Writers of Pro Football Prospectus 2008

Most Recent FO Features


» The Deep Ball Project

Guest columnist John Kinsley breaks down the tape of every deep pass in the NFL in 2017 and comes away with a shocking conclusion: even without Andrew Luck, the Colts had the best long-ball quarterback in the league.

20 Dec 2005

Week 16 DVOA Ratings

by Aaron Schatz

Lots of good stuff in this week's commentary for the math-o-phobic, which you will find here on FOXSports.com:

  • I once again try to explain to people the concept of "considering the entire season," which is probably a fool's errand
  • That home-field advantage stuff I've been promising for a couple weeks
  • The history of great defenses, one year later
  • Seattle drops two spots because personally I can't stand anything having to do with the Pacific Time Zone, it all smells like hippies
  • Minnesota's Hanukkah miracle
  • The Texans are ahead of not one, not two, but six other teams
  • The most hated team in the NFL ranks way too high, thus guaranteeing an all-time hate mail record

Hey, wait a minute -- I thought I had "east coast media bias?" What the hell is Oakland doing at #16?

I have no idea, really. I went back looking at each game to figure out how DVOA could keep kicking out two-touchdown losses as "close" and I can't quite put my finger on the reason. They've beaten the Cowboys and Redskins, but that's just two wins. Oakland doesn't seem to be trending upwards -- for crying out loud, they lost to the Jets last week. I know to most people who read the Internet, this means that the DVOA system is messed up and every single result is stupid, but FO readers know differently. Hard cases make bad law.

Once again, in an effort to reduce trollery, the first of the two tables is actually in order of WEIGHTED DVOA, i.e. the FOXSports.com Power Rankings. The offense, defense, and special teams ratings on that table, however, are total-season numbers, not weighted numbers. This week we also say adios to FUTURE SCHEDULE from the second table. If you can't look up the ratings for the last two opponents of any team, well, therapy may be required.

To save people some time, there is no need to post comments of this nature:

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

Although I considered just running the entire power rankings commentary as 32 iterations of this complaint.

* * * * *

These are the Football Outsiders team efficiency ratings through 15 weeks of 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. SPECIAL TEAMS DVOA is adjusted for type of stadium (warm, cold, dome, Denver/Mexico City) and week of season.

WEIGHTED DVOA represents an attempt to figure out how a team is playing right now, as opposed to over the season as a whole, by making recent games more important than earlier games. This is the statistic used for the FOXSports.com Power Rankings. WEIGHTED DVOA for offense, defense, and special teams is available on those separate pages.

IMPORTANT: Beginning with Week 12 of 2005, DVOA is based on second-order opponent adjustments. Until other years are updated with this new system, only first-order DVOA (second table) can be compared to previous seasons.

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

1 IND 44.3% 1 46.6% 1 13-1 33.9% 1 -15.8% 4 -5.4% 31
3 DEN 30.4% 3 38.0% 2 11-3 26.0% 5 -5.1% 14 -0.7% 22
2 CIN 39.2% 2 30.6% 3 11-3 28.2% 3 -8.2% 11 2.8% 9
5 NYG 27.3% 5 28.8% 4 10-4 7.6% 9 -12.8% 7 6.9% 3
4 SD 28.1% 6 26.8% 5 9-5 29.0% 2 2.5% 22 1.7% 12
6 SEA 26.2% 4 26.3% 6 12-2 26.8% 4 0.4% 18 -0.3% 18
9 KC 20.5% 8 22.7% 7 8-6 22.2% 6 0.1% 17 -1.7% 25
10 WAS 17.6% 11 18.7% 8 8-6 3.2% 12 -14.6% 5 -0.2% 17
12 CAR 11.0% 13 17.8% 9 10-4 -7.3% 18 -15.8% 3 2.4% 11
8 PIT 21.9% 9 15.2% 10 9-5 7.6% 8 -16.5% 2 -2.2% 27
7 JAC 22.0% 7 14.1% 11 10-4 4.8% 11 -14.1% 6 3.0% 8
11 CHI 15.0% 10 12.1% 12 10-4 -18.2% 28 -31.8% 1 1.4% 13
13 NE 6.0% 19 5.7% 13 9-5 17.9% 7 12.3% 27 0.4% 15
20 BAL -4.0% 24 4.1% 14 5-9 -19.2% 29 -10.9% 8 4.3% 7
21 MIN -6.9% 17 0.7% 15 8-6 -13.2% 25 -7.2% 12 -1.0% 23
19 OAK -3.6% 20 -2.3% 16 4-10 1.9% 13 1.9% 20 -3.6% 29
14 DAL 2.9% 14 -2.9% 17 8-6 -0.5% 15 -2.4% 15 0.9% 14
18 ATL -3.4% 18 -4.3% 18 8-6 6.3% 10 9.6% 26 -0.1% 16
15 TB 2.2% 12 -4.7% 19 9-5 -7.2% 17 -10.0% 9 -0.6% 20
16 PHI -0.8% 15 -6.8% 20 6-8 -9.0% 20 -8.8% 10 -0.6% 19
22 CLE -7.5% 21 -8.0% 21 5-9 -7.9% 19 2.1% 21 2.5% 10
17 MIA -0.9% 16 -9.5% 22 7-7 -13.0% 24 -5.9% 13 6.2% 4
23 TEN -13.7% 25 -11.4% 23 4-10 0.2% 14 19.1% 30 5.2% 5
24 DET -19.4% 23 -17.0% 24 4-10 -15.6% 26 1.7% 19 -2.1% 26
25 ARI -20.3% 26 -17.3% 25 4-10 -12.6% 22 6.5% 24 -1.2% 24
31 HOU -33.2% 31 -21.3% 26 2-12 -16.2% 27 25.7% 32 8.7% 1
28 STL -25.5% 30 -23.4% 27 5-9 -6.7% 16 18.1% 29 -0.7% 21
27 NYJ -24.3% 29 -23.4% 28 3-11 -23.0% 31 -1.2% 16 -2.5% 28
29 GB -25.6% 22 -24.3% 29 3-11 -12.7% 23 5.1% 23 -7.7% 32
26 BUF -22.6% 28 -27.6% 30 4-10 -20.8% 30 9.5% 25 7.7% 2
30 NO -28.3% 27 -30.5% 31 3-11 -11.3% 21 12.5% 28 -4.4% 30
32 SF -72.0% 32 -71.5% 32 2-12 -52.8% 32 23.7% 31 4.5% 6

  • 1st ORDER DVOA is the "older style" DVOA from before Week 12 2005 which only runs opponent adjustments once. This stat should be used to compare with previous seasons.
  • NON-ADJ VOA is total VOA without any adjustment for opponent strength, luck in recovering fumbles, or the effects of weather and altitude on special teams.
  • 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.
  • 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.
  • VARIANCE (VAR.) 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).

W-L 1st ORD
1 IND 44.3% 13-1 40.5% 1 48.8% 13.0 1 -4.5% 26 6.5% 32
2 CIN 39.2% 11-3 37.5% 2 40.7% 11.3 2 0.2% 17 18.2% 17
3 DEN 30.4% 11-3 28.3% 3 27.9% 10.4 3 6.3% 6 18.3% 20
4 SD 28.1% 9-5 27.6% 4 20.5% 10.4 4 8.2% 1 11.5% 29
5 NYG 27.3% 10-4 27.2% 5 32.2% 9.8 6 -1.9% 21 20.8% 13
6 SEA 26.2% 12-2 24.8% 6 40.0% 10.3 5 -15.5% 32 14.5% 26
7 JAC 22.0% 10-4 23.0% 7 20.6% 9.6 7 0.4% 16 21.6% 11
8 PIT 21.9% 9-5 20.2% 8 22.7% 9.5 8 7.1% 5 18.4% 19
9 KC 20.5% 8-6 17.5% 10 13.5% 9.4 9 3.8% 11 14.1% 27
10 WAS 17.6% 8-6 18.8% 9 7.7% 8.6 11 2.4% 14 25.8% 12
11 CHI 15.0% 10-4 13.2% 11 11.2% 8.7 10 -6.8% 27 26.6% 8
12 CAR 11.0% 10-4 13.0% 12 27.1% 8.5 12 -11.0% 31 19.9% 24
13 NE 6.0% 9-5 7.0% 13 3.1% 8.4 13 3.8% 13 18.4% 18
14 DAL 2.9% 8-6 2.0% 14 0.6% 7.3 16 5.6% 8 22.9% 9
15 TB 2.2% 9-5 1.6% 15 9.6% 7.0 18 -10.2% 30 26.3% 5
16 PHI -0.8% 6-8 -2.1% 18 -6.0% 7.1 17 3.8% 12 26.8% 10
W-L 1st ORD
17 MIA -0.9% 7-7 -0.9% 16 -1.6% 7.4 15 -2.7% 24 17.7% 21
18 ATL -3.4% 8-6 -1.4% 17 8.1% 6.8 19 -6.9% 28 13.5% 30
19 OAK -3.6% 4-10 -6.3% 21 0.0% 6.6 21 6.0% 7 10.1% 31
20 BAL -4.0% 5-9 -4.3% 19 -9.8% 6.0 22 7.9% 3 31.4% 4
21 MIN -6.9% 8-6 -5.3% 20 -9.2% 7.6 14 -2.7% 23 24.2% 14
22 CLE -7.5% 5-9 -8.1% 22 -17.6% 6.6 20 5.3% 9 15.5% 25
23 TEN -13.7% 4-10 -14.1% 23 -7.6% 5.4 23 -0.1% 18 17.2% 23
24 DET -19.4% 4-10 -17.5% 24 -22.6% 5.3 24 -1.1% 20 28.5% 6
25 ARI -20.3% 4-10 -20.3% 25 -14.0% 4.8 25 -9.2% 29 13.7% 28
26 BUF -22.6% 4-10 -22.9% 27 -19.9% 4.3 27 0.7% 15 33.0% 3
27 NYJ -24.3% 3-11 -23.8% 28 -31.4% 4.1 29 4.0% 10 17.3% 22
28 STL -25.5% 5-9 -22.4% 26 -21.3% 4.2 28 -2.3% 22 16.9% 16
29 GB -25.6% 3-11 -24.5% 29 -22.8% 3.4 31 -0.5% 19 28.8% 2
30 NO -28.3% 3-11 -26.2% 30 -28.0% 4.7 26 -2.9% 25 26.8% 7
31 HOU -33.2% 2-12 -30.5% 31 -39.1% 3.8 30 7.9% 2 21.6% 15
32 SF -72.0% 2-12 -67.0% 32 -72.4% 0.9 32 7.4% 4 41.4% 1

Posted by: Aaron Schatz on 20 Dec 2005

313 comments, Last at 25 Dec 2005, 1:51pm by admin


by Ryan H (not verified) :: Tue, 12/20/2005 - 7:50pm


by bill tetrault (not verified) :: Tue, 12/20/2005 - 7:54pm

You obviously have a screw loose keep your eyes on the New England Patriots

by DGL (not verified) :: Tue, 12/20/2005 - 7:59pm

Personally, I love the idea of doing the entire commentary using the Canonical Football Outsider Hater Comment Template.

We could try to preempt the trolls by each picking a team and posting a comment of this nature (said in knowing disbelief, of course), so that it all gets said before they get here...

by DGL (not verified) :: Tue, 12/20/2005 - 8:00pm

Damn, missed it by five minutes.

by admin :: Tue, 12/20/2005 - 8:04pm

New England is clearly ranked too low because they know how to beat the Colts so the first 11 games of the season don't mean anything. ESPN, which has the Patriots fifth while beating Kansas City doesn't help the Giants at all is way better than this. You have a screw loose and should keep your eye on the Patriots, possibly by turning on the television or the radio since you live in Boston and have been rooting for the team since the late 80's and wear a Richard Seymour jersey and own all the Three Games to Glory DVDs and therefore must really hate this team. U suk. Brady iz god.

by Potomac Drainage Basin Indigenous Person (not verified) :: Tue, 12/20/2005 - 8:09pm

This week while watching the Cowboys embarass themselves, I was thinking "I wonder how much this will raise Washington's DVOA." One spot was the answer!

by Ferg (not verified) :: Tue, 12/20/2005 - 8:12pm

Re 5: Bravo! I'm sure you've saved many trolls a lot of time and effort.

OK, here is my irrational conspiracy theory about the Texans. They are trying as hard as they can to lose the Bush Bowl-- i.e., to beat SF and get out of the top draft pick-- because the man they really want is cousin D'Brickashaw, or whoever the top OT prospect is, but they don't want to have a fan revolt for passing up Reggie Bush. This has resulted in their skyrocketing WDVOA.

(Hey, it makes at least as much sense as Kris Brown intentionally shanking a FG to keep the top pick.)

by thad (not verified) :: Tue, 12/20/2005 - 8:18pm

the Cowboys have a worse rushing offense than the Lions by dvoa
god they are pathetic.

by Bencoder (not verified) :: Tue, 12/20/2005 - 8:21pm

#6 - Most likely as Washington moves into the top ten of teams it will be harder for them to crack additional spots unless most of the teams above them play significantly poorly. You might want to look at the change in their weighted DVOA and compare it to other teams near their ranking. This will give you more of an idea of the effect of the Cowboys game.

by Catfish (not verified) :: Tue, 12/20/2005 - 8:22pm

Uh oh, I see trouble with 5-9 Baltimore ranked 14th, ahead of Dallas and Tampa. With that and the Raiders at 16th, will Aaron be accused of a pro-criminal bias?

by ji\'s apple pie (not verified) :: Tue, 12/20/2005 - 8:24pm

How hard is it to move DVOA this late in the season? I was interested to see how the Charger's defensive rating would be improved by the Colts game, and it was 2.7%. Is that a lot to move after 15 games? How well does DVOA say they actually played against the Colts? It seemed like a monster game for that unit, especially since everybody said the Colts would put up 30+ points on that secondary.

by Staubach12 (not verified) :: Tue, 12/20/2005 - 8:29pm

Wasn't the troll template first suggested by a reader (not one of the outsiders)? If so, I think FO should at least give that person credit when using his template every week. It's become such a good way of discouraging stupid posts that the guy should at least get recognition for it.

by Gatts (not verified) :: Tue, 12/20/2005 - 8:30pm

the Patriots move up six spots, which won’t do anything to appease the Brady-lovin’ masses, and the Raiders are inexplicably high, guaranteeing that my e-mail box is totally screwed.

YOUR e-mail box? You're lucky I nuke the stuff before it gets to you.

I estimate 2/3 of the mail the last 3 weeks has been one-liner Pats hate mail.

by ElAngelo (not verified) :: Tue, 12/20/2005 - 8:41pm

Isn't constantly calling the Fox Sports DVOA chart for the math-o-phobic a bit wrong? I love reading the numbers chart, but also enjoy the Fox page because of the notes and commentary, which are much more interesting than just reading a set of numbers.

by Richard (not verified) :: Tue, 12/20/2005 - 8:48pm

Cincinnati is clearly ranked too high because their quaterback hasn't won anything yet. We all know looking at a team's quarterbacks and subjectively evaluating their ability to lead teams to playoff wins is way better than this. You clearly spin your draydel counter-clockwise, and by that I mean you saw the furst showing of Blokebrack Mountin when it opened at you're local theatre.

by Rob (not verified) :: Tue, 12/20/2005 - 8:49pm

I agree with #14. The writing done by FO is what keeps me coming back to the site again and again. You guys are far and away the smartest, most amusing, and most insightful NFL writers on the net. Except for Simmons (exclusively in the humor category) it's not even remotely close.

by Bencoder (not verified) :: Tue, 12/20/2005 - 8:50pm

#15 - anti-semitic comment sighting!

by M (not verified) :: Tue, 12/20/2005 - 8:52pm

(This was also posted on the Boller outplays Favre thread - may be more appropriate here)

This might be more of a question for the DVOA rankings, but has there ever been a team involved in two more lopsided games from the opposite side of the spectrum than this year’s GB packers? The swing between their NO victory & Baltimore loss is 94 points. I have never heard of anything like this before. I also am curious if this year has had the highest AMV of all games in recent memory (average margin of victory). This year seems to be close to 12 points. The last year I remember being that high was 1990. Any takers on my question(s)?

by admin :: Tue, 12/20/2005 - 8:53pm

An intern speaks! Wow, that much Pats hate mail I didn't even see, huh?

I've been calling it "commentary for the math-o-phobic" since 2003, when we had like five readers. It's sort of a tradition, and has nothing to do with FOX.

Washington doesn't go up in weighted DVOA because -- remember from last week's commentary -- the San Francisco win dropped in weight the same week that they blew out Dallas.

And I think #15 is joking, so it's all good.

I keep forgetting which reader suggested the template and I should know that. If it was you, speak up.

by Richard (not verified) :: Tue, 12/20/2005 - 8:55pm

re: #17

Don't know if you were kidding or not. Excuse me if you were.

Didn't mean it as anything but a parody of a troll, but I also tried not to say anything hateful. I don't know what's wrong with either spinning a dradle counter-clockwise or seeing Brokeback Mountain. Both of those things sound OK to me. I just wanted to use them in a silly context.

by Smeghead (not verified) :: Tue, 12/20/2005 - 8:59pm

17: I'd say 15 was heavier on the gay-bashing than the anti-semitism. But hey, it's a big tent. Extra points for heavy flurry of diverse misspellings, plus a UK affectation, "theatre".

Very complete effort. Too complete. I smell agent provocateur.

by Smeghead (not verified) :: Tue, 12/20/2005 - 9:00pm

20: Yes!

by Bencoder (not verified) :: Tue, 12/20/2005 - 9:02pm


Do you use a macro to generate your final tables or do you have to plug the numbers in by hand? I used to generate a daily library of run charts for a manufacturing process and was able to reduce the generation time from hours to a minute once we developed a good macro.

by Catfish (not verified) :: Tue, 12/20/2005 - 9:02pm

Any perceived and/or joking semitism/homophobia in #15 is nothing compared to the email Aaron posted in Audibles at the Line on Monday. That one even included a boatload of [redacted] words too.

re: 12
I believe Aaron has cited the creator when he's posted the template in the past (I believe it was zlionsfan). He probably just forgot this week.

by MJK (not verified) :: Tue, 12/20/2005 - 9:05pm

It was zlionsfan. I can't find the original comment (it was presumably prior to Week 13), but in Week 13, Comment 9, Aaron makes reference to it and to zlionsfan.

by DavidSEA (not verified) :: Tue, 12/20/2005 - 9:12pm

So it's going to be interesting how far the Colts fall, when they get their clocks cleaned against Seattle this week. Also them benching there starters for the next couple weeks as well.

Seattle has history to make still with SA breaking the touchdown record.

by fromanchu (not verified) :: Tue, 12/20/2005 - 9:16pm

last year is way more important than this season. there needs to be a column for previous superbowl wins. or is that already included in swagger?

by lummox (not verified) :: Tue, 12/20/2005 - 9:20pm

Just wanted to extend advanced condolences just in case this week's rankings trigger an avalanche of irate responses from angry Miami fans.Also, interesting to note how low the Patriots's DVOA is even after Saturday's win. Could it be the team is not as great as Peter King thinks? Le gasp.

by Dustin (not verified) :: Tue, 12/20/2005 - 9:20pm

so when the seahawks beat the colts next week do we get the number two spot with an asterisk? yeah i didn't think so. but i still think it is funny that the seahawks are going to finish with the best record in the nfl and three weeks ago i saw clowns on espn giving them the 4th seed in the playoffs. p.s. it only smell of hippie for about a 10 mile radius around the space needle so don't damn us all in the pacific nw to that lowest of low.

by jojo (not verified) :: Tue, 12/20/2005 - 9:25pm

Is that a Propagandhi reference in the Bills commentary?

by David Brude (not verified) :: Tue, 12/20/2005 - 9:30pm

I noticed that only about 10% points separates Manning from Palmer in DVOA yet the total team pass offenses are about 20% apart. What causes the difference in the team DVOA since those guys have basically played almost all of the snaps.

by DGL (not verified) :: Tue, 12/20/2005 - 9:31pm

Seattle: Seattle is clearly ranked too low because they're going to beat the Colts next week and finish with the best record in the NFL.. ESPN, which ranks the top teams like the NCAA coaches' poll, is way better than this. three weeks ago the clowns on espn were giveing them the fourth seed in the playoffs and now there going to clean the colts clocks.

Just doin' my part.

by Richard (not verified) :: Tue, 12/20/2005 - 9:33pm

Weighted DVOA has Cincinnati so much closer to the Giants than the Broncos. No doubts about DVOA, just that all season I haven't considered the Giants that strong while I've been having this debate with myself as to who I would take in a winner-take-all situation between Denver and Cincinnati.

Does anybody else think that the Giants might be overvalued, and if so, why do you think it's happening?

by Richard (not verified) :: Tue, 12/20/2005 - 9:34pm

re: #32


by zip (not verified) :: Tue, 12/20/2005 - 9:39pm

Is that a Propagandhi reference in the Bills commentary?

The commentary is up? Where? (omg!)

by NF (not verified) :: Tue, 12/20/2005 - 9:43pm


Zlionsfan is currently in a cave in the wastes of northern Michigan, and won't come out until Matt Millen is fired.

by Maltodextrin (not verified) :: Tue, 12/20/2005 - 9:48pm

I'm sure the Seattle hippie comment was not meant to be taken seriously, just an exaggeration for comedic effect.

It is really funny how often sarcasm and irony goes over people's heads on this site.

by bobman (not verified) :: Tue, 12/20/2005 - 9:51pm

RE: #15... how exactly do you spell dreydel/dreidl...? just wondering. This question is officially semantic, rather than anti-semitic. Of course if I did not care, would it be anti-semantic...? Is that so wrong?

RE: #26. Ahem... Ahem. Where to start? I'll adopt the tone I use with my 5 year-old (who, just two months ago told me "The Patriots always beat the Colts."): Well, it's reasonable to expect that with half the 1st teamers starting, the Colts will struggle against a team that has a fairly poor record vs winning teams but is nonetheless 12-2 and has HFA to play for and is at home.

Depth is an issue in Indy, and you saw what happened with 3-4 starters missing last week; they clawed back from a 16 point deficit to lead in the 4th qtr, but could not hold serve. They'll do the smart thing and rest the walking wounded and a few more this week. So yes, I expect you to get the W. Who wouldn't?

Good luck Shaun--when you get to 49 TDs in one season give me a call--I'll buy you a beer at the Roanoke on Mercer Island (now that it's smoke free). Maybe Manning, who sat on the bench for 8 quarters last season, will buy you a steak at St Elmo's when you pass 49 and sit for two whole games.

I prefer to think of history being made in week 20 (21?). After all, how many Lombardis did Manning's 49 TDs last year get him? When the well-rested Colts face whoever they face in the playoffs, things will look a little different than they will on 12/24 at the Q.

by NF (not verified) :: Tue, 12/20/2005 - 9:54pm


How many times have Cincinnati receivers fumbled the ball this season?

by bobman (not verified) :: Tue, 12/20/2005 - 9:55pm

Hey, there are PLENTY of hippies out here. Punks, too, and... probably a few unreconstructed beatniks. I am sure I am missing some major social movements... Yuppies? yup. Preppies? Not so much. One of the great things about Seattle is, we all mix in pretty well.

by Staubach12 (not verified) :: Tue, 12/20/2005 - 9:55pm

The Redskins comment needs to be updated.

by David Brude (not verified) :: Tue, 12/20/2005 - 10:03pm


4 fumbles for their top 3 WR's. That can't cause a 10% difference can it.

by Dustin (not verified) :: Tue, 12/20/2005 - 10:05pm


32: the meat of my comment was the bit about the asterisk... like the colts had a few weeks ago. and i don't know how you mistook me as thinking espn ranks are better. i was just pointing out how seattle will finish with the best record in the nfl and for so many seem to be only an after thought. and this doesn't even make me bitter... it just makes me laugh. now just because you suck at reading comprehension don't be "giveing" me a bad name.

37: It is really funny how often *irony* and sarcasm go over people's heads. that hippie comment by adam was ironic as rain on your wedding day.

by VarlosZ (not verified) :: Tue, 12/20/2005 - 10:07pm

What the hell is Oakland doing at #16?

The point spread on their game with Denver this week is 13 points, while wDVOA thinks it should be 12.3 points. In other words, I don't think there's anything wrong with Oakland's DVOA rating, though it's possible that there's something wrong with the ratings for the three teams below them (Dallas, Atlanta, and Tampa Bay). I don't think there's a problem with any of those teams' rankings.

by BillWallace (not verified) :: Tue, 12/20/2005 - 10:10pm

I'll do it.

The Redskins are clearly ranked too low because Joe Gibbs is a proven winner with 3 Super Bowl rings, and 4 of the coaches ranked higher have NONE! Michael Wilbon's power rankings based on how many black players each team has is way better than this. Chris Cooley is gong to rape your wife and eat yuor kids.

by DavidSEA (not verified) :: Tue, 12/20/2005 - 10:12pm

RE #38: Say what you want throwing 49 touchdowns is far easier then running 27 or they wouldn't even have two records. Not saying 49 TD's is not alot as it is and I was routing for him to get 50. However as it has been said the Modern era Colts have proved nothing until they win it all. As it was shown last week they are vulnerable and frankly it will be interesting if they can beat NE this year when it counts. I would love to see a Seahawks vs. Colts SB but the AFC looks mighty tough to get through in the playoffs.

by BillWallace (not verified) :: Tue, 12/20/2005 - 10:20pm

re: 46 I would say the odds of the Colts making the Super Bowl are the same or slightly higher than Seattle. Both are the best team in the conference, but not a lock.

by dfarrar777 (not verified) :: Tue, 12/20/2005 - 10:24pm

(Bangs head agaist wall and weeps for those in Seattle who were using the template and didn't even realize it).

If Seattle beats the Colts, it's looking more and more like they'll do it against a very large number of scrubeenies. That will do a lot for their record, and a great deal for their post-season chances, but I doubt it will lift their DVOA anymore than Martin Grammatica could lift Ted Washington. Tiki Barber ran for 220 yards against a defense that used to be good against the run - in fact, they were good until Tiki got a hold of them. That's part of how you rise in these particular rankings, no? Bring statistical excellence against teams who tend not to allow that excellence?

The Seahawks almost lost to the Titans. After THAT performance, I'm happy with 12-2. Rankings be darned, darn it!

by Maltodextrin (not verified) :: Tue, 12/20/2005 - 10:26pm

#43: Aaron's comment was in fact "ironic" because he intended it to mean the opposite of what he was saying. He didn't really downgrade the Seahawks because of anti-hippie bias against the West Coast.

Believe it or not people actually use the word ironic correctly once in a while.

by Sean D. (not verified) :: Tue, 12/20/2005 - 10:28pm

This is the second time I've seen comments about Brees' short passes to Gates. I know at least 1 (if not 2) were 3rd down conversions. Also, I remember earlier in the year Gates would break some of those because the defender couldn't tackle him. I don't see anything wrong with making a 200 lb safety try to tackle Gates in the open field. Gates will typically try a little spin move (like a post up move in basketball) and can get extra yards because of his size. Also, some of those passes are checkdowns when Brees sees pressure coming and knows he can get a 3-4 yard pass instead a run stuffed in the backfield, an incomplete pass, or worse a sack/fumble. Lastly I don't see a problem with making the D watch gates over the middle because it sets up the deep ball. I'd love to look at the McCardell catch and see if the safety stepped up and let him go by because he was watching Gates over the middle.

by Richard (not verified) :: Tue, 12/20/2005 - 10:33pm

re: #38

It's spelled "dradle". I had to look it up to make sure I spelled it incorrectly for the post :).

by Richard (not verified) :: Tue, 12/20/2005 - 10:37pm

re: #45

Made me laugh.

At first I was like "I don't think Wilbon's that bad" and "Uh, oh - rape", but then I remembered what the point was :)

by SM (not verified) :: Tue, 12/20/2005 - 10:46pm

I found the commentary on FOX - link in the name.

by B (not verified) :: Tue, 12/20/2005 - 10:48pm

Does Seattle still have something to play for next week? I hope they wouldn't risk thier starters for something as meaningless as the rouching TD record, although considering how upset Shaun was last year when he missed the rushing title, maybe they have to.

by Devin McCullen (not verified) :: Tue, 12/20/2005 - 10:58pm

Aaron, if you wanted to be hip and youthful (and reference a Fox TV Show), you should have called the Baltimore forecast a "Chrismukkah dome team miracle".

BTW, can I get a confirmation that the Green Bay defense posted the worst DVOA of the year? I know they had to, I'd just like to hear it.

by Matthew Furtek (not verified) :: Tue, 12/20/2005 - 10:58pm

Or conversely for the Redskins...

Washington is clearly ranked too high because I never see their highlights on NFL Primetime or hear any ESPN announcers talk about them in a positive light, so they must suck. Dr. Z's Power Rankings, which he comes up with after sleeping through most of the games on Sunday afternon is way better than this. You should open up your eyes and see how evil Dan Snyder is. Just the fact that he is the Redskins owner should be a 16 loss penalty against them.

by spenceKarl (not verified) :: Tue, 12/20/2005 - 10:58pm

Aaron, your opening paragraphs of commentary on the Patriots makes perfect sense, to explain - in simple terms - their current ranking. And yet, I'm sure many will decide not to read, and instead, scroll down to find the "horror" and "disrespect" of a 13th ranked team.

by Paul (not verified) :: Tue, 12/20/2005 - 11:00pm

On the little charts to the left you have BAL ranked 3rd and 4th for special teams (Neat trck) it should be Giants 3rd...and I'm pretty sure that this post is very anti-syntaxic

by Mr.X (not verified) :: Tue, 12/20/2005 - 11:01pm

Denver is clearly ranked too high because thin air makes me brethe farts. Your momma is way better than this. All base's are belong to us.

by WAHawksfan (not verified) :: Tue, 12/20/2005 - 11:05pm

Seattle still needs one more W or a Bears loss to clinch HFA. So they do have something to play for still

by spenceKarl (not verified) :: Tue, 12/20/2005 - 11:08pm

RE: 58 (Paul)

Actually, it should be Buffalo 4th in Weighted DVOA with 6.4%. The Giants have actually fallen to 10th in Weighted DVOA with 2.3%.

by admin :: Tue, 12/20/2005 - 11:09pm

Yep, yep, commentary went up:


Unit pages are now up also for offense, defense, and special teams. I'm running all the individual stat macros and table builders while watching House. Oh, House, you devil you.

by Staubach12 (not verified) :: Tue, 12/20/2005 - 11:11pm

And the Redskin's comment on the fox commentary still needs to be updated from last week. Unless you intended to be ironic (insulting Cooley?).

by Purds (not verified) :: Tue, 12/20/2005 - 11:41pm

I've always believed in the DVOA rankings, but as a Colts fan I can't believe a couple of them this week:

1) The Colts defense jumped up to a #4 DVOA (or #5 weighted DVOA) ranking from #8? I'd like to believe they are that good, but...

2) The Colts have all of a sudden become respectable (#14) against TE's? Again, I hope you're right, but ...

by Catholic Samurai (not verified) :: Tue, 12/20/2005 - 11:44pm

I thought all the godless commie heathens celebrated Festivus? :D

by thad (not verified) :: Tue, 12/20/2005 - 11:49pm

re 39
The Bengals wide recievers and tight ends have fumbled 7 times and lost three.

by B (not verified) :: Tue, 12/20/2005 - 11:55pm

65: Us godless heathen commies celebrate "seasonal holiday"

by Jay B. (not verified) :: Tue, 12/20/2005 - 11:55pm

zlionsfan's DVOA commentary template appeared during the Attack of the Atlanta Fan Trolls after week 8 (right before the Falcons lost to the mighty Packers).

by Harry (not verified) :: Tue, 12/20/2005 - 11:55pm

C'mon, the Pats are clearly now ranked TOO HIGH because all Pats fans know that the Patriots thrive on disrespect. If Aaron were really a Pats fan he would rank them 28th and mock them mercilessly in order to urge them on to victory.

by mshray63 (not verified) :: Tue, 12/20/2005 - 11:56pm

Since the Colts already lost I expect that the hype for Saturday's Seahawks game to be a little less than it might have been. And in an effort to bring that hype back to where it should be, can anyone tell me when, if ever, there's been a regular season game between teams that combined for 25 wins going in? I haven't yet found a site with every team's complete schedule for years past, but based on just the final records it was impossible in about half the years since the 16 game schedule.

by Catholic Samurai (not verified) :: Wed, 12/21/2005 - 12:00am

RE #67:

Thanks, but I was asking the "godless commie heathens", not the "godless heathen commies". Can't you read?!? :D

by Oswlek (not verified) :: Wed, 12/21/2005 - 12:03am

It's one thing to lose to Kansas City on the road — that's going to happen most of the time. But they let Buffalo stay in that game too long, barely beat Baltimore on the road, and were a blown field goal away from overtime on Thanksgiving.

Sorry to be the anal, corrective guy, but the Denver was a blown FG away from losing that game. It did go to overtime.

by spenceKarl (not verified) :: Wed, 12/21/2005 - 12:05am

RE: 64 (Purds)

Well ignore the jump from 8th to 4th and look at take a look at the change in actual DVOAs. They went from -11.6% last week to -15.8% this week. Not as dramatic a jump as the actual ranking might suggest. It's just the other teams around them that make the change seem so dramatic.

by admin :: Wed, 12/21/2005 - 12:08am

Individual pages now updated: QB, RB, WR, TE

by MRH (not verified) :: Wed, 12/21/2005 - 12:12am

Home-field advantage absolutely exists. I've estimated in the past that it is worth roughly 17 percent of DVOA — in other words, a team at home is even with a team that is 17 percent higher in DVOA. (Chicago at the New York Giants would be one example, Minnesota at Carolina another.)

Shouldn't that be NYG at Chicago and Carolina at Minnesota? The better team by DVOA on the road is equal to the weaker team at home? Or am I confused?

Otherwise, great article as always.

by admin :: Wed, 12/21/2005 - 12:22am

Erg. Apparently today I was sampling the crackpipe. Washington comment is fixed and I just sent in fixes on 72 and 75. I fixed the special teams box too.

Believe it or not, GB did NOT have the worst defensive game of the year. SF scores as ever so slightly worse against PHI in Week 2 and WAS in Week 7. Favre had -7.0 DPAR. Rodgers had -8.8 DPAR. Boller had 17.1 DPAR which is of course completely ridiculous because the guy was at -12.9 for the season prior to that game. What a colossal fluke.

Just to clarify the comment by Doug Farrar above, DVOA doesn't consider when teams pack it up for the season and send in the backups. So if Seattle plays starters and Indy plays backups, and Seattle slaughters Indy, Seattle will in fact zoom up the charts, and I will have to pull my hair out trying to explain it to the masses.

Speaking of the masses, we keep checking the Atlanta message board to see if somebody has posted anything along the lines of "Gee, maybe those FOX guys were right." So far, no such post.

by JonL (not verified) :: Wed, 12/21/2005 - 12:25am

He caught two passes for 48 yards and people are already saying he's going to be one of Carolina's main weapons for the rest of the year. Two catches, folks. This is how desperate the Panthers are to find a receiver who can play opposite Steve Smith.

But...what about the inanimate carbon rod? In Rod we trust!


by JonL (not verified) :: Wed, 12/21/2005 - 12:30am

Oh, and that home-field advantage note was great.

by CaffeineMan (not verified) :: Wed, 12/21/2005 - 12:32am

The Specials roool. 8)

And 13th, are you kidding? What's to complain about? The Pats are moving up in the world. But look, their defense is only up to 27th overall and their pass D is still 29th, so there's still lots of fuel for the disrespect fire. Hot diggity d*mn. :D

by NF (not verified) :: Wed, 12/21/2005 - 12:38am

Great comment on the Jaguars. The Jaghs are lucky their ending schedule is so easy and they were so good before the late-season collapse they were still a decent team afterward.

by Jay B. (not verified) :: Wed, 12/21/2005 - 12:44am

Re: 68 (myself)

zlionsfan's template actually appeared in the Week 10 DVOA thread, Shot to Hell version. See link in my name.

by Devin McCullen (not verified) :: Wed, 12/21/2005 - 12:49am

Thanks, Aaron. I guess the opponent adjustments aren't as big as I thought.

by David Brude (not verified) :: Wed, 12/21/2005 - 12:51am

re 66

So I ask again 7 fumbles causes an additional 18% difference in team passing DVOA vs the 4% difference between Palmer and Manning. I'm just curious how the two QB's can be so close yet the passing games be so different in team DVOA. I realize CIN primarily goes to their two WR's whereas INDY gets it to their TE's and RB's as well.

by MRH (not verified) :: Wed, 12/21/2005 - 12:55am

Re 76: anyone who watched the game could clearly see that Vick's special talents are hidden by the way the NFL keeps "Score" or counts passes completed only if they were caught. How many games has Grossman won anyhow? Vick is better because he's won more games and Orton is better than Grossman for the same reason. Any correct spelling is purely random.

by Becephalus (not verified) :: Wed, 12/21/2005 - 1:00am

Anyone see the most recent Sport Illustrated cover where it asks "Who is this?" or something like that about S.A.. I feel pretty safe in saying that any halfway serious football fan knows who S.A. is and anyone who doesn't know who he is wouldn't really care if they did know? So who is this article for an audience of 15 or 20 people or what?

With that headline I was expecting an O lineman or an assistant coach or something, not someone who is universally regarded as one of the 5 best at the second highest visibility position in the game.

I mean christ I could come up with 20 better cover articles than that in about 5 mins, just about the NFL.

by Adam (not verified) :: Wed, 12/21/2005 - 1:01am

Tom Brady has said and I believe FO did an article or dedicated part of an article about how much trouble he had against the Dolphins secondary, and now half of that secondary is in KC. I'd say they've obviously got his number and have gotten KC to go along with that they know. I believe that very basically it's just tight man to man on the WRs and then both safeties in deep zones.

by Adam (not verified) :: Wed, 12/21/2005 - 1:02am

How do you quote something in your post, I just tried to quote the part of the article on foxsports.com that said: "Something made Tom Brady throw those four interceptions against Kansas City, and you can bet every defensive coordinator facing New England in the AFC playoffs will be looking at that tape."

by Matthew Furtek (not verified) :: Wed, 12/21/2005 - 1:02am

Comments on the individual stats:
1) What is McCardell doing so well that he jumps up in DPAR?
2) How come no one is talking about Warrick Dunn for MVP? I think the Falcons would be nothing without him... actually, maybe they would be the Ravens?

by Matthew Furtek (not verified) :: Wed, 12/21/2005 - 1:05am

#85 (Becephalus)
I thought the same thing when I saw that cover... maybe they should have put up a picture of Walter Jones or Steve Hutchinson (sp).

by Dan Riley (not verified) :: Wed, 12/21/2005 - 1:07am

Did I miss something? Nearly 100 comments and I don't see even a whimper from a Pats fan. Is the Whining Pats Fan an FO urban legend? If most are like me, yes it is. We're happy to relinquish power rankings no matter how cleverly devised to the Colts and their ilk. We know--from lots of experience, I might add--that nothing matters until January...and if it doesn't work out for us, we have 3 Games to Glory, Vols 1-3. On a cold winter night, the only thing better than flannel is ooooooh smugness.

by dfarrar777 (not verified) :: Wed, 12/21/2005 - 1:08am

"Just to clarify the comment by Doug Farrar above, DVOA doesn’t consider when teams pack it up for the season and send in the backups. So if Seattle plays starters and Indy plays backups, and Seattle slaughters Indy, Seattle will in fact zoom up the charts, and I will have to pull my hair out trying to explain it to the masses."

Hmmm. That would be difficult to explain either way.

by Pat (not verified) :: Wed, 12/21/2005 - 1:09am

Believe it or not, GB did NOT have the worst defensive game of the year. SF scores as ever so slightly worse against PHI in Week 2 and WAS in Week 7.

Yah, but how much of that is due to the adjustment for Philly's offense? I mean, yah, sure, Baltimore is less beat up now than they were earlier, but c'mon. McNabb and Owens versus SF, or McMahon and Lewis? Yah. Thought so. So GB's awful, awful performance will forever rank as the "honest-to-god, we were really trying!" lowest defensive DVOA of the year.

And on that...

by Larry Richards (not verified) :: Wed, 12/21/2005 - 1:10am

Aaron,Re: Uh-oh, here come the angry e-mails. C'mon Seattle fans, you know you are supposed to beat Tennessee by more than four points, right? Wrong!!!!! You writers seem to forget this is Pro Football…………The Pro players are the best of the best! Seattle did what they had to do....Win! It does not make any differance if its 1 point or 43, it still is a win!

by Pat (not verified) :: Wed, 12/21/2005 - 1:12am

Philly is clearly ranked too low because McNabb will possess McMahon and make him not suck and Greg Lewis will kill Owens and usurp his power. Dr. Z's rankings are way better than - oh, crap, the Eagles suck there too. What about... suck there too? And... oh, crap. Yah. Uh. Oh, god, I can't even keep this up. Where's the nearest bar so I can drown my sorrows... is it 2006 yet...?

by JonL (not verified) :: Wed, 12/21/2005 - 1:15am

RE: #87

It depends on how you want to do it. You could just throw quotes around the quote, or if you want to get super-crazy you could use HTML tags and italicize it. That's (Shift+comma)i(Shift+period)quote(Shift+comma)/i(Shift+period).

You can also bold the text, substituting b for i.

Anyone know if the blockquote tags work? I'd try it, but I'm afraid to blow up the site.

by Jay B. (not verified) :: Wed, 12/21/2005 - 1:19am

Re: 93-

The Tennessee Titans are not the "best of the best." Or at least, one might say that they're among the worst of the best of the best.

Aaron did an article recently indicating a very strong correlation between teams' abilities to beat up on weak teams and their performance in the postseason. So, Seattle's inability to trounce them doesn't bode particularly well for them.

by spenceKarl (not verified) :: Wed, 12/21/2005 - 1:21am

RE: 93 (Larry Richards)

But wouldn't a dominant win suggest a possible better chance at future success over a close win? I think it does. And this is part of what DVOA helps us learn - How a team actually played within those Ws and Ls.

by Pat (not verified) :: Wed, 12/21/2005 - 1:23am

Anyone know if the blockquote tags work? I’d try it, but I’m afraid to blow up the site.

Yah, look just above the "name" field. Lists the HTML which works.

by JonL (not verified) :: Wed, 12/21/2005 - 1:24am

I know a lot of people have criticized the Washington defense for not forcing turnover's but in addition to currently being ranked 5th, they also rank as being pretty consistent (6.2 var, good for 25th). Although they do have a 7.8% EPR (experimental pass rush) rating, or 28th in the league.

Just to throw another (spring/summer) research idea out there, it might be interesting to look at teams that predominantly have their corners cover a side of the field vs. a particular receiver, and whether there's a correlation to a team's defensive ranking vs. #1, #2 and other receivers.

by Andrew (not verified) :: Wed, 12/21/2005 - 1:25am


Speaking of Atlanta, whatever happened to our friend Se7en_Dust and his compadres?

You guys apologized about pegging Atlanta as a not very good team too soon.

by JonL (not verified) :: Wed, 12/21/2005 - 1:26am

Yeah, look just above the “name� field. Lists the HTML which works.

I guess that would help, wouldn't it?

by Travis (not verified) :: Wed, 12/21/2005 - 1:30am

Re: 70

The Seahawks-Colts game will be the first NFL regular season game between teams with 25 or more combined wins. 5 others had 24 wins:

Week 15, 1983: Cowboys (12-2) at Redskins (12-2). Basically for HFA throughout playoffs (though WAS would also have to win the following week). Redskins won 31-10.

Week 16, 1999: Jaguars (13-1) at Titans (11-3). JAX would have clinched AFC Central with win, but lost 41-14.

Week 16, 1984: Rams (10-5) at 49ers (14-1). 49ers had clinched HFA, Rams (I think) had already clinched wild card berth. 49ers won 19-16.

Week 16, 1984: Broncos (12-3) at Seahawks (12-3). Winner would clinch AFC West and #2 seed, while the loser would get the #4 seed. Broncos won 31-14.

Week 17, 1991: Lions (11-4) at Bills (13-2). Bills had already clinched HFA, Lions needed win and Bears loss to win NFC Central. Lions won 17-14.

The best percentage matchups (halfway through season or later):

Week 12, 1969: Vikings (10-1) at L.A. Rams (11-0). Vikings won 20-13.

Week 10, 1969: Cowboys (8-1) at L.A. Rams (9-0). Rams won 24-23.

Week 10, 1991: Oilers (7-1) at Redskins (8-0). Redskins won 16-13.

Week 13, 1990: Giants (10-1) at 49ers (10-1). Had the potential to be the most hyped game ever, but oth teams lost the week before. 2nd highest MNF game of all time. 49ers won a snoozer, 7-3.

Week 11, 1962: Packers (10-0) at Lions (8-2). Lions won 26-14.

Week 11, 1977: Baltimore Colts (9-1) at Broncos (9-1). Broncos won 27-13.

by Andrew (not verified) :: Wed, 12/21/2005 - 1:41am

Dustin and Co. on the Left Coast:

Remember all those Lombardi trophies won by: The 2004 Steelers? The 2002 Eagles? The 2001 Rams? The 2000 Titans? The 1999 Jaguars? The 1998 Vikings? The 1997 Chiefs and 49ers? The 1995 Chiefs? The 1993 Oilers? The 1992 49ers? The 1990 49ers? The 1988 Bears? The 1987 49ers? The 1986 Bears? Oh yeah, these teams won diddly squat, despite having the honorific of "best record in football".

Better to be the strongest team than the team with the most wins. More often than not in the past 20 years, the team with the most wins has gone home empty handed at the end of the season, without even a Super Bowl loss to show for it.

by Mikey (not verified) :: Wed, 12/21/2005 - 2:01am

Really good comment on the TB-NE game being the Bucs third straight road trip. I had missed that.

Travel fatigue, I think, is a much underrated factor in all sports. I wonder what the winning percentage of NFL teams playing their third straight on the road is?

by DGL (not verified) :: Wed, 12/21/2005 - 2:18am

Re: #51 and #38: Actually, I've seen it spelled both "dreidel" and "draydel". I think the former is the traditional spelling. Although you could go with the Yiddish, dreydl.

I like the Colbert Report shoutout on the Bears.

by Fnor (not verified) :: Wed, 12/21/2005 - 2:21am

This column is clearly overrated because I am not drunk enough
This column is clearly ranked too high because it disagrees with THE SYSTEM. Me throwing darts at my jersey-clad Bellichick voodoo dolls is way better than this, because it involves alcohol. Like I heard the Jets tailgating party did once on that show Mike & Mike and it was green and it tasted like cough syrup but that's only because Greeney is one of those metro losers and probably likes the washington redskins because of their suggestive name and that is why Minnesota is underrated because Aaron sucks and this article is useless I'm going to go gamble against it.

by Fnor (not verified) :: Wed, 12/21/2005 - 2:22am

The lack of semicolon after the first line ruins the effect :(. Or... does it...?


by Bruce Dickinson (not verified) :: Wed, 12/21/2005 - 2:24am

regarding the continuity of the Bears D:

perhaps they won't be a #1 defense next year, but the stats you presented a while back about Lovie Smith's defensive ranks while coordinating for the Rams might indicate (he had them in the top 10 all 3 years) that they'll remain at an elite level (esp if they stay healthy)

by bobman (not verified) :: Wed, 12/21/2005 - 2:26am

RE: 102: whoo, Travis, some nice work there. I remember the last game, Orange Crush crushing Bert Jones and my suddenly inept Colts at Mile High. The game was not as "close" as the score made it appear and the (redacted!) who walked around Mile High with a sandwich board sign stating "10-1, alone at the top" was all too correct. In fact, 28 years later, I am still holding a grudge. Not (redacted-ing) healthy.

RE: #64: Purds, I am pretty sure the quality of the opponent has something to do with it. i.e., holding Gates to no TDs and 60+/- yards is considered a big deal, comparatively.

RE: #90: Mr. Riley, I am a Colts fan and I do not like the Patriots, (Sam I am....) but your post caused me to burst out laughing. I re-read it with the last line being said in the voice of "Comic Book Guy" from the Simpsons. I'm hoping to enjoy some smugness in a couple months. Bartender, I'm buying for the house: extra smugness all around. Cheers.

by ems (not verified) :: Wed, 12/21/2005 - 2:50am

Non-troll Seattle fan here. Is anyone interested in actually talking about why the Seahawks/Titans game was so close, and what it means for the future? Total DVOA aside, in retrospect, I'm not surprised at the way the game turned out. DVOA doesn't account for injuries, and here's my take on that missing set of information.

We've had a fair number of injuries on defense this year, and in this particular game, we were missing two starting linebackers, our #2 and #3 corners, our starting SS, and a starting DT. I don't mean to use injuries as an excuse for poor play; I think our backup players have been playing above and beyond all expectations. While we've played against poor offensive teams the last two weeks, they've shown themselves to be a talented bunch, and I am hopeful that they'll improve next year.

In this game though, we weren't able to establish a pass rush, and Steve McNair picked our young, inexperienced secondary apart. I hadn't been keeping up with his numbers, but I guess he's ranked 8th in DPAR for the year, and of course ranked #1 this last week. He's a good QB, a veteran QB, and I think he proved more than able to abuse our young, inexperienced secondary. More to the point, he's the highest rated QB we've played this year in DPAR. While our young defense was able to pound San Fran and Philly with their terrible QBs, the Titans had the player(s) to attack our current weakness.

What does this mean for the future? Specifically, I'm really glad Indy lost this week to SD. If we have the same problems with the pass rush as we did against the Titans, Indy's starters could put up 35 on us in one half. If that actually happens, hopefully our offense will keep it close enough to close out the second half (against their second string) and keep homefield advantage (even though Aaron said our homefield advantage ranks 24th in the league, I'll take it).

According to what I've read, we're supposed to have all our defensive starters back for the playoffs except for Jamie Sharper and (of course) Ken Hamlin. I'm feeling pretty optimistic that they will play well. Specifically, that they will be better than where DVOA currently rates them.

[Sorry you guys have had so much trouble with trolls this year, and that Aaron has got so much hate mail. FWIW, I'm new to the website this year and I love that Seattle is ranked #6 in a totally objective, statistical ranking. We could be doing a lot worse, all things considered. I WILL enjoy watching SA break the rushing TD record this year though, and hopefully watching us tie Indy for the best record in the NFL. Best Seattle team ever, by far! :)]

by Travis (not verified) :: Wed, 12/21/2005 - 2:59am

Re: 104:

Since 1970, NFL road teams have won 42.1% of regular season games. Teams playing their 3rd or more straight road game have won only 39.8% of those games.

All Road games:
70-79 814-1086-32 43.0%
80-89 906-1212-10 42.8%
90-99 939-1387-2 40.4%
00-04 541-722-1 42.8%

1995 96-144-0
1996 91-149-0
1997 93-145-2
1998 89-151-0
1999 100-148-0
2000 110-138-0
2001 112-136-0
2002 108-147-1
2003 99-157-0
2004 112-144-0

Road game 3+:
70-79 62-84-1 42.5%
80-89 23-40-0 36.5%
90-99 31-44-0 41.3%
00-04 5-15-0 25.0%

1995 4-3-0
1996 4-3-0
1997 5-6-0
1998 1-5-0
1999 4-2-0
2000 2-6-0
2001 2-2-0
2002 0-5-0
2003 1-1-0
2004 0-1-0

Other things I found out while researching this:

The 1990 Bengals were the last to play 4 (and 5) straight road games, doing so in Weeks 4 through 8. They shared Riverfront with the Reds; when the Reds made the playoffs and World Series, home dates were swapped with the Oilers and Browns.

Sharing stadiums with baseball teams meant that the Raiders and Jets would often open with 4 or more road games.

The 1982 Jets, counting the playoffs, ended the season with 6 straight road games.

by mshray63 (not verified) :: Wed, 12/21/2005 - 3:14am

Hey Travis, muchas gracias for #102! Did you get that data online someplace?

by Matthew Furtek (not verified) :: Wed, 12/21/2005 - 3:34am

Do we need to have a page to discuss the individual DPAR/DVOA rankings? I'm just gonna post my random observations

Looking at Thomas Jones' success rating, Chicago must also have a very bad offensive line as well. Looks like his line gets a huge benefit from a number of 10+ yard runs. Conversely Corey Dillon has benefitted from good line play.

Washington has made a significant improvement in 10+ yard runs this year.

Arizona should run the Run and Shoot next year.

Steven Jackson, overrated?

Micheal Vick's peers: Trent Dilfer, Charley Frye, David Garrard, and Kyle Boller.

Jacksonville has the best drive:int ratio???

by Ferg (not verified) :: Wed, 12/21/2005 - 4:21am

Re 113: Steven Jackson, overrated?

Not necessarily-- he's been very respectable catching passes, and both he and Faulk had significantly better DVOA last year. It seems like it's more of general malaise on the Rams.

Micheal Vick’s peers: Trent Dilfer, Charley Frye, David Garrard, and Kyle Boller.

Don't forget to add in Vick's rushing-- that brings him up to Kerry Collins/Eli Manning territory. Not earth-shattering, but nothing to be ashamed of either.

by the K (not verified) :: Wed, 12/21/2005 - 4:50am

Buffalo is clearly ranked too high because I have to sit through another blowout in another hopeless season, and they never win any of their few national TV appearance games. My system of throwing darts at an NFL dartboard is clearly better than this because it gives me something to do while watching another blowout. pls put bills at 32nd lolol k thx i think ur site rox usally.

by big_adventure (not verified) :: Wed, 12/21/2005 - 5:35am

Excellent math-o-phobic commentary, as always. I haven't had the chance to digest the (!) 115 comments in 9 hours that hit the site here, but the Fox site, in the sidebar, has a link to a brilliant article:

"Sherman confident one loss won't affect his future"

Umm, Mike, what about the other 10?!?!?!?


by Gatts (not verified) :: Wed, 12/21/2005 - 5:44am

Is the Whining Pats Fan an FO urban legend?

As the keeper of the mailbox, NO!

Actually so far there've been more Seattle mails this week.

It's maddening because most of the hate mail we get comes from people that think this is a subjective ranking, no matter how much Aaron says it isn't.

by big_adventure (not verified) :: Wed, 12/21/2005 - 6:16am

#103 -

Better to be the strongest team than the team with the most wins. More often than not in the past 20 years, the team with the most wins has gone home empty handed at the end of the season, without even a Super Bowl loss to show for it.

Andrew, this is unfortunately obvious. Let's assume that the team with HFA has an 80% chance of winning their 1st playoff game and a 60% chance of winning their second matchup (conf champ level). The chance of them making the big game in that case is 48%.

Thus, "more often than not" is exactly how often the team with the best record will miss the World's Largest Roman Numeralicized Marketing Spectacular.

It's still better than NOT being the top ranked team in the conference. If we figure that the number 2 team probably has a 75% chance in their first game, and a corresponding 40% chance in the CC game. They only wind up with a 30% chance of getting to plan a 3 games to glory DVD set.


by bobman (not verified) :: Wed, 12/21/2005 - 6:45am

RE: 110

ems, as a Colts fan living in the 206 area code, I started to get concerned about the Indy/Sea game two weeks ago when the Colts had a 26-3 4Q lead over Jax and relaxed anough to let up two scores in the final frame to the immortal David Garrard (!)--really they ended up needing a 1st down with less than a minute left to secure the game, while the Hawks were beating the living (redacted!) out of the 49ers, a game they easily could have let up and stumbled like the Colts. I grumbled to the missus "the Hawks are hungier and better focused than the Colts." Then came last week, when the Hawks shifted into reverse for the Titans game (a team Indy beat twice by a total of 66-13). At the same time, the Colts let up early, got punched around, and still almost beat a quality SD team that had a lot more riding on the game. If both were meeting at full strength, I'd now feel a lot better about the Colts' chances.

Injuries aside, I'd focus on team maturity, emotions, and the coach's experience and ability to guide them. Dungy has passed the test so far with a lot of stupid distractions and Holmgren, whle allowing a lot of late-season and late-game letdowns the past few years, HAS won it all for Pete's sake, so he should know how to keep the ship headed in the right direction. Maybe last week was their mulligan and they start picking up steam for the playoffs starting 12/24. The same may apply to the Colts, but they'll be playing fewer stars.

Plus, the Colts have a recent history of throwing in the towel on meaningless late-season games only to wax the same opponent in the playoffs shortly thereafter (Mr. Shanahan, I'm looking in your direction!). I don't think it's their plan to lose as much as it is to save the players and show nothing. Hawks might do the same (though less likely) making this week's game a potentially truly weird affair. (That's one reason I turned down TWO offers to go.)

So we won't REALLY know until Feb when, hopefully, they play again in the friendly confines of Ford Field, where Manning put up 6 TDs in 3 qtrs on T-giving day last year. Okay, okay, I know, it was against the Lions. But even after opponent adjustments, it's probably 3 or 4 TDs-worth of work, no?

by admin :: Wed, 12/21/2005 - 9:32am

OK, there's a hate mail from a Seattle fan that says "You have the Giants what third? And they lost to the 49ers." Did I miss something? When did this happen?

I got another Seattle e-mail calling me an "East Coast Nazi" which I must admit is a bit of a change from the antisemitic e-mails.

To be serious for a moment, I have no problem with non-troll fans discussing their teams. I'm sorry if certain teams create waves of trollery and I not only have nothing against those teams, I have nothing against intelligent fans of those teams who want to seriously discuss those teams. If we poke fun at the trolls, we hope you understand we're not poking fun at all of that team's fans. (No, no, not even Atlanta.)

I knew about the Seattle injuries this week, but were those guys (other than Sharper and Hamlin) in the game against San Francisco? Or are the 49ers just that bad? I've heard conflicting reports on whether Dyson will return for the playoffs... also, the defense is losing guys to injury but the offense is getting them back (Darrell Jackson).

Oh, and Steven Jackson has been really inconsistent this year. Sometimes he shows that great ability and sometimes you end up with six yards against the Cardinals, possibly the weirdest individual game of the year until Boller went nuts on MNF.

by spenceKarl (not verified) :: Wed, 12/21/2005 - 9:51am

RE: 44 (VarlosZ)

Where did you come up with 12.3?

by Jamie (not verified) :: Wed, 12/21/2005 - 9:53am

Seattle fans need only to look at who they have beaten and then they will get a glimps of what their chances are of a world championship. Numbers do not lie. A .407 opponent win percentage is not going to get it.

by dryheat (not verified) :: Wed, 12/21/2005 - 10:54am

My favorite commentary thus far, nice work Aaron. And finally a music reference I get. It's nice to know I'm not the only one quoting the Specials when Rudi gets a nice run.

I'm not sure what to make of the Pats comment. While I really only think you can fault Brady for 1.5 of the four INTS (as opposed to the 3.5 you've written), it's at least the third, and I think fourth time he's been charged with four picks in a game. It's not like DCs have figured him out after the other bad games he's had -- the next week he goes out and performs nearly flawlessly. So I don't think it's a defense causing him to have those games about once a season. Maybe the receivers have a case of the dropsies that day, maybe his shoulder was bothering him enough that he couldn't get his usual zip on the ball, maybe he was overcompensating and throwing everything too high, maybe it was abnormally windy. If there were was a fundamental flaw in his game, it would have been exposed by now.

Great research on historical data Travis.

You can't spell Tuiasosopo without "so so". Actually, I'm not sure I can spell it at all.

by DGL (not verified) :: Wed, 12/21/2005 - 11:17am

Fnor (107),

No, trolls wouldn't know what to do with a semicolon if it bit them.

Oh, and Aaron, the time the Giants lost to the 49ers was in the 2002 NFC Wild Card game - remember, the one with the botched snap on the FG as time expired, and the illegal man downfield penalty on the Giants that should have been a pass interference penalty on the 49ers? Well, OK, it was a few year ago, but still...

by Chris (not verified) :: Wed, 12/21/2005 - 11:22am

Aaron or anyone else. I am looking for stats for redzone offense and defense. I can't seem to find them anywhere on this site. I haven't had enough coffee yet so am I just not seeing them or do you not put those up? I noticed on the Bengals commentary you included them and wanted to know where the Colts ranked. Appreciate any help.

by Kuato (not verified) :: Wed, 12/21/2005 - 11:23am

RE: 111 (3 or more Road games)

Don't forget that a few years ago, the Bears played a full season of 16 road games and this year the Saints have done the same. Both teams as it turns out have been horrible. I know this is not the same thing exactly, but just wanted to throw that on the fire.


by Smeghead (not verified) :: Wed, 12/21/2005 - 11:26am

I'm mystified in general at the emotional investment some fans have in X rating system, where X = some algorithm, some guy's off-the-cuff opinion, some poll, drawing names out of a hat, or whatever. Who the hell remembers where Dr. Z had the Broncos rated after week 11 last season or -- much love to FO -- where the Bucs were in DVOA at the end of October? Who cares? (I guess one could say the same about sports generally, if one had one's priorities in life straight.)

Having said that, as a Seattle fan, I'm particularly baffled that we could emit any form of indignance. I mean, these people must be bandwagon-hoppers or northwest transplants or something. No honest sports fan in the Pacific Northwest with any sort of historical perspective could greet the likelihood of home field advantage in the playoffs with anything but trepidation. The last two Seattle teams that had their respective leagues' best records going into the playoffs meekly folded up in the ALCS (2001 M's) and got bounced in the first round by Dikembe Mutombo (93-94 Sonics). The Seahawks have never had a first-round bye and haven't won a thing in January since Dave Krieg was handing to Curt Warner; certainly Mike Holmgren's track record should not induce a swagger.

Little humility might be called for. Plenty of time in the offseason to trash-talk if they DON'T choke. Heed me now and let your psychologist thank me later, fellas.

by spenceKarl (not verified) :: Wed, 12/21/2005 - 11:30am

RE: 125 (Chris)

The DVOAs for specific scenarios are not available on the site, but Aaron's great about responding to inquiries.

by Harry (not verified) :: Wed, 12/21/2005 - 11:39am

#123 - I agree, I don't think Brady's game against the Chiefs is all that indicative. "Something made him throw 4 picks" - yeah, his shoulder. 2 were clearly just overthrown and the last was a desperation throw into a tight spot that Dwight probably should have held on to. If the Chiefs did anything special, it was putting on the pressure up front, although usually Brady has shown he usually does not get rattled. I do worry that Brady may be starting to develop a case of the Favres - a belief that he can make throws most QBs wouldn't and shouldn't. He's had a bit of luck this year too, I can remember 5 or 6 occasions this year when DBs dropped Brady balls they should have had.

by Rich Conley (not verified) :: Wed, 12/21/2005 - 11:45am

RE: The patriots

Aaron, any chance you can publish the DVOA numbers for each of the last 3 or 4 games? (or do I just not know where they are?). In the last 4 or so games, they've moved up about 5% overall DVOA each week pretty consistently. I'm just curious how high the performances have been individually.

by Andrew (not verified) :: Wed, 12/21/2005 - 11:46am

dryheat #123:

Miami has Brady's (and the Patriot's) number, and they play them tough every time they meet.

One of these days, the rest of the league might wake up and play the Miami Two-Safeties deep scheme against them too, just like some teams have finally woken up and copied what Tampa and Philly do against Michael Vick.

by Chris (not verified) :: Wed, 12/21/2005 - 11:48am

RE: 128 (spenceKarl)


by Andrew (not verified) :: Wed, 12/21/2005 - 12:02pm

Brady and Interceptions:

Over 1/3 of Tom Brady's career interceptions (22 of 64) have come in 6 games where he has thrown 3 or 4 picks. Denver Week 7 2001, Green Bay Week 6 2002, Buffalo Week 1 2003, Washington Week 4 2003, Miami Week 15 2004, Kansas City Week 12 2005. Another 1/3 (22 of 64), came in 11 games where he has thrown 2 picks. In the other 54 games of his career, he has just 20 interceptions. Brady seems to have a tendency to have annual meltdown games, where other Quarterbacks, like Manning or McNabb, are more consistent, but also have more games with a single pick.

McNabb for example, has just one game with 3 picks, and 12 games with 2 picks out of 94 games in his career, leaving 39 picks to be spread over 81 games (a worse ratio than Brady minus his meltdowns).

Philosophical Question: Is it better to meltdown twice per year in a gory pickfest, but otherwise take care of the football well, or to be more consistently sloppy, but with an overall lower number of picks?

by James, London (not verified) :: Wed, 12/21/2005 - 12:05pm

Re #131

It's fine to copy the 'Miami two Safeties Deep Scheme' if you have the Dolphins personnel. What Miami have been very good at in recent seasons is getting pressure from the front seven, and in particular, the front four. If you can't do that, then keeping the safeties deep isn't going to work.

In fact, if you can get pressure from your D-line while being able to keep your safeties in a deep zone, you're going to cause most QBs trouble...

by Dan (not verified) :: Wed, 12/21/2005 - 12:07pm

B, The Seahawks are still playing for home field advantage in the NFC. A Seattle win or a Bears loss makes it happen and of course Seattle plays on Saturday while the Bears play on Sunday.

by Oswlek (not verified) :: Wed, 12/21/2005 - 12:18pm


You beat me to it. It was just a bad game, nothing more, nothing less. Although, I disagree with you that Brady should only be faulted with 1.5 of them. I think he was responsible for all of them.

Two were terribly overthrown that the receiver was lucky to get a hand on it. One was throw behind the receiver and into double coverage (plus is hung up long enough to let a third receiver get into the play). And the last one I can't remember right now, but I remember thinking it was another QB mistake.

Brady was just trying to do too much as the D hadn't yet gelled and Faulk, Dillon and Givens were out.

What I found funniest is KC's safety (can't remember his name now) talking after the game about how they knew exactly what the Pats were doing with every play, thus leading to the picks. I found myself muttering aloud, "why did you leave guys open 20-30 yards down the field if you knew what was happening? Did you really predict as well that the pass would be overthrown?"

by Fnor (not verified) :: Wed, 12/21/2005 - 12:20pm

I had a really horrible nightmare last night. We were lighting everyone up in the playoffs, winning at CIN and then at IND. I was so happy! We were the last wild card, so the Heinz Field AFC Championship Curse couldn't rear its ugly head!

But then I thought, and in my dream NE had beaten JAX and won at DEN. Plus, they lost to Miami, resting their starters, making them 10-6 to PIT's 11-5. Which means it would be NE at PIT in the AFCC. I woke up screaming "NOOOOOOOOOOOOOO!!! WHY WON'T YOU LEAVE US ALONE?!"

What a scary dream.

by Oswlek (not verified) :: Wed, 12/21/2005 - 12:21pm


"third receiver" should say third defender.

by dryheat (not verified) :: Wed, 12/21/2005 - 12:21pm

129 Harry...you bring up a great point that I meant to but forgot: His last two INTs vs. KC (remember because it was so recent), plus his last two in Denver in 2001 (remembered because I was there), were desparation heaves coming late in the fourth quarter down by more than one score. I'm not sure about the other games, but the same scenario could apply. In those cases, where the defense can drop everybody into coverage, I find it hard to fault the QB for Hail-Mary type INTs.

131. Miami is a divisional game (obvious statement of year). I'm of the opinion that division games are almost always closer that talent would indicate, owing to familiarity and the offsprung contempt. Miami and New England have always played each other tough, even when Miami was elite and New England a doormat. By my quick count, Miami is 2-6 vs. Brady. I'm not sure that qualifies as "having Brady's number." In the same timeframe the Bills are 1-8, the Jets are 1-7, and the Colts are 2-6. If the Bates defense were particularly effective vs. Brady, other teams would be copying it. It's not all that difficult to play. Miami has had excellent football players in its secondary, which probably is more relevant than the scheme they play in.

133, obviously, you'd rather have a qb who passes the ball very accurately for 14 games and very inaccurately for 2 that one who is mediocre for 16.

by Sophandros (not verified) :: Wed, 12/21/2005 - 12:23pm

What about analysis of the Saints' home v. away performance?

I want to know whether my team sucks more at "home" or away?

I think that you guys have an anti-Saints bias because their name is rooted in Christianity. ;)

BTW, love the Vikings reference.

by Independent George (not verified) :: Wed, 12/21/2005 - 12:23pm

This thread is almost a miniature version of 'Best of FO'. I think the only thing we're missing are the irrational Manning vs. Brady comments.

Aaron, have you ever considered having a separate email link in the Fox rankings open a pre-formatted generic hatemail form?

by Oswlek (not verified) :: Wed, 12/21/2005 - 12:24pm


for what it's worth, the game would still be in NE if Pitt plays at Cincy. Home/away is based on seeding, not record.

So keep your dream alive. :)

by Andrew (not verified) :: Wed, 12/21/2005 - 12:27pm

big_adventure #118:

Better to be the strongest team than the team with the most wins. More often than not in the past 20 years, the team with the most wins has gone home empty handed at the end of the season, without even a Super Bowl loss to show for it.

Andrew, this is unfortunately obvious. Let’s assume that the team with HFA has an 80% chance of winning their 1st playoff game and a 60% chance of winning their second matchup (conf champ level). The chance of them making the big game in that case is 48%.

For simple purposes of in-conference play, yes. But we were discussing the best team record in the NFL - both NFC and AFC conferences together. More often than not, the team with the best overall record in the NFL has gone home empty handed.

The following teams with the "best record in Football" took something home in the past 20 years.

2003 Patriots - Lombardi and Hunt Trophy. 2001 Rams - Halas Trophy. 1996 Packers - Lombardi and Halas Trophy. 1994 49ers - Lombardi and Halas Trophy. 1991 Redskins - Lombardi and Halas Trophy. 1989 49ers - Lombardi and Halas Trophy.

6 teams in 20 years with a trophy for the best record in football, is really not a very good record. AFC top dog teams (and the 49ers) have fared especially poorly in this regard. Home field throughout the playoffs should translate into a 80% chance of winning the divisionals and a 60% chance of winning the championship, for an overall 48% chance of the top team taking home a trophy. But, its been just 30% instead.

by Andrew (not verified) :: Wed, 12/21/2005 - 12:41pm

Oswlek #136:

"What I found funniest is KC’s safety (can’t remember his name now) talking after the game about how they knew exactly what the Pats were doing with every play, thus leading to the picks."

Probably Sammy Knight, ex of the Dolphins and several humiliations of Brady.

by MJK (not verified) :: Wed, 12/21/2005 - 12:47pm


You hit the nail on the head in response to Andrew's comment. Yes, playing two deep safeties workes against Brady...IF your team has good tacklers (as Miami does). Otherwise, Brady dumps off quick to Kevin Faulk or Daniel Graham or Patrick Pass, or throws a quick hitch to David Givens at the sideline, and the reciever promptly breaks/dodges tackles and runs 4-15 more yards. Example--last year, KC tried playing two deep safeties against NE, and the NE recievers caught short and ran wild with YAC, because KC is a bad tackling team. Not many teams have both two canny safeties that can play deep effectively AND CB's and a front seven that can play tight and tackle short stuff well. (Denver this year can, and look at all the trouble the Pats offense had in that game). This is why the Patriots run the quick hitch play at the goal line so often, because it works.

by B (not verified) :: Wed, 12/21/2005 - 12:49pm

Miami's formula for beating Brady:
1) Have two very good corners in tight press coverage.
2) keep two safties deep.
3) Have Jason Taylor play DE
4) Have Zach Thomas at MLB.

by jebmak (not verified) :: Wed, 12/21/2005 - 12:53pm

It is about a 48% chance of making the superbowl with those figures. You would want to multiply another .60 for chances of winning the big game. .8 x .6 x .6 = 28.8% I would say that the numbers are on target.

by Mikey (not verified) :: Wed, 12/21/2005 - 12:55pm

Travis, wow. Thanks for doing the legwork to answer that question about teams playing three straight on the road. Interesting stuff.

Notable to me that the current win % of teams playing 3+ away from home is the worst of any decade, and that these extended road trips are also a lot less common than they used to be (thanks mostly to more football-only facilities). Is it possible that today's players are less equipped to play three straight on the road partly because it's so rare that they're asked to do so?

by admin :: Wed, 12/21/2005 - 12:57pm

Hi ho.

Even if the Brady interceptions were just his shoulder, or come from the fact that KC has two old Miami defensive backs, or came in the fourth quarter with the team losing, the fact is that they happened. What Patriots fans seem to want to do now is believe that not a single event took place on the planet Earth between mid-February of 2005 and the moment the Patriots took the field against the Jets three weeks ago. Not only that, but they get viciously angry at anyone who dares suggest that any event that may have taken place during that time might be used as evidence in analyzing future events.

Anyway, the DVOA system counts every play during the season because it is a great way to measure teams. Just because I'm a Patriots fan I can't suddenly wave my hand and make all those losses go away, placing a healthy Rodney Harrison back in the secondary and a healthy Matt Light back on the line. If I could, why couldn't a Colts fan just say that last Sunday meant nothing and that the Colts really weren't trying and therefore in reality the Colts are still undefeated?

There's something else people aren't noticing. Let's say that the Patriots, in fact, do have some magic hold over the Colts and they have their mojo back now (and I would love it if that were true). There's another team possibly looming for the AFC Championship game, a team that they seem to play every year, and that always gives them fits, even if New England escapes with a win. I have yet to hear any Patriots fan talk about what happens if the Patriots beat the Colts ... and then go to Denver.

by B (not verified) :: Wed, 12/21/2005 - 1:01pm

Re 143, 147: Going back to 1998, the team with the highest Weighted DVOA won 3 of the last 7 superbowls, and the team with the highest estimated wins won 4/7. This includes the fact that many of the top teams rested players in the last one or two weeks.

by Todd S. (not verified) :: Wed, 12/21/2005 - 1:04pm

#50 That's exactly how it looked to me. Sanders seemed to be sitting on the short route to Gates, and McCarddell went right by him. Good play call and execution. Bad coverage by Sanders.

by Andrew (not verified) :: Wed, 12/21/2005 - 1:11pm

dryheat #139:

Brady Stat-Lines vs. Dolphins:

05-10 21 of 36, 275y, 2t, 2i, 77.9
04-15 18 of 29, 171y, 3t, 4i, 73.3
04-5 7 of 19, 76y, 2t, 1i, 62.6
03-14 16 of 31, 163y, 0t, 0i, 67.0
03-7 24 of 34, 283y, 2t, 0i, 115.2
02-17 25 of 44, 221y, 1t, 1i, 68.5
02-5 17 of 31, 240y, 2t, 2i, 74.7
01-15 11 of 19, 108y, 1t, 0i, 91.6
01-4 12 of 24, 86y, 0t, 0i, 58.7

BTW, the record for Miami in this series is 3-6, not 2-6, with most of the Patriots wins being quite ugly and very low in DVOA. That's 3 of 19 losses by Brady in 5 years. And they've really been lucky to not lose several other of these games. If the Brady lead Patriots lost 1/3 of the time to the rest of the league, they'd have compiled 25 losses since Week 3 of 2001, not 19.

For a QB who has been well above average in most of his outings, to look consistently mediocre against one team says somthing about that team. Brady himself admits it, and attributes it to the defensive scheme they play.

I also wouldn't attribute this to overwhelming pressure. Miami has sacked Brady 20 times in 9 games, while his career total is 153 sacks in 72 games, so Miami has just 1 sack over average. He has 5 fumbles against Miami in those games, which is under his career average coming from 47 fumbles in 72 games - we'd expect 6 fumbles.

It seems like scheme more than anything.

by dryheat (not verified) :: Wed, 12/21/2005 - 1:15pm


For my part, I'm not dismissing the early-mid part of the season for the Patriots. It happened. I witnessed an ass-whuppin' by San Diego, another by Indianapolis, and a team that absolutely (and I can't stress this enough) Q-U-I-T at Denver by halftime. It's definitely a concern for me. I think DVOA has them right where they should be. My point concerns the commentary on Fox, where you suggest that defensive co-ordinators are going to study the KC tape to see what they did to force Brady into throwing interceptions. The short answer is nothing. It did happen. But it happened (I believe) for reasons unrelated to the Kansas City defensive scheme that day.

As for Denver, didn't the Pats win the last two times out there before this year? Granted, that means Denver has about an 11-2 edge in Denver in the last 20 years, but Brady-led Patriot teams are 2-2 out there. If they match up, I'm worried about Denver's current talent, not past history between the two squads.

by B (not verified) :: Wed, 12/21/2005 - 1:19pm

Aaron: If you really want to dishearten me and the other Pats fans, just point out that there record is 3-1 vs the NFC South, 4-0 vs the AFC east and 2-4 against the rest of the AFC, with all four losses coming against teams who are still in the playoff hunt. And they were all by 8 points or more. Yikes.

by Mikey (not verified) :: Wed, 12/21/2005 - 1:19pm

117 -

Gatts, isn't it true that DVOA *is* subjective in that it makes subjective judgements on how to weigh various elements that go into producing the final number?

I'm not a DVOA expert, so correct me if I'm wrong on any of the following examples of what I would say is subjectivity in the math:

--Offense, defense, and special teams are all weighed equally in total team DVOA. A reasonable fan could argue that the three should be weighed 40/40/20, 45/45/10, or any other combination. Making them each a third of the total is a subjective assessment of how much each contributes to winning.

--It's subjective to assess success rates based on individual plays rather than each series. For example, a set of downs starting from first-and-ten that results in gain of six, gain of two, no gain, punt earns a success rate of 66% (the punt is excluded, yes?) while a set that results in no gain, no gain, gain of 12 produces a success rate of 33%. Obviously in that case you'd rather have the "less successful" series. Again, I'm not making a case against DVOA here, just saying these are decisions someone has to make that aren't entirely objective.

--Weighing the more recent games more heavily I think is another example of making a subjective judgement as to what makes for a truer read of a team's strength.

I hope this doesn't read like a rant against DVOA. I'm a big fan of DVOA, this site, and all efforts to bring better analysis to understanding sports. I guess my point is that when DVOA is criticized it's better to stand up and say This is our system and we believe in it than to hide behind the idea that the system is truly objective. Any analysis that synthesizes multiple categories of data is subjective to some degree.

by Andrew (not verified) :: Wed, 12/21/2005 - 1:23pm


If we consider the opportunity to win the Conference Championship trophy but lose the Super Bowl, 48% of teams with the best record should at least get one trophy, regardless of winning or losing the Super Bowl. However, there have been too many chokes by #1 AFC teams and the 49ers for this to happen.

You are correct that when we make the criteria winning the Super Bowl, then the numbers are on target. But to win it, you must get there, and #1 teams have underperformed in the Divisional and Championship games.

by Oswlek (not verified) :: Wed, 12/21/2005 - 1:25pm


The sack numbers are deceiving. The biggest part of the "scheme" is the fact that Miami was able to get significant pressure with just 4 rushers. That allowed the CBs to play up tight and the safeties to play over the top. The scheme has never been complicated, it was great players that made Miami's D tough.

by dryheat (not verified) :: Wed, 12/21/2005 - 1:31pm

Okay, in the interest of accuracy, I've actually done some research to look at teams vs. Brady.

Dolphins 3-6
Bills 1-9
Jets 2-7
Colts 1-6
Broncos 3-1

I regret the inaccuracies, but don't change my POV because of them.

by White Rose Duelist (not verified) :: Wed, 12/21/2005 - 1:32pm

Andrew, the one very obvious point that you're missing is that 12 teams qualify for the playoffs. All else being equal, each would have a 1/6 chance of making the Super Bowl, since 2 of the 12 move on. If the first seed is making it more than 16.7% of the time, then there is a positive correlation between seed and winning the conference.

Of course the team with the best record doesn't always win the Super Bowl. Otherwise, why bother with wild cards, or even the playoffs? But the team with the best record is more likely to than any other team.

If being the #1 seed is not optimal - which seed is best?

by RCH (not verified) :: Wed, 12/21/2005 - 1:32pm

RE: Brady's picks against KC, I think that any time a defense can arrange to have Heath Evans as Brady's feature back thats the start of a good scheme.

by Oswlek (not verified) :: Wed, 12/21/2005 - 1:36pm

Aaron #149,

I have to assume you are refering to your email box, because I have not witnessed much of the vicious anger you are speaking of on these boards.

For the record, yes, I do believe that Brady's bad day can be largely ignored because he has had one in every other SB year; Denver in 01, Buffalo/Wash in 03, Pitt/Miami in 04. Yet, Brady has the lowest int/attempt ratio in the playoffs of any QB playing today (with a certain minimum, of course).

As to the season before the Jets game, of course it matters. If something was exposed once, it can be again. And the secondary won't have to face a good QB until the playoffs. But, it is not outside the realm of possibility that the D has just finally gotten healthy and learned to trust one another. We cannot know until the chips are on the table, but they've certainly earned my faith.

by admin :: Wed, 12/21/2005 - 1:37pm

Mikey -- all variables involved in DVOA have been improved over time in order to create numbers that balance a strong correlation with winning and a strong correlation from year-to-year and from game-to-game to try to measure the inherent quality of teams. The only time the variables were subjective was the first time, three years ago -- since then, all improvements are based on research and experimentation.

And the most important part of the statement "DVOA is objective" is that it does not change based on which team I root for, or which team's fans piss me off, or whether I like a certain player, or think he tries hard, or whether I think a team quit on their coach in the last game, or all those other things that Peter King and Dr. Z and other people base their rankings on. It doesn't include a "faith" component.

Oh, and yeah, the Pats stuff is mostly e-mail box. But there was some pretty bad stuff last week, mean stuff about the town where I grew up (and those of us in Boston know what the subtext of that is).

by B (not verified) :: Wed, 12/21/2005 - 1:39pm

158: How come nobody ever talks about Shannahan having Brady's number?

by MRH (not verified) :: Wed, 12/21/2005 - 1:47pm

#155 - I'll take a stab at answering. I believe the "weightings" (success, off/def/st in total DVOA, and in establishing the weighted DVOA) are arrived at by testing combinations to see what gives the best predictive/descriptive outcome. This is an objective approach to assigning values. They are not made up because Aaron or FO thinks one set of numbers is better and then used without testing that hypothesis (that would be subjective). When someone pointed out that 2nd order DVOA (if I'm calling that correctly) might be a better measure, Aaron tested it and adopted it: objective.

#102 and 111 - great research Travis. I'll throw one other game in there, final regular season game between teams with the fewest losses: 1967 Baltimore Colts (11-0-2) at Los Angeles Rams (10-1-2). Rams won 34-10 to take division title and loser Colts stayed home with one loss in the pre-wild card days (they tied in their 1st matchup so Rams won H2H tiebreaker). Rams were #1 in PF and PA; Colts were #2: 398-196 and 394-198.

Oh, and the best record team in the NFL (Rams) lost in the 1st rd to the Packers who of course then won the Ice Bowl and the Super Bowl (over the best record in football 13-1 Raiders).

by ems (not verified) :: Wed, 12/21/2005 - 2:03pm

#119 - I've been thinking for weeks that our upcoming game with Indy will have the looks of a preseason game, with neither coach wanting to show off very much, lest we meet again in February. Since we need the win more than you do for seeding purposes, I take solace in the fact that you went 0-5 during the preseason. :)

On the other hand, I can't say I'm looking forward to playing you guys in Detroit.

#120 - We were missing Dyson and Herndon for the SF game, so I'd venture to say that they ARE that bad. Jordon Babineaux has made some big plays this year and shows promise, shifting back and forth between safety and CB. I certinly don't want to denigrate the effort he's put out. He certainly did well enough against SF, but my conjecture was that he won't hold up for a full game against the likes of a Steve McNair (or even worse, Peyton Manning) without some help from the pass rush, which he didn't get. I'd like to think that having Dyson and Herdon back would have made our victory a little more decisive.

Now, why didn't we get any pass rush? I'd love to know. I saw the comment someone made that Tennesee had the first offensive line we've gone up against that didn't have injury replacements from the last week, and we had the injury to Chuck Darby. Just looking at the adjusted sack rates, Tennesee is ranked 8th for protection, and Seattle is ranked 9th for production, which seems like a pretty even match up. I'm new to the site, so I'm not sure where else to look for information on this issue, but those rankings seem to indicate to me that our lack of pass rush wasn't a huge surprise. Without the pass rush, our secondary got torched, and despite our huge lead in Total DVOA for the year, I think this result wasn't entirely surprising. I'm thankful that we pulled it out in the end.

I also think there's some truth to the sentiment that the talking heads are passing around: Last year, Seattle would have dropped that game. Hopefully we'll get our CBs back, and won't have this problem during the playoffs.

by Mikey (not verified) :: Wed, 12/21/2005 - 2:07pm

To the extent that "objective" means you're not doing any of the tweaking that angry fans might accuse you of, yeah, I get it.

If anybody here follows horse racing, did you know that the Beyer figures in the Daily Racing Form are routinely tweaked based on the observations of the guys who produce the numbers? I've followed ponies for quite a while and didn't know that until this year. Geez, if you're going to do that why bother doing the math at all?

by admin :: Wed, 12/21/2005 - 2:24pm

I didn't mean to imply that having faith in the Pats is bad, by the way. I have faith in them -- I certainly think they'll whip Jacksonville. But I try to keep the home-town faith out of my writing. It's strange to get so many e-mails accusing me of wanting to be Ron Borges, only looking for the negative with the Pats, since the whole reason this site exists is that I wanted to prove Borges wrong in 2002 when he said the Pats were a one-year wonder.

by Nolan (not verified) :: Wed, 12/21/2005 - 2:27pm

The only thing the 164 comments show is how pathetic the DVOA is! Is NE that good in beating TB 28 to nothing or is TB just that bad sometimes? The stats don't take into account a team giving up on a season or a team running up a score or not running up a score. If you ask me the games late in the season matter less than the early and midseason do because so many teams are out of playoff contention or as in the case of INDY resting players. Let it be known the DVOA sucks and doesnot work

by Ricky (not verified) :: Wed, 12/21/2005 - 2:32pm

Just wanted to check back in. I see much has been said since last week, but I bet the knucklehead who was calling for his Chiefs to come to Gillette this January has changed his tune a little (along with the Tampa guy saying Galloway was going to torch the Pats and the guy who discounted the fact that Simms has NEVER performed in a big game, but anyways, no I told you so's here).

Since the Pats don't play a meaningful game until the play-offs (though I guarantee they'll finish 11-5 winning their final two tune-ups), I have nothing to say until the play-offs. Happy Holidays.

by big_adventure (not verified) :: Wed, 12/21/2005 - 2:40pm

149 -

Aaron - it could be like eBay! There could be a time limit, and the best bid late gets to pick the outcomes!

Even better, the bids would have to be in the form of the RandomWebsite.com Football Outsiders' Ultimate Generic Hate Post v1.0 (care of zlionsfan). Unfortunately, that would mean that Pittsburgh wins out, as Fnor topped the list already with his delightful prose.

There is always next year.


by MME (not verified) :: Wed, 12/21/2005 - 2:40pm

#155 (Mikey):

Aaron already explained himself, but I thought I'd point out what I believe are misunderstandings on your part (Aaron, of course, can correct me if I'm wrong).

–"Offense, defense, and special teams are all weighed equally in total team DVOA. A reasonable fan could argue that the three should be weighed 40/40/20, 45/45/10, or any other combination. Making them each a third of the total is a subjective assessment of how much each contributes to winning."

Take a look at the numbers up at the top. If you notice, the DVOA numbers for off/def get much larger than the ones for special teams (which generally stay less than 10%). I believe Aaron has stated that this is because the formula for ST ratings takes into account that there are far fewer ST plays than Off/Def plays, and thus the weighting you speak of is in the numbers, thus it makes sense to add them up equally in the overall rating. (Also, I believe Aaron has stated that ST ratings are about 1/7 of total weight, with Off/Def getting 3/7 each)

–"It’s subjective to assess success rates based on individual plays rather than each series. For example, a set of downs starting from first-and-ten that results in gain of six, gain of two, no gain, punt earns a success rate of 66% (the punt is excluded, yes?) while a set that results in no gain, no gain, gain of 12 produces a success rate of 33%. Obviously in that case you’d rather have the “less successful� series. Again, I’m not making a case against DVOA here, just saying these are decisions someone has to make that aren’t entirely objective."

Yes, the choice of what constitutes 'success' is subjective, but is mostly based on fan's understanding of football. However, I believe you are confusing Aaron's Running Back Success Rate stat with DVOA:

In the example you gave, the first series would be broken down so that the first play gets a little bit more than one 'success point' because it got more than 45% of yards needed on 1st down (say, 1.5 points). The second play would get less than one, as it failed to make 60% of yards needed on 2nd down (say, .75 points). The 3rd down play of course gets nothing. Thus, the total 'success' for the series is about 2.25.

The sencond series is different. The first two plays obviously get 0 success points. The third play actually gets a little over 3 success points (Aaron, somewhere, pointed out a first down is worth 3 success points regardless of what down it is gained on). So this series has 3 points worth of 'success', which is more than the first series.

And just to make it clear, the RB Success rate stat for both series (assuming all plays are running plays) would be equal, as only the 1st down play of the first series is considered 'success' and only the 3rd down play of the 2nd drive.

I hope that makes the math a little more clear (probably not though...). Also, I kind of posted this so Aaron could confirm if my understadning is correct. Knowing how DVOA works is what makes me believe in its legitmacy.

by zlionsfan (not verified) :: Wed, 12/21/2005 - 2:43pm

Doesn't matter to me whether or not I'm cited as the author of MadLibs: DVOA ... just thought it was pretty cool that it got used! Now if it would only help some more ... the suggestions by Aaron and DGL are good ones. I think we could probably get every team represented; even some of the less-frequently supported teams have fans here.

Just to help out: Detroit is clearly ranked too low because the Wings are still playing well enough to lead their division easily. ESPN's NHL rankings are way better than this. Wingz rule!!!11111111

Back to my cave ... :)

by CaffeineMan (not verified) :: Wed, 12/21/2005 - 2:44pm

Re: Brady's picks - "Seems like scheme more than anything" and Pats chances in the playoffs in general.

I agree with the comments that say it's scheme AND personnel. Miami's defenses against the Pats during the Wannstedt era have never been complicated, just well executed man-to-man by good players and pressure by good players up front. Throw in a dollop of fierce division rivalry and there you have it.

I'm not saying you can't identify tendencies but exploiting them is not as easy as designing a scheme, as Andrew seems to imply, although this year it's easier because the defense hasn't been able to get off the {redacted} field.

Pats passing offense is up to #3 in DVOA, behind only Indy and Cincy. Plus, of the teams with the top 5 passing offenses, NE has the worst running game (14th), with Indy the next worst (8th) and the other 3 teams in the top 5 in rushing DVOA as well as passing, so the Pats passing game has been doing it without a net this year. Brady does have a tendency to lay an egg periodically, but it's not fear of the Kansas City game film in the hands of playoff opponents that keeps me up late at night thinking of the Pats chances.

No, it's still the high-powered offenses that stomped their defense in the past and that they will face again if they get past their likely wildcard game against Jax. And I don't think most of the Pats regulars on this site are forgetting those games, either. It's just that in the last couple weeks, their chance have improved from "They're crawling to the playoffs and are gonna have tire tracks down their backs from getting run over by Harrison/Wayne/Stokely/Clark" to "Look! A semblance of a defense has appeared. We've got a chance." Which is enough to at least speak the possibility out loud again. :D

Personally, I'm clinging to the hope inspired by that little smile by Richard "The Quiet Man" Seymour in the locker room at the end of the Bucs game.

by James, London (not verified) :: Wed, 12/21/2005 - 2:47pm

If #168 is not a joke, (I'ts getting hard to distingush between the Trolls and the 'Troll Baiters'), then would you like to explain how Tampa, who were 9-4 and top of the NFC South before playing NE, and who, despite losing, are 9-5 and slap-bang in the middle of the playoff race, have "quit on the season"?

by brasilbear (not verified) :: Wed, 12/21/2005 - 3:06pm

The Bears are clearly ranked too low because now that Glassman..er Grossman has returned, the offense will be way better. Filling your mouth with jello and spitting it on a list of all the teams and then raking them by which team has the most jello stuck to it is way better than this. Ron Riveria will find a way to get the offense to score more that 15 points while Norv Turner keeps the defense running.

by Wes M (not verified) :: Wed, 12/21/2005 - 3:08pm

Seattle is clearly ranked too low because of your obvious East Coast Bias. How many of you statnerds are west of the Mississippi? The Rockies? I thought so. The BCS, though also riddled with ECB - the Ducks got screwed, is way better than this. All u Seattle hatres R just jelus of us having Microsoft and Starbucks HQ's here. Millinares per capita, baby!

Oh, and : Fi-re Mil-len clap clap clapclapclap

by spenceKarl (not verified) :: Wed, 12/21/2005 - 3:21pm

Arizona is clearly ranked too low because Rackers is the best kicker ever. Seriously, how can u put them below Tennese??? Who the hell have they beat?? Nobody. That's right. My grandmother with alltimers is way better at making rankings than this stupid DVOgAy garbage. Everyone nows stats are worthless anyways. Lets see where your rankings are when my Cards give a Xmas eve spanking to the Beagles. Rankers will kick all you're stupid assez.

by TomS (not verified) :: Wed, 12/21/2005 - 3:29pm

So if estimated wins is the key predictor, can we just take the top eight, have a straight up, seeded tournament and the winner take all? Or should we just say the Indy is the victor and forego all the very useless actual games?


by Nolan (not verified) :: Wed, 12/21/2005 - 3:33pm

#174 I didn't say TB quit on the season it was a general comment on the fact that at this time of the season you have some teams not playing as hard as they did earlier in the season be it because they are out of the playoffs or as in Indy's case just trying to stay sharp for the playoffs with out gettting hurt

by squintsp (not verified) :: Wed, 12/21/2005 - 3:36pm

TomS, for the sake of my sanity, please tell me you're joking.

by spenceKarl (not verified) :: Wed, 12/21/2005 - 3:37pm

RE: 179 (Nolan)

Are you suggesting that Indy was just trying to stay sharp last week, rather than win the game?

by admin :: Wed, 12/21/2005 - 3:37pm

Yeah, I will try to clear up everything from MME's comment #171 later today or in mailbag or something ... so many things to respond to, so little time... Arizona, of course, is one of the teams that I have not yet been accused of bias against. Come on Arizona fans, where are you? (Except Jerry P.)

by TomS (not verified) :: Wed, 12/21/2005 - 3:38pm

wha duh ya mean, joking? I think that is as good an idea as any other I have heard (read, actually).

by spenceKarl (not verified) :: Wed, 12/21/2005 - 3:47pm

Aaron, any chance of moving all the previous Mailbags to their own section listed in the drop-down Archives menu? They provide some really great info, but they're such a pain in the arse to search for. It would really be helpful to see them all listed in one place.

by Nolan (not verified) :: Wed, 12/21/2005 - 3:47pm

#189 NO I am refering to the game this week against Seattle. If the game is lopsided because Indy is not playing at full strength. Then that doesn't mean Seattle is better than Indy. but the DVOA will not take that into account

by Nolan (not verified) :: Wed, 12/21/2005 - 3:49pm

I mean #179

by spenceKarl (not verified) :: Wed, 12/21/2005 - 3:51pm

You mean #181

by Mshray (not verified) :: Wed, 12/21/2005 - 3:53pm

The 49ers are clearly ranked too low because we beat Tampa. Who did Houston beat?. Dr. Z is way better than this & he also appreciates many of the fine wines from around here...and...and...I'm sorry Mr. York, it's really impossible. No one can do this, I don't care how many season tickets you give me.

by Nolan (not verified) :: Wed, 12/21/2005 - 3:54pm

Yeah #181

by steve (not verified) :: Wed, 12/21/2005 - 3:55pm

objectives - what defines success?

The explanation of DVOA on this site, clearly identifies that each play has a different value based on its situation (e.g. gaining 8 yards on 3rd and 10 is not as valuable as gaining 2 yards on 3rd and 1). DVOA tries to standardize this situational valuation - I think that is good. However, is the valuation comprehensive enough? And should it (could it) vary from team (or head coach) to team? Should it vary from game to game (interconference worth less than intradivision)? Should it vary based on a mathematical probability of whether winning/losing the game will affect the liklihood of making the playoffs, gaining HFA, getting a bye?

For example one head coach may be more inclined to hide his team's true strengths than another coach. And in doing so, the true power of the team may need a different calculation than other teams. I believe this to be the case, but have no idea how one would quantify such an attribute.

by Rich Conley (not verified) :: Wed, 12/21/2005 - 4:05pm


Because certain teams have first round byes, you can't just say you've got a 1/6 chance of making it to the SBowl.

A team with a bye has a 50% of winning its first game, and a 50% chance of winning the conf championship. So anything over 25% making the superbowl is a correlation.

Teams without a bye have a 12.5% chance.

Seed 1,2. x25% 50%
Seed 3-6. x12.5% 50%

Thats saying pure 50% chance of winning a game, not actual chances of winning games.

by spenceKarl (not verified) :: Wed, 12/21/2005 - 4:10pm

RE: 185 (Nolan)

Then feel free to refer to the DVOA ratings up through Week 14. Ignore the remaining two weeks...I do understand what you're saying though. And injuries are a problem that DVOA is currently unable to track; But does this fact alone really suggest that "DVOA sucks and doesnot work"? Not in my mind anyhow.

by B (not verified) :: Wed, 12/21/2005 - 4:12pm

190: DVOA treats all plays as having equal value, weighted DVOA treats recent games as more valueable, and EW treats plays in certain situations as more valueble than others (Red Zone, Offense in 2nd Quarters, Defense in second half of close games). Your idea is interesting, but first we'd have to figure out the optimal strategy. Is it more important to win a divison game than it is to win a game against a playoff team in your conference? The Guts vs. Stomps article suggests that thrashing weak opponents is more important than squeaking past strong ones. Maybe teams should "protect thier house" and try to win all home games, but since it's more difficult to win on the road, perhaps teams should work harder to win on the road.

by spenceKarl (not verified) :: Wed, 12/21/2005 - 4:14pm

RE: 190 (steve)

I'm confused. Are you suggesting that gaining 8 yards on 3rd and 10 is more valuable to one team over another? Or of more value to one coach over another?

by steve (not verified) :: Wed, 12/21/2005 - 4:22pm

re 194: i'm suggesting that not only does each situation in a game vary in its value to the eventual outcome of the game and has a different definition of success/failure, but the game as a whole has different value and definition of success/failure. Also, the value and definition likely varies from coach to coach (same as team to team - using "team" and "coach" to represent the same thing).

by David Brude (not verified) :: Wed, 12/21/2005 - 4:32pm

The RB numbers did not get updated even though it says they are updated through week 15.

by spenceKarl (not verified) :: Wed, 12/21/2005 - 4:32pm

RE: 195

How would you approach changing the success value for each play? I want to go back to the 8 yards on 3rd-10 situation for a second. The goal was to achieve a 1st down. Obviously the play was unsuccessfull since the resulting gain was 2 yards short. And obviously the play holds some percentage of success since at least 8 yards were gained; But why do you believe some coaches value those 8 yards more than other coaches?...I'm not trying to be a smartass. Just confused still.

by Rich Conley (not verified) :: Wed, 12/21/2005 - 4:37pm


Different coaches value that differently because some are more aggressive than others.

One coach may think about going for it on 4th and 2, while another punts away without it crossing his mind.

by Kevin (not verified) :: Wed, 12/21/2005 - 4:40pm

re: 194
I belive steve (190, 195) is saying that each play should be taken in the context of the season, not just the game. He's also supposing that DVOA should take into account how much of one's arsenal a coach decides to lose. Maybe Lovie Smith has 5 dynamite passing plays for Kyle Orton that he hasn't used yet because he doesn't need them. I think the points are interesting, but the first involves some subjectivity and probably does not have a large enough sample size (there are so many possible combinations and a limited number of games) and the second is completely subjective. DVOA is not the best possible measure to determine a team's success or likelyhood of success, it's just another step in the continuous striving toward such a measure.

by brasilbear (not verified) :: Wed, 12/21/2005 - 4:40pm

RE: 197

There are some who would argue that this years Bears team would be happy with 8 on 3rd and 10. Playing a field position and TOM game means not taking risks (throwing to the marker when the defense is expecting it) and taking what the defense gives (the underneath throw) in an effort to run off clock and limit turnovers. In this case, a three and out is not a bad thing, provided there were no turnovers and a positive gain in field position.

Its quite debated on the Bears board. The anti-Grossman side is worried the Bears will opt for more passing yardage meaning more risk and less TOP.

by brasilbear (not verified) :: Wed, 12/21/2005 - 4:42pm

Sorry, TOM should be TOP which means Time Of Possession.

I will now beat myself about the head.

by Oswlek (not verified) :: Wed, 12/21/2005 - 4:49pm


I think what he is saying is that coaches have different views as to how many tricks they need to pull out of the bag.

For instance, maybe Denver is up on Cleveland 17-3 in the second quarter and the defense is playing great. Shanahan may decide to do absolutely nothing special the rest of the game. Maybe he has a play that would work really well on 3rd and 12, but doesn't use it. If he sacrifices a little offensive efficiency, so be it. The game is well in hand.

by Nolan (not verified) :: Wed, 12/21/2005 - 4:52pm

The problem is that no one on this site is only taking into account certian weeks

by spenceKarl (not verified) :: Wed, 12/21/2005 - 4:59pm


I think the problem I have with steve's ideas, is what Kevin suggested. That there is subjectivity in determining which coach values a particular gain more than another.

by spenceKarl (not verified) :: Wed, 12/21/2005 - 5:01pm

RE: 203 (Nolan)

Each team has 16 weeks of football. The sample size of data increases drastically when you perform analysis on a per play basis.

by steve (not verified) :: Wed, 12/21/2005 - 5:09pm

re 197: I'm going to get fictional here:
let's suppose i'm a 9 and 5 team, strongly expect to win my division and make the playoffs, but will need some unlikely help to earn a bye. I'm playing a team in the other conference. the situation is 3rd and 10. In such a situation, If I value protecting my true strengths much, then I may define success of the play as (1) not divulging my true tendency in the situation and (2) gaining the 10 yards with (1) being most important. Whereas someone who does not value protecting strengths as much may simply consider success as being whether they gain 10 yards or not. Therefore an 8 yard non-disclosing gain is more important to my team (a partial success) than another team in that same situation (a total lack of success).

I believe there are teams that simply "do what they do" (I remember a couple years ago, pats vs. colts last play of the game mcginist penetrates and stops the run in goal line stand - he said after the game, we studied that on film, we knew what they were going to do in that situation.) And I believe there are teams that save the things that they do best for the right situation. I just think it would be difficult to assess these attributes - one would need to know the objectives of each season, game, play, player; what defines success and what defines failure. I think the best that "outsiders" can do is to apply a set of standardized values for these things. But the value of a game could probably be calculated based on the likely impact of a win/loss to the team's season objective (hopefully that is to win the SB, but what happens when that is no longer a possibility?)

by Jonathan (not verified) :: Wed, 12/21/2005 - 5:47pm

These rankings are complete nonsense. The Giants ahead of the Seahawks?

1) The Seahawks have 2 more wins than the Giants.
2) The Seahawks are on a 10 game winning streak.
3) The Seahawks just annhilated the Eagles who the Giants barely squeaked by.
4) The Seahawks BEAT the Giants head to head.
5) Hasselbeck's QB rating is 20 points higher than Manning.

I understand these rankings are based solely on statistics. Ok, because they make absolutely no sense to anyone who actually FOLLOWS the NFL.

by Rich Conley (not verified) :: Wed, 12/21/2005 - 5:52pm

re 207:

The giants play much tougher competition than the seahawks.

Also, theres a such thing as teams matching up differently, so head to heads are useless. IE saying the seahawks beat the eagles badly, and the giants barely beat them means nothing.

QB rating is retarded.

What it comes down to is the giants have beat better teams, so theyre considered a better team.

by DGL (not verified) :: Wed, 12/21/2005 - 6:03pm

#207: Actually, we already covered Seattle in #32. Please pick another team, reference the zlionsfan template, and try again. Thanks for playing the FO Message Board Curse game.

by ChrisFromNJ (not verified) :: Wed, 12/21/2005 - 6:04pm

The Giants are clearly ranked too low because Eli Manning is a clutch passer with a brother who's god so he's got to be god too and statistics be damned just because he can't throw straight or smart doesnt mean he doesn't WIN GAMES because the QB is always the reason a team WINS GAMES.. ESPN's ranking is way better than this because hold on, they have us lower... um... CBS, because Pete Prisco... no, he has us lower, too. Um, Beatpaths is better because- what you say? they have us nine spots lower??? Yahoo- um, damnit. Sagarin? Anyone?? I give up.

Yeah, Aaron, it's kind of hard to whine when DVOA has had us as a) the best team in the conference and b) an average of two or three spots higher than everyone else out there for most of the season. So damn you and your numbers for trying to lull us into thinking we can make the Super Bowl and making us more dissapointed when Eli throws five picks in our first playoff game. D@mn you and your Bostonian anti-New York bias shenanigans trying to get our hopes up like that. The Sox Sux!

You say you don't get any hate mail from Arizona fans. I'll bet you don't get any from the Giants or Bengals (another DVOA darling), either.

by Mikey (not verified) :: Wed, 12/21/2005 - 6:05pm

MME - Thanks for your response. The special teams factor was not clear from what I had read earlier.

I probably would have been more clear if I had only made the last point in my original post, which was that any analysis (of anything) that combines multiple categories of data has some element of subjectivity in it.

The answer to the question "Who has the most passing yards in the NFL?" is objective fact. The answer to the question "Who is the best QB in the NFL?" can't help but be subjective, even if the method of reaching an answer is run through rigorous checks.

DVOA is interesting BECAUSE it's subjective. It takes an inherently subjective question and answers it with a much greater level of nuance and rigor than we can get from any truly objective reading of stats.

That's all I was getting at.

by ChrisFromNJ (not verified) :: Wed, 12/21/2005 - 6:07pm

Damn, I messed up with the HTML. It should be:

The Giants are clearly ranked too low because Eli Manning is a clutch passer with a brother who’s god so he’s got to be god too and statistics be damned just because he can’t throw straight or smart doesnt mean he doesn’t WIN GAMES because the QB is always the reason a team WINS GAMES. ESPN’s ranking is way better than this because hold on, they have us lower… um… CBS, because Pete Prisco… no, he has us lower, too. Um, Beatpaths is better because- what you say? they have us nine spots lower??? Yahoo- um, damnit. Sagarin? Anyone?? I give up.

Yeah, Aaron, it’s kind of hard to whine when DVOA has had us as a) the best team in the conference and b) an average of two or three spots higher than everyone else out there for most of the season. So damn you and your numbers for trying to lull us into thinking we can make the (redacted) Super Bowl and making us more dissapointed when Eli throws five (redacted) picks in our first playoff game. D@mn you and your (redacted) Bostonian anti-New York bias shenanigans trying to get our (redacted) hopes up like this. The Sox Sux!

You say you don’t get any hate mail from Arizona fans. I’ll bet you don’t get any from the Giants or Bengals (another DVOA darling), either.

by Jay B. (not verified) :: Wed, 12/21/2005 - 6:18pm

Nolan #203: The problem is that no one on this site is only taking into account certian weeks


by Jay B. (not verified) :: Wed, 12/21/2005 - 6:19pm

Argh! Stupid html tags. Is there anything in the works to fix the preview feature, so that what you see is actually what you get?

by Nolan (not verified) :: Wed, 12/21/2005 - 6:20pm

#208 Who are you kidding The Giants are an above average team at best that shows flashes of being real good at times. It is a lot easier to be in real close games when you have the benefit of poor officiating on your side. Lets get real! This is the same rating system that had to fake some of its own numbers to get undefeated Indy in first place instead of 6th ealier in the year. The best thing about this discussion is that everything will come out in the wash won't it

by B (not verified) :: Wed, 12/21/2005 - 6:21pm

213: Nolan thinks the ratings are wrong because they count games that haven't happened yet (Like Indy vs Seattle when Indy will probabally be resting thier starters, even though Dungy hasn't indicated that he will).

by Independent George (not verified) :: Wed, 12/21/2005 - 6:40pm

#212: Hey, when did Chris Simms start posting on these boards? Welcome, Chris - how are your ribs/head/neck/everything?

Seriously, though, I've noticed that Giants and Steelers fans have been consistently saying that their teams are being overrateed by DVOA. I think it's because Steeler fans are haunted by Tommy Maddox, and Giants fans are haunted by the 2000 Super Bowl. At least, I am haunted by the 2000 Super Bowl.

by Jay B. (not verified) :: Wed, 12/21/2005 - 6:42pm

Nolan, if it makes you feel any better about DVOA, when FO does the playoff previews, their analysis will account for "scrub games" where appropriate. For instance, last year when previewing Eagles playoff games, they only considered the Eagles' DVOA through week 15, so their two losses with non-starters playing didn't hurt their rating. Also, when they only have 12 teams to think about instead of 32, they can be a little more granular and attempt to estimate the impact of specific injuries on different teams' chances.

by Cruzer (not verified) :: Wed, 12/21/2005 - 6:55pm

I used to find the weekly power rankings an interesting read however now since it is all mathematically derived I find it is irrelevant and misleading. The subjective element was necessary to adjust for obvious mis-rankings that could arise from simply moving some numbers around.
I understand the concept of looking at the whole season but obviously during the course of a 17 week schedule things change and to include what a team did several weeks ago to calculate what their standing is today can be completely wrong if any one or more of several variable that you do not include in yor calculations have changed. Too bad.

by Nolan (not verified) :: Wed, 12/21/2005 - 7:01pm

My only point is that this system (DVOA) is not accurate at all and the only evidence I need is everyone on this site who are complaining about it! Remeber this is not like college football where the teams are ranked in a poll when all is said and done the only fact the will be needed is who did what in the playoffs, and who won on the superbowl. tell me how many of you guys have had no problem with the power rankings at all this year? the system is a sham it is no better at picking order of power in the NFL than the BCS will tell us what teams are best in CB.

by Falco (not verified) :: Wed, 12/21/2005 - 7:11pm

Just wanted to check back in. I see much has been said since last week, but I bet the knucklehead who was calling for his Chiefs to come to Gillette this January has changed his tune a little

I guess I am the knucklehead? Let's see, I questioned whether the defense and the Pats were really surging when just 3 games earlier the Chiefs had moved the ball very successfully, and then asked Santa for a 5th seed and a trip to Gillette. My tune hasn't changed. I still want those things, they just aren't going to happen after blowing a game against Dallas then yakking on themselves in New York. Let's break down the notes of my tune:

Do- I want my team to make the playoffs (duh);

Ra- I want them to make the playoffs seeded as highly as possible (yes);

Mi- I would rather play the #4 seeded division champ in the playoffs (yes, since NFL switched formats, #4 seed division winners are 3-3 at home in first round of playoffs (1-2 in AFC));

Fa- Given the matchups, I would rather play the Patriots than at either 13-3 Bengals/Broncos in the playoffs;

So- Contrary to a post last week that misunderstood DVOA/home field, DVOA + 17.0% HFA has performed very well SU and ATS in playoffs (complete data in later post when it is in front of me). When HFA adjusted DVOA favors a different team than the point spread, the DVOA pick is 7-3 SU, 7-2-1 ATS in playoffs;
La- Who wouldn't want a chance to knock off the defending champs? Isn't that what it is all about? Oh, and with all the "math sucks", "intangibles", invincibility talk, it happens more often than you might think. Flashback to 1996 FootballOutsiders Message board:

Cowboyzrool: Dallas is clearly ranked too low because they have won 3 of the last 4 Super Bowls and have all the intanjibles, the theory of the Cowboys divine right to continue as champions forever as America's team, is way better than this; who are they going to lose to an expantion team like the Panthers? The Packers! they choke in all the big games with that lose cannon Favre. Please!!!!! You all are a bunch of #$#%#$^, monkey- #%$^#$&, pieces of #$%&$.

Ti- but KC can't tackle Tiki, which will bring us back to . . .


by Jay B. (not verified) :: Wed, 12/21/2005 - 7:14pm

#220 (Nolan):

Evidence? People complaining is evidence that DVOA is not accurate? Um, no.

Even if you choose to accept anecdotal complaints as valid evidence, you must compare the number of complaints to the number of compliments. Or do you believe every single reader must think DVOA's ranking is perfect for it to be of any value?

by Rich Conley (not verified) :: Wed, 12/21/2005 - 7:23pm

Dont be so excited about playing the pats, Falco.

Having Ashworth, Seymour, et al return in the buffalo game changed this to a different team. Its like playing the colts with Josh Sorgi, and playing them with Peyton.

by BigggK (not verified) :: Wed, 12/21/2005 - 7:24pm

Are you kidding!? The bears are ranked lower than the Jags,,,,even though they could not get out of a wet niner bag last week! The facts are Pittsburgh is 9th and the Bears are 10th,,,,then panthers,,,then jags,,,the Bears have every offense and especially QB's shakin tumultously! Who cares if we score more w/D than O! We win!

by Ray (not verified) :: Wed, 12/21/2005 - 8:05pm

Seriously, if anyone thinks that the DVOA power ranking is invalid because people complain about it, then EVERY power ranking is invalid. Have you never read those articles where every author talks about the hoard of hate mail they get over their rankings? It'll happen ALL the time no matter how accurate your rankings are.

Oh, and Falco #221, awesome. The only problem is that I had troubles fitting it to the melody in my head (especailly 'So').

by Falco (not verified) :: Wed, 12/21/2005 - 8:08pm

Re: 223

First, lets get the facts straight. Ashworth and Seymour both started in the game against KC. (source, my memory, and if that is not good enough for you, http://www.nfl.com/gamecenter/gamebook/NFL_20051127_NE@KC). If you went with door number 3, "Dillon did not play", I can accept that, but get your excuses right. And I dont think he plays defense.

KC-420 total yards of offense, 24 first downs, lead 26-3, only punted 3 times (twice in 4th quarter when running clock with lead),1 td, 3 fg in red zone.

Let's play one of these things is not like the other: KC offense, NYJ offense, BUF offense, TB offense. Now let's look at the AFC offenses in the playoffs. Jacksonville as the 5th seed is about as good a matchup as New England could want in that first round, but after that, you better hope Ellis Hobbs and Asante Samuel and the like really have improved, as opposed to the alternative, the pass defense has looked better just because they faced Burke Bollinger, J.D. Losman, and Curt Simms. Oh, and I disagree that the dropoff is the same as Peyton Manning vs Josh Sorgi.

by thad (not verified) :: Wed, 12/21/2005 - 8:44pm

re 208
Why do you think QB ratings are retarded?
They have an excellent correlation with winning.

by B (not verified) :: Wed, 12/21/2005 - 8:53pm

Since the KC game, the Patriots got back Givens, Dillon, Kevin Faulk, Pass (on offense), and moved Ashworth from RT to LT. On defense Seymour, who played in the KC game, is now back to 100%, and Vrabel moved from OLB to ILB, so Beisel and Chad Brown are no longer starting. It might be less than the difference between Peyton Manning and JIM Sorgi, but I think it (plus home field advantage) is enough to make up a 10 point differential.

by zip (not verified) :: Wed, 12/21/2005 - 8:54pm

His name isn't Josh Sorgi!

by B (not verified) :: Wed, 12/21/2005 - 8:57pm

Hmm, my mistake Vrabel was ILB in the KC game.

by BillWallace (not verified) :: Wed, 12/21/2005 - 9:18pm

Re: Chris Simms and Big Games.

I would argue that at Home against the Skins was a bigger game than away at the Patriots at the tail end of a 2-0 road trip. And Simms was unfortunately very very good in that game. very...

by Dan Riley (not verified) :: Wed, 12/21/2005 - 9:21pm

Aaron raised the question, and the Rich-Falco debate has inspired me to answer...a Pats Fan's Wish List of Playoff Opponents from most ardently desired to most awesomely feared:

by Dan Riley (not verified) :: Wed, 12/21/2005 - 9:27pm

On edit, I just realized the Colts have Peyton Manning at QB:

by Paul (not verified) :: Wed, 12/21/2005 - 9:33pm

First of all, Travis you ARE the King. Great work. Now, can you tell us if Home Field is worth more early in the season, or late in the season, or no significant difference?
Second, I can't believe that a Seahawks fan (#207)would try the "compare scores" game with the Giants...Washington 20 Seattle 17; Giants 36 Beloved Skins 0... I can only wish the Skins were playing Seattle this week.

by spenceKarl (not verified) :: Wed, 12/21/2005 - 9:54pm

RE: 227

"They have an excellent correlation with winning."

Correlation does not necessarily indicate causation.

by Travis (not verified) :: Wed, 12/21/2005 - 10:05pm

Re: 234

Thanks for the compliment. To answer your question, road teams do seem to play worse as the season goes on.

NFL road teams since 1977 (start of the 16 game schedule):
Week W-L-T Pct
1 178-212-0 45.6%
2 179-210-1 46.0%
3 151-193-0 43.9%
4 156-193-0 44.7%
5 135-211-2 39.1%
6 157-188-1 45.5%
7 117-228-1 34.0%
8 143-203-1 41.4%
9 157-199-0 44.1%
10 149-211-2 41.4%
11 175-212-1 45.2%
12 146-243-1 37.6%
13 161-228-2 41.4%
14 155-232-1 40.1%
15 158-232-1 40.5%
16 163-228-0 41.7%
17 85-152-0 35.9%
18 6-8-0 42.9%
First Round 30-60-0 33.3%
Divisional 26-82-0 24.1%
Championship 19-35-0 35.2%

Total 2646-3760-14 41.3%
Regular Season 2571-3583-14 41.8%
Playoffs 75-177-0 29.8%

1-4 664-808-1 45.1%
5-8 552-830-5 40.0%
9-12 627-865-4 42.0%
13+ 728-1080-4 40.3%
Playoffs 75-177-0 29.8%

by Travis (not verified) :: Wed, 12/21/2005 - 10:18pm

One minor correction: the 16 game schedule began in 1978, not 1977.

The road win statistics for 2000-2004 are more dramatic:

Week W-L-T Pct
1 35-43-0 44.9%
2 37-40-0 48.1%
3 33-37-0 47.1%
4 32-39-0 45.1%
5 32-38-0 45.7%
6 34-35-0 49.3%
7 27-43-0 38.6%
8 26-44-0 37.1%
9 35-36-0 49.3%
10 32-39-1 45.1%
11 34-44-0 43.6%
12 33-45-0 42.3%
13 31-47-0 39.7%
14 31-47-0 39.7%
15 27-51-0 34.6%
16 34-44-0 43.6%
17 28-50-0 35.9%

by Paul (not verified) :: Wed, 12/21/2005 - 10:26pm

Wow, I type a sort of tongue in cheek request, not really expecting results, go tuck in my kids, come back and there it is! Looks to me like the trend reflects a few teams packing up early every year...

by Jon (not verified) :: Wed, 12/21/2005 - 11:22pm

Not sure where to ask this , but were the RB stat pages updated? The yardage totals don't seem to include last week's games. Thanks.

by Keith Cockrell (not verified) :: Wed, 12/21/2005 - 11:23pm

The Dallas Cowboys are clearly ranked too low because they’re my favorite team. My ranking system in which the Cowboys are always at the top is way better than this. The Cowboys are better because they …. Uh …. Because they …. Uh …Nevr mind. I think I meaned to sin this n last weak. (And if there weren’t already 200+ posts so that this will end up getting mainly ignored it wouldn't be sent it this week. Darn those Redskins.)

By the way, how do you guys get emphasis on certain words when you post?

by DGL (not verified) :: Wed, 12/21/2005 - 11:26pm

#219: ...obviously during the course of a 17 week schedule things change and to include what a team did several weeks ago to calculate what their standing is today can be completely wrong if any one or more of several variable that you do not include in yor calculations have changed.

Similarly, to not include what a team did several weeks ago, or to weight too highly what they did recently, can be completely wrong. Anything can be completely wrong, but DVOA has been shown to correlate rather well with team performance. I can't find data more recent than VOA V4.0 (introduced in May, 2004), but with that version, the correlation coefficient between VOA and wins for the 2001-2003 season is 0.861, and every time Aaron updates the formulas, he tests to ensure that the new version gives at least as good correlation across a number of variables including wins, year-to-year performance, and point differential.

The point is that any power rankings will fail to take into consideration something. If you prefer purely subjective power rankings, read Dr Z, Peter King, or Prisco. If you prefer the committee approach to subjectivity, read ESPN. The proponents of DVOA prefer a more mathematical approach.

by DGL (not verified) :: Wed, 12/21/2005 - 11:37pm

#220: ...this system (DVOA) is not accurate at all and the only evidence I need is everyone on this site who are complaining about it!

That evidence proves nothing except that people disagree with what DVOA says. To use that as evidence that DVOA is not accurate requires the assumption that what the complainers are saying is accurate. Like the Cowboys fans last week who were complaining that DVOA was wrong about the Cowboys. And the Falcons fans after Week 10 who were positive that DVOA was wrong about the Falcons. We've seen how that turned out.

...when all is said and done the only fact the will be needed is who did what in the playoffs, and who won on the superbowl.

OK. Let's all petition Fox, ESPN, SI, and CBS Sportsline to take down their power rankings, because they're all obviously unnecessary.

And by the way, you forgot to use the zlionsfan template.

Every team is clearly ranked too high or too low because there are a bunch of people on this site who are complaining about DVOA. Waiting until the Super Bowl is way better than this. the system is a sham it is no better at picking order of power in the NFL than the BCS will tell us what teams are best in CB.

by admin :: Thu, 12/22/2005 - 1:27am

Hey. Sorry I was out of commission for the evening. I guess I forgot to update the RB sheet, so I'll do that tomorrow. Benjy added FO Mailbags to the ARCHIVES drop down menu above. I think I mentioned the "home field advantage gets stronger late in the season" thing in one of my BP chats earlier this year ... Two Minute Warning also did research that showed this to be true. The next question would be, is this true for all teams, or is the effect of weather and the advantage of northern teams helping HFA numbers for the whole league in November and December?

Here's the thing about the complaints about the system: At the beginning, they were distressing to read. Was I just not explaining things enough to people? Was the system greatly flawed?

Gradually, however, I began to realize something. All the hateful e-mails, all the troll comments, only roughly one out of 200 complains about the system without complaining about a specific team. In reality, there are very few people who think DVOA doesn't work if DVOA has their team higher than the other power ratings on the Web. That's all people care about and that's why all that hate mail people send me just gets laughed at by this point. The actual, intelligent readers of this site offer valid, interesting ideas about how to improve the numbers. That's what I care about.

Although the Seattle obsession with time zones is really getting overwhelming. Seriously, is that entire city paranoid?

Oh, and I should add that the folks at FOXSports.com are very happy with the relationship with FO, so they don't take the troll complaints seriously either.

by Fnor (not verified) :: Thu, 12/22/2005 - 1:43am

#233: I wouldn't be too happy as a Pats fan to draw Pittsburgh's number. Bellichik can't turn the field into goo, because that would make Bettis jump for joy. Plus, it would enable them to avoid the aforementioned Heinz Field AFC Championship Game Curse. Plus, we're better this year. Finally.

Re: QB rating
Part correlation, part causation. Obviously, throwing TDs and not throwing INTs will help you win (or so announcers tell me, about 20 times per game). However, the values associated in this model with these two numbers and completition percentage are, in the words of a poster whose idetity eludes me, "wiggedy-wack." INTs are bad, sure, but is Favre's wannabe punt as bad as Maddox's OT meltdown? What if it's a tipped ball? Is a QB that never throws an INT but fumbles every 5 snaps really better than one that never fumbles and throws an INT every game? These are serious problems with the system, and you see a clump of people at the top and at the bottom with NO ONE in the middle, because you either score a boatload of TDs with few INTs or you don't score many TDs and therefore suck hardcore. It's not very useful.

#217: In the words of the great scarved one: "Ah, my arms! Agh, my legs! ARGH! My everything!"

MANDATORY BONUS FAMILY TIME FUN: The winning question tonight at trivial pursuit for me was "What sci-fi show was able to run for 26 seasons after the BBC realized it could "regenerate" its title character?"

It was a thing of beauty.

by Jay B. (not verified) :: Thu, 12/22/2005 - 1:45am

Aaron, I imagine you've also gotten some hatemail from more realistic fans of bad teams, who instead of blasting DVOA's low rating of their teams, instead blast it for rating rival teams "too" high.

Am I right?

by Thok (not verified) :: Thu, 12/22/2005 - 2:07am

San Francisco is clearly ranked too high because we couldn't score on a Pop Warner team even if they fumbled in their endzone. Your momma throwing darts is way better than this. You'll see how much we suck when we draft Reggie Bush, clone him 16 times, and still lose 42-0 .

You know, this is a great way to vent about really sucky teams.

by Andrew (not verified) :: Thu, 12/22/2005 - 2:10am

On the playoffs.

The 6 seeds do not have an equal chance in the playoffs. Roughly, home teams win 70% of wild card games, 80% of divisional games, and 60% of championship games. #5 and 6 seeds have almost no chance.

In the 25 years since the 16 game schedule in 1978, and the start of the wild card round:

In the NFC, the #1 seed made the Super Bowl 15 times, the #2 seed 8 times, the #3 seed twice, the #4, 5, and 6 seeds never.

In the AFC, the #1 seed made the Super Bowl 11 times, the #2 seed 8 times, the #3 seed never, the #4 seed 5 times, the #5 seed once, the #6 seed never.

In 15 years since 1990, with the 12 team format:

In the NFC, the #1 seed made the Super Bowl 9 times, the #2 seed 5 times, the #3 seed once, the #4, 5, and 6 seeds never.

In the AFC, the #1 seed made the Super Bowl 6 times, the #2 seed 5 times, the #3 seed never, the #4 seed 4 times, the #5 and 6 seeds never.

But what is the best way to predict Super Bowl winners? Leading team in Pythagorean Wins. This has worked 10 out of 15 times since 1990, with the misses being some of the biggest upsets in NFL history: the 2001 Rams, the 1998 Vikings, the 1995 and 1992 49ers, and the 1990 Bills. Obviously, that indicator is predicting Colts this year. And when we consider just the two teams that made the Super Bowl, the leader in Pythagorean wins has won 13 out of 15 times since 1990, with the only two misses being games decided by a Field Goal - 1990 Bills and the 2001 Rams. Could you ask for a surer proposition for a straight-up bet than something that is 13 for 15?

by scarface (not verified) :: Thu, 12/22/2005 - 2:39am

2 Questions for Aaron and FO about the fact that the entire season's performance is reflected in DVOA, and the "weighting" of the most recent 8 games is pretty high:

1. How did you arrive at the number "8"? Was that subjective? Or did you try a whole bunch of numbers for predictive value, and 8 worked best?

2.While I get your reason for using the whole season's performance in DVOA in the manner you do, and while that might be the best predictive system ON AVERAGE, is there some way to specifically account for injuries/recoveries which might dramatically skew a team's performance from last week to this week? After all, our goal is PREDICTIVE value CURRENTLY. Clearly, this only makes a difference in special cases, and not the average case: I'm talking about some case where three stars have returned to the team in good shape, or Tommy Maddux happens. I realize this might be easier to do with subjective tinkering...or maybe you could adjust the weighting of recent games based on how "similar" the team was then and now. (Of course, one would have to define "similar": something reflecting the difference in the playing roster, just in shere number of players, or sum of individual values above replacement or something like that.)

Sorry if you answered this somewhere, I've spent a lot of hours on this site and over the last year and haven't found it:)

by spenceKarl (not verified) :: Thu, 12/22/2005 - 9:40am

RE: 243 (Aaron)

Wow, that was fast. Thanks for throwing the Mailbags into their own archive page.

by Larryz (not verified) :: Thu, 12/22/2005 - 9:51am

#247 should be really encouraging to Seahawks fans (among others). The one anomaly that jumps out: who are these #4 AFC seeds whose success has been completely off the map, outperforming AFC #3 and NFC #3 and #4 by a huge margin in both the post-1978 and post-1990 stats? Perhaps the Patriots have been secretly planning all along to end up as #4 seed?

by big_adventure (not verified) :: Thu, 12/22/2005 - 10:30am

243 -

Aaron, you don know that Seattle has the highest suicide rate in the country, right? Perhaps it's all the persecution anxiety they feel thanks to the dastardly FO Anti-West-Coast bias!

Or, in proper format:

Seattle is clearly ranked too low because Aaron hates rain and my dentist is jewish and from New York and he says the Seahawks are better than Tom Brady and Eli Manning all rolled up into one I dont care if Payton Maning playz just one qorter the Hawks will kill him and eat their children with Shon Alexandar and Walter jones suicide is way better than this. Soon as im finisht with my tripple banana-mocha-kleenex frappacino, Im gonna right a sucide note, reed it at Starbicks and jump off the space kneedel.

by Falco (not verified) :: Thu, 12/22/2005 - 10:57am

Re: 250

Keep in mind that these seedings represent the following: prior to 2002, the #3 seed was the worst of the three division winners, and the #4 seed was the best wildcard team and got a home game (usually playing in the same division as the #1 or #2 seed); since 2002, the #4 seed is the worst division winner, and the #5 seed is the best wildcard team and has to play on the road.

The four #4 seed wildcards from AFC since 1990 to make Super Bowl were Baltimore (2000) who was second in total DVOA behind the #1 seed from its own division, Tennessee; Tennessee (1999), third in total DVOA behind the #1 seed from its own division, Jacksonville, as well as St. Louis; Denver (1997), which when those numbers are done will likely be very high in DVOA, and finished behind KC in the regular season with the second best record in AFC; and Buffalo in the year of the greatest comeback of all time.

I think looking at the #4 seed and comparing to this year's NE is false hope. You could phrase it a different way--since 1990, no lowest seeded division winner has made the Superbowl, and only Carolina (#3 out of 4) made it as a division winner who did not have a first round bye.

by spenceKarl (not verified) :: Thu, 12/22/2005 - 11:06am

RE: 248 (scarface)

I'm sure someone else could probably answer your questions with more detail, but for now, I'll refer you to two links as a quick response.

For the first question: Aaron changed the values of the weights earlier this season. It's not subjective. I believe it used to be though. He talked about them briefly in the Week 9 Ratings.

For the second question: DVOA does not account for injuries. But a similar question was addressed in a recent FO Mailbag.

Someone else might be aware of other links that address your questions, but those are the most recent that I recall anyways.

by Ricky (not verified) :: Thu, 12/22/2005 - 11:49am

Aaron, I generally pay no attention to "power rankings" as in general I think it is a pretty stupid concept. I like to think that I can identify the good, the bad, and the ugly in the NFL without necessarily giving them a numerical order. However, as a Pats fan, I don't disagree with where your rankings have had them listed considering their performances over the year. Also, I appreciate how your approach has a firm formula behind it rather than a paragraph explaining, "Cadillac is rolling in Tampa," or something ridiculous like that.

As for the Chiefs fan, the Pats play an extremely team oriented defense. A large part of what they do depends on continuity and famialiarity with one another on the field as well as the overall scheme. Don't dismiss this. If you listen to Belicheck, Seymour, Colvin, Harrison, Bruschi, McGinest, Vrabel, et al., they all speak to how important this aspect of what they do is. I had nothing but positive things to say about the Pats defense after that loss to the Chiefs (despite the total yards and passing yards, etc.) because it was the beginning of their road back. That loss was primarily on the offense, and it the only time I could ever say that Brady's play was a major reason they lost (even though his starting offense was comepletely depleted). I hate to sound like I am continually making excuses because the bottom line is the Pats got beat in KC, but my point is that this a different team than it was then. Seymour's healthy and Wilfork and the rest of the line have fallen in line as a result. The linebackers, the lifeblood of Bill's defenses, are now fully healthy (colvin and bruschi) and adjusted to some changes (Vrabel) and they are playing liek they did all last year. The secondary hasn't been truly tested quite yet, but they are playing better and Ellis Hobbs is going to be a good corner for a long time, he has that make-up and potential. The safety play has been the most inconsistent on the team, but Atrell Hawkins is physical and falling into line and Eugene Wilson is proven. Prior to this stretch starting against the Jets, the Pats had six different starting strong safetys to begin the year. That's insane in this type of defense. Anyways, sorry for how long and probably obvious this is, but I'm bored at work.

by Andrew (not verified) :: Thu, 12/22/2005 - 1:04pm

Falco (and LarryZ):

Keep in mind that these seedings represent the following: prior to 2002, the #3 seed was the worst of the three division winners, and the #4 seed was the best wildcard team and got a home game (usually playing in the same division as the #1 or #2 seed); since 2002, the #4 seed is the worst division winner, and the #5 seed is the best wildcard team and has to play on the road.

The 1992 Bills and 1997 Broncos, would have been #5 seeds in the current playoff format. The 1999 Titans would have been #5 or #6 seed. The 2000 Ravens would have been a #2 or #3 seed. The 2003 Panthers would have been a #4 seed before 2002.

Part of the poor showing by NFC #3 seeds comes from an inexplicable strength for upsets by NFC #6 seeds (8 victories - 1991, 1992, 1993, 1994, 1997, 1998, 2002, 2004). #6 seeds tending to advance then means that #4 NFC squads often face the #2 seed, while the #1 NFC squad often faces an extremely weak #6 opponent, so both #1 and #2 NFC teams end up with a better chance of advancement than if they faced the #4 and #3 teams respectively.

In the AFC, the #6 seeds almost invariably lose, with just two upsets in 15 years (1995, 2004).

#5 seeds in both conferences have advanced one round 4 times. NFC (1990, 1991, 1992, 2004). AFC (1996, 1999, 2001, 2003).

I think looking at the #4 seed and comparing to this year’s NE is false hope. You could phrase it a different way–since 1990, no lowest seeded division winner has made the Superbowl, and only Carolina (#3 out of 4) made it as a division winner who did not have a first round bye.

However, I think that is missing the point. In the AFC, 4 teams in 15 years have gone from having to play in the wild card round at home (no bye), then winning two games on the road to make the Super Bowl. Carolina then did this in 2003 in the new format, to make it 5 teams overall in the NFL in 15 years.

That 1 of 4 of the #3 and #4 seeds makes the Super Bowl is a 1:3 proposition over 15 years, meaning any one of those teams has a theoretical 1:12 shot.

To summarize, 15 of 30 #1's have made it over 15 years (1992 was the only year at least one #1 didn't make it to the Super Bowl, 1993 was the only year both did), 10 of 30 #2's, and 5 of 60 #3's and #4's.

by admin :: Thu, 12/22/2005 - 1:23pm

RB page finally correct. Why yes, that is Larry Johnson on top.

by B (not verified) :: Thu, 12/22/2005 - 1:38pm

As a Patriots fan, I'll take those odds. I feel confident that right now they're playing at a level that's on par with the 2003 Panthers, at least. And if they don't make it, at least they'll go out in a blaze of glory at Indy or Denver.

by Chucky (not verified) :: Thu, 12/22/2005 - 1:52pm

Random question from someone who often reads and rarely posts:

Is there an FO-approved adjustment from DVOA differential to actual NFL points? VarlosZ (#44) seems to think so, but I don't remember seeing it.

I know there's an established PAR relationship... just not sure if there's one for the team stats.

by spenceKarl (not verified) :: Thu, 12/22/2005 - 2:04pm

I wondered the same thing back in #121, but inquired with less eloquence.

by Falco (not verified) :: Thu, 12/22/2005 - 2:19pm

Re: 254, I have no problem with any of those comments, NE is likely better now than at week 8, just like the Chiefs are better now than when they lost to Buffalo without Willie Roaf, and Green was hit on every pass. I have never said NE sucks, or the result would be the same. NE this year is not NE 2003 or NE 2004 with the losses that are not coming back this season. The AFC also a tougher conference up top, and DVOA confirms that with IND, DEN, and CIN all above 30.0% in total DVOA and 10.0% in total DVOA - variance.

Re: 255, the difference is Baltimore was #2 DVOA as a wildcard having to win two road games, Tennessee was #3 DVOA as a wildcard, and I suspect when 1997 is calculated Denver will be top 3. They won on the road but were the close to or the better team in those road games. However, the new format certainly does make it harder for a wildcard to make the Superbowl, having to now win 3 straight.

Which segues to earlier, when I said I would post DVOA playoff data. Here it is.

First, methodology: I used the total DVOA numbers from 1998-2004 that are available (no weighted available yet), and added 17.0% to the home team DVOA across the board. Superbowl was straight up DVOA of each team on neutral field. I converted to point spread using 17.0%=3.4 pts to calculate DVOA point spread. I then compared to point spread and final results as published by the Goldsheet. I include the point spread numbers because that tells us how DVOA is doing compared to public perception. If DVOA was less than 0.5 points away from actual point spread, I did not consider it in the point spread numbers.

Straight up picking the winner:
56-21 in all playoff games
20-8 in wild card round
24-4 in semifinals
7-7 in championship round
5-2 in Super Bowl

Against the Spread:
42-25-3 (7 No Plays) in all playoff games (62.7%)
13-12-1 (2 No Plays) in wild card round
14-9-1 (4 No Plays) in semifinals
10-3 (1 No Play) in championship round
5-1-1 in Super Bowl

Compare that to just taking the "better team" as measured by higher seed (best record)
55-22 Straight Up
39-35-3 ATS (52.7%)

In games where DVOA had a 2.76 pt or more (FG or more) difference compared to the point spread, the ATS record is 27-15-3, and excluding Denver 1998 (see comments below), it is 27-12-3 ATS. That's a pretty good percentage unless your name is Stu Feiner and you claim to win 90%. In games where DVOA actually disagreed with the favorite to win, irrespective of point spread, DVOA was 7-3, 7-2-1 ATS, and 7-2, 7-1-1 ATS excluding Denver v. Atlanta.

A few comments: Denver 1998 is clearly misrepresented by total DVOA, and is a word of warning regarding this year's Colts. That team shut it down after losing to the Giants, and DVOA picked against them using only total DVOA in all 3 games against the spread, and even had ATL as a slight favorite. Using only the games where Denver played to win that year, I highly suspect Denver was the top team by DVOA, rather than 3rd behind NYJ and ATL.

Second, while the highest seed was only 7-7 in conference championship games, DVOA was quite successful in picking underdogs that covered the spread vs those that did not in that round.

by Travis (not verified) :: Thu, 12/22/2005 - 2:34pm

Re: 243

Warm weather teams do seem to play worse at cold weather teams as the season goes on, but not vice versa. Similarly (see next post), Mountain/Pacific teams play worse at Eastern/Central teams as the season goes on, but not vice versa.

Warm/Dome: ARI (1988-), ATL, CAR. DAL, DET, HOU, IND, JAX, MIA, MIN (1982-), NO, oAK (includes L.A. Raiders), SD, SF, SEA, STL (includes L.A. Rams), TB, TEN (includes Houston Oilers)
Cold: BAL (both Colts and Ravens), BUF, CHI, CIN, CLE, DEN, GB, KC, MIN (1978-81), NE, NYG, NYJ, PHI, PIT, St.L. Cardinals (1978-87), WAS

Since 1978:
All Road Games
W-L-T Pct
Aug/Sep 646-817-1 44.2%
Oct 631-890-5 41.5%
Nov 679-949-6 41.7%
Dec/Jan 615-927-2 39.9%
Playoffs 75-177-0 29.8%
Total 2646-3760-14 41.3%
Reg. Season 2571-3583-14 41.8%

Warm/Dome at Cold
Aug/Sep 186-224-0 45.4%
Oct 136-238-0 36.4%
Nov 147-248-1 37.2%
Dec/Jan 132-219-0 37.6%
Playoffs 19-58-0 24.7%
Total 620-987-1 38.6%
Reg. Season 601-929-1 39.3%

Warm/Dome teams away (all)
Aug/Sep 357-441-0 44.7%
Oct 323-475-1 40.5%
Nov 329-495-1 39.9%
Dec/Jan 307-457-0 40.2%
Playoffs 41-99-0 29.3%
Total 1357-1967-2 40.8%
Reg. Season 1316-1868-2 41.3%

Cold at Warm/Dome
Aug/Sep 137-184-0 42.7%
Oct 152-197-2 43.6%
Nov 180-249-1 42.0%
Dec/Jan 184-250-0 42.4%
Playoffs 18-45-0 28.6%
Total 671-925-3 42.1%
Reg. Season 653-880-3 42.6%

Cold teams away (all)
Aug/Sep 289-376-1 43.5%
Oct 308-415-4 42.6%
Nov 350-454-5 43.6%
Dec/Jan 308-470-2 39.6%
Playoffs 34-78-0 30.4%
Total 1289-1793-12 41.9%
Reg. Season 1255-1715-12 42.3%

All Road Games
W-L-T Pct
Aug/Sep 123-143-0 46.2%
Oct 135-173-0 43.8%
Nov 141-177-1 44.4%
Dec/Jan 142-229-0 38.3%
Playoffs 18-42-0 30.0%
Total 559-764-1 42.3%
Reg. Season 541-722-1 42.8%

Warm/Dome at Cold
Aug/Sep 32-33-0 49.2%
Oct 26-45-0 36.6%
Nov 22-48-1 31.7%
Dec/Jan 28-62-0 31.1%
Playoffs 6-14-0 30.0%
Total 114-202-1 36.1%
Reg. Season 108-188-1 36.5%

Warm teams away (all)
Aug/Sep 68-76-0 47.2%
Oct 72-97-0 42.6%
Nov 59-103-1 36.5%
Dec/Jan 71-133-0 34.8%
Playoffs 11-26-0 29.7%
Total 281-435-1 39.3%
Reg. Season 270-409-1 39.8%

Cold at Warm/Dome
Aug/Sep 26-35-0 42.6%
Oct 29-35-0 45.3%
Nov 43-46-0 48.3%
Dec/Jan 39-46-0 45.9%
Playoffs 4-10-0 28.6%
Total 141-172-0 45.0%
Reg. Season 137-162-0 45.8%

Cold teams away (all)
Aug/Sep 55-67-0 45.1%
Oct 63-76-0 45.3%
Nov 82-74-0 52.6%
Dec/Jan 71-96-0 42.5%
Playoffs 7-16-0 30.4%
Total 278-329-0 45.8%
Reg. Season 271-313-0 46.4%

by Travis (not verified) :: Thu, 12/22/2005 - 2:35pm

Since 1978:
Eastern/Central at Mountain/Pacific
Aug/Sep 59-106-0 35.8%
Oct 59-107-0 35.5%
Nov 72-125-0 36.5%
Dec/Jan 70-124-0 36.1%
Playoffs 15-37-0 28.8%
Total 275-499-0 35.5%
Reg. Season 260-462-0 36.0%

Eastern/Central away (all)
Aug/Sep 519-675-0 43.5%
Oct 499-747-4 40.1%
Nov 556-812-6 40.7%
Dec/Jan 511-778-2 39.7%
Playoffs 65-148-0 30.5%
Total 2150-3160-12 40.5%
Reg. Season 2085-3012-12 40.9%

Mountain/Pacific at Eastern/Central
Aug/Sep 99-93-1 51.6%
Oct 96-97-1 49.7%
Nov 83-85-0 49.4%
Dec/Jan 70-95-0 42.4%
Playoffs 9-25-0 26.5%
Total 357-395-2 47.5%
Reg. Season 348-370-2 48.5%

Mountain/Pacific away (all)
Aug/Sep 127-142-1 47.2%
Oct 132-143-1 48.0%
Nov 123-137-0 47.3%
Dec/Jan 104-149-0 41.1%
Playoffs 10-29-0 25.6%
Total 496-600-2 45.3%
Reg. Season 486-571-2 46.0%

Eastern/Central at Mountain/Pacific
Aug/Sep 12-22-0 35.3%
Oct 12-20-0 37.5%
Nov 16-20-0 44.4%
Dec/Jan 15-26-0 36.6%
Playoffs 4-5-0 44.4%
Total 59-93-0 38.8%
Reg. Season 55-88-0 38.5%

Eastern/Central away (all)
Aug/Sep 104-121-0 46.2%
Oct 116-137-0 45.8%
Nov 120-154-1 43.8%
Dec/Jan 123-188-0 39.5%
Playoffs 18-35-0 34.0%
Total 481-635-1 43.1%
Reg. Season 463-600-1 43.6%

Mountain/Pacific at Eastern/Central
Aug/Sep 15-15-0 50.0%
Oct 13-30-0 30.2%
Nov 13-14-0 48.1%
Dec/Jan 13-30-0 30.2%
Playoffs 0-7-0 0.0%
Total 54-96-0 36.0%
Reg. Season 54-89-0 37.8%

Mountain/Pacific away (all)
Aug/Sep 19-22-0 46.3%
Oct 19-36-0 34.5%
Nov 21-23-0 47.7%
Dec/Jan 19-41-0 31.7%
Playoffs 0-7-0 0.0%
Total 78-129-0 37.7%
Reg. Season 78-122-0 39.0%

by Andrew (not verified) :: Thu, 12/22/2005 - 2:43pm

Ricky #254:

and it the only time I could ever say that Brady’s play was a major reason they lost (even though his starting offense was comepletely depleted).

Brady was the sole reason for the loss in Miami in 2004, where he literally threw the game away in the final minutes through two extremely poor decisions, some of the worst outcome decisions I had ever seen a skilled QB not named Drew Bledsoe or Brett Farve make (right up there with McNabb's TD throw to Roy Williams this year, and the Tommy Gun-Jacksonville game).

by spenceKarl (not verified) :: Thu, 12/22/2005 - 2:43pm

RE: 261 (Falco)

"I converted to point spread using 17.0%=3.4 pts to calculate DVOA point spread."

Where did you come up with that? And how has it been validated?

by Ricky (not verified) :: Thu, 12/22/2005 - 2:48pm

That's fair Andrew. So two insignifcant losses versus three super bowl wins which he most directly won. Not a bad ratio.

by Jon (not verified) :: Thu, 12/22/2005 - 3:15pm


Just asking for a clarification: are the stats at the bottom for the team with the better seed or the team with the better record? Wild card teams have occasionally had more wins than division winners (e.g., 10-6 Indy at 9-7 Jets in 2002). Also, the Ravens and Giants were both 12-4 when they played in Super Bowl XXXV. Finally, how often did favorites win over the same period?

by Andrew (not verified) :: Thu, 12/22/2005 - 4:00pm

Ricky #265:

Those two examples are hardly the only time Brady threw the game away. They are the most glaring recently.

by Falco (not verified) :: Thu, 12/22/2005 - 4:03pm

SpenceKarl, Aaron earlier posted that home field advantage was roughly 17.0%, and also said that equated with a number that was somewhere between 3.25 and 3.40 pts. For ease of calculation, I rounded to 3.4, thus, every 1.0% = 0.2 points. Sorry, I cant remember when that was posted.

Jon, sorry, you are right. I should have just said, higher seed/home team. I didnt compare every record to assure that the home team had the better or equal record. As for the question about favorites, I dont have it in front of me now, but its not going to change a whole lot, because the vast majority of higher seeds were also favored. Off the top of my head, last year, Pittsburgh was a home underdog to NE and lost, and in 2000, PHI was a home underdog vs TB and won, NO was a home underdog vs STL and won, and Miami was a home dog to Indy and won, all in the first round. I may be forgetting a couple. DVOA, incidentally, had PHI and MIA as the favorite to win, and NO/STL as a pickem.

by spenceKarl (not verified) :: Thu, 12/22/2005 - 4:30pm

RE: 268

Thanks. I'd like to hear from Aaron then about where the 3.25 to 3.40 range comes from.

by Ricky (not verified) :: Thu, 12/22/2005 - 4:42pm

Andrew, you are honestly saying that Tom Brady loses games for the Patriots? Have you ever watched the Patriots? If so, you just don't understand football. There is nothing more to say about that.

by Jacob Stevens (not verified) :: Thu, 12/22/2005 - 4:59pm

I read the home-field advantage stuff, but it doesn't look like the results are indicative of anything substantial to me. And the results are incomplete at best.

Sure, you're measuring DVOA, so you are taking the strength of the opponent into consideration, but by "ranking" home-field advantage for 2005 based on the difference of the teams' performance between home and road games, you are not considering the difference of the opposing teams' performance, between their own home games, and their games at any other venue.

Furthermore, the reason there seems to be no consistency whatsoever for measuring home-field advantage is because you're measuring things that really are determined by many other factors than home-field advantage. You're pointing to San Francisco as having the best because their performance differential was the greatest. But the weaker teams (San Francisco, currently #32 DVOA) have more room for improvement in their performance, particularly in a system that measures teams' performance compared to a floating league average. I'm sure you would point to San Francisco's #1 rank in variance as evidence of their difference in performance between home and road games, but I would point to their great variance as a characteristic of their performance in general that cannot be accounted for in the system you've used to measure home-field advantage.

The more consistent teams will not show a marked improvement when playing at home. Better teams breed consistency. Thus you haven't adequately measured the value of teams' home-field advantage.

Even taking a larger sample, of 10 years, as you said, to more accurately measure home-field advantage, would not be a better-controlled measurement because it wouldn't take into account the coaching and personnel changes.

But I don't know of anything that is a more accurate measurement of home-field advantage. So, without adjusting for the strength of the opponent, or anything else, I'd still go back and say a teams' W-L record is still a great way to measure home-field advantage, and in that case I know that Seattle and New England are still two of the best around.

by Starshatterer (not verified) :: Thu, 12/22/2005 - 5:01pm

Ricky (#270 )--

Try to forgive Andrew -- he has Mike McMahon losing games for his team this year, so he's not his usual chipper self. ;-)

And he does have a point -- Brady has completely tossed away a game or so, pretty much every year he's started. O'course, he's also pulled more than a few wins out of thin air, so it's easy to exaggerate this impact of his annual pick-fest on New England's fortunes. To date, he hasn't coughed up a hairball like that in the playoffs, so he's either super-clutch or he's due, depending on what team you root for.

Regarding playoff seeding and the Patriots, I think they have an advantage this year most #3-4 seeds do not: they're locked in with two weeks (the exact seed may not be locked, but they're 100% certain to host a wild-card-round game).

So expect to see a lot of the Doug Flutie Eperience, the Heath Evans Experience, the Dan Klecko Experience, and so on. They won't be as fresh, rested, and injury-free as they would be if they had a playoff bye, but they'll likely be as fresh, rested, and injury-free as anyone playing in the wild card round can be.

by Catfish (not verified) :: Thu, 12/22/2005 - 5:07pm

Re: 270

Andrew wasn't saying that Tom Brady always loses games for the Pats, just that he has in the past. But this isn't an insult. I'm pretty sure every quarterback who has been a starter for a decent period has thrown away a game. Tom Brady, Brett Favre, and Peyton Manning have all lost games for their team, but that doesn't mean that they are bad quarterbacks. They've all obviously caused far more victories than losses.

by Ricky (not verified) :: Thu, 12/22/2005 - 5:22pm

I still completely disagree, though. Andrew said "hardly" and to me it implied there were many more. In 03-04 and 04-05, the Pats lost a total of four games out of a possible 38. One of those was the Miami game. How does that constitute "hardly"? In 2002-03, the Pats went 9-7 and missed the playoffs, but statistically, that was Brady's best year (excluding this year). I see what you guys are saying, I know QB's are going to throw picks and that sometimes they are going to make decisions that wind up losing the game, that's the nature of the position; however, I'd say Brady has done this far less than any qb in the league since he took over for Drew Bledsoe.

by C Mu (not verified) :: Thu, 12/22/2005 - 5:28pm

I'm offended for all hippies...a Seahawks Fan who lives in Mexico!

by Rich Conley (not verified) :: Thu, 12/22/2005 - 6:43pm


There have been 5 or 6 games in the last 4 years that brady has lost for the patriots, and then the Defence has done something to win the game.

I'm a staunch NE fan, and brady is great, but he typically has 2 or so games a season where he comes out looking like drew bledsoe... IE too much anti-histamine or something... that glazed eye look, and throws 3 or 4 picks, or just can't hit open recievers for his life.

by jonnyblazin (not verified) :: Thu, 12/22/2005 - 6:46pm

Anyone else curious as to how one game this late in the season against a lower eschalon team can change the Ravens ranking as much as it did? While I would certainly agree with Aaron that they aren't as good as their ranking suggests, doesn't that conflict with the stomps/guts theory that teams who whip bad teams to inflate their dvoa are legitimately as good as teams who hang with good teams? Its also problematic to say that the game was a fluke because GB's secondary quit, it is a lot harder for me to swallow a secondary quitting as opposed to a front 7, who will generally 'quit' because of a smashmouth rushing attack and having nothing to play for.
Also, looking at the individual breakdowns, the Ravens offense moved from being ranked 31st to 29th and their defense moved from 15th to 8th. What this tells me is that their offense still stinks and their defensive dvoa is finally in line with where I expected it to be the entire season. I think this has something to do with the fact that they finally got a few INTs in a game. Somehow I think that INTs are being somewhat overated in defensive dvoa, there might be a luck factor involved that has allowed teams with mediocre defenses to inflate their numbers when an errant pass happens to be picked off instead of falling incomplete. Last year the Ravens had plenty of picks, and this year they upgraded their secondary and have fewer.

by Falco (not verified) :: Thu, 12/22/2005 - 6:50pm

Re: 271

You make some good points, and some not so good points. And I dont think the conclusion is supported at all.

Justice Potter Stewart, unable to define obscenity with a hard line rule, once said "I know it when I see it."

I know home field advantage exists in football, it is just hard to define on a single game basis. Your point about other variables is true, in fact, I suspect that San Fransisco's outrageous number is due to other variables as well (Simms' first start; cross-country travel to PHI and WAS; Alex Smith's spectacularly bad performances and random chance as to when he will have only 3 turnovers vs 6). However, to make some of the broad statements, you would have to believe all statistical analysis is worthless. Over a large enough sample size, other variables would even out (home team starting QB hurt vs road team starting QB hurt; advantageous matchup for home team vs advantageous matchup for road team).

I do notice one trend in these numbers:

The top 8 teams in home field difference are all "defensive teams," that is teams with a better relative defensive DVOA than offensive DVOA. In fact, the top 4 in difference between D & O-- usually due to horrific offenses--appear here (Chicago, Baltimore, San Fransisco, NY Jets). Conversely, of the 7 teams with a better road performance to date, 5 have better relative offenses than defenses (Cin, Hou, Ind, SD, ATL).

by TomC (not verified) :: Thu, 12/22/2005 - 6:56pm

Ah crap. After #224, I can't make any more smug comments about Bears fans' being of higher intelligence and class than the rest of the world.

I guess I could claim that it was a joke, and he just forgot to use the template...

by Jacob Stevens (not verified) :: Thu, 12/22/2005 - 8:43pm


Good pickup, on the defensive teams showing more improvement at home, and the offensive teams yielding a better road record. I hadn't noticed that. It does make sense.

I don't believe statistical analysis is worthless, though. I am way into crunching numbers. It just looked to me like the home-field number crunching was just a net performance improvement for teams home and away, which is not even comprehensive enough to me to say who has the best home-field advantage. So I guess I'm complaining, but I definitely am not trying to say that stats don't mean anything.

by Stiller Fan in Cle (not verified) :: Fri, 12/23/2005 - 4:48am

Cleveland is clearly ranked too low because if you throw out all the games trent "the suk" dilfer playd, they owned everyone. Besides next year lebron will be playin WR and beating everone all by himself. Even peter king's coffee rating system is way better than this. U just wait til frye ownez everyone me and my boyz wont ever let you forget u hater.

Anyway, now that I've got that out of my system, let me say that the Browns do look disturbingly better. The defense is short a playmaker or two, but everyone generally plays diciplined and stop the run. You'd think that if they play well the last two weeks, they're DVOA might start leapfrogging some legitimate teams.

BTW, is anyone else entertained by the LeBrons commercials like I am? Or is it just because I'm a Cavs fan?

by Chris (not verified) :: Fri, 12/23/2005 - 5:43am

Baltimore rose several spots, but the reasons for that are probably either A) One of their worst games of the season is being weighted less now as well as B) Them performing unexpectedly well against a Packers Secondary that wasn't attrocious.

by B (not verified) :: Fri, 12/23/2005 - 11:39am

282: I think it's more accurate to describe the Baltimore passing game as "preformed suprisingly well against a pass defense that had previously not been atrocious but unexpectedly quit trying."

by jeff (not verified) :: Fri, 12/23/2005 - 12:19pm

Aaron; It is time to rethink your weighting system. If you look at the Pats now and the Pats in 2001 you will see alot of similarity. The last 8 games of the 2001 season the Pats had the best defense in football (by a phenomenal 5+ points per game) and no "expert" picked up on it. Almost no one picked them to win any games in the playoff and only Joe Theisman picked them to win the Super Bowl. What good is your highly technical system if it can't put all of the teams most likely to win in the playoffs at the top by the end of the season? The system needs more flexibility to allow a reaction to serious injury or recovery like the Pats.

by B (not verified) :: Fri, 12/23/2005 - 12:22pm

Jeff: Weighted DVOA has picked almost 50% of the last 7 superbowl winners. Show me a system that's done better.

by spenceKarl (not verified) :: Fri, 12/23/2005 - 12:58pm

RE: 284 (jeff)

If you look at the Pats now and the Pats in 2001 you will see alot of similarity. The last 8 games of the 2001 season the Pats had the best defense in football.."

I believe Weighted DVOA does put the majority of it's emphasis on the last 8 games. And in the past 8 weeks, is there anyone that really believes the Patriots "had the best defense in football"? Or would you rather the Weighted DVOA emphasize only last week's game?

As for the 2001 Pats, I don't know if I'd say they had the best defense in football at the end of the 2001 regular season, but according to Weighted DVOA for defenses, theirs was certainly better than the AFC teams they faced in the playoffs:

NE -9.6%
PIT -7.0%
OAK -3.1%

by spenceKarl (not verified) :: Fri, 12/23/2005 - 1:16pm

RE: 284 (jeff)

I not sure how dramatic the difference will be, but Aaron mentioned that in the off-season he would try to update all the previous season's DVOAs to use the new 2nd-order adjustments that he's now been using since Week 12. They do provide better accuracy than the 1st-order adjustments, so maybe you'll see a difference in the 2001 Weighted DVOAs at that point.

by Rich Conley (not verified) :: Fri, 12/23/2005 - 1:21pm

I think the argument hes trying to make, and I agree, is that DVOA and weighted adjusts slowly because it takes so much into account.

The patriots have moved from -9% overall, to -5% to -1% to 6% in overall dvoa in the last couple weeks. That means they have been playing consistently above that 6% for the last month+, but DVOA takes time to catch up. That means when you think you're playing a 6% team, you're playing a team thats been playing at 20%+ for a couple weeks.

by spenceKarl (not verified) :: Fri, 12/23/2005 - 1:38pm

RE: 288

Well one of the great things about DVOA is that you can take the numbers in various situations and use them to expand your overall evaluation of a team, beyond what the total or Weighted DVOA tells you (just as you did in looking at their improvement in the last few weeks). So in analyzing the 2001 Pats, I wouldn't look soley at their final Weighted DVOA. There was more to the story.

by Starshatterer (not verified) :: Fri, 12/23/2005 - 1:48pm

B (#285 )--

Isn't "almost 50%," 3 of 7? And, since weighted DVOA reflects the entire season, the 12 playoff teams are fixed.

I mean, it's much better than throwing darts, or most expert picks, but I still wouldn't bet the mortgage on it.

by B (not verified) :: Fri, 12/23/2005 - 1:54pm

My analysis of the 2001 Pats is they were a team with an average offense, great special teams, very good pass defense and lousy run defense. Clearly the achilles heel of the team was the run defense, but fortunatly for them, each the three playoff games had a situation which affected the opponent's ability to run (Snow in Foxboro, Bettis playing in his first game in over three months, and Mike Martz). I think this is the secret to thier success. The special teams really helped out, too. Not only with Vineteri's heroics, but a punt return for a TD and a blocked field goal. Other factors were Kordell Stewart turning back into a pumpkin, injury problems on the Rams (espicially Warner's thumb). So, as long as all those things happen, the Patriots really can win the superbowl. Wooo!

by B (not verified) :: Fri, 12/23/2005 - 2:20pm

Star: I was going to comment on your post, but first I looked up the odds for various teams to win the Superbowl, and the third best odds, behind Indy and Seattle, is New England. Man that team is totally disrespected. Only a 4/1 shot to win the AFC championship? Outrageous. Anyways, if you feel that Indy won't win the AFC championship, WDVOA says Denver (13/2) or Cincy (6/1) are great bets.

by Starshatterer (not verified) :: Fri, 12/23/2005 - 2:23pm

B (#291 )--

The specific case of this year's Patriots, seeems more to me ike the case of the 2003 Panters, who were very average for the first half of the season, then improved dramatically.

In the Panthers' case, the improvement came with the insertion of Ricky Manning, Jr. at corner. I think the comparable point in the Patriots this year was a double: insertion of Artrell Hawkins at safety, and Ellis Hobbs at corner.

Will that be enough for them to even make the Superbowl? We'll find out. I sure hope so, but I'm not nearly as sanguine about it as the fans haunting the Boston Herald's message boards these days.

by admin :: Fri, 12/23/2005 - 2:24pm

Read comment 284.

Now read comment 243.

Do you see why I can't take comment 284 seriously?

by B (not verified) :: Fri, 12/23/2005 - 2:32pm

293: Okay, but which AFC team has WRs equivelent to Pinkston and Thrash?

by Starshatterer (not verified) :: Fri, 12/23/2005 - 2:57pm

B (#295 )--

That 2003 Panther team also beat a little team with Isaac Bruce and Torry Holt at receiver. Came within a Cardiac Cat's whisker of beating one with Deion Branch and David Givens. Some of these guys are considered to be pretty good.

Believe me, I'm not concerned about DVOA's respect or lack thereof, for the Patriots or any other team. It's statistical analysis, and stats are notorious for their lack of respect. ;-)

All I meant was, picking three of seven winners, from a field of twelve, may be better than any other system, but that's not all the way up to "good."

There may not be a good, statistically based way to pick the Superbowl. Much of the "any given Sunday" charm of the NFL comes from the small number of games, and the finality with which the odds-on favorite to win the Superbowl, can be gone for a single bad game in the playoffs.

by admin :: Fri, 12/23/2005 - 3:08pm

Let's try this again. You can't judge a rating system on how it picks a specific set of individual games. Individual games are all about matchups. If DVOA itself could pick individual games, our playoff preview articles would consist of nothing but a sentence: "Team X had the higher DVOA and should win." DVOA is a great tool, but it is great because it can be broken down to analyze matchups, not because it is an all-in-one rating.

by B (not verified) :: Fri, 12/23/2005 - 3:25pm

You know, there is one "system" for playoff predictions I've always wanted to try. Take the 12 playoff teams, pick the team with the fewest injured players. Does anybody know if we can find historical data on this?

by spenceKarl (not verified) :: Fri, 12/23/2005 - 3:36pm

RE: 298 (B)

I know this isn't exactly what you're talking about, but Two Minute Warning featured a system recently that makes weekly picks based on injury report evaluation. Apparently it's picked correctly ATS 67% of the time in the past three seasons.

by spenceKarl (not verified) :: Fri, 12/23/2005 - 3:38pm

Sorry, 67% of their best picks.