly as I do about that group is one of my other favorite football topics - the Hall of Fame. I would like to see players other than Elway get into the Hall of Fame (ok, I'm way off topic from the current DVOA ratings now - I think this was all started when I got mad at Andrew's one-hit wonder comment). I know Sharpe has a good chance, but that seems to be about it. Gary Zimmerman's been denied twice now, and I would like to see more attention on Steve Atwater.

by james (not verified) :: Fri, 10/28/2005 - 2:53pm

We can start a poll here

1. Denver/GB
2. Carolina/NE
4. SF/Cin(the second one)
5. Ne/Stl

The other four were great because they basically ended on the last play of the game. Denver/GB is the greatest because of the sheer amount of great plays made by both teams. Then a chance for GB to mount a game winning drive at the end. Besides Car/NE its the only SB I remember where neither team left their offense at home.

by Tony D (not verified) :: Fri, 10/28/2005 - 3:32pm

Ok, using 1% DVOA offense = 0.215 points scored, 1% DVOA defense = .717 points prevented, and adding 3 pts for home field, I get the following predictions for game scores. Tell me if they look right:

DAL 25

CLE 47
HOU 27

JAC 36

CAR 26


GB 16
CIN 39

OAK 35
TEN 31

WAS 31
NYG 22

MIA 29
NO 17

KC 22
SD 31

PHI 22
DEN 18

TB 45
SF 0

BUF 22
NE 30

PIT 24

Now obviously scores like 18 and 22 aren't likely football scores, but you see the meaning.

This also doesn't take into account points scored by the defense (INT for TD, for example), but I think DVOA reflects fumble numbers (although I don't think it can predict scores from them).

by Tony D (not verified) :: Fri, 10/28/2005 - 3:34pm

By the way, if these turn out close, we're all going to be Vegas millionaires next week.

by James Gibson (not verified) :: Fri, 10/28/2005 - 3:50pm

Tony D - 18 and 22 are more football scores than 1-0 for Detroit/Chicago is. If that one turns out to be correct, you can pick my lottery numbers any time.

by Tony D (not verified) :: Fri, 10/28/2005 - 3:53pm

Hahaha, point taken.

by Pat (not verified) :: Fri, 10/28/2005 - 3:55pm


Unfortunately, you can't do that - you're completely and utterly neglecting an entire portion of the game - special teams.

That's the first point. The second point is that it's not quite as easy as "offensive DVOA+defensive DVOA convert to points predicts score" because the defensive DVOA above is the average of all defensive plays. Teams usually face 50/50 run/pass over the year, but individual teams run and pass more, so it's not as easy as matching up "offensive+defensive".

DVOA wasn't meant to do what you're trying to do.

by Fiver (not verified) :: Fri, 10/28/2005 - 3:56pm

I dunno about any of the others, but CHI 1 DET 0 looks about right to me.

by Brian (not verified) :: Fri, 10/28/2005 - 4:04pm


Now that's funny. Most of the scores all seem reasonable. I doubt JAC and TB will win quite that big though.

by Mshray (not verified) :: Fri, 10/28/2005 - 4:54pm

Aaron, Ray is absolutely right, leave the repulsion of trolls like #111 to your regulars. I was hoping someone else might have said this by now but...

To whit: Hey, AitD, just for the sake of historical accuracy, the term 'jump the shark' comes from a late-run ep of Happy Days in which Wisconsinite Fonzi goes surfing & masters the sport to the extent that he can defy basic properties of gravity & fluid dynamics, by elevating himself & his surfboard out of the water & over a shark, like he was Evel Knievel in a stunt car. This term was then widely used in TV criticism to describe any show on its last legs that turned to unbelievable & unsupportable hijinx in an attempt to retain viewer interest (c.f. the Dynasty season-ender when it was conceivable that they had gunned down every single cast member).

So not only has this term became one of the biggest cliches of the nascent 21st century, but accusing the Outsiders of becoming all "media insidery on us" would seem to indicate your use of this term for it's exact opposite intent.

And yes, in my opinion, we should all be reading a lot of Salon.com. Why not?

by B (not verified) :: Fri, 10/28/2005 - 5:27pm

Did you see the Arrested Development episode where Herny Winkler's character jumped the shark?

by Andrew (not verified) :: Fri, 10/28/2005 - 5:37pm

james #152:

How about Phl/NwE, StL/Ten, and Pgh/Dal (versions 1.0 and 2.0 in 76 and 79)?

by R.J. (not verified) :: Fri, 10/28/2005 - 5:44pm

Yes I did. A little too aware of itself as a precious comedy to really be funny, don't you think?

by james (not verified) :: Fri, 10/28/2005 - 5:53pm

Can't comment on Pburg/Dallas because I wasn't alive yet.

I thought the other two were good games, but nothing compelling came out of either.

I guess you could sub any of those for Carolina/New England.

