Writers of Pro Football Prospectus 2008

Most Recent FO Features


» Four Downs: AFC West

There's a serious need for defensive help in Kansas City, Los Angeles, and Oakland. In Denver, meanwhile, the Broncos must determine whether or not Case Keenum can really be a long-term solution at quarterback.

10 Oct 2006

Week 6 DVOA Ratings

by Aaron Schatz

Here's a look at this week's DVOA and DAVE ratings, with commentary now available on FOXSports.com. Chicago is now dominating every stat imaginable, whether you're talking FO stats or conventional stats. While the preseason projection still counts for 46% of the DAVE rating, you'll see below that the DAVE ratings are very much in line with the DVOA ratings with just a few big exceptions:

  • Seattle is playing far below its projection, due in large part to that single blowout by the all-powerful Bears.
  • Indianapolis is playing far below its projection, despite being undefeated; see Any Given Sunday for more on that issue.
  • St. Louis and New Orleans have continued to play well after surprising early wins -- but check out the colossal gap between the Rams' DVOA and their non-adjusted VOA.

Late Wednesday update: While DVOA is no longer the "official" FOXSports.com power rankings, we have created a keyword at FOX that will allow you to always access the latest DVOA commentary. Simply use keyword "DVOA" at FOXSports.com.

* * * * *

These are the Football Outsiders team efficiency ratings through Week 5 of 2006, measured by our proprietary Defense-adjusted Value Over Average (DVOA) system that breaks down every single play and compares a team's performance to the league average based on situation in order to determine value over average. (Explained further here.)

OFFENSE and DEFENSE DVOA are adjusted based on strength of opponent as well as to consider all fumbles, kept or lost, as equal value. Opponent adjustments are currently set at 50% and will increase each week until they are full strength after Week 10. SPECIAL TEAMS DVOA is adjusted for type of stadium (warm, cold, dome, Denver/Mexico City) and week of season.

DAVE is the new early-season formula that combines early-season performance with our preseason projection to get a more accurate picture of how well teams will play over the course of the entire season. This is the rating used to rank teams at FOXSports.com. After Week 5, the preseason projection counts for 46% of the rating. (DAVE stands for "DVOA Adjusted for Variation Early.") The projection will count for 32.5% next week, 15% the week after that, and then will disappear after Week 8.

To save people some time, please use the zlionsfan template for all complaints:

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

1 CHI 64.7% 1 40.4% 1 72.5% 5-0 19.5% 3 -34.2% 2 11.0% 1
2 PHI 45.5% 4 30.4% 3 40.0% 4-1 24.2% 1 -20.0% 7 1.3% 11
3 SD 41.5% 2 32.5% 2 56.6% 3-1 13.0% 5 -24.2% 5 4.3% 4
4 JAC 28.0% 14 17.4% 5 33.9% 3-2 3.3% 11 -25.4% 4 -0.6% 16
5 KC 25.2% 5 17.3% 6 24.0% 2-2 -4.0% 16 -20.5% 6 8.7% 3
6 BAL 24.6% 3 13.5% 9 39.8% 4-1 -16.0% 27 -38.0% 1 2.5% 9
7 NYG 17.8% 16 10.8% 11 14.9% 2-2 20.6% 2 2.5% 19 -0.3% 14
8 STL 16.2% 8 -2.2% 18 31.6% 4-1 11.6% 8 -4.6% 14 0.0% 13
9 DAL 14.6% 6 7.1% 15 11.6% 2-2 4.2% 10 -15.0% 10 -4.5% 27
10 NE 11.9% 11 9.9% 12 20.7% 4-1 12.6% 6 4.4% 21 3.7% 5
11 NO 11.1% 9 -6.0% 19 17.9% 4-1 4.4% 9 -3.0% 17 3.7% 6
12 ATL 9.8% 15 9.5% 13 23.3% 3-1 -8.9% 22 -26.2% 3 -7.5% 32
13 CIN 8.2% 13 13.5% 8 8.4% 3-1 -0.1% 13 -5.3% 13 3.0% 8
14 WAS 7.3% 7 5.7% 17 0.9% 2-3 12.5% 7 7.1% 23 2.0% 10
15 DEN 6.0% 22 12.0% 10 -1.7% 3-1 -4.8% 17 -14.1% 11 -3.3% 25
16 IND 4.6% 10 15.5% 7 13.1% 5-0 17.5% 4 11.6% 27 -1.3% 18
17 MIN 4.4% 19 -6.2% 20 -0.7% 3-2 -8.5% 21 -16.1% 9 -3.2% 24
18 CAR 3.6% 17 7.4% 14 3.8% 3-2 2.7% 12 -3.7% 15 -2.8% 22
19 SEA -0.8% 21 18.2% 4 -5.2% 3-1 -9.8% 23 -10.5% 12 -1.4% 19
20 PIT -1.2% 20 6.9% 16 -13.4% 1-3 -12.5% 25 -17.0% 8 -5.7% 29
21 BUF -10.1% 12 -15.8% 25 -13.2% 2-3 -10.0% 24 0.5% 18 0.4% 12
22 GB -20.6% 26 -16.9% 26 -28.4% 1-4 -6.1% 18 8.7% 25 -5.8% 30
23 CLE -21.0% 24 -15.1% 23 -25.3% 1-4 -22.3% 28 8.4% 24 9.7% 2
24 DET -21.6% 25 -15.3% 24 -29.3% 0-5 -7.2% 19 13.9% 28 -0.5% 15
25 SF -22.2% 27 -23.4% 27 -22.9% 2-3 -3.4% 15 17.0% 29 -1.7% 20
26 TB -22.4% 28 -14.8% 22 -31.9% 0-4 -14.4% 26 4.9% 22 -3.1% 23
27 MIA -22.9% 23 -12.2% 21 -8.0% 1-4 -22.5% 29 -3.4% 16 -3.9% 26
28 NYJ -27.3% 18 -26.7% 28 -26.5% 2-3 -7.4% 20 18.9% 30 -1.0% 17
29 ARI -32.9% 30 -27.2% 29 -27.0% 1-4 -23.4% 30 3.0% 20 -6.5% 31
30 HOU -39.2% 29 -30.5% 30 -36.9% 1-3 -2.7% 14 39.6% 32 3.1% 7
31 OAK -51.9% 31 -32.4% 31 -60.1% 0-4 -38.3% 32 8.7% 26 -4.8% 28
32 TEN -51.9% 32 -36.1% 32 -44.2% 0-5 -26.5% 31 23.6% 31 -1.8% 21

  • ESTIMATED WINS uses a statistic known as "Forest Index" that emphasizes consistency as well as DVOA in the most important specific situations: red zone defense, first quarter offense, and performance in the second half when the score is close.  It then projects a number of wins adjusted to a league-average schedule and a league-average rate of recovering fumbles.  Teams that have had their bye week are projected as if they had played one game per week.
  • PAST SCHEDULE lists average DVOA of opponents played this season, ranked from hardest schedule (#1, most positive) to easiest schedule (#32, most negative). It is not adjusted for which games are home or road.
  • FUTURE SCHEDULE lists average DVOA of opponents still left to play this season, ranked from hardest schedule (#1, most positive) to easiest schedule (#32, most negative). It is not adjusted for which games are home or road.
  • VARIANCE measures the statistical variance of the team's weekly DVOA performance.  Teams are ranked from least consistent (#1, highest variance) to most consistent (#32, smallest variance).

1 CHI 64.7% 5-0 4.8 1 -12.1% 28 -10.0% 30 10.6% 15
2 PHI 45.5% 4-1 4.7 2 -9.9% 24 2.7% 11 0.5% 32
3 SD 41.5% 3-1 3.9 3 -16.1% 30 -4.7% 22 8.1% 17
4 JAC 28.0% 3-2 3.2 7 8.9% 7 -10.0% 31 36.6% 4
5 KC 25.2% 2-2 3.1 10 -10.2% 26 0.7% 16 43.5% 3
6 BAL 24.6% 4-1 3.4 5 -12.0% 27 -1.6% 20 17.6% 10
7 NYG 17.8% 2-2 3.2 6 14.2% 4 7.2% 6 7.2% 20
8 STL 16.2% 4-1 3.2 8 -18.2% 31 3.5% 9 10.6% 14
9 DAL 14.6% 2-2 2.6 18 5.8% 11 0.1% 17 21.6% 7
10 NE 11.9% 4-1 3.1 9 -9.2% 23 -8.3% 28 6.3% 23
11 NO 11.1% 4-1 2.9 13 -10.1% 25 7.8% 4 13.4% 11
12 ATL 9.8% 3-1 3.6 4 -8.1% 22 6.0% 8 48.0% 1
13 CIN 8.2% 3-1 2.7 17 3.0% 15 2.7% 13 12.2% 12
14 WAS 7.3% 2-3 2.9 14 6.4% 9 7.9% 3 21.1% 9
15 DEN 6.0% 3-1 2.8 15 15.6% 3 -5.5% 24 10.6% 13
16 IND 4.6% 5-0 3.0 11 -14.5% 29 -0.2% 18 7.5% 19
17 MIN 4.4% 3-2 3.0 12 11.0% 6 -6.3% 26 4.3% 28
18 CAR 3.6% 3-2 2.7 16 -3.6% 21 11.9% 2 8.0% 18
19 SEA -0.8% 3-1 2.5 19 5.6% 14 -5.7% 25 5.9% 24
20 PIT -1.2% 1-3 2.4 20 13.7% 5 -0.3% 19 5.6% 25
21 BUF -10.1% 2-3 2.1 21 6.2% 10 -6.6% 27 25.3% 6
22 GB -20.6% 1-4 1.8 25 23.2% 1 -4.8% 23 4.6% 27
23 CLE -21.0% 1-4 1.3 28 -1.1% 19 2.0% 15 4.0% 29
24 DET -21.6% 0-5 1.5 26 16.0% 2 -2.5% 21 3.4% 31
25 SF -22.2% 2-3 1.9 22 0.4% 18 6.1% 7 43.7% 2
26 TB -22.4% 0-4 1.8 24 1.3% 17 14.5% 1 7.2% 21
27 MIA -22.9% 1-4 1.4 27 -22.6% 32 2.7% 12 5.2% 26
28 NYJ -27.3% 2-3 1.3 29 -3.5% 20 -11.7% 32 28.9% 5
29 ARI -32.9% 1-4 1.8 23 5.7% 13 2.8% 10 10.3% 16
30 HOU -39.2% 1-3 1.2 30 8.6% 8 -9.1% 29 3.4% 30
31 OAK -51.9% 0-4 0.4 32 5.7% 12 2.2% 14 6.9% 22
32 TEN -51.9% 0-5 0.5 31 2.1% 16 7.2% 5 21.4% 8

Posted by: Aaron Schatz on 10 Oct 2006

351 comments, Last at 02 Jan 2007, 11:47pm by Jeff


by David (not verified) :: Tue, 10/10/2006 - 3:47pm

So, my gut reaction was right - the Eagles really do seem to be the second-best team in the conference, if not the league, and the Bears should still scare the living daylights out of me. Good to know.

by Murr (not verified) :: Tue, 10/10/2006 - 3:48pm

So, the top 3 are pretty much inline, the bottom 12 are pretty much inline - and everybody between 4 and 20 is a big ol' jumbled mess. Okay, got it.

I think I'm more surprised by Detroit being ranked above SF more than anything else.

by K. Derek7 (not verified) :: Tue, 10/10/2006 - 3:49pm

The Bears are just... scary.

by Pat (not verified) :: Tue, 10/10/2006 - 3:50pm

The only thing that gives me hope as a Philly fan versus the Bears is that close win over Minnesota. Philly's still got far, far less variance than any team in the league - they've never not looked like one of the best teams in the league for any significant amount of time.

Well, that, and the fact that Philly won't play Chicago until the playoffs (thank God!). And only time will tell how long it takes defensive coordinators to figure out Grossman.

by Ryan (not verified) :: Tue, 10/10/2006 - 3:50pm

Yikes..The Bears' variance jumped 6% from last week. I'm curious as to what their variance is when excluding the Minn game? I can't imagine it being very high without that game included.

by RJ (not verified) :: Tue, 10/10/2006 - 3:53pm

I think it's interesting that on a play-by-play level DVOA is not altogether impressed by the Broncos -- every team they've played is higher ranked.

How much of this is due to the schedule adjustments only being at 50% strength?

How much of this is hidden indicators and flaws in the team that we don't know about? (Read: beyond "Jake Plummer is unpredictable")

Anyway, great job! Are the Bears the best 5-0 team in DVOA's history?

by Jesse (not verified) :: Tue, 10/10/2006 - 3:54pm

Denver's defense is clearly ranked too low because they are clearly the best defense in the NFL, having allowed 16 points in the past three games, and 1 touchdown so far this season. Common sense is way better than this. w00t! go bronx! superbowl '07!!!!

by White Rose Duelist (not verified) :: Tue, 10/10/2006 - 3:55pm

Doesn't excluding a game kind of defeat the point of variance?

by jonnyblazin (not verified) :: Tue, 10/10/2006 - 3:57pm

Ugh, after a couple of weeks moonlighting as a mediocre offense, the Raven's O is right back to where everyone expects them to be. I'm hoping that with a little more Musa Smith and Mike Anderson, along with more familiarity of the offensive system for McNair can get them to at mediocre once again.
Its a very interesting picture so far in the AFC, a bunch of quality teams all with noticeable flaws. I don't see any team asserting their dominance over the conference any time soon.

by Pacifist Viking (not verified) :: Tue, 10/10/2006 - 3:58pm

When I saw the rankings earlier in the season I thought, "If I understand this system, and I probably don't, the Vikings will continue to win or lose close games and bounce around #16 or #17 regardless of their record." So far, that's looking about right.

by Ryan (not verified) :: Tue, 10/10/2006 - 4:01pm

Well, yeah. But suppose for example the absoulutely totally unlikely and impossible event that the Bears keep on winning by 20 points for every game(I'll eat my cap if that happens; luckilly I dont own any caps). If so, that one game could just be considered a statistical anomaly that drags the numbers down, (such as using the average salary of 100 people in a company, with 99 people making between 10K and 50K and the CEO making 10,000,000 dollars.) Should we start including median DVOA numbers to see what a team does on a game-by-game basis? Because, KC got that HUGE jump by winning a single game, but if one were to look at their median game results, one could possibly see that KC isn't quite as good as their numbers would suggest. (No offense to KC at all. Good job for surviving one freak accident and one close-to-horribly freak accident.)

by Pat (not verified) :: Tue, 10/10/2006 - 4:01pm

Nice to see that Baltimore and San Diego's defense are falling back to reality now that they can't play the Raiders every week. Not that they won't end up good - they will, most likely - but they won't end up far and away the best defense in the league, like they were earlier.

by Fnor (not verified) :: Tue, 10/10/2006 - 4:01pm


by Crushinator (not verified) :: Tue, 10/10/2006 - 4:02pm

Just for Comparison, I went to past years to see what the best team in the league had at seasons end. I know the season is still early and a lot of the outliers will trend more towards average, but the closest team in DVOA history to this years Bears is the 1999 Rams - and the Bears are ranked 20% higher then them. Most years, the Bears are ranked per just raw DVOA percentage as nearly twice as good as most of the past #1 teams in the league.

There are 3 teams this year that rank higher than most of the previous highs. The 1999 Rams were 45.8% and currently hold the record as "best team in DVOA" but Philly and SD rank higher than every other #1 DVOA team as well.

I'm glad Indy's rated where they are. They're lucky to be 5-0, they could VERY easily be a 2-3 team right now. and the fact that they have a harder future schedule than J-ville might allow Jax to take the division.

Also, talked about it last night, but the Jets with their schedule have a really good chance of snubbing some better team for a Wild Card spot.

by Ryan (not verified) :: Tue, 10/10/2006 - 4:03pm

RE #10. Yeah, the Vikings are like the mini-Bears from last season. Their offense isn't quite as horrible and their defense isn't quite as dominant as the Bears from last season. But look at how mediocre CHI really was last year. They struggled to put away teams in the bottom half as well as "played up to their competition" against teams in the top half. I don't think numbers can really predict how a team like that will perform from one week to the next, except to predict that there's no way to accurately predict that team.

by max (not verified) :: Tue, 10/10/2006 - 4:12pm

Well, week 6 will be a nice test for DAVE. The Seahawks at Rams will pit one team with a +15 [DVOA-DAVE Rank] against a -10 [DVOA-DAVE Rank].

Now, if it comes to pass that the Rams somehow beat the Seahawks and DAVE still ranks the Seahawks higher, I wonder what we should all think about DAVE?

by Pat (not verified) :: Tue, 10/10/2006 - 4:12pm

but the closest team in DVOA history to this years Bears is the 1999 Rams

Better to look at Week 6 ratings from previous years. The top team is always around ~50% or so, and by the end, was more like ~30%. That puts the Bears about as good as the 1999 Rams. Great.

by David Ferrier (not verified) :: Tue, 10/10/2006 - 4:14pm

Sweet! An excuse for my Packers. According to these stats the Packers have had the hardest schedule in the NFL so far.

by beedubyuh (not verified) :: Tue, 10/10/2006 - 4:26pm

I am having difficulty typing this as tears of joy run down my face. I hope my keyboard is waterproof.


by COINFLIP (not verified) :: Tue, 10/10/2006 - 4:29pm

Bears = Number one on the threatdown. Again.

by JonL (not verified) :: Tue, 10/10/2006 - 4:30pm

Maybe this has been answered before, and I guess it's implied by #16, but if a team's DVOA is higher than its DAVE, does that mean it's outperforming expectations (it's actual value is greater than it's preseason projections, which will eventually go away)? Or is that simplifying it too much?

by Independent George (not verified) :: Tue, 10/10/2006 - 4:33pm

Pat, besides the raw numbers, I'd also be worried about matchups with the Bears. Subjectively, I'd say that the Eagles' defense should do well, but their pass-happy offense plays right into the strength of the Chicago defense. These guys are ridiculously, frighteningly fast; the best strategy against them seems to be to pound them with the running game - not because it's particularly effective, but because it keeps your QB from getting drilled on every down.