The other 4 SB's I talked about were games where I felt like I was watching history being made. You had Elway claiming greatness. You had Norwood with the biggest miss in SB history. You had the Montana drive. You had Brady going from backup to MVP. I chose Carolina/New England because Carolina is the worst really good team of all time. They didn't do anything well yet they were almost impossible to beat.

by B (not verified) :: Fri, 10/28/2005 - 6:12pm

I'd take NE/Phi or StL/Tenn over NE/StL, and I'm a Patriots fan. The NE/StL only had about 2 good quarters of football in it.

by James Gibson (not verified) :: Fri, 10/28/2005 - 6:25pm

I guess the number of good Super Bowls after Den/GB has in part relegated Den/GB downward. St. Louis - Tennessee might have had the most dramatic ending. I was alive for PIT/DAL, but not watching football yet. For my enjoyment:

1. Den/GB (yes, I'm biased, and yes, all those years of frustration as a Broncos fan came out after that game. Plus it was the end of the streak and the blowouts had gotten so bad that people were calling DAL/PIT two years before and NE/GB the year before good Super Bowls.)
2. NE/Car
3. NE/StL
4. SF/Cin (boring first half)
5. Ten/StL (also boring first half)
6. NYG/BUF - Good Super Bowl in general, but I hate it to see it end on a missed kick. Plus, by this time I was sure the streak would end and it didn't.

by pcs (not verified) :: Fri, 10/28/2005 - 9:19pm

Mshray #160:

Fine explanation, but one correction: Fonz was on water skis, not a surfboard. and OH MY GOD i am a nerd.

by james (not verified) :: Fri, 10/28/2005 - 10:24pm

james gibson,

We've definitely been spoiled with good superbowls as of late. Every game except falcons/denver and baltimore/giants was pretty good.

It's been so competitive that I'd actually forgotten about how many straight games the afc had lost and how many of the games were over before halftime.

by Andrew (not verified) :: Sat, 10/29/2005 - 12:17am

James Gibson #166:

Some recent Super Bowl memories of mine:

1) 1993 - not again! Well, at least the commercials are interesting.

2) 1994 - thank God for once since 1985 we weren't going to have to watch either the Broncos or Bills get blown out by the NFC champion again! We could watch it happen to the Chargers instead. At least its something different.

3) 1995 Pgh/Dal Super Bowl Version 3.0 was the first one I had an inkling the AFC might win since the Norwood Bowl. Very disappointing to watch in Pittsburgh and see the Steelers lose fairly closely to the disgusting Cowgirls gang of thugs.

4) Hey! Its someone's chance besides the Skins, Cowgirls, Giants, and 49ers to destroy the AFC!

5) OMG! The AFC won! 1997 and 1998 Denver repeats - warm and fuzzy feelings to see Elway go out on top as the Champ and MVP. Marino, you suck! Screw your records! Young, its tiring to watch you choke against Farve. Confirmed my feeling that Elway was the real deal like Montana, and Marino and Young were not.

6) 1999 - Good for you Coach Vermiel! Wish you could have done it in 1980! But did anyone even notice that McNair almost overturned Simms' "untouchable" completion percentage record from 1986?

7) 2000, 2002 - BORING! These Quarterbacks are both horrible! And didn't we get rid of these blowouts with free agency and the cap?

8) 2003 NwE/Car - how did they possibly pack all that action into just 15 minutes? Yes, it really was all just 15 minutes of action! How is Carolina even in this game? Why won't they just lose? Was New England really that vulnerable?

9) 2004 - Damn you McNabb!

by james (not verified) :: Sun, 10/30/2005 - 11:57pm

re: no 153

a couple of conclusions that could be useful that could be useful

against the total the projections were 2-10 so far

It could be useful to know that the DVOA projection are always wrong.

Secondly, so far out of 11 games DVOA predicted 11 teams within 10 points of their projection.

Its would be useful to know how that these projections will usually get one team pretty close and one team very wrong.

So if we know that the total is probably wrong that gives us a clue as to how to adjust one team or another.

Also seems like most teams with better estimated wins at home cover (5-1 so far today with pitt and buff left)

So using all that info I'll guess

Buf 13
N.E 28

Bal 10
Pitt 28

by Tony D (not verified) :: Mon, 10/31/2005 - 4:13pm

I've revised the estimates to include special teams, and it changed the MNF game to:

BAL 13
PIT 27

by DavidH (not verified) :: Mon, 10/31/2005 - 5:27pm

Ok, using 1% DVOA offense = 0.215 points scored, 1% DVOA defense = .717 points prevented

Why would the conversions for offense and defense be different?

by james (not verified) :: Tue, 11/01/2005 - 1:55pm

david H,

I think it's right to make offense and defense have different adjustments. After all doesnt a really good defense like Ravens 00' shut down just about every offense ever.

i like the way he did it. 2 and 12 on totals is amazing. Could be valuable for us gamblers.

Anyone out there have any clue on where I can find red zone stats on college teams?