Looking at their Div I-AA schedule for the rest of the year, @NE might be the only thing standing between the Bears and 16-0 (unless you believe Miami really does have some mystical power against undefeated teams... Somehow, I just don't think Joey Harrington can pull a Marino like in '85).

by JasonR (not verified) :: Tue, 10/10/2006 - 4:34pm

Am I mssing something?

I think the schedule rankings might be wrong.

Den: (16.2+25.2+11.9+24.6)/4 = 19.5 (not 15.6)

Det:(-.8+64.7-20.6+16.2+4.4)/5 = 12.8 (not 16.0)

If you divide the Denver total by 5 and the Detroit total by 4 you get the chart totals.

I was curious how Denver only had the third hardest schedule when the teams they played are ranked 5,6,8 & 10. So unless I am wrong (which is always a possibility) I think the numbers are messed up.

by Nate (not verified) :: Tue, 10/10/2006 - 4:36pm

22 - I think @STL could be quite tough as well. They are a very good team at home.

by Stillio (not verified) :: Tue, 10/10/2006 - 4:38pm

Jacksonville's easy remaining schedule is a bit of an illusion. They have four games left against top ten DVOA teams (PHI, KC, NYG, NE) and four games against bottom three DVOA teams (HOU, TEN 2x each). It's mostly just the complete suckiness of the bottom feeders in the AFC South, and the fact that Jax hasn't played them yet that makes the schedule look so easy. The team that I feel most sorry for is Tampa Bay. 0-4 and the hardest remaining schedule...ouch.

by Joel (not verified) :: Tue, 10/10/2006 - 4:40pm

How much are the '06 Broncos like the '05 Bears? Smothering defense, barely legal offense. Any theories *why* the Denver offense is so so bad this year? Are the rumors about defenses figuring out how to nullify the bootleg true?

by Crushinator (not verified) :: Tue, 10/10/2006 - 4:44pm


Plummer just looks inaccurate. The playcalling is similiar, the routes are similiar. Players are open. He's just missing them.

by Ryan (not verified) :: Tue, 10/10/2006 - 4:46pm

RE #26 You've helped make my point there. As I stated earlier, very large or very small numbers drag will always drag down the average. For example, you're working at a job where you make 35K per year and another company is trying to sway you to change jobs. They tell you that your Average employee makes only 131500 dollars and that their Average employee makes over 1 million dollars! Each company has 10 employees. What they're NOT telling is that your CEO makes only 1 million dollars while their CEO makes 10 million dollars, and that when eliminating both CEOs we find that the median salary in your company is 35 thousand dollars per year and the median salary in their company is only 30 thousand dollars per year. Thus, if you were to change jobs you would expect to make 5 thousand dollars less per year at the other company. So, to accurately predict wins, it would be unwise simply to go with the average strength of schedule ahead, and to judge the wins on a case-by-case basis, (which I've seen some people here do when projecting wins on the last week's DVOA article).

by Bill Barnwell :: Tue, 10/10/2006 - 4:46pm

I'm shocked and dismayed Aaron missed a Colbert reference -- poor, poor chopped up Keary Colbert player comment in the book.

by Ryan (not verified) :: Tue, 10/10/2006 - 4:51pm

So, based solely on less than 1/3 of the schedule, I'm going to subjectively eliminate teams from winning the superbowl based on the numbers and what I subjectively believe a team requires to win the Superbowl. My criteria is pretty simple. To win the superbowl, a team needs to rank in the top 10 in both offense and defense, and might not want to suck horribly on special teams. So as of right now, everybody except for the Bears, Eagles, and Chargers are eliminated from winning the Superbowl, and the Chargers are eliminated because of Martyball. SO the Superbowl will be PHI vs. Chicago. Oh wait...they're in the same conference....(I'm only basing these numbers on the non DAVE Dvoa, btw.)

by J (not verified) :: Tue, 10/10/2006 - 4:54pm

I'm not quite as statistically savvy as many of you -- can somebody please explain to me how the Pats' seemingly struggling passing game can be ranked #4?

by admin :: Tue, 10/10/2006 - 4:54pm

I did the Colbert reference last year in a DVOA ratings commentary. By the way, the answer to question 16 is: "We should all think that nobody is supposed to judge the validity of a statistic by a single game result."

by Independent George (not verified) :: Tue, 10/10/2006 - 4:57pm

I think we've overlooked something, here... Who would win - Urlacher or Hurricane Ditka?

by Ryan (not verified) :: Tue, 10/10/2006 - 4:57pm

RE #23. They prorate the numbers so that every team plays the same number of potential games. Take your 78 and divide that by 5 and you'll get 15.6 exactly. I can't say what the Detroit numbers mean, because I'm getting the same results as you.

by Jason (not verified) :: Tue, 10/10/2006 - 4:58pm

Re #5 The whole game? Heck, just exclude the Winfield INT return (which Grossman says he was trying to throw away) and re-analyze. 19-9 (not that that would necessarily be the score)is much different than 19-16. Besides, I think their main trouble in that game (inside pass blocking, especially on the blitz) has been shored up pretty well. Heck, it was shored up at halftime of that game. The Bears aren't likely to average 25-point margins of victory, but it's surely not going to be 3 either. Credit the Vikes, but as far as the Bears are concerned that was the outlier of their performances so far. I wouldn't lend as much credence to it as some people seem to want to do.

Also, the Bears have given up 3 TDs this year, but really two are on the offense, not the D. And while the defense has set up a number of short fields for the offense, they still haven't actually scored themselves. Overall, though, I think in terms of point differential they're about where they should be because Grossman's been lucky on some throws so far (see my post in the Quick Reads thread). I don't see him continuing to throw high off receivers hands or off his back foot into three defenders or having Davis out jump four guys on a tipped ball and be successful. The Bears have been fumble lucky, too. While still very good, their underlying performance hasn't been quite as otherworldly as most people seem to think.

This is not to say I'm not ecstatic about the season so far, but if you listen to all the coaches and players when they say they still can play better, you should realize they're not just blowing smoke.

by Pat (not verified) :: Tue, 10/10/2006 - 5:00pm

but their pass-happy offense plays right into the strength of the Chicago defense.

I don't know. The Bears haven't really faced a well-rounded pass offense yet, so it's tough to say if they've got a weakness in pass defense or not. That is, a team with a serious pass catching threat short (RB/TE), middle (TE/WR), and long (WR). And Philly is an obscenely balanced pass offense.

Yes, they've faced the Seahawks, but the Seahawks didn't have an RB who could catch or a TE worth talking about. Ditto with Buffalo, Detroit, or Green Bay. Minnesota is the most well balanced team they've faced, and they don't have a serious threat at TE, either.

They do face a balanced offense in the Giants in a few weeks - but unless Grossman reverts horribly, they'll shred the Giants secondary into tatters.

by DGL (not verified) :: Tue, 10/10/2006 - 5:04pm

So I don't know whether to view the Steelers as the best crappy team in the league (highest DVOA among teams with less than 2 wins) or as only the worst mediocre team in the league (a 9 point DVOA dropoff from them to Buffalo).

by zlionsfan (not verified) :: Tue, 10/10/2006 - 5:04pm

Re 2: I believe there are a couple of reasons for that, and any help on this would be appreciated.

One is that the ways in which Detroit has hurt themselves are not always reflected in DVOA. For example, if I recall correctly, the length of an interception return is pretty much random, so a team that tends to throw interceptions that are returned for touchdowns is hurt more on the scoreboard than in DVOA.

The other is that with the exception of being throttled by the #1 DVOA team, the Lions have been in the games they've played pretty much the whole way. They've just Millened those games.

Mil-len, v. tr.
1. To affect negatively through inadequate performance, specifically through inexperience and/or incompetence: Looks like they're going to Millen their chances.
2. To lose a football game as described in 1.

cf. Ford

Ford, v. tr.
1. To allow something of value to depreciate through decades of incompetence while insisting that one's best effort is being put forth: That used to be a nice house, but he's really Forded it, so I don't see why anyone would buy it now.

by Fnor (not verified) :: Tue, 10/10/2006 - 5:07pm

#36: I think it's mostly Berrian (and the threat of Berrian). He's really turned on this year, and his progression has helped their offence far more than any progression Rex has made (still bouncing passes off linemen's heads).

The amusing thing is that everyone's still so scared of Muhammad. He's pretty good underneath, but he's nowhere near the top threat. And he drops a lot, anyway.

by zlionsfan (not verified) :: Tue, 10/10/2006 - 5:08pm

Re 36: Actually, both Shawn Bryson and Dan Campbell have posted pretty good numbers so far, although there's no way to tell if that's just a fluke of sample size or if it's an indication that the wrong people are getting the ball ... they're both first in receiving DVOA among non-qualifiers so far this season.

by Ryan (not verified) :: Tue, 10/10/2006 - 5:08pm

RE #36 From what I've seen from the Bears in the past two seasons, their biggest weakness is blown coverages 40 yards down the field, usually caused by miscommunication between the Safeties and CB's. Take the SEA game, for example. SEA had an EASY TD pass down the right sidelines in the third quarter where no Bear was within five yards of the WR and the ball was right on his hands. That wouldn't have changed the result, but it did expose the Bears' weakness. Basically, the Bears are good for one or two blown coverages deep down the field a game, from what I've seen, (and I've watched about 240 of the 300 minutes the Bears have played this year.)

by CaffeineMan (not verified) :: Tue, 10/10/2006 - 5:10pm

St. Louis and New Orleans have continued to play well after surprising early wins — but check out the colossal gap between the Rams’ DVOA and their non-adjusted VOA.

The Pats have a similarly large difference: 11.9% DVOA, 20.7% Non-adj. VOA. And just when I was about to get excited about them outperforming their DAVE. Unw00t.

by Jason (not verified) :: Tue, 10/10/2006 - 5:15pm

Re #36 Agreed. I still don't trust Tillman and if you get a line that can stall the pass rush and have an accurate QB, I fear they could exploit the secondary. I don't think Indy's at all a better team, but facing Manning scares me. He has weapons and the ability to do a lot of damage if the line gives him any time at all. I fear the Foxboro game because Belicheck has had a lot of success patching together no name lines and of course if Brady has time, he eventually finds someone. When the line got all over Delhomme last year, the Bears won impressively, but come the second time around, they found away to slow down the rush (and Tillman was inexplicably stupid) and that's when they got burned. No opponent so far has had that combo and so our passing D looks fantastic, but I'm certain that's far more the lines doing than the secondary's. If Arizona had any kind of line (and Fitzgerald wasn't out), I might be nervous about Leinart having time to find two very good receivers. As it is, I just fear he'll be injured. This is also why I agree the Eagles could be dangerous, not only do they have the line, the receivers and the QB, they have the D. I still think the Bears would prevail, but it wouldn't be a foregone conclusion.

by joel in providence (not verified) :: Tue, 10/10/2006 - 5:16pm

so is aaron donning the fake moustache and ditka sweater now? "my prediction: Even with Ditka with one arm tied behind his back... Bears 182 Lions -31."

by jonnyblazin (not verified) :: Tue, 10/10/2006 - 5:16pm

"Nice to see that Baltimore and San Diego’s defense are falling back to reality now that they can’t play the Raiders every week"

Ummm, last week if my memory serves correct the Ravens defense was at -41% DVOA, this week they are at -38% DVOA. I'm not sure losing three percentage points counts as coming back to reality.

by Pat (not verified) :: Tue, 10/10/2006 - 5:18pm

#40: True, but Detroit doesn't use Campbell, so it doesn't really matter if he's on the team. The only pass that was thrown his way in the Chicago game was caught for 23 yards. Maybe they should've gone there more often.

Detroit's much more of a "WR+RB" passing game, which is what Martz has always done. I think in that sort of a game, the WRs and RBs have to be very good passing threats, and Detroit's aren't.

by wrmjr (not verified) :: Tue, 10/10/2006 - 5:18pm

#39, but Muhammad has currently caught 65% of passes thrown to him, vs. 53% for Berrian.

by Crushinator (not verified) :: Tue, 10/10/2006 - 5:19pm


Well, it helps that the Bears gameplan was, and they pretty much said this, "We'll let Steve Smith get his yards and shut down the rest of the offense and force him to beat us". and then of course he did.

That's like saying "We'll shut down Dominic Rhodes and force Peyton Manning to beat us".

by Crushinator (not verified) :: Tue, 10/10/2006 - 5:20pm

and 47

Berrian usually goes deep, and Grossman rarely throws a ball away. If he has to throw a ball away, he tends to just heave it far on his deep route and hope his WR can make a play. I wouldn't look too deeply into his catch %.

by Pat (not verified) :: Tue, 10/10/2006 - 5:22pm

Week 3: -82%
Week 4: -60%
Week 5: -41%
Week 6: -38%

I'm expecting Baltimore to come out higher than San Diego, but Baltimore has been crashing back to reality.

by Jersey (not verified) :: Tue, 10/10/2006 - 5:22pm

Philly's low variance is a good sign for Eagles fans. They havn't had the hardest schedule yet, but outside 1 quarter against NYG, they'd be doing exactly what they should be doing - demolishing easy teams, and outplaying talented divison rivals. I'm really looking forward to the rest of the season.

RE: Denver's offense
Is it really all Plummer's fault? I havn't really seen any Denver game's this year, I missed parts of last night's game . I've always been a Plummer fan, but even if you hate the guy, you could never have seen it coming that he would play this bad.

by Ryan (not verified) :: Tue, 10/10/2006 - 5:23pm

RE #47 Which comes to the great debate. Is 12% less catch percentage worth the +9 yards per reception? In my subjective viewpoint, I'll take the extra nine yards per catch and the lower catch percentage on most occasions, unless it's 3rd and 8 and I need a first down or I lose the game. I would also say that Berrian's lower catch percentage is because of the fact that Grossman is passing to him deep far more often than to Moose, and I think everyone can agree that shorter passes usually have a higher completion percentage than deeper passes in general.

by dbt (not verified) :: Tue, 10/10/2006 - 5:27pm

Given that peanut tillman's usual coverage is underneath and the safety's job is over the top, and chris harris is no longer the deep safety (not to mention mike brown being hurt last year in the playoffs), I think their deep coverage is going to be better.

by zlionsfan (not verified) :: Tue, 10/10/2006 - 5:31pm

46: That's what I mean, Pat. Bryson's been fairly good out of the backfield for a while, but he seems to be mostly a blocking back now, while Jones is getting a huge number of passes thrown his way (some of which work pretty well; many of which do not).

I don't know why they bothered to sign Campbell if they weren't going to look to him more often. I thought they made an effort to find guys like Pollard (was) and Campbell so they could get more production out of the TE spot, but as you've pointed out, Kitna's not been looking that direction.

Guess we'll have another chance to see how the Bears decide to hold them in check. I'd expect Chicago to use a 3-3-3 against them. No sense in exposing guys to injury unnecessarily.

by Will Allen (not verified) :: Tue, 10/10/2006 - 5:34pm

Gosh, I wish the Vikings had Seattle at home, and the Patriots on the road, over their next two games, because I get the feeling that Seattle is a lot more beatable out of their stadium than the Patriots are out of Kraft's sandbox.

by wrmjr (not verified) :: Tue, 10/10/2006 - 5:36pm

49 & 52, agreed...I was simply noting that saying M.M. drops a lot isn't really indicated in his stats this year. For most of his career he has been in the 50's, so we might reasonably expect him to drop more as the season goes on.

by Ryan (not verified) :: Tue, 10/10/2006 - 5:41pm

Yawn..So, to an FO who may be reading this, when can we expect to have Off. and Def. lines updated, and drive stats as well?

by ammek (not verified) :: Tue, 10/10/2006 - 5:42pm

Quick note: the Packers are 1-4, not 1-3.

by tunesmith (not verified) :: Tue, 10/10/2006 - 5:43pm

Denver's also played some pretty good defenses so far. Hopefully their offense will get a little well against Oakland.

I've never really understood how people use the DVOA rankings here to predict wins, anyway - are they just doing it by seeing if one team is ranked ahead of another in the DVOA rankings?

by MAW (not verified) :: Tue, 10/10/2006 - 5:44pm

So, considering that Tampa has an 0-4 record and the hardest remaining schedule, what are they odds they run the table and become the Bucs of old?

by MJK (not verified) :: Tue, 10/10/2006 - 5:47pm

I argued against this last week, but I'm starting to come around. It is really weird that DVOA thinks that NE has an elite offense and a slightly below average defense. Their last win was almost entirely due to defense, and defense-produced turnovers. Their run offense has looked good 3 times and poor twice, their pass offense has looked only average to poor every time, and their defense has looked slightly below average to well above average. What is skewing the DVOA?

by Ryan (not verified) :: Tue, 10/10/2006 - 5:47pm

#60. I'd imagine the single game against the Bears accounts for a great deal in that hardest remaining schedule for TB. I will say this, though. It's always easier to lose than it is to win, so it's more likely for a team to go winless than undefeated.

by Ryan (not verified) :: Tue, 10/10/2006 - 5:51pm

I retract my statement before. Unless it turns out that NO really is a pretender, or that Sea has locked up its standing before week 17, they very well could go winless.

by Will Allen (not verified) :: Tue, 10/10/2006 - 5:56pm

Philly at this point really is the only credible threat to Chicago in Soldier Field come January, and even then it isn't a great threat, but at least it isn't a forgone conclusion. I don't follow the AFC as closely, but maybe the the guy who masterminded the downfall of the 2000 Rams could do something, but, then again, last year's divisional game against the Panthers aside, Lovie Smith is not Mike Martz. In any case, it is useful to remember how close the Titans came to beating the '99 Rams.

It seems to me that to beat the Bears in the playoffs, a team will have to have a very physical performance by the offensive line, even if it doesn't provide obvious results, and a quarterback who handles pressure exceedingly well, and a defense that pressures Grossman into bad throws. The list of teams that fit that description isn't very long.

by Mannie Fresh (not verified) :: Tue, 10/10/2006 - 5:56pm

When was the last time a team dominated (I know it's real early) its team like the Bears? They score the most points per game and allow the least points. That's domination on both sides of the ball. I know the 99 rams were pretty dominant (Kurt Warner was the best QB, Faulk was very good on both rushing and receiving with +1000 yards in both categories, and Kevin Carter leading the NFL in sacks).

When do you think the Bears will lose?

They played a passing heavy team in Seahawks, AND WON. They played a run heavy team in Buffalo, with McGahee coming into the game with the most rushing yards, AND WON.

I think thier main challenge isn't New England, which I think they can win on the road. They should have a fight in New York against the Giants. They have a strong running game in Tiki Barber and a dynamic passing game in Eli and Plaxico/Toomer. That is the game I'm circling as a game in which will test the Chicago's ability to not only win with attacks in passing but also in running. Winning on the road is another spot I think the Bears should work on before the playoffs.

by James C (not verified) :: Tue, 10/10/2006 - 6:00pm

Tillman played the Panthers playoff game with an injured left shoulder, if the Bears had the depth (Manning Jr) then that they have now, Tillman wouldn't have played. If you look at his poor play during the game it is a lot easier to understand if you realise that he was effectively playing with one arm - ie. Smith being able to push him over on the first TD, Smith being able to wrestle the ball off him near the endzone. The remainder of Smith's demolition of the Bears secondary I would put down to single covering him with Chris Thompson (yes thats 'The Mighty Chris Thompson') without deep help and Mike Brown getting hurt on some of the worst turf I have ever seen.

by Wanker79 (not verified) :: Tue, 10/10/2006 - 6:01pm

Aaron, are the top teams' DVOA above their DAVE and the bottome teams' DVOA below their DAVE by design?

by Mannie Fresh (not verified) :: Tue, 10/10/2006 - 6:01pm

Re: 64

Bro u read my mind about the 99 Rams

by Will Allen (not verified) :: Tue, 10/10/2006 - 6:06pm

Yeah, the Giants could definitely ruin their bid for 16-0; it really takes only one less than maximum intensity performance on defense to produce a loss, and if the Giants get their pass rush untracked, they could keep the Bears from scoring a lot. The Bears have a nice offensive line, but we're not talking about the '72 Dolphins in that regard.

It might require an offensive line performance like that of the only undefeated team to beat the Bears in the playoffs, however.

by navin (not verified) :: Tue, 10/10/2006 - 6:06pm

Hey tunesmith, why does Arizona have a beatpath over San Fran? Shouldn't the SF>STL, STL>ARI, ARI>SF take care of that?

by NF (not verified) :: Tue, 10/10/2006 - 6:07pm

Without 100% opponent adjustments, I'd take the DVOA rating of any team that played Chicago with a grain of salt. Buffalo was probably too high last week, but following the loss to Chicago they plummeted from 12 to 21 in current DVOA. And their DVOA is being adjusted upwards, so it isn't just that they are getting less credit for playing well against bad teams.

Also, it appears that Washington's offense has got it together at the same time the defense has collapsed.

A stats question: is Estimated Wins useful for predicting who will win games based on the data as the season is ongoing? I know that pythagorean wins is very poor at predicting winners because it is not opponent-adjusted and relies on entirely retrospective information, but does Expected Wins not have that problem because it is opponent-adjusted and is based on DVOA, a predictive stat?

by navin (not verified) :: Tue, 10/10/2006 - 6:10pm

Keep in mind the Bears only decent road opponent was the Vikings, and the Bears needed a late drive to win that game. I could see @NYG, @NYJ, @NWE, and @STL to all be tough matches based on the earlier Minnesota game.

Also the Bears play the Giants and Jets back to back, whatever that means.

by Independent George (not verified) :: Tue, 10/10/2006 - 6:15pm

#65 - As much as I'd love it if the Giants were to knock off the Bears, I just don't see it happening. The Giants offense should put up a good fight against the Bears D, but unless Grossman regresses in a major way, I just don't see the Giants secondary holding up (even if the pass rush improves).

Plus, I don't want to jinx them.

by Will Allen (not verified) :: Tue, 10/10/2006 - 6:18pm

It is wise to pay more attention to special teams rankings as well. I 'd have a lot more confidence in the Vikings' ability to contend for a playoff spot if they didn't surrender so much field position on punts and kicks.

by johnt (not verified) :: Tue, 10/10/2006 - 6:22pm

Well, let's just give the Super Bowl to the Bears. It would appear they have no weaknesses and are unlikely to be beaten all season.

Seriously, it's way too early for this. I have some degree of faith in the Bears D but I am not convinced the success of the Bears O will continue. Their ridiculously easy schedule will make it difficult to judge, but I am unconvinced on Grossman.

That being said, I really don't think the blueprint for beating them is Carolina, which was basically a fluke game caused by Lovie Smith's refusal to put 10 men on Steve Smith when needed. If anything I would say it is the Steelers win from 05 that is a better guide, where their big offensive line manhandled the Bears to the tune of almost 200 yards rushing.

by Mannie Fresh (not verified) :: Tue, 10/10/2006 - 6:22pm

Cot Damn! Bears rank #2 in Defense, #3 in Offense and #1 Special Teams...

What are some good hotels in Miami?

by Will Allen (not verified) :: Tue, 10/10/2006 - 6:27pm

George, I wouldn't bet on it either, of course. If the Bears lose, it'll likely come on the road when their defense has a little less than full intensity, and when Grossman gets some pressure. I don't know the timing of the Bears schedule, but it is most likely to happen when they get home field advantage wrapped up, so if any of the games navin mentions above come in week 15-17 or so, those would be the best candidates.

by Ryan (not verified) :: Tue, 10/10/2006 - 6:31pm

DVOA is clearly rankd 2 hi becuz Every football expirt knows that the AFC is way supur betor than the NFC. Monkeys flipping coins iz way bettir then this. Since CHI and PHI are the top 2 teams, and IND izn't even in the top 2, this sistem must be uber reterded. You guyz r ridikulus and shuld be kicked off the net. (Wonders if this will get through the filters)

by Nate (not verified) :: Tue, 10/10/2006 - 6:32pm

Will Allem (77) - the last three games are Bucs, @Lions, and Packers. Note exactly an imposing stretch. But we'll lose a few of @NYJ, @NYG, @NWE, and @STL.

by Gerry (not verified) :: Tue, 10/10/2006 - 6:33pm

I know the numbers is what the numbers is, but, damn, the Giants sure do not seem to me to be a team that should have a below average variance. IIRC, they did last year, too, and as a fan I found them to be maddeningly inconsistent.

by Catfish (not verified) :: Tue, 10/10/2006 - 6:33pm

43: Like someone said already, since the Bears play Cover 2 the safety will be covering deep the majority of the time. Also, while Manning is always dangerous, he is most dangerous because he eats blitzes alive by audibling (audible-ing?) and throwing to the hot receiver. When a defense can get good pressure rushing only 4 and leaving 7 in coverage, Manning is much more human. The Bears are such a team.

As for the Eagles, they have a bunch of good threats, but don't have a huge gamebreaker at WR. Last year, at least, the Bears success against the pass came from completely neutralizing receivers 2-5. Against #1's they were good but not great. They lost in the playoffs because Steve Smith can beat teams by himself and the Bears refused to adjust.

65: I posted this elsewhere, but on the pfr blog (linked), they list the biggest point margins of all 5-0 teams since 1970. The 99 Rams are first, with a margin of 123 points and 4 wins by at least 20. This years Bears are second @120 and 4 20 point wins (The Bears and Rams are the only 2 teams to do this). The 91 Skins were third with a margin of about 110. Of course this doesn't guarantee anything; the 74 Pats rank 6th or so and they finished 7-7, but it does indicate good things.

Of course by typing these things I have guaranteed that the Bears will lose a bunch of games stupidly and miss the playoffs. Such is life.

by Will Allen (not verified) :: Tue, 10/10/2006 - 6:35pm

Oh, I think you're right, johnt. It'll take a superior offensive line performance to beat the Bears in the playoffs, along with getting pressure on Grossman. There's certainly a non-trivial chance that it could happen, but it has to recognized how historically dominant their performance has been through five games. It'll be fun to watch.

by Kyle W (not verified) :: Tue, 10/10/2006 - 6:36pm

Am I the only person remotely impressed with the Rams' spectacularly average special teams rating?

by DavidH (not verified) :: Tue, 10/10/2006 - 6:44pm

So, based solely on less than 1/3 of the schedule, I’m going to subjectively eliminate teams from winning the superbowl based on the numbers and what I subjectively believe a team requires to win the Superbowl. My criteria is pretty simple. To win the superbowl, a team needs to rank in the top 10 in both offense and defense, and might not want to suck horribly on special teams.

Seeing that you didn't list Dallas this year because of their -4.5%, 27th ranked ST DVOA, I'm guessing that -3.8%, 25th ranked is also not good enough. I'm going to define sucking horribly as being in the bottom 10. So...

Week 6, 2005:
Cincinnati (6/4/6)
Indianapolis (5/5/21)
[Pittsburgh is out at 3/7/25]

Week 6, 2004:
Philadelphia (3/9/3)
New England (4/3/22) [BARELY]

Week 6, 2003:
Seattle (4/4/10)
Denver (8/10/18)
[New England is out at 17/15/15]
[Carolina is out at 22/23/2]

1. DVOA is way better now than in 2003.
2. Early season special teams ratings do not seem to be important at all (in predicting Super Bowl winners).

by Fnor (not verified) :: Tue, 10/10/2006 - 6:45pm

#83: I actually had to double-take at it more than the offence, so yes.

I'd like the Bears more but I 1) see too many problems that good teams can exploit (provided they actually play any good teams), and 2) their fans are talking about the super bowl in week 5 (3, actually, if I recall correctly). I'm sure the football gods will not take kindly to this.

by admin :: Tue, 10/10/2006 - 6:54pm

Lightning round.

General: The Bears will lose. It's not a question of picking the team that will beat them, but there will be many teams with some chance of beating them and one of them will do it. This isn't college football, and nobody gets Temple on the schedule.

37: The Steelers are the worst mediocre team in the league. (This decision is where the preseason projection is useful.)

57: I dunno. Tonight or tomorrow?

59: I'll say this again, since it seems I have to say it at least four times a year: DVOA is not meant to be taken as a sure-fire method of predicting that one team will beat another. There are just way too many issues involved in deciding which team should be favored in a game: individual matchups, strategic matchups, injuries, home field, etc. If it was always "higher DVOA should be favored," I wouldn't write those weekly game previews that drill down into specifics.

61 and 7: I'm also curious about the Patriots pass offense and the Broncos defense, and I'll work on a blog post about that.

67: The DVOA through four or five games will naturally have a much larger variation than the preseason projection, which is designed to represent 16 games, or DAVE, which combines the two. Also, the distance between the best and worst teams usually drops as the opponent adjustments get larger.

by HLF (not verified) :: Tue, 10/10/2006 - 6:55pm

The Bears look absolutely spectacular so far; however, since the '85 Bears it appears that a late undefeated team has taken their first loss from a more lowly opponent, rather than a game you'd circle on your calendar at this point in the season. Let's all take Winston Wolf's advice, the Bears ain't there yet.

by Independent George (not verified) :: Tue, 10/10/2006 - 6:57pm

#85 - Then there's always the freak injury factor. Not that I wish it upon anyone - and the Bears seem to have better than average depth - but how often do teams go an entire season without any major injuries?

by Wanker79 (not verified) :: Tue, 10/10/2006 - 6:58pm

Just to follow up on #78 (btw, pretty funny):

Judging by DVOA, it seems the NFC East is clearly the best division in the NFL. The AFC West is a close second with the top three being pretty equivelant to the NFC East, but Washington (#4 in the NFC East) kicks the crap outta Oakland (#4 in the AFC West).

And the NFC as a whole is the better conference. Their top 2 teams (Chicago and Philly) are significantly better than their AFC counterparts (San Diego and Jacksonville), and their bottom 3 teams (SF, TB, & Ari) are significantly better than their AFC counterparts (Hou, Ten, & Oak). And, with the exception of the #3 and #4 teams (KC/NYG & Bal/StL), every NFC team is higher than their AFC counterpart (if only slightly in some cases).

by HLF (not verified) :: Tue, 10/10/2006 - 6:59pm

Oh, and as the Steelers should have lost the SB last year, it appears Fate has treated them as the SB losers, not the Seahawks, and the Steelers will miss the playoffs and likely finish sub-.500 this year. Good times. Oh, and FI-RE MIL-LEN!!

by Pacifist Viking (not verified) :: Tue, 10/10/2006 - 7:03pm

I woke up in some corner of hell where the Bears are the talk of the league.

by TBW (not verified) :: Tue, 10/10/2006 - 7:04pm

The Bears have played one team with a positive DVOA and squeaked out a win.

They have four more games with teams positive DVOA's, three on the road.

They have a brittle QB, which everyone seems to have forgotten. A guy with 8 games of experience and 4 TD passes prior to this year.

I'm not saying the Bears aren't good, clearly they are, but I think if we come back to this thread in late December it will be good for a LOT of laughs.

by Ryan (not verified) :: Tue, 10/10/2006 - 7:11pm

The 99 Rams had a QB with no NFL experience who was stocking groceries before he got his NFL job....So..It comes to show you cannot accurately predict how a player will perform until you actually see the player play for more than 2 games in a row.
and btw
Chicago is clearly ranked too high because everybody knows one single road game in week three decides the outcome for the entire rest of the season. Ranking a team only by their worst game is way better than this. IND and BAL and SD and PHI are way better because they haven't played a bad game yet against a team with postive OR negative DVOA..(oh wait.sorry..they have.) (btw..all four of those teams are pretty kickass, even IND.) You all are idiots and should turn off your computers forever.

by Pat (not verified) :: Tue, 10/10/2006 - 7:12pm

Seattle's negative DVOA right now is likely entirely due to the fact that they played the Bears and got crushed, though. If the opponent adjustments were at full strength, Seattle might be positive as well. DAVE clearly has a lot of faith in them.

Detroit and Green Bay do suck, though.

by Jason (not verified) :: Tue, 10/10/2006 - 7:13pm

No way the Bears lose to the Jets or the Rams. Are you kidding?

The Giants seem so schizo, I couldn't possibly predict what they might do in a given game. Worst case, I could see Eli getting lucky on his heaves and the D line getting pumped up to pressure Grossman. I don't think a Giants win is likely, but one or two big plays can undermine any actual quality discrepency. I could just as easily see the Giants looking worse than they did against the Hawks in the first half, too. Weird team.

I expect the Bears to beat the Pats, too, but again no guarantee. If the Pats pulled it off, I would expect them to win differently, though, not with a couple of big plays or takeaways, but with a smart efficiency on both sides that prevents the Bears from making their own big plays.

But if we're talking about going undefeated, I don't see it, not that they couldn't but because they'll have bigger things on their minds. They're extremely likely to have the division and home field wrapped up with 3 or 4 games remaining. They'll have little reason to play their starters more than they did preseason, so as soft as their end of season schedule is, they'll be playing second and third stringers. Ok, maybe that'll still be enough, but I wouldn't put money on it.

Re: Tillman
I didn't know he was hurt. How is this the first time I've heard of that? Anyway, the first CAR game, they made sure everything was in front of them. Smith was very active, but didn't score. In the second game, Tillman let Smith get behind him wide open, even tripping on his own feet the first time to set an very ugly tone for the game (not that I buy into that stuff too much). Injury or not, that just seemed like an individual not playing within (and undermining) the defensive scheme. And it ended their season unceremoniously. The Bears put up a bunch of points that game and those Smith TDs singlehandledly undid what should have been a much more successful season. Plus, my disdain for him isn't based just on that game. Any big pass play against the Bears, this year or last, is invariably on his side of the field. It's why I loved the Hester drafting. I'd love to see him evolve into the elite DB he's expected to be. I'd be light years more confident in our secondary. Incidentally, having heard that after Manning Jr, Todd Johnson is the next nickel on the depth chart ahead of Hester, you have to wonder how poorly he's developing on that front. How is this not a bigger story? Or does Smith view the nickel as requiring somebody more like a strong safety?

by Pat (not verified) :: Tue, 10/10/2006 - 7:14pm

PHI are way better because they haven’t played a bad game yet against a team with postive OR negative DVOA..(oh wait.sorry..they have.)

Philly hasn't played a bad game yet this year. They lost a game, but it wasn't because they played poorly.

by HLF (not verified) :: Tue, 10/10/2006 - 7:14pm

TBW (#92),

You could say pretty much exactly those same things about the Rams in 99 after game 5 or 6, and tons of folks were (a nobody for a QB, lousy schedule, coming back to Earth soon now, etc...). I don't think most here are awarding the Bears anything but "best looking team so far this season, bar none".

by Ryan (not verified) :: Tue, 10/10/2006 - 7:18pm

And if you look at the Bears game against MINN, only the offense had a bad game, but their defense showed up for the entire game. People have to judge a team's performance based on what the Entire team does, and not judge based on one young player who still needs to learn not to throw a weak pass off a fifteen step drop. And no, I wasn't saying that any of those teams played a "bad" game, just that Any team no matter how good or bad can play better or worse than usual.

by aj (not verified) :: Tue, 10/10/2006 - 7:22pm

92, Ahh, but the Seahawks have a negative DVOA only because of the Bears...Their only win against a team with a positive DAVE score was against the Seahawks, and that was a blowout...

by Alex (not verified) :: Tue, 10/10/2006 - 7:32pm

Re 30:"To win the superbowl, a team needs to rank in the top 10 in both offense and defense, and might not want to suck horribly on special teams."

Umm...only 3 of the last 8 super bowl winners ranked in the top ten in both offense and defense, so I think you might need to adjust those criteria of yours.

Teams that were ranked in the top ten in both categories were only 3-3 in super bowls against teams that were not. The other 2 super bowls featured no teams that met your criteria.

The 3 teams that won the super bowl, and met your criteria: 1999 Rams, 2004 Patriots, 2005 Steelers.

by TBW (not verified) :: Tue, 10/10/2006 - 7:40pm

We'll just ignore the fact that Seattle was missing the NFL MVP from last season for that game ?

Yes, Seattle probably would have a positive DVOA if you threw out the Chicago game, but it was also the only game they have played minus Alexander. Given Seattle's play this year it's not unrealistic to think that Seattle minus Alexander is a very slightly negative DVOA outfit.

Any chance we could have a repeat of the Colts DVOA fiasco of last year, where their schedule was soooooo easy, that it became impossible for them to be ranked #1 ? That Chicago schedule is farily ridiculous. I'm also curious about the playoff performance of teams that skate through the regular season without being tested.

by Sean (not verified) :: Tue, 10/10/2006 - 7:42pm

Look, the Bears have been playing very well in all phases of the game and the numbers prove it. That said…I would like Philadelphia to walk into Soldier Field and win come playoff time. I would like Seattle to walk in and win. I would like Carolina to walk in and win. I can think of another team or two that I’d be sorely tempted to take. And that’s all assuming Chicago finishes as the #1 seed in the NFC, which I don’t think they will. (Expect Seattle to finish with that top seed, even though they’re not firing on all cylinders right now.) Look, Rex Grossman is playing over his head and he’s getting lucky breaks to boot. Come playoff time, the Bears will have to beat a team with a much, much better quarterback than their own. And then they’ll likely have to beat two more if they plan on hoisting the Lombardi Trophy. It happens, but it’s a rarity.

by Ryan (not verified) :: Tue, 10/10/2006 - 7:44pm

Trent Dilfer.

by Jesse (not verified) :: Tue, 10/10/2006 - 7:47pm

RE: Denver's offense

Denver's offense has been spluttering this year for several reasons (I'm going to exclude the Rams game as a fluke, similiar to the Miami game last year).
Jake Plummer is part of the problem, but not all of it. While he has shown flashes of brilliance, he has made many more bonehead throws (the INT to Rolle being one of the worst). He has also been extremely inconsistent with his throwing accuracy. However, he hasn't gotten a ton of support from the running game, which has been very good at times, but useless at other times. The last part is that they have been starting in very bad field position for much of the season, part of which is on the defense in not getting enough turnovers (excepting the 2nd half of the Ravens game). Also, our special teams return unit has been mediocre at best, which doesn't help the field position at all.
I think it was also mentioned earlier, but the Broncos have gone up against some tough defenses - the Rams and Chiefs defenses have improved greatly over last year, and the Patriots and Ravens defenses are consistently strong. Look for the Broncos O to light it up against the Raiders.

by Fnor (not verified) :: Tue, 10/10/2006 - 7:49pm

#101: I think "fairly" is a bit generous.

by BB (not verified) :: Tue, 10/10/2006 - 7:50pm

92: I think the Bears heavily contributed to the negative DVOAs of 2 of the 4 teams that have negative DVOAs. Both Seattle and Buffalo were at least 10% positive going into the game and left the week with negative overall DVOA.

As 94 said, Detroit and Green Bay just plain suck. But to retrospectively think the Bears haven't played anyone because they basically singlehandedly gave 40% of their schedule negative overall DVOA is going a bit overboard.

by Pat (not verified) :: Tue, 10/10/2006 - 7:52pm

Any chance we could have a repeat of the Colts DVOA fiasco of last year, where their schedule was soooooo easy, that it became impossible for them to be ranked #1 ?

Doubtful. DVOA was updated with second-order opponent adjustments to compensate for just that.

by Jason (not verified) :: Tue, 10/10/2006 - 7:54pm

Re #92
Brittle? I hate this argument. It's not even an argument. It was a foolish attempt to use facts poorly to build a case to discount the Bears chances before the season and even more foolish after five games of excellent health and passing results (Yes, this coming from the same person who above thinks Rex still isn't as good as his rating). He's had two injuries in four years. Just two. They just both happened to be severe ones that happened at times of the year to miss significant chunks of the season. He made a cut on turf and his ACL blew. It happens to a lot of athletes. A lineman collapsed on his ankle and it snapped. Not really hard imagine that would happen to a lot of guys. It's not like he get's a bump and then misses two games here for a bruised sternum and then rolls his ankle running out of bounds and misses another three. That's McNair. You don't hear he strained his groin in practice or bruised his shoulder trying to throw a bomb. That's brittle. That's injury prone. He's not one of the guys you always hear is questionable or probable, struggling with some new nagging injury every week. He's healthy and producing. To think he's spontaneously not going to be healthy is, to be as nice as possible, a reach.

As for squeaking out a win, first on a play by play basis, they outplayed the Vikes by more than 3 points. Also, maybe everyone else's DVOA is negative because of the beating the Bears gave them and Vikes is positive because they happened not to be quite so dominant that game. Five games isn't much to say a certain performance or numerical analysis exactly pegs their ability. Sixteen isn't even enough. That's why you look on a play-by-play basis, not a game-by-game basis. You can even go further and see that some plays that were successful were lucky to be and discount a team's ability to repeat such a play. The point is, DVOA is a nice metric, but it's not exact, and certainly not so after 5 games. The Vikes DVOA is positive and they hung with the Bears that game, but take it for what it is and look deeper. Look not only at the individual plays, but also how each team has progressed since that game. The Bears have dominated two teams since. The Vikes lost to the Bills and needed to two defensive TDs to beat the Lions. So maybe that tete-a-tete should be seen more as anomaly than evidence the Bears are something to be laughed at come December.

The Bears have been historically dominant so far in just a handful of game. However, temper that with an easy schedule and also what I feel is more good luck than bad and I don't mind discounting them a little (but not a lot) at all. But they're still the class of the entire NFL. And I say that as someone who still sees a lot of plays that could have been better. But, the players and the coaches have earned the confidence that they will address those mistakes and correct them. They're going to get better, which is scary.

by ZS (not verified) :: Tue, 10/10/2006 - 7:55pm

THE G-MEN is clearly ranked too low because they beat the Eagles!!!!! An American Idol style voting system is way better than this. I SHOULD SMACK YOU UPSIDE YOUR HEAD, SYSTEM.

by Ryan (not verified) :: Tue, 10/10/2006 - 7:55pm

Since I had my 10/10 theory hurt..I'm going to challenge the whole "best QB always wins" theory. starting from 2000 in terms of DVOA let's look at how many times the "best" QB won in the Super Bowl. In 2000 Trent Dilfer's DVOA was 39th in the League. In 2001 Tom Brady's DVOA was 17th in the League to Kurt Warner's 1st. In 2002 Brad Johnson's DVOA was 4th in the League to Rich Gannon's 3rd. In 2003 Tom Brady's DVOA was 14th to Jake Delhomme's 22nd. In 2004 Tom Brady's DVOA was 2nd to Ben's 3rd. In 2005 Ben's DVOA was 3rd to Hasselback's 5th. So over the last six Superbowls it's 3-3. So to summarize, the team that plays best wins, not always the best QB.

by Jake (not verified) :: Tue, 10/10/2006 - 8:00pm

This really is torture for us Chiefs fans. Weren't we done after Trently went down and the Bengals anhilated us? Now I have to live and die at least a few more weeks.

by CA (not verified) :: Tue, 10/10/2006 - 8:02pm

In order to have a 50% chance of going undefeated, you would have to give the Bears a 94% chance of winning each game, assuming the same chance of winning every game, from here on out (0.5^[1/11]=0.938931). If one makes the much more realistic although still arbitrary assumption that the Bears have a 75% chance of winning each game, that gives them a 4% chance of going undefeated (0.75^11=0.042235). Note also that, assuming the Bears get a playoff bye, if you give them a 75% chance of winning each playoff game, there is still only a 42% chance that they win the Super Bowl (0.75^3=0.421875). I am making multiple major simplifying assumptions here, but the point is that it is highly unlikely that any team, no matter how good, is going to go undefeated, and if they do, it probably has at least as much to do with luck as with skill.

by Independent George (not verified) :: Tue, 10/10/2006 - 8:02pm

#101 I’m also curious about the playoff performance of teams that skate through the regular season without being tested.

Er, isn't that what Guts vs Stomps are all about? While I agree we're probably jumping the gun on the Bears (and I readily admit my guilt in bringing it up), the combination of the Bears early dominance and their cupcake schedule points towards an historically good season.

by Ryan (not verified) :: Tue, 10/10/2006 - 8:02pm

darn..my silliness put brady and Ben together in 2004...although you could argue that WAS the SB ;). It was sposed to be MCNabb..but it doesn't change the results anyway.

by Jake (not verified) :: Tue, 10/10/2006 - 8:08pm

Does anyone know how much, if any, regression to the mean DVOA does? I've heard Aaron talk about reducing the importance of factors that are wonkier, but I can't seem to find anything detailed.

by Ian (not verified) :: Tue, 10/10/2006 - 8:09pm

The Colts run defense is obviously a concern. It's never been strong and right now its been terrible. And it is likely they'll lose sometime in the near future to a strong running team.

I think it's a mistake though to judge them based on their early season performance. For one thing, this is a team that's been dominant the last several seasons with little to show for it, and is probably getting a little lackadasical about the regular season. Additionally, they are breaking in a new running attack (which has been improving steadily as the year goes on) and several players on defense, and have also had a number of injuries.

If the Colts can get their run defense up to just the bottom of average, the rest of their play should come around.

As to the writer who said that the Bears could handle Manning by getting pressure with just their front 4. What? I have never seen a team consistently pressure Manning just rushing 4. The teams that have dealt with him the best are 3-4 teams that throw a lot of complex blitz packages at him and feature smart, veteran defenders, like the Patriots of the past and the Steelers.

by James (not verified) :: Tue, 10/10/2006 - 8:11pm

To all those people who discount teams playing versus the Bears this year, remember the Chargers from last year versus the Colts.

Did anyone give the Chargers a chance with the "worst" secondary in the NFL against the great Colts aerial attack?


by Peter (not verified) :: Tue, 10/10/2006 - 8:15pm

110: I'm sorry, did anybody actually suggest that the better QB always wins? Maybe next you can investigate whether interceptions reduce the chance of winning.

by Jake (not verified) :: Tue, 10/10/2006 - 8:16pm

A couple more things:

1. Huge props to KUBIAK for predicting the rise of the Falcon D. Except for the NO game, they've been a major fantasy and RL D.

2. I have to wonder how much having Stokely out hurt the Colts so far. Granted, his first game didn't go well for the team, but I want to see how they play once they get back to the offensive set they played in so much last year. 1-3-1.

by Pat (not verified) :: Tue, 10/10/2006 - 8:18pm

darn..my silliness put brady and Ben together in 2004…although you could argue that WAS the SB ;).

Yeah, because when the Patriots beat Pittsburgh by two touchdowns, that clearly meant that it was a closer game than when they only beat Philly by a field goal in a game that was tied at the beginning of all four quarters.

by admin :: Tue, 10/10/2006 - 8:21pm
by Ken (not verified) :: Tue, 10/10/2006 - 8:28pm

Re 95:

I haven't been following the inside sccop to closely, but last I heard they were going to have Hester focus on ST exclusively, and there was as good a chance he would play on offense as on defense.

I wouldn't get too down on Tillman. He can be inconsistent, but he's a pretty good player overall. He's a bigger cornerback, so I don't think he matches up well against the Steve Smiths of the league. You'd probably want him if Terrell Owens (well, TO of 3 years ago) came to town, though. I still don't understand why they didn't have Vasher on Smith last year.

by putnamp (not verified) :: Tue, 10/10/2006 - 8:28pm

I'm really hung up on Seattle's home and road splits and how they're going to play out this year. It's clear that without Alexander (and he was injured in earlier games, so that accounts for some of his lackluster numbers in my mind), on the road, they've got a lot of ground to make up. Are they necessarily any different than they were last year, though? Keep in mind they played very well in 2 playoff games - at home - last year, then made a number of stupid mistakes on the road in Detroit against an arguably inferior team.

I'm not sure Seattle's all that different this year than last, sans the injury to Alexander and the adjustment phase for the OL.

If I were a Chicago fan I'd definitely be hoping that Grossman doesn't fade appreciably or get injured for the rest of the season, too. As long as he's around they look pretty good.

by Ryan (not verified) :: Tue, 10/10/2006 - 8:36pm

Read Statement 102. He said it's a Rarity for the "worst" QB to win. So yes somebody did say "the better QB always or almost always wins".
120. You have to not take remarks too seriously, man. I wasn't even trash talking, I was simply adding humour to my silly error.

by David (not verified) :: Tue, 10/10/2006 - 8:57pm

He said it's a rarity for a team to beat three quarterbacks "much, much better than their own." that's a bit less extreme than saying the better quarterback almost always wins.

by Lou (not verified) :: Tue, 10/10/2006 - 9:02pm

Thanks for including Northwest Indiana, which pretty much considers itself part of Chicagoland. The Region loves da Bears.

RE 119: The colts still run the 1-3-1 most of the time, only with Clark in the slot and Utecht at TE. Though having Stokely back may help.

by tunesmith (not verified) :: Tue, 10/10/2006 - 9:20pm

re #70:

I introduced a "beatfluke" variant to beatpaths.

When you have a beatpath like ARI=>SF=>STL=>ARI, normally the beatloop would take out all of the victories.

However, the beatfluke variant notices that STL has developed an alternate beatpath to SF.

When an alternate beatpath is formed in a way that contradicts (not reinforces) an existing beatloop, it breaks the beatloop at that part of the chain.

So therefore, STL=>ARI=>SF is retained in the graph.

Backtesting has shown that the beatflukes variant is slightly more accurate at picking game results than not using it.

by tunesmith (not verified) :: Tue, 10/10/2006 - 9:23pm

Fox commentary:

And it's silly to rank teams based on X beating Y — for crying out loud, San Francisco beat St. Louis, which beat Denver, but does anyone on the planet think San Francisco is better than Denver?

hee hee.

The beatpaths answer to this is that STL is in a beatloop with SF and ARI, so the SF=>STL victory doesn't count anymore.

Although, I offer no argument against beatpaths being silly. Click on my name to see! Beatpaths was 10-4 this week, which wasn't great (although you could argue that my power rankings picks were 12-2).

by admin :: Tue, 10/10/2006 - 9:25pm

Ah, but tunesmith, you don't rank teams based on X beating Y. You rank based on X beating Y and Z but not W or Q or whatever. More complicated than most people are willing to think about. No offense meant.

by Jim (not verified) :: Tue, 10/10/2006 - 9:26pm

Any team winning games as easy as the Bears are must be a strong team. I've only watched the Bears a few times with Grossman as QB and he just strikes me as the type of QB who will make some truly horrible decisions. Against Minnesota he almost lost the game with a bad pick in his own end, but he also had another sure int for a touchdown dropped by Sharper and 3 other passes that could have easily been picked.

I think Grossman will be the undoing of the Bears when they get into the tough games in the playoffs.

by Jim (not verified) :: Tue, 10/10/2006 - 9:35pm

Just a thought here, but isn't some of this analysis a little superfluous. I looked at last years stats and the top 4 teams in yards per play differential - yards/play gained on offence vs. yards per play given up on defence:

Seattle .9
Pittsburgh .8
Indy .6
Carolina .5

The four conference finalists. Who were the top 4 DVOA teams from 2005?

Arbitrary breakdowns, like performance in the red zone, given subjective value tend to sku real statistical value. It's like giving value to clutch hitting in baseball when in truth it's been proven over and over again it doesn't exist.

by MDD (not verified) :: Tue, 10/10/2006 - 9:39pm

Seattle also hasn't played with Jerramy Stevens yet this year. Not that losing the top 2005 DVOA reciever from a team would have much of a differece.

by Pat (not verified) :: Tue, 10/10/2006 - 9:41pm

Aaron, tunesmith:

Plus, of course, what you were really saying is that rankings should never be influenced solely by the fact that X beat Y. Especially if it occurred in the previous week. You can't ignore the rest of the season, which is what most people who do power rankings do.

by Pat (not verified) :: Tue, 10/10/2006 - 9:42pm

given subjective value

What makes you think that anything here is given subjective value, rather than objectively determined weights?

by BlueStarDude (not verified) :: Tue, 10/10/2006 - 9:45pm

Good job as always Aaron. But, regarding Pat Watkins, wasn't he beat because of mental mistakes not physical limitations? It seems to me he has been better than adequate so far in coverage this year - and for a second-day draft pick who wasn't expected to come in and start right away (though FS was a spot up for grabs), that's not so bad, and it should only get better from here. As for the coaching: on one play you have the perfect blitz called, but Ellis whiffed (yes, McNabb moved, but barely - Ellis should have disrupted that play); and on the other, Williams should have been there to help but for his bad angle (say what you want about Williams in coverage, one thing he has been good at over the years is getting over to someone else's guy to make a play).

by Scott de B. (not verified) :: Tue, 10/10/2006 - 9:56pm

Patriots making slow but steady improvement. That's good to see.

When was the last time a team dominated (I know it’s real early) its team like the Bears? They score the most points per game and allow the least points. That’s domination on both sides of the ball.

According to SportsCenter, the last 5-0 team to be #1 in the league in points scored and allowed is the 1964 Bills.

by Scott de B. (not verified) :: Tue, 10/10/2006 - 9:58pm

Ah, but tunesmith, you don’t rank teams based on X beating Y.

Tunesmith does. His whole project is a test to see how well a ranking system like that does. It's completely objective. The answer: it doesn't do too bad.

by Will Allen (not verified) :: Tue, 10/10/2006 - 10:06pm

Aaron, I'm going to risk the FO curse by noting again that PFP 2006 had the Vikings with 5.9 projected wins (with a 30% chance of winning four or less), which did factor in a projected schedule strength of 27th, so the projection was not that they would be average, but really very lousy. Thus when you say "The Vikings are an average team gifted in the past by early-season luck and gifted in the future by an easy remaining schedule.", it should be noted that the easy 2nd half schedule was part of the win projection, as were the tough remaining schedules part of the win projections for the Rams and Saints.

My point? To be honest, I dunno. Maybe that your projection for the Vikings is already as off target as it is for the Rams and Saints, or more so, once remaining schedules are factored.

As to the luck factor, is an average team which has played five close games really lucky when they win three, and two of those at home? I'm too lazy to check right now, but their fumble luck at this point seems about average, but I think their opposition perhaps has missed more field goals. It just doesn't strike me that they have had unusually bad or good luck. Just average, like their performance.

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

Seattle .9
Pittsburgh .8
Indy .6
Carolina .5

The four conference finalists.

Only if "Indy" is short for "Denver."

by Ali Nagib (not verified) :: Tue, 10/10/2006 - 10:13pm

Where did everyone get this idea that A) you have to have a good QB to win the Super Bowl or B) your QB has to play well in the Super Bowl to win? Did anyone watch the last one? I mean, Big Ben is a Proven Veteran QB, and he played great to boot, right? Oh, that's right....neither of those things is true.

The Bears have a GREAT defense and right now the offense and special teams are matching it. Even if they regress to average in offense AND ST, they'll still be a playoff team. They certainly aren't going to win it all, and even with the #1 seed I wouldn't give them more than a 20% chance to win once they get there, but they are definitely the best team right now, and can certainly win the Super Bowl even if they aren't at the end of the season. The TEAM needs to play well for 3 games in Jan/Feb, and one player (especially Grossman) could tank the team, but even if he plays poorly, they can still win. They won 11 games last year with Kyle Orton. KYLE ORTON!

by andrew (not verified) :: Tue, 10/10/2006 - 10:15pm

I don't suppose one of the things on the horizon would be some kind of java applet listing the ratings that lets us sort them by clicking on headings or something... (barring them just being avail as a spreadsheet, which I don't know ift hat's been asked before).

by D (not verified) :: Tue, 10/10/2006 - 10:17pm

"Should Chicago fans be worried that Grossman’s completion percentage has gone down every game this season? He’s gotten away with it so far because he completes enough long passes to keep his YPA relatively high but it kind of reminds me of Eli last year. Eventually those incomplete passese are going to fall into defensive backs’ hands instead of to the ground.
That said right now he’s playing like the second coming of Dan Fouts so if he keeps his number ups he’s an MVP candidate."

I posted that in last week's DVOA discussion thread, and after the Bears-Bills game (which I was at) I can say that I still stand by my statement. There were three dropped INTs and Grossman is continuing to make questionable decisions. He is ignoring recievers that are open short in favor of going for the deep ball. Given the Bears schedule, it shouldn't really matter since they will continue to dominate opponents, but the more he succeeds with his go-for-broke style the more bad habbits he is going to develop. That will cause him problems in the playoffs.
(It was still a really fun game though. GO BEARs!)

by D (not verified) :: Tue, 10/10/2006 - 10:21pm

All that said, I have been wrong before and certainly could be in this case.

by tunesmith (not verified) :: Tue, 10/10/2006 - 10:32pm

I guess the best way to say it is that beatpaths very much does rank based on X beating Y unless there's one of those contradictions. But when there's no contradiction, it's entirely based on who beat who. I think there are only three games (sf=>stl, was=>jac, and one of the new orleans games) out of the entire season so far that will contradict this week's power rankings.

No offense taken, if I've ever phrased beatpaths as a competition with dvoa it's been tongue-in-cheek. Besides, comparing DVOA to beatpaths is kind of like a smart guy (dvoa) entering a political debate with an idiot. I joke. But I think beatpaths is a fun complement for dvoa because dvoa seems to pay attention to everything BUT the wins/losses. Mine pays attention to wins/losses and... uh, nothing else.

Aaron, I'm still confused about dvoa predicting games. I've read before that the power rankings really aren't supposed to do that, but you do make reference once in a while to dvoa's season record. So does King Kaufman. What picks are those? I'd love to follow those week to week. Beatpaths is something like 12th right now in King's panel o' experts.

by Lou (not verified) :: Tue, 10/10/2006 - 10:53pm

RE D, I'm definitely concerned about Grossman's play, and the risks he's willing to take. Also I'm concerned about the running game against a good D, and I'm concerned anytime we try to return a punt. Lastly, I'm very concerned the Bears start to believe their own hype. But, theres plenty of time for them to learn and improve.

by Becephalus (not verified) :: Tue, 10/10/2006 - 10:57pm

Hmmm, I am wondering what MIN future looks like. If I was to be pessimistic I would try this.


So 5 more wins? Does this seem plausible for a pessimistic estimate?

by D (not verified) :: Tue, 10/10/2006 - 11:00pm

By the way if anyone wants to relive one of the best SNL sketches ever, click my name.

by Phil (not verified) :: Tue, 10/10/2006 - 11:09pm

Not that it matters that much, but the first player that I saw rockin' the mohawk was Freddie "I'd like to thank my hands for being so great" Mitchell. I think it was during the NFC Championship Game two years ago.

by Sean (not verified) :: Tue, 10/10/2006 - 11:39pm

"Where did everyone get this idea that A) you have to have a good QB to win the Super Bowl or B) your QB has to play well in the Super Bowl to win? Did anyone watch the last one? I mean, Big Ben is a Proven Veteran QB, and he played great to boot, right? Oh, that’s right….neither of those things is true."

Yes, but before that game, Roethlisberger was lights out against Cincinnati, Indianapolis and Denver. Pittsburgh doesn’t get a sniff of the Super Bowl if Ben didn’t play his best football in the AFC playoffs. Yes, in any given game the best team wins, yada yada yada, but I have a really hard time seeing Rex Grossman outplay Matt Hasselbeck or Donovan McNabb in a playoff game, regardless of the venue. Again, Grossman is playing over his head and he’s having every break go his way. He’s also been in the best possible situation all year long, with a defense and a special teams that has been giving him all the support in the world. He’s played nothing but downhill football. I don’t know. I watched him be absolutely dreadful all preseason long, and I’ve watched him throw bad passes while falling backwards that somehow hit off defenders and get caught by receivers for 30 yard gains. Either he’s going to improve or the offensive performance is going to decline. I would bet heavily on the latter.

I’m not suggesting that a young quarterback can’t be very good and/or very successful. I think San Diego is the best team in football and that Philip Rivers is a part of the reason why. I just don’t think that Rex Grossman is very good, regardless of how he is playing at the moment.

by dbt (not verified) :: Tue, 10/10/2006 - 11:42pm

I've watched Grossman very closely all 5 games, and while I'm not willing to say his peak is going to be Favrian or Marinish, I think his decision making has already gotten a lot better from the Minnesota game.

Also, keep in mind that crowd noise was a huge factor in the performance of the offensive line and the audible ability of Grossman.

The only two games left in domes (assuming home field in the playoffs) are St. Louis and Detroit in the last 4 weeks of the season, and by then he had better be ready to deal with it.

People mentioned injuries. The team has already lost their starting free safety but he was already being demoted and (I think) being groomed as Mike Brown's replacement at strong safety. The secondary is already better.

The unit that worries me most injury-wise is the offensive line. They play well as a unit but none of them are particularly strong standouts and most of them are old.

by Will Allen (not verified) :: Tue, 10/10/2006 - 11:46pm

Becephalus, I'd definitely say they could end up with eight wins. Heck, give'em a few more injuries and less that eight is a real possibility. I said from the beginning, however, that nine wins was the most likely outcome. I saw a team with above average lines on both sides of the ball, injuries notwithstanding, and vastly improved coaching on both sides of the ball. Given a weak schedule, it was hard to see where the PFP 2006 projection of a 30% chance of four wins or less, and a 30% chance of five or six wins, and only a 16% chance of nine wins or more, was coming from. From Aaron's comments, I think injury prediction played a large part, and this still may be proven correct; the Vikings have already suffered season ending injuries to three players expected to start or have significant playing time. I have no way of agreeing or disagreeing with injury predictions, because I don't have any way of predicting them myself, other than to say that it is not unreasonable to think that a guy like Brad Johnson, who has never been Mr. Durable, ain't going to improve in this department as he gets older. I really wish the Vikings had a better back-up than Brooks Bollinger, and kid who didn't get most of his college snaps at a big school.

Aaron also rated the Vikings as having one of othe worst coaching staffs in the league, I'm sure mostly due to the fact that it was untested. What was perhaps missed was that it was extremely unlikely that a new staff would not be a huge improvement on the mess the previous ownership had created. That isn't a harsh criticism; it simply is difficult to quantify the effects of ownership.

As it turned out, I think they hit the jackpot in defensive coordinator Mike Tomlin. I think it likely he'll be head coach well before he turns 40. I am less enthused about their offensive coaching, but that may just be a reflection on not having too many playmakers on offense.

Anyways, there's still plenty of time for PFP 2206 to be proven more acccurate than I in regards to the Vikings, but I like my chances right now with them at 3-2, and by far the toughest part of their schedule over after seven games. I won't be totally shocked to be having crow for Christmas, however.

by dbt (not verified) :: Tue, 10/10/2006 - 11:50pm

Sean, he's been in the league for 4 years, practiced with the same OC and receiver corps for the first time in even longer, has good arm strength and has been getting more accurate as the season goes on.

He also has to outplay hasselbeck or mcnabb playing against the Bears defense, which isn't has hard as in a vacuum.

San Diego is a pretty good football team but I don't think they have the right coach to make it to the super bowl. I'd be surprised if they make the postseason, even as talented as they are.

by BillWallace (not verified) :: Tue, 10/10/2006 - 11:59pm

Excellent gratuitous Michael Irvin insult.

I think that the long term results of Seattle and St. Louis will be a decent commentary on DAVE.

by Marko (not verified) :: Wed, 10/11/2006 - 12:06am

Sean, if you're going to criticize Grossman based on his preseason performance, how can you be so high on Rivers? Rivers was awful against the Bears in the preseason. It took only three plays for him to throw a pick six right to Brian Urlacher.

by Will Allen (not verified) :: Wed, 10/11/2006 - 12:07am

I wonder if the home field in the playoffs, including the conference championships, does not convey a greater advantage when the home field quarterback has fewer than a couple of season's worth of starts. That is, is getting the home field in the playoffs more important when the qb is inexperienced? My gut says yes, but that's never been a solid gold indicator.

by CA (not verified) :: Wed, 10/11/2006 - 12:12am

Re: 138

Will, to be fair, to the extent that the Vikings are outperforming the expectations set for them in PFP 2006, they really aren't doing so for the reasons that you and other Vikings fans suggested that they might.

The much-ballyhooed left side of the offensive line has been a tremendous disappointment, with McKinnie, Birk, and especially Hutchinson looking grossly overpaid. I say especially Hutchinson not because he has been especially bad -- he has been just mediocre -- but because he's got a $50 million contract as a freakin' guard. I just can't get over that. Anybody else think the Vikings should try moving him to RT, where Marcus Johnson has struggled, to put it kindly?

Brad Childress continues to succumb to the fallacy that prompted the formation of this great site, that you need to "establish the run" to win. The rushing attack has been subpar, certainly far from the dominant running game that you thought the Vikings were building. The passing game also is anemic. Brad Johnson every week is demonstrating that he is washed up. The lack of a true number one receiver is obvious. All in all, the offense developed by Childress, the leader of the coaching improvement that you have trumpeted, what's the word? ah yes, sucks!

Surprisingly, the strength of the team has been the defense. Smoot has improved from godawful to merely bad, and Winfield seems to have broken out of his 2005 funk as well. Kevin Williams has returned to his 2004 form. Although many people, myself included, thought that the current personnel was ill-suited to the Cover 2, the new defense seems to be working quite nicely, especially considering how early it is in the season. Most shockingly, the linebackers, who seemed coming into the year to be among the worst in the NFL and for whom I heard very few Vikings fans holding out much hope, have played quite well.

I guess my point is that if the Outsiders were wrong about the Vikings, so was everybody else, because the Vikings have failed in many of the areas that their backers cited before the season as strengths and have succeeded in areas that nobody expected. And oh by the way, they're still not very good.

by mediator12 (not verified) :: Wed, 10/11/2006 - 12:28am

Really interested in the Blog on Denver's defense. They have given up 34 points in four games and 12 of those have been in FG range from TO's. They are also the best team in the redzone at preventing TD's. They have caused two Red zone TO's and blocked a red zone FG attempt as well.

It really seems like their bend but do not break style has been penalized for giving up 9,6,7,and 3 points per game.

BTW, I am still trying to understand anyone who says the number 4 Rushing offense in the league is not helping Plummer ??? That is after playing 3 of the top 10 rush defenses in 4 games to boot.

by Will Allen (not verified) :: Wed, 10/11/2006 - 12:45am

Uh, CA, I'm certainly not surprised that defense has been the strength of the team. I thought it was pretty obvious that when they signed Hutchinson and Richardson, with Brad Johnson for a qb, that they were going to try to run the ball a lot, pass conservatively, and play very sound defense. If healthy, they had well above-average interior defensive linemen, and defensive ends who had a chance to be above average.

I can't stress enough how poorly coached this team has been in recent years, and not just because of the head coach. I was saying this before the start of the season, and it really has shown up in how the linebacking and defensive backfield has improved.

On offense, they haven't been good, but it is entirely premature to declare the Hutchinson and Mckinnie signings mistakes. Mckinnie is definitely an above-average olt in pass protection, and you just wait and see how that position is going to sky-rocket salary wise in the next year or two, with the bigger cap numbers. Mckinnie won't be on top for very long. Hutchinson hasn't been great, but he hasn't been nearly as bad as you indicate. Offensive linemen often take more than five games to begin working well together. In two years, I predict the Hutchinson signing won't be viewed as an error. People are failing to adjust their salary expectations to the new reality of the much larger cap.

Their biggest problem on offense has not been Johnson or line play, but receivers who have been terrible. Besides not making any notable plays, they have been dropping passes with great frequency. Drops have cost Johnson at least two touchdown passes, and about 250 yards. Pretty hard for playcalling to adjust for that. If they fall short of eight or nine wins, it'll be the receivers who are the primary cause. Relying on a drunk like Koren Robinson was engaging in too much optimism.

I predicted the Vikings would be well above average on defense, and as to whether they are a good team, I'd rather not debate the semantics of that, but simply just estimate how many wins a team will get, given a particular schedule.

by Will Allen (not verified) :: Wed, 10/11/2006 - 12:53am

Oh, also, when a team lacks playmakers at wide receiver, running the ball gets a lot more difficult. Now, if you want to fault Childress because he doesn't have those playmakers, in part due to counting on Koren Robinson, fine, but don't fault playcallng which reflects the personnel on the field. The Bears' playcalling didn't suck last year; their quarterbacks did.

by jules (not verified) :: Wed, 10/11/2006 - 12:58am

157. Mediator I'm on board with your explanation of Denver's defensive DVOA ratings, but I am looking forward to any insight from Aaron on this, lets call it an anomaly.

I am absolutely puzzled at Plummer's performance. I realize Rod Smith isn't the supra weapon of previous years and Walker isn't the deep threat that Lelie was and he is still learning the system, but there is no explanation I can think of to explain Plummer's performance.

Jake was known to fall apart when he was pressured, but now he has regressed into rookie Kyle Orton circa 2005/06.

I don't want to jump on the "Start Cutler Now!" bandwagon, but it does make you wonder. If Pitt could make a serious run with the rookie Big Ben (2005) and Rivers seems like the real deal, could Cutler really do any worse?

by Crushinator (not verified) :: Wed, 10/11/2006 - 1:03am

The Cover 2 D seems to work with teams that overcomplicated their Defensive play book. Most Cover 2 Ds have a pretty bare-bones play selection.

When Dungy went to Indy, the Colts defenders already felt they understood the D and they rarely do seem to be caught out of position. Not that the Colts D is really a model of success...

by navin (not verified) :: Wed, 10/11/2006 - 1:18am

Re: 127/tunesmith,
I can understand having STL>SF because of the beatfluke, but I still think that ARI>SF should be cancelled out, unless there's something else about Arizona's season that shows that Arizona beating San Fran wasn't a "fluke" (which DVOA says *was* a "fluke").

by Will Allen (not verified) :: Wed, 10/11/2006 - 1:20am

All I know, Crushinator, is that the Vikings' linebackers in recent years couldn't even get lined up right with any consistency, given how poorly they understood the defense that Cottrel wanted to run. There's a reason why Cottrel is drawing his coach's pension, and no longer coaching.

by Mannie Fresh (not verified) :: Wed, 10/11/2006 - 1:23am

91: God you're stupid

92: God you're stupid

102: God you're stupid

109: God you're stupid

78: Somebody needs to smack your English teacher.

by Fnor (not verified) :: Wed, 10/11/2006 - 1:27am

#164: Man, awesome argument. I'm convinced.

by dbt (not verified) :: Wed, 10/11/2006 - 1:30am

164: 109 was a parody. :)

by Mannie Fresh (not verified) :: Wed, 10/11/2006 - 1:45am

must be a comedian....

by Mannie Fresh (not verified) :: Wed, 10/11/2006 - 1:54am


did you mean #78 was the comedian?

by Pat (not verified) :: Wed, 10/11/2006 - 2:11am

Mine pays attention to wins/losses and… uh, nothing else.

It'd actually be interesting to compare the results of the beatpath rankings with other pure win/loss systems, like Elo's and Laplace's method. I'm especially fond of Laplace's method given the simplicity and intuitiveness of the idea. Might actually give an idea for a better beatpower ranking, too.

by Sean (not verified) :: Wed, 10/11/2006 - 2:12am

Re 152: That’s true, but I also don’t see any reason to expect this Bears defense to be an immovable object for a Hasselbeck or a McNabb. Much of the hype for this Bears team is coming from their performance against Seattle and rightfully so. That said…I’ve seen that game before. When the up-and-coming team gets a nationally televised home game early in the season, they have a tendency to put the wood to whoever it is they play. Seattle isn’t in form yet, things snowballed early and that was that. It happens. It was every bit as predictable as that the Eagles would beat Dallas in the TO game. It doesn’t mean they’ll beat Dallas again. It just means that the circumstances were right. I’m not saying this Bears defense isn’t very good, because obviously they are. But I wouldn’t draw the conclusion that 16-0 is around the corner or that no one can beat them and beat them good. Carolina looked totally overmatched in the regular season, then they turned around and humiliated that defense in the playoffs. A good team can beat them, at home or on the road.

Re 154: Rivers had one bad game in the preseason; Grossman was soundly beaten out by Brian Griese. Rivers was also a much better college player, which factors into my expectation that he is more likely to sustain good play throughout the season. And while I haven't seen every snap of either player, from what I've seen, Rivers has been notably better in the regular season with his decision making.

If Grossman keeps it up for another four games or so, I'll change my tune. But not yet.

by Mannie Fresh (not verified) :: Wed, 10/11/2006 - 2:13am


These experts sure know the difference between Rex Grossman and Brian Griese...

In case this NFL front page changes, they have Brian (#14) on the front page as a link to thier "expert" power rankings instead of Rex...I know that the Bears have great depth, but a backup Qb who only completed 5 passes on the front page?

Please tell me why Football Outsiders are with Fox Sports...

by jurb (not verified) :: Wed, 10/11/2006 - 2:49am

Lot's of Super Bowl Shuffle references this week. Even as a Bear fan, I'm not ready to go down that road yet.

True fact: The Super Bowl Shuffle won a Grammy.

by D (not verified) :: Wed, 10/11/2006 - 3:29am

The Super Bowl Shuffle was nominated for a Grammy but lost to a Prince song.

by Moose (not verified) :: Wed, 10/11/2006 - 4:34am

Okay, don't laugh...but you heard it here. Upset pick: the Bears will lose to the Cardinals this weekend.

by Moose (not verified) :: Wed, 10/11/2006 - 4:43am

I forgot to mention. I am not a Cardinals fan.

by James C (not verified) :: Wed, 10/11/2006 - 8:05am


You do know that Fitzgerald is injured.

by James C (not verified) :: Wed, 10/11/2006 - 8:19am


The reason that you had not heard about Tillman being injured is that ut was far easier for talking heads to spout on and on about how they needed to upgrade the secondary, draft a corner, change defensive coordinator or shove their heads so far up Steve Smith's behind that they can taste what he is eating (although I do think he is a great player). It isn't as flashy a headline, 'Top wideout dominates beatup defense without adequate depth'. Similarly , 'Manning signed for (realistically) $10.5m for three years to provide depth' doesn't quite grab the attention as well as, 'Manning signed for $23m (most of which he will never see) to challenge Tillman.'

by Jim (not verified) :: Wed, 10/11/2006 - 8:56am

139 - David,

Sorry my mistake. However, Denver was tied for 4th with Carolina at .5.

I think the point is that yardage per play differential is very significant.

by Jim (not verified) :: Wed, 10/11/2006 - 9:07am

134 Pat...from the DVOA definition section:

"We then expand upon that basic idea with a more complicated system of "success points." A successful play is worth one point, an unsuccessful play zero points. Extra points are awarded for big plays, gradually increasing to three points for 10 yards, four points for 20 yards, and five points for 40 yards or more. There are fractional points in between."

How in any way is that objective?

by Starshatterer (not verified) :: Wed, 10/11/2006 - 10:10am

Please tell me why Football Outsiders are with Fox Sports.
Fox Sports is paying them for more articles than ESPN did.

They used to have a column (part of a column?) a week on ESPN's Page 2. Now they have about six a week on Fox. Even if the money was exactly the same (which I doubt), writers and analysts tend to prefer getting more of their work published.

by Fnor (not verified) :: Wed, 10/11/2006 - 10:26am

re #179

I think we should define a success as 5 yards, because that makes good sense to me.

After fiddling with the numbers, it seems that defining 5 yards as a success gives us the best correlation between our statistics and actual quality, when only adjusting that variable.

See the difference? One is whimsical, the other is done because, regardless of the author's feelings, it gives the best result.

by Fnor (not verified) :: Wed, 10/11/2006 - 10:27am

#180: Not to mention that they had to share a column with Skip Bayless.

by admin :: Wed, 10/11/2006 - 10:54am

With all of this discussion of whether DVOA is or is not more accurate than other ratings systems, I wanted to point out something important, something I mention every so often.

DVOA is not valuable because it is the most accurate total power rating. In fact, I would not be surprised if there was another stat-based rating out there that was slightly more accurate when it came to determining that Team X was slightly better than Team Y. (That's not to say I don't do a ton of work to make it as accurate as possible, of course.)

DVOA is valuable because it breaks down every single play. And because it breaks down every single play, it can be broken down into every single play -- and into groups of plays, and situations. Thus, you can use it to analyze matchups, to analyze specific players, to look at how teams do on different downs, or in the red zone, or in the second half. You can use it to figure out why certain teams are winning instead of just saying "Team X is the best team in the league." You can use it to figure out trends that help you to predict how teams will develop in the future (i.e. the third down off-season thing).

No game in the NFL is decided by the fact that Team X is, in total, better than Team Y. It's ALL about matchups and specifics. DVOA can analyze that. Sagarin's rating, Beatpaths (no disrespect to you, Tunesmith) and the other stat ratings can not.

Sometimes, the fact that we run a list of teams from 1-32 each week will obscure this important fact.

by Kaveman (not verified) :: Wed, 10/11/2006 - 10:55am

#157: I'm waiting for that particular blog quite eagerly myself. Denver's ratings do not jibe with how they've played, IMO.

#160: Oh no please, stay off that bandwagon. The Steelers offense can be dumbed down and continue to work, apparently, but the Broncos offense is different. Shanahan schemes for formations, matchups, past plays... no rookie QB is going to understand the whole playbook, much less the gameplan. Can Cutler send a receiver in motion and read the defense from its response? Recognize all the matchups he needs to? Plummer was no rookie, and he only got comfortable in the Broncos offense in his third year. And as a clinching argument: our schedule still has the Colts, Steelers, Bolts (twice), Hawks, Bengals and KC in Arrowhead. No rookies please.

Regarding the Broncos' play so far, the Rams game was bad on several fronts:
- The gameplan was apparently very complicated. Cutler looked and sounded very relieved that he wasn't the one who had to run it. And we saw how well Plummer did with it. Shanahan out-schemed himself.
- The O line played horribly. 4 sacks? After giving up, what was it, 15, all of last season?
- The D line had gap coverage issues in the first half and gave Steven Jackson too much room early on. This got fixed later in the game though.
- Besides Paul Ernster, special teams has been terrible, as everyone has noticed. They have (hopefully) now returned to average against the Ravens.

The other three games... that KC defense is pretty good, hey? And we beat the Pats reasonably authoritatively, and those unbeaten Ravens too. So yeah, I don't think DVOA is doing justice to the Broncos at the moment. But that's okay. We haven't done anything much yet.

by Pat (not verified) :: Wed, 10/11/2006 - 10:56am

How in any way is that objective?

Fnor's right.

Incidentally, yards per play has been looked at before. It's even in the FAQ. It doesn't work as well as DVOA does, especially for defense.

Plus it completely and totally ignores special teams.

by DGL (not verified) :: Wed, 10/11/2006 - 11:00am

#179: It's "objective" because, for example, every 40-yard play is worth 5 points. A 40-yard play isn't worth 6 points for the Pats (becuz T0m Brady Rul3z!!) and 3 points for -- oh, who is Aaron supposed to hate this week? Let's go with the Rams. The system is "undistorted by emotion or personal bias; based on observable phenomena."

The assignment of success points to play results requires some judgement, but merely requiring judgement does not make something subjective. If Aaron were to have said, "Well, the Pats always seem to get about 3 yards on first down, and the Pats are great, so let's call 3 yards on first down a success," that would be subjective. On the other hand, tweaking the numbers in the model until the DVOA ratings show the strongest correlation with actual results is pretty objective.

by Independent George (not verified) :: Wed, 10/11/2006 - 11:04am

#180 - Even if the money was exactly the same (which I doubt), writers and analysts tend to prefer getting more of their work published.

That's because writers are stupid:)

Hextuple (is that a croumlent word?) my workload for exactly the same amount of money? Brilliant!

by Kaveman (not verified) :: Wed, 10/11/2006 - 11:12am

On Jake Plummer: I think other teams have keyed on the fact that he has been mediocre or worse in the pocket. Shanahan has tried to address this: he called straight dropback plays for a full preseason game, trying to get Plummer more comfortable in the pocket. Didn't work. He just seems to be more inaccurate; the strangest thing.

So defensive coordinators have noticed and have taken away the bootleg. Now if I understand it correctly, doing this leaves a cutback lane open for a running play. But the Broncos don't seem to have been able to exploit this. It does seem like they have been calling a lot of end arounds and reverses, but I don't know if this has anything to do with the bootleg. Anyone know?

by Drew (not verified) :: Wed, 10/11/2006 - 11:23am

The Bears could win every game this season. No, I don't think they will go undefeated. What I mean is there is no single game on the schedule where you can definitively state that they have a poor chance. They will go into every game with a good shot at winning.

I don't think, as a fan, you can hope for better than that.

by Wanker79 (not verified) :: Wed, 10/11/2006 - 11:31am

Re: 179

Because those gradations weren't just randomly pulled out of a hat. They were found to create the strongest correletion to points(?).

by Wanker79 (not verified) :: Wed, 10/11/2006 - 11:32am

I should learn to refresh before responding.

by Mannie Fresh (not verified) :: Wed, 10/11/2006 - 11:32am

I'm thinkin twice all the time that the Bears will go undefeated...probably too much love for my bears...although there isn't ONE game that looks unsurmountable, one of them little critters will beat us this year. Look at last year, we lost to Cleveland. We hit rockbottom then, and won a like 6 games in a row. This year, a team like St. Louis on the road...on Monday Night Football (man I really don't feel jittery when I say it after it moved to it new house). I think they can surprise us. We'll see if the St. Lunatics have it when they face Seattle this week...

by rk (not verified) :: Wed, 10/11/2006 - 11:42am

The Bears are a lot like the 2004 Steelers. They have a young QB who is inexplicably playing way better than anyone expected, a dominant defense, a vertical passing game, and a sound power running game. They were both good early in the season then exploded to the top of everyone's power rankings by blowing out an undefeated conference rival at home. The Steelers kept winning despite a decline in performance then got knocked out by New England, the team they blew out in the regular season, at home. I could see the Bears declining in performance (they almost have to) but still winning just about every regular season game then losing to Seattle in the playoffs.

by Mannie Fresh (not verified) :: Wed, 10/11/2006 - 11:44am

Here's a question. What IF, WHAT IF *please no jinx* Grossman got injured (I know unbelieveable right?), how far do you think the Bears can make it this year with Brian Griese (ya that QB the Bears wanted to start ahead of Grossman in the preseason)? They will definitely do better than last year (11-5). Brian is a good QB (not that it means anything, but he's 5 for 5 this year passing). He's A LOT better than Orton, so do you think they can reach thier ambitions of playing late in the post-season?

by TomC (not verified) :: Wed, 10/11/2006 - 11:55am

On MNF, Theismann predicted (at least half seriously) that the Bears would go 19-0 (including playoffs & SB). That fact alone guarantees a regular season loss and seriously dampens hopes of playoff success.

by Tom S (not verified) :: Wed, 10/11/2006 - 12:20pm

So da Bearse could generate a good study considering the theory that big wins over weak opponents are a stronger indicator than close wins over strong opponents. Bears have the weakest schedule in the NFL this year (based on last year's winning pct) AND the teams are not showing any improvement over last year's record. So is this "The Perfect Storm?" Two waves that coincide to make the biggest wave possible?

by Goathead (not verified) :: Wed, 10/11/2006 - 12:28pm

I think its interesting that everyone assumes da Bears would rest starters if they got to say 14-0, since Indy indicated they would. Did last season work out really well for the Colts? I think if you know there's a bye week coming, go all out don't wuss out like the Colts...

by Marko (not verified) :: Wed, 10/11/2006 - 12:38pm

Mannie, the Bears didn't want to start Griese over Grossman in the preseason. Some Bears fans and many in the Chicago media wanted Griese to start, but the Bears (players, coaches and front office) were solidly behind Grossman the entire preseason. So far, their faith in Rex has been rewarded.

by dbt (not verified) :: Wed, 10/11/2006 - 12:40pm

#193: the 2004 steelers were starting a rookie quarterback playing over his head. Grossman is a 4th year quarterback who is developing on schedule and was projected to do well. He has good touch on the deep ball, good accuracy, rapport with his receivers and is working on his decision making.

I do think it's hilarious that someone actually argued that Grossman is basically a rookie and would clearly suck, then say that the Chargers are the best team in football.

(They're clearly the best team in the AFC right now, but let's wait and see how the coaching staff progresses.)

I was enthused, by the way, to see Jamie Dukes on the "NFL-N" last night raving about Rivers because of his awesome college completion rate. Nice to see that someone gets it.

by Mannie Fresh (not verified) :: Wed, 10/11/2006 - 12:41pm

Ya...Grossman came into Minnesota, thier last game last year with the hot arm. He sat out that game, even without a lot experience. He then faced Carolina, and got ripped in the playoffs. I think the Bears will stay competitive in all the games to stay on the edge at all time. They'll only rest those who are injured. They'll atleast play one half of the last game.

by Marko (not verified) :: Wed, 10/11/2006 - 12:43pm

Goathead, I don't assume the Bears would rest starters if they got to 14-0 or wrapped up home field advantage early. They did that last year for the final game against Minnesota because they were locked into the #2 seed, and then they came out flat in the playoffs. I think they learned their lesson from that, and from the way the Colts played last year after resting starters. If there are players who are banged up, sure, they'll rest them. But otherwise, I don't think so.

by Goathead (not verified) :: Wed, 10/11/2006 - 12:55pm

Marko, Agreeed, resting Grossman last year was a huge mistake for the Bears, and was obvious to a lot of people at the time. This year who knows, but I'd like to see someone have the guts to go all out regardless. Just ask yourself, what would Ditka do?

by Yakuza Rich (not verified) :: Wed, 10/11/2006 - 1:02pm

Good stuff on the Cowboys. Although you missed that Watkins was also responsible for the LJ Smith 60 yard pass, which Parcells alluded to on Monday. Dallas was playing a 4 across short zone which is used by teams, especially Dallas, to cut off slants and take on screens. Watkins has deep centerfield responsibility and thought Smith was going to run a slant or find a spot in between the zone, something he'd been doing all game long at that point. Smith went deep and Watkins had already taken a couple of steps to help out Anthony Henry.

I'm not too upset with Watkins on a WR as FS are called to occasionally cover WR's one on one. My bigger gripe is that they sent Anthony Henry on a blitz instead of the faster Terence Newman. That being said, outside of the slot corner blitz...I've never been a big fan of the corner blitz anyway.

What's really killing the Cowboys is Bledsoe, especially the turnovers. He's on pace for 32 turnovers this season and so far his turnovers have given opponents an average field position of being at the Dallas 40 yard line. Tough to prevent the opposing team from scoring points when your QB is continually giving them outstanding field position.

by James C (not verified) :: Wed, 10/11/2006 - 1:13pm

Ditka would go bright purple before putting his name to hair spray, personal loans, special pills to 'help you feel like a man again' and assorted other crap.

by Marko (not verified) :: Wed, 10/11/2006 - 1:17pm

I know what Ditka would do because I know what he did. In 1985, the Bears didn't rest players at the end of the season. They cruised into the playoffs by playing the same way they had played all year.

by Crushinator (not verified) :: Wed, 10/11/2006 - 1:20pm

Chicago is a relatively sub-average team in coverage of #1 WRs, on the DVOA by receiver page.

The Panthers still have hope!

by Crushinator (not verified) :: Wed, 10/11/2006 - 1:21pm

Well, hope if they can do well in their next 3 games of @ Bal, @ Cincy, and Dallas.

by Bobman (not verified) :: Wed, 10/11/2006 - 1:23pm

Bears! Bears! Bears! That's all I hear about on this damn website. Clear Great Lakes bias on this damn site! (Well, ignoring all the Favre bashing and Cleveland ignoring and Millen--is he still employed? Oh, where was I?)

Oh yeah, it's like one or two years it's all NE just because they won, like, two consecutive SBs (yawn), then the governor of Indiana puts Aaron on the payroll (aha! You thought we didn't know!) so it's all about Indy and the streak and why Manning might win a 3rd consecutive MVP (double-yawn), and now the Chicago mob is blackmailing FO and it's all Bears all the time. Guys, we all did stupid things in college--don't knuckle-under to the pressure!

I paid good money to access this site and I want --no, DEMAND-- more fair and even coverage--let's see some in-depth discussions of the AZ OL and why Edge will win MVP this year, of Houston's D, of Tennessee's... are they still in the league? Of Oakland's.... okay, I'll stop. The laughing is hurting my ribs.

Yes, I am kidding, but you just mark my words, next season, somebody is going to serously complain about just this issue: "Whine whine whine, you guys love NE, Indy, and Chi. whine whine whine. why don't you talk about team X ever? whine whine whine."

One thing I have yet to see is the mathematical probability breakdown of a 16-0 season in Chicago. You know, like last year we started seeing in week 12 or so for Indy--90% chance they win game 12, 75% for game 13, 75% game 14..... Just offhand and without firing up Excel, I'd say Chi looks likey have a better shot at it than Indy did last year. I really can't say how Indy is winning, but they are--who wants to put money on a SB between two 18-0 teams? How about if they are 90 minutes from each other? Will all casual fans in the country throw their TVs out the window and commit suicide? Not sure, but I guess that's why they play the games.

by doktarr (not verified) :: Wed, 10/11/2006 - 1:31pm

90 minutes from each other? How fast do you drive?

by Fnor (not verified) :: Wed, 10/11/2006 - 1:31pm

#199: I'd disagree with a lot of what you said about Grossman. His decision-making hasn't improved at all, he's merely hit a mix of bad defence and Manning (the younger) luck. I think that during the Minnesota game alone, there were about 6 passes that should have been interceptions but were dropped. Not "should have had that one," but "If that were a WR I wouldn't blame the QB" dropped.

Detroit was better, but not by much. His inaccuracy was made up for by a fantastic effort by Berrian (who is, by the way, awesome). He still one-hopped passes to the flat, threw into bad coverage (I'm still impressed he found bad coverage to actually throw into), and at one point bounced a pass off a lineman's head.

He is still consistently off, usually more over- than under-thrown, which better for avoiding interceptions on the long ball than the short one.

The worst part is, he has a schedule full of patsies to play against and could possibly have all of these very bad habits stick, as others have commented on. I don't think he'll be a great QB, but he could be a pretty good one. If he thinks he can keep playing the way he has thus far, he's going to end up as a bad QB. After all the man-love Muhammad gave him (going so far as to phone in games last year because he was upset Orton was starting) and the general attitude of the city's media, this is a real possibility.

Since I'm gathering ALL the flames unto me, I might as well also say that from what I've seen the Bears have also had some excellent fumble-luck early in the games. I don't have the numbers, though, so it's just a suspicion.

by zip (not verified) :: Wed, 10/11/2006 - 1:37pm

I really can’t say how Indy is winning, but they are–who wants to put money on a SB between two 18-0 teams?

No way in hell Indy is going undefeated. Read AGS this week... or just look at the score against the Titans.

by Travis (not verified) :: Wed, 10/11/2006 - 1:39pm

Re: 208

One thing I have yet to see is the mathematical probability breakdown of a 16-0 season in Chicago.

According to http://www.nfl-forecast.com, the Bears have something like a 1% chance of going 16-0. However, the power rating it uses probably underrates the Bears.

by Becephalus (not verified) :: Wed, 10/11/2006 - 1:41pm

we did this same excercise with IND last year. LET ME REVIEW.

Even if we are going to be generous and say that CHI has an 85% chance of winning EACH its remaining 14 games (assuming superbowl birth).

They have basically NO chance of running the table. So you need to be really really really confident to claim they will run the table. Hell with an 85% expected win % odds would be they would lose 1 of their next 4 or 5 games.

by Pat (not verified) :: Wed, 10/11/2006 - 1:46pm

90 minutes from each other? How fast do you drive?

Depends what you define as "Chicago." Stadium to stadium, no way in hell you're doing it in 90 minutes. But some definitions of Chicago extend virtually all the way to Indianapolis.

by Will Allen (not verified) :: Wed, 10/11/2006 - 2:13pm

I once got tucked in between two nutcases with radar detectors and cb radios, and drove from the Minneapolis city limits to Chicago in less than five hours, as measured by the Cook County line, so it it is that demarcation that I've always considered the beginning of Chicago.

by Jason (not verified) :: Wed, 10/11/2006 - 2:22pm

Re 93 Inexplicably? The guy's a first round QB, been studying NFL defenses and his offense for four years. How is his success inexplicable?

Re 200
Ripped? He put up 21 points. It was the defense (Tillman) that got ripped.

And regarding resting. They'll rest. They're not going to risk injury over some 'intangible' like being flat. Lovie Smith isn't so weakminded to put that above team health. You want your talent to be on the field, not on the sidelines.

Some of these comments are ridiculous. Grossman's success is not inexplicable. Preseason is 'glorified practice' to quote Smith, so let's not weight it as actual performance because it helps your non-case. Despite not being perfect (and who is), the Bears are a dominant team. Despite that, the chances of running the table are still very small. Grossman and everyone else have shown game to game improvement. Barring injury, they'll be favored in every game. If Grossman goes down, not that there's any significant reason he would, Griese's a very capable back-up. That's why they signed him. That's the value of depth. They have it everywhere except maybe LB.

by admin :: Wed, 10/11/2006 - 2:23pm

Click on my name for the Pats commentary on the blog. Denver commentary to come later today.

by mactbone (not verified) :: Wed, 10/11/2006 - 2:31pm

Re 214:
Well, Lafayette's a little over an hour on 65 from downtown Indy, so I don't know how you define "virtually." I'd say that Chicago doesn't extend very far south into Indiana, it's mostly east into Indiana.

by Jason (not verified) :: Wed, 10/11/2006 - 2:31pm

Oh, and the Bears were 11-5 (and that includes punting the final game at Minny) last year with a dominant D, a terrible offense and middling special teams. This year they improved the defense (wow) and both the offense and special teams are top 5 (and that might be understating it), yet some people out there are still trying to justify their preconceived inclinations that this team doesn't have the inside track to Miami? Unbelievable. Anything can happen, of course, but that unknown is the only (weak)argument against them getting there that isn't ridiculous. Every other objective measure so far shows them to be elite.

by James C (not verified) :: Wed, 10/11/2006 - 2:33pm

#188 Kaveman

If a defense decides to try to stop the bootleg it will either send the defensive end on an outside stunt or blitz an outside linebacker to try to prevent the quarterback escaping the pocket. This should make it easier for a tight end or blocking back to isolate the guy on the back side of the play - unless the defender just runs himself out of the play alltogether - this should allow the remaining blockers to open up the cutback lane more easily.

If defensive ends are heading upfield on a regular basis I would have thought that this would scupper the chances of most reverses as he would be stood right in the way of the oncoming receiver. I suspect Shanahan is putting those in to ensure that opposing teams have to account for it. I would look for cutback runs and screen passes to try to take advantage of defenses trying to take away the waggle (bootleg) with their DEs.

by dbt (not verified) :: Wed, 10/11/2006 - 2:53pm

#216: They don't fall off much at sam from Hillenmeyer to Ayenbedejo, but the other two guys are all-pro caliber and their backups aren't much. Hillenmeyer slides over as backup mike.

I have no idea what they do if Briggs is out. Leon Joe? They were grooming Jamar Williams for will (and he may well start there next year if Briggs walks as I expect) but he tore a muscle on special teams and is on IR.

by Will Allen (not verified) :: Wed, 10/11/2006 - 2:54pm

As Aaron noted above, the Temple Owls don't play in the NFL, which makes it exceedingly unlikely that the Bears will go 16-0, but that also means that the Bears completely crushing four out of five opponents so far very significant, in terms of predicting the future, even considering the strength of those five opponents. I still think Philly matches up best in the playoffs, but I probably shouldn't read too much into Seattle's performance against the Bears.

by Travis (not verified) :: Wed, 10/11/2006 - 2:57pm

According to Mapquest, Soldier Field and the RCA Dome are 182 miles away from each other.


The Chicago Metropolitan Statistical Area extends all the way southeast to Jasper County, IN, while the Indianapolis MSA extends northwest to Boone County, IN. MSA definitions Jasper and Boone Counties are less than 70 miles apart.

by Wanker79 (not verified) :: Wed, 10/11/2006 - 3:23pm

Wow. I've just been playing around with the individual numbers. I'm mixing stats regardless of number of touches, so I figure it makes more sense to look at DPAR rather than DVOA, otherwise I get stuff like the top 5 WRs being Maurice Hicks, Shawn Bryson, Dan Campbell, David Tyree, and Billy McMullen and there are a few things that jump out.

1) There are 233 players that appear in receiver stats (WRs, RBs, & TEs). Roddy White is dead last in DPAR (Michael Jenkins is 126th and Ashley Lelie is 143rd). So I guess the answer to the preseason "who is going to be Atlanta's breakout receiver" debate was 'd: non of the above especially not Roddy White'.

2) Chris Chambers is 3rd from last (#231) in receiving with -4.30 DPAR (actually he's tied for 2nd to last but I'll give him the edge over DeShaun Foster because of his higher DVOA), while Ronnie Brown is ranked as the 98th receiver with 1.80 receiving DPAR.

3) But Chambers is ranked 30th out of 75 (RBs & WRs) with 3.3 rushing DPAR. Ronnie Brown is ranked dead last with -7.4 DPAR. Maybe Chambers and Brown should switch positions???

4) I knew T.O. was having a pretty bad year, but I had no idea it was this bad. He's 191st in receiving DPAR. Actually he's in a tie but I gave him the edge since he has a higher DVOA than...wait for it...wait for it...Cory Schlesinger. He's the 101st WR in DPAR, but there are 48 RBs and 41 TEs ahead of him also. BTW, Reggie Brown in 28th, Greg Lewis is 29th, Brian Westbrook is 34th, Donte Stallworth is 55th, LJ Smith is 65th, Correll Buckhalter is 78th, Matt Schobel is 112th, and Hank Baskett is 116th. The only Eagle that makes the list ranked below Owens is Thomas Tapeh, so he has that going for him...which is nice.

5) There are 29 offensive rookies that show up in the tables. When ranked by combined DPAR (rushing plus receiving for RBs and WRs, and passing plus rushing for QBs, not counting Special Teams DPAR) Maurice Jones-Drew and Marques Colston are neck-and-neck at the top with 8.8 and 8.6 DPAR respectively. The top 10 is rounded out by Norwood, Addai, Maroney, D.Williams, Gradkowski, Klopfenstein, Leinart, and Leon Washington tied with Chad Jackson. FWIW, the esteemed Reggie Bush is 23rd with -1.1 DPAR and Vince Young is dead last with -4.6 DPAR (thanks to -3.6 rushing DPAR).

by Goathead (not verified) :: Wed, 10/11/2006 - 3:28pm

Exactly. Ditka understood... Don't mess around, don't overthink it. Take your guys, play hard every week. If you're good enough you'll win the big one. Most dominant team I've ever seen was the 85 bears. I'm way sick of all this talk about resting starters. The ONLY time when I approve of something like that is when you're locked into playoff position and playing a likely 1st round opponent on the last week of the season. Then you hire a bunch of thugs for a week and hope to hurt as many of the other guys as possible :-)

by Fnor (not verified) :: Wed, 10/11/2006 - 3:34pm

re: Resting starters

Somewhere on the site (too busy to look for it now) was a study that looked at the effect of resting starters and subsequent playoff performance. I believe the results of the study were that the decision was irrelevant, but I can't recall off the top of my head.

by TomC (not verified) :: Wed, 10/11/2006 - 3:45pm

Fnor (#210) -

Like a lot of people on this thread (particularly Bears fans), I think your assessment of Grossman is a bit pessimistic --- he's definitely prone to bad throws under pressure, but when you give him time he's actually damned accurate --- but you're dead on about the fumble luck. Things started to even out a bit last Sunday (4 total fumbles, 2 recovered by each team), but the Bears are still way ahead for the season. If all of the payback comes in a single week, the Bears could end up losing a game everyone thought should be a cakewalk.

by Wanker79 (not verified) :: Wed, 10/11/2006 - 4:39pm

Re: 227

You just described Drew Bledsoe. Take that however you'd like.

by Kal (not verified) :: Wed, 10/11/2006 - 4:41pm

On DVOA giving points for 5, 10, 20, and 40 yard plays - why not make it fractional points instead of integral? It seems like this would achieve a bit more overall success in terms of translating to good players than bad. Or were those breakdowns just the breakpoints, and not actually indicative of the overall process?

While the Bears may not run the table - and there are a ton of obstacles for doing so - they certainly look ridiculously strong right now, especially in comparison to the rest of the league.

by turbohapy (not verified) :: Wed, 10/11/2006 - 4:55pm

Re: 119

Stokely rarely saw the field at all last year, so I'm not sure having him out is a huge difference. You're thinking of 2004 when they ran 3 wides a lot.

by Pat (not verified) :: Wed, 10/11/2006 - 5:04pm

On DVOA giving points for 5, 10, 20, and 40 yard plays - why not make it fractional points instead of integral?

It is. It actually says it in the description, right there.

by Wanker79 (not verified) :: Wed, 10/11/2006 - 5:04pm


Extra points are awarded for big plays, gradually increasing to three points for 10 yards, four points for 20 yards, and five points for 40 yards or more. There are fractional points in between.�

by Kal (not verified) :: Wed, 10/11/2006 - 5:08pm

Ah. Thanks, #231 & 232. hooray for reading comprehension!

I am surprised that there's such a big difference between a 1 point play (5 yards) and a 3-point play (10 yards), but I know that this is mitigated by those 3-point plays not being that special when they don't actually count.

by JoeSmith (not verified) :: Wed, 10/11/2006 - 5:09pm

While most of the Grossman criticism here is insiteful some is clearly sour grapes by fans of divisional rivals. He does still make some questionable throws, I don't think anyone can deny that. This is something he has to work on and in my opinion, he clearly has been.

But let's not get carried away. Every time the Vikings game is brought up, the numbers of dropped INT goes up. First it was 3, then it was 4, now it's 6? Why make things up? 6 dropped INT isn't even close to being accurate in any possible way. Are we counting deflected passes at the line of scrimmage as dropped INT's? Saying he wasn't much better and was still inaccurate in the Lions game is patently absurd. He completed 75% of his passes in that game. How can that possibly be construed as being inaccurate? It is impossible to contribute all of that one play that Berrian made in that game?

He does make questionable throws but attributing all of his success this year as luck is a little naive. Also saying things like "He is consistently off" is not quantifiable and likely entirely incorrect. He's 4th in league passer rating so he's likely not "consistently off." These kind of comments add absolutely nothing to the discussion.

I do appreciate some of the insight provided by the comments on this site but there is some axe grinding going on in this topic. It's petty. Grossman is playing well. It's alright to say so, even if you are a Vikings fan.

by Fnor (not verified) :: Wed, 10/11/2006 - 6:01pm


I charted the first half of both the Minnesota and Detroit games. I watched the second half of each of them twice (first on gameday, second in fast-motion when Aaron requested comments so I knew what I was talking about).

The rest of your argument is kind of fluff. We saw Eli Manning last year be innaccurate and have a high passer rating- something we still talk about and a very plausible explanation for what myself and others have suggested in this thread and others.

He did have a high completion percentage in the Detroit game. I said above, however, that this was party due to Berrian generally being awesome, and also partly due to the Lions' atrocious defence. His completion percentages for the games after DET were 56%, 54% and 55%, in succession.

I'm a Steelers fan.

by Jake (not verified) :: Wed, 10/11/2006 - 6:21pm


Last year the Colts had 3 wideouts on the field 80% of the time.

So there's that.

by Ryan (not verified) :: Wed, 10/11/2006 - 6:32pm

Not to say that McNabb isn't awesome...but has anybody looked at his completion percentages the last three games? hmmm Looks almost exactly like Grossman's doesn't it? Also, Grossman has played against the 11th,12th,and 13th rated passing defenses the last three weeks while McNabb has played against the 25th, 24th, and 16th ranked pass defenses the last three weeks. Not all of those poor rankings can be attributed to McNabb's play.

by Pat (not verified) :: Wed, 10/11/2006 - 6:44pm

What about the completion percentages? McNabb's averaging ~16 yards per completion - his completion percentage is bound to be low. Grossman's only averaging ~13 yards per completion.

Plus, poor accuracy for a mobile QB isn't as bad as poor accuracy for a traditional QB. Mobile QBs avoid having to throw into tight spots (and thus throw interceptions) by moving.

by Ryan (not verified) :: Wed, 10/11/2006 - 6:56pm

I'd like to ask a question. Do you believe that DVOA has any merit whatsoever? The stats indicate that Grossman and Mcnabb have played very close to equal this season. So who's to say that any one of them will be better or worse as the season goes on? People on here don't acknowledge the statistics set right before their eyes. Look at the passing DVOA of McNabb's opponents and the passing DVOA of Grossman's opponents. It's not even close. I think we can all agree that MINN, SEA, and BUFF have better pass defenses combined than SF, GB, and DAL, and the numbers indicate that through the first five weeks, that is true. And I like how you say that 13 yards per completion is low, when Grossman is second to McNabb in yards per attempt and second to McNabb in yards per completion. And if Berrian gets his second foot down on that one 30+ yard TD catch in the 3rd Quarter last week, that jumps up Grossman's ypc even slightly more. Btw, as has been stated before Grossman almost never dumps the ball off or throws it away, which lowers his completion percentage. I'll take 60% and 8 Yards per pass over 70% and 6 yards per pass like you see with dumpoff quarterbacks. In conclusion, completion percentage is the most overrated statistic of all QB stats in my opinion.

by tunesmith (not verified) :: Wed, 10/11/2006 - 6:58pm

Re: 162. There really isn't a reason to suggest that ARI=>SF was a fluke, either, though. ARI did beat SF, and there is no alternate reason to suggest SF is better than ARI, so ARI=>SF should remain. There is SF=>STL=>ARI, but it's already pretty clear (given the season so far) that SF=>STL was the fluke. So ARI shouldn't be penalized for that when they've beaten SF head-to-head.

by admin :: Wed, 10/11/2006 - 7:00pm

Click on my name for the blog post on why Denver's defense ranks third in yards allowed but 11th in DVOA. You are going to kick yourself when you realize what the reason is.

by Ryan (not verified) :: Wed, 10/11/2006 - 7:04pm

Well duh Denver's going to be 3rd in Total yards allowed..considering they've only played four games.

by Ryan (not verified) :: Wed, 10/11/2006 - 7:05pm

On the same tone as your blog...the Cubs have gone undeafeated in over a week.

by David Crabbe (not verified) :: Wed, 10/11/2006 - 7:09pm

Your rankings and your comments about the Denver Broncos TOTALLY contradict each other.

Perhaps you didn't notice that the Broncos defeated BOTH the Baltimore Ravens and the Kansas City Chiefs both of whom are SUPPOSEDLY ranked ahead of the Broncos and better(??) than the Broncos according to YOU!

I guess you also closed your eyes when the Broncos defeated the Patriots....IN BOSTON!

by admin :: Wed, 10/11/2006 - 7:14pm

You know, I'm finding the hate mail and angry comments a lot funnier this year. I think getting a normal amount of sleep really helps my mood. The Denver stuff is hilarious. Should I dig out the angry Indianapolis e-mails from the middle of last year, when I had the Broncos ranked number one ahead of the undefeated Colts? Ah, the fickle opinions of fandom.

by Pat (not verified) :: Wed, 10/11/2006 - 7:21pm

And I like how you say that 13 yards per completion is low, when Grossman is second to McNabb in yards per attempt and second to McNabb in yards per completion.

Uh, I didn't say it was low? I said it was lower than 16. So a low completion percentage (for McNabb) means less than for Grossman.

Plus, it's not like Grossman's yards/catch are that far different than Manning (12.0 yards/catch), and Marc Bulger (12.4 yards/catch), especially if you only consider his low completion percentage games (the last three) where his yards/catch dropped to 12.5 yards/catch. And Manning and Bulger are in the low 60s in completion percentage, with Grossman at 55%.

McNabb's are clearly way freaking different, and so having a lower than normal (say, 60% being normal) isn't surprising.

Btw, as has been stated before Grossman almost never dumps the ball off or throws it away, which lowers his completion percentage.

Yeah. But that kind of thing is what gets you interceptions if you're not a mobile quarterback, which Grossman has gotten a little bit lucky with.

In conclusion, completion percentage is the most overrated statistic of all QB stats in my opinion.

I'm not trying to attack Grossman, mind you - but Grossman's low completion percentage can't be explained away by pointing to McNabb.

by Ryan (not verified) :: Wed, 10/11/2006 - 7:26pm

(Resurrecting a supposedly debunked hypothesis of mine) Yesterday I hypothesized that a team that's in the top 10 in both offense and defense would have the best chance of winning the Superbowl, and a better chance than other teams. It was attacked, and I let it go, but then I went through the numbers of the DVOA era and found that 19 teams ranked in both the top 10 in off and def and 6 of these teams won or made it to the Superbowl. That's an average of just 2.375 teams per season ranking in the top 10 in both categories. The percentage of these teams that made or won the superbowl was 6/19 or just less than 1 out of 3. Among the rest of the teams, roughly 1/25 teams that didn't rank in the top 10 in both categories made or won the superbowl. So, it would appear by a small sample size of eight seasons, that a team that ranks in both the top 10 in offense and defense has a greater chance of making or winning a superbowl than a team that does not. So I'm sticking with my pick of either PHI or CHI versus SD. That's right..I'm picking the superbowl after five weeks, before any injuries or chokejobs have happened. Unfortunately...Marty is STILL SD's coach, but I guess I'll let that pass. And I can't be held accountable for my pick, since I'm only using a statistic to make my decision and not making a judgement call. It's bulletproof....

(wishes there were a sarcasm smiley)

by Pat (not verified) :: Wed, 10/11/2006 - 7:31pm

There really isn’t a reason to suggest that ARI=>SF was a fluke, either, though. ARI did beat SF, and there is no alternate reason to suggest SF is better than ARI, so ARI=>SF should remain.

I'm always really surprised at some of the results out of the beatpath rankings which totally disagree with Elo or Laplace type rankings, which are designed to do exactly the same thing (rank teams with only wins and losses). There's something weird going on there.

by Marko (not verified) :: Wed, 10/11/2006 - 7:33pm

"Unfortunately…Marty is STILL SD’s coach, but I guess I’ll let that pass."

Unlike Marty, who won't let anyone pass.

by Ryan (not verified) :: Wed, 10/11/2006 - 7:37pm

Well I'm not saying that you in person are attacking him; and at least you're making relevant arguments as opposed to saying "Grossman sukz..lol..hez bad." which is about half of the arguments against him so far. Once again though, Peyton has played against just as inferior passing defenses as McNabb, and they've both boosted their stats against Houston, which I hadn't even mentioned until now, and Bulger has played against GNB, Det, SF, and ARI, while he played relatively poorly against the pretty darn good Denver defense. That's why I think we should just stick to the DVOA's for now instead of just speculating about how horrible or ungodly amazing a QB will be by the end of the year. I've watched all of Grossman's games this year, and have noticed every single weakness everybody has mentioned, btw, but I also think it's irrational to predict a player's performance based on one poor game against a close to top ten passing defense which is what about 1/3 of the people arguing against Grossman are doing. They list the entire season's numbers for a reason, otherwise somebody like Tom Brady would look horrible if we only judged his performance based on his poor game/s.

by Ryan (not verified) :: Wed, 10/11/2006 - 7:42pm

And..remember...who..the Bears had last year?...Kyle...freaking..Orton...Does anybody think Orton would be doing as well as Grossman this year, or would he have a 63% completion percentage..and average...negative 5 yards per pass? Yeah I think that line looks about right to me.

by dbt (not verified) :: Wed, 10/11/2006 - 7:48pm

oh well. I guess i'll stick with the guy with the #1 DVOA and #3 DPAR despite coming out for at least two full quarters of clock killing backup reps in blowouts.

by Ryan (not verified) :: Wed, 10/11/2006 - 7:51pm

Man...Is that the only layout Blogsport has? Shanoff has the exact same layout...Maybe they should add a few different layouts to add some variety..

by Franco (not verified) :: Wed, 10/11/2006 - 8:22pm

How on earth can KC be number 5 by ANY
measuring stick? They lost to two of
the better teams in the league (Cin
and Den) and beat SF and Az, two of the
worst in the league. They are only 2-2.
They should be ranked right around 15,
not 5! But the Colts are lower than KC
because "the system docks them for
an easy schedule"! You guys are just
making these rankings up, right?

by tunesmith (not verified) :: Wed, 10/11/2006 - 8:27pm

248: I’m always really surprised at some of the results out of the beatpath rankings which totally disagree with Elo or Laplace type rankings, which are designed to do exactly the same thing (rank teams with only wins and losses). There’s something weird going on there.

Well, the strict win/loss handling only leads to the beatpath graph, a directed graph. There's usually a ton of possible ways to create a power rankings system that doesn't contradict the beatpath graph. The tiebreaker I use for the power rankings has to do with how high up in the graph (sorta) each team's immediate beatwins are. That's seemed slightly more accurate at "predicting" results (through backtesting) than other tiebreakers. It's quite possible there are more accurate tiebreakers.

I don't know much about Elo or Laplace beyond their thumbnail descriptions and can't find good technical descriptions of how they work, but it sounds to me like they're probably a lot more complicated than just paying attention to wins or losses. Elo actually has decimal points!

This week's beatpath graph simply is ordered in a way that respects the win/loss outcome of every single game so far this season wherever possible. Right now it's ignoring only three: SF=>STL, CAR=>NO, WAS=>JAC. The current power rankings are also 71-3, but there are probably plenty of other 71-3 power ranking orderings that will do a better job at predicting next week's games.

This week's ELO rankings, however, disrespect the following game outcomes: ARI=>SF, STL=>DEN, SF=>STL, DEN=>NE, NO=>ATL, BUF=>MIN, HOU=>MIA, CAR=>NO, WAS=>JAC. They're 65-9 at respecting wins and losses. So they're clearly paying attention to something other than wins and losses.

by Peter (not verified) :: Wed, 10/11/2006 - 8:43pm

247: What you just said is that teams that play well tend to win the Superbowl.

Here's a better one. The team that leads the league in total DVOA will win the Superbowl. Of 8 seasons, that's happened 4 times: the 1999 Rams, 2004 Pats, 2002 Bucs, 1998 Bengals.

Too tight? How about the top 3? That adds in the 2000 Ravens, 2003 Pats, and 2005 Steelers, so 7/24, slightly less than a third WINNING the Superbowl.

So basically... I'm sorry to crush your theory, but there's no reason to talk about balance. What matters is being really, really good at Offense + Defense + Special teams, and teams can "pick" what they want to specialize in, from 1st Offense/19th Defense (1998 Broncos) to 21st offense/1st Defense (2000 Ravens and 2002 Bucs).

As a side note, it was very silly for you to compare teams that are in the top 10 in both O and D to all other teams... how could the average of all other teams POSSIBLY compare? Half the teams in the leagues are below average and don't even make the playoffs.

by Kaveman (not verified) :: Wed, 10/11/2006 - 8:46pm

Aaron: There are obnoxious Broncos fans? Surely you're joking! :-)

(Yes, I'm ignoring #244.)

I still have this dissatisfied feeling, because in my (subjective and biased) perception, Denver is playing better than this ranking and better than some of the teams ranked higher. It'll be interesting (to me, anyway) to see when DVOA matches my perception.

by CaffeineMan (not verified) :: Wed, 10/11/2006 - 8:54pm

Aaron, thanks for those blog entries on Denver and NE. You're right, the strength of DVOA is it's ability to be broken out by groups of plays. I was fascinated to see the comments on the Pats defense. I've had the impression that it has sucked rocks, although slightly less than last year. But I know others impressions have been that it's actually decent. All I could think of was those endless drives. But others impressions must be heavily influenced by the performance in the second half of close games (Top 10) and in the 4th quarter regardless of score (Top 5).

by Pat (not verified) :: Wed, 10/11/2006 - 9:23pm

I don’t know much about Elo or Laplace beyond their thumbnail descriptions and can’t find good technical descriptions of how they work, but it sounds to me like they’re probably a lot more complicated than just paying attention to wins or losses. Elo actually has decimal points!

Nope. That's all they do. Elo's method was developed to rank chess players, where wins and losses is all you have! For a good mathematical description of what Laplace's method is, look here.

There's enough information there to actually completely calculate the rankings yourself.

So they’re clearly paying attention to something other than wins and losses.

Not really. Elo's method (really, a Bradley-Terry model) is just sacrificing some previous accuracy for smaller error bars. Thinking more about it (riding the bus is good for things like that), I think the beatpaths ranking is probably most like (and is probably nearly identical to) a maximum-likelihood determination using a step-function.

Start with a number - we'll call it a 'likelihood'. Then, come up with some number (we'll call it P), from 0 to 1 for each team. If team A beats team B, and P(A) is greater than P(B), then that was a "likely" result, and we add 1 to the likelihood function. If team A beats team B, and P(A) is less than P(B), that's an "unlikely" result, and we add nothing to the likelihood function.

Then the rankings consist of finding P for all teams such that the likelihood is at a maximum. In this case, it would mean finding rankings such that the maximum number of games are predicted correctly.

Other rankings don't use a step function for the likelihood ranking, which is why they don't backpredict the games as well - they accept that the output of two closely ranked teams can go either way. So if P(arizona) and P(SF) are both identical, then you add 0.50 to the likelihood function regardless of who wins, because it was a toss-up. However, they can actually end up with a *higher* overall likelihood because they don't get burned as hard as a step function does.

Which means they have less overall error in their rankings, and are overall more stable.

This week’s ELO rankings, however, disrespect the following game outcomes: ARI=>SF,

See, that's the example: ARI/SF are virtually identically ranked. ARI beating SF adds less to the likelihood function than moving ARI above SF in the rankings does.

Really, the beatpath rankings aren't that much different than the Bradley-Terry models. All rankings really have two inputs: 1) the wins and losses of a team, and 2) the likelihood function : i.e., the chance that team A with strength P(A) will beat team B with strength P(B). The beatpath rankings assume that function's a step function. The Elo rankings assume it's a logistic curve. Then, it's just a matter of finding the strengths to maximize the likelihood.

by Seattle Doug (not verified) :: Wed, 10/11/2006 - 9:24pm

#187 The cromulence of words is oftenwise subjectagable, but hextuple doesn't stand a chance against "sextuple".

by Pat (not verified) :: Wed, 10/11/2006 - 9:42pm

This week’s ELO rankings, however, disrespect the following game outcomes: ARI=>SF, STL=>DEN, SF=>STL, DEN=>NE, NO=>ATL, BUF=>MIN, HOU=>MIA, CAR=>NO, WAS=>JAC.

Actually, let me be a bit simpler, since I was way too wordy there. The beatpath rankings are ignoring 4 games. The Elo rankings (Sagarin's a dope: it shouldn't be uppercased, since it's a guy's name, not an acronym) are ignoring 9. Why? Because it considers the 4 games that are being ignored for the beatpath rankings (SF=>ARI, NO=>ATL, PIT=>CIN, WAS=>JAC) to be more likely outcomes (between two evenly matched teams, someone's gotta win) than some of the other outcomes (like STL=>DEN), purely on the basis of who the other teams have beaten.

by Bill (not verified) :: Wed, 10/11/2006 - 9:52pm

256 - The 1998 Bengals?

by Reinhard (not verified) :: Wed, 10/11/2006 - 10:06pm

Pennington vs MIA
Plummer vs OAK
Rivers vs SF

Which one do I start?

by Kevin (not verified) :: Wed, 10/11/2006 - 11:38pm

bears fans: beware of injury.

by thad (not verified) :: Wed, 10/11/2006 - 11:54pm

Pat, how would you define a mobile qb?
Franco, the Colts rush defense is a disgrace.

by Ben (not verified) :: Thu, 10/12/2006 - 12:08am

Re: 236 Where did you find that stat? I could buy that for '04, but the Colts ran much more two tight end last year. Now, Clark lines up in the slot a lot, so you might be able to call that 3-wide, but Stokley saw the field a lot less in '05 then in the year before.

by Peter (not verified) :: Thu, 10/12/2006 - 12:09am

262: I believe that qualifies as typo of the century. Well, I guess I could have put in "Cardinals" for "Rams" or something.

by D (not verified) :: Thu, 10/12/2006 - 12:21am

That number (I'm guessing) came from the "Strategic Tendencies" portion of the Colts' chapter in the PFP 2006. I don't know if that includes the times when Clark lined up in the slot or not though.

by Lou (not verified) :: Thu, 10/12/2006 - 12:31am

RE: 266
According to PFP 06 they ran 3+WR sets 83% of the time last year. I've done charting for a few halves of their games this year, and I'd say thats about the percent they've run it this year with Clark lined up in the slot.

by Fnor (not verified) :: Thu, 10/12/2006 - 12:35am

For the fiftieth freaking time, Orton's performance at the end of his run was roughly replacement level. Grossman came in and played at... replacement level. Wahey!

Bigger differences than QBs this year: game planning, competent recieving performances, Muhammad not throwing temper tantrums.

by tunesmith (not verified) :: Thu, 10/12/2006 - 1:13am

re 259, 261: beatpaths currently ignores three games, not four. cincinnati beat pittsburgh.

I remember laplace's method now - I looked at it about a year ago and ignored it because it uses math. :) It's not just wins and losses and relationships - it's also numerators and denominators and stuff. That's no fun! Beatpaths doesn't use math at all, it's all logic. Well, except for the tiebreakers. But the graph itself is 100% math-free.

I'm being deliberately obtuse here, but I really don't like the numerator/denominator because it does make an arbitrary judgment of how much better a team is from another. Even though it's consistent, I just don't like that.

There's no value judgment of how much better one team is over another when there's a beatpath. It's ordinal, not mathematical. It's just "better", not twice as better or five times as better or anything like that. If there's a beatpath chain of five teams, they could be widely different in quality, or they could be all scrunched up together in quality. There's no need to think about probability, or how to reduce the power of a team that is five links away from another, or anything like that. The system doesn't care - it just relies on the other relationships in the graph to make it all shake out. It's way, way simpler than all that math stuff.

Wish I could find an archive of weekly power rankings for elo and laplace; I'd love to see how much worse beatpaths is at the weekly picks.

by Dave (not verified) :: Thu, 10/12/2006 - 1:53am

Hi, I'm new here. I'm a long time Baseball Prospectus reader. I've been reading this site for a few weeks now and enjoy it, but after reading #270 I had to post for the first time.

You say that you've repeated this point 50 times... that Grossman came in and played at replacement level. I don't mean to be a smartass, but are you seriously making a point out of that? He played in two games last year, which also happened to be only the 7th and 8th games he'd EVER played in.

Now I know baseball is different than football, but isn't sample size discussed around here? How can his performance in 2005 or really any of his pre-2006 performance be brought up in a discussion about this season?

Has a study been done about what you can tell about a QB after his first 8 games? How about if those 8 games were spread out over three years? How about by looking at the stats for any random 2 games in a QBs career?

I would guess 'not much' would be the answer to all of those questions, which means we really had no idea how much Grossman might add or even take away from this offense. However, through visual, statistical, and anecdotal evidence - we can probably take a very educated guess that while certainly lucky to have a few INTs dropped, he has added a whole lot.

by Bobman (not verified) :: Thu, 10/12/2006 - 2:03am

182 miles, 90 minutes... you guys got a problem with that? Okay, I sit here disgraced, exposing my east coast upbringing and west coast hometown, with precious little time spent west of Hackensack and east of Kirkland (current home of the Seahawks). I assume that as TV markets go, Chicago extends at least to Gary, Valparaiso and S Bend, and Indy extends up to the lake? The TV markets might even touch? Champaign Ill looks closer to Indy than Chicago--what TV market is that considered to be in? Living in Seattle, where the closest team is SF 700 miles away, and then Denver at 1,000, 180 miles is a short trip. (Oakland is excluded for obvious reasons.)

On an actual NFL note, what is a typical NFL rate for third down conversions? Indy is currently at 58.5% with the next closest team (Houston!!??) at 44.2%. The next closest playoff type team is Philly at 40.3. Throwing out the outlier Colts and Raiders, the average of the remaining 30 teams is 36.8%. Indy is more than one and one-half times better than the NFL average.

How can we reconcile Indy's poor stats, actually struggling offense (compared to years past) with what looks like a record pace for 3rd down conversions? Manning's rating on 3rd downs so far is 133.3 (on 3rd and 8-10 it's 143.9). I'd venture to guess that's unheard of. In his freakshow 2004 season, 3rd down was his worst and last year he was consistent on 1/2/3rd downs between 100 and 106. (and only 76.9 on 3rd and 8-10). I'll guess I should wait til they play some better D's, but somehow in a year when they're 5-0 but rightfully thought of as crummy on D and not terribly explosive on O, he seems to be doing better under pressure (4th qtr comebacks and 3rd downs) than ever.

McNabb, who is higher in DPAR has a 3rd down QB rating of 83.6 and Rex with a higher DVOA is 91.0. Manning = Mr Clutch? The former Mr. Clutch is 66.2 on 3rd down, but he's missing his old WRs, so he's got an excuse. I know clutchness doesn't exist statistically, but it is a favorite meme here.

Maybe the regular season for Indy is finally becoming just a testing ground for the playoffs, so long as the D actually enables them to make it that far.

by Bobman (not verified) :: Thu, 10/12/2006 - 2:04am

Sorry to go on so long and then double post, but stats were from ESPN, FWIW.

by Pat (not verified) :: Thu, 10/12/2006 - 2:11am

Pat, how would you define a mobile qb?

Bill Barnwell's study did a good job - something like 4 rush attempts/game average. McNabb isn't there anymore, but he was for most of his career, and you don't forget the ability to move around in the pocket.

beatpaths currently ignores three games, not four. cincinnati beat pittsburgh.

You listed a 56-4 record: what's the fourth game? I could look it up, but I'm lazy. :)

Beatpaths doesn’t use math at all, it’s all logic.

Laplace's method is logic. You should read Colley's description of it - the craps example. It's very explanatory.

Hell, math is logic. I can write down the requirements that you're placing in the system in mathematical form pretty easily.

Well, except for the tiebreakers.

But the tiebreakers are exactly how you determine the rankings. They're the entire bread and butter.

because it does make an arbitrary judgment of how much better a team is from another.

Not really - at least, not any more than you're doing. You're ranking teams from 1 to 32. If 1 is 1.0, and 32 is 0.0, then all you're doing is equidistantly placing each team. Since the likelihood function you've got only has a derivative at zero, there's no penalty in equidistantly distributing the rankings. But that's just due to your choice of likelihood function.

Elo's method is doing the exact same thing - except the likelihood function is a logistic curve, which means you can shove teams slightly closer to each other, and slightly farther away if it improves the overall likelihood of the given set of data.

Basically, it's like this: you throw away games as beatflukes if they're contradicted by the graph via a different path. But why couldn't the graph be contradicted by the beatfluke win instead? In your method, it wouldn't make a lot of sense - you'd be adding tons of mispredicted games, and only gaining 1, so it'd be a net loss. But if you allowed close losses (like, SF=>ARI) to be fractional mistakes, you could end up winning out overall.

Note that I'm not suggesting that - but that's the essential difference between the system you've got and Elo's rankings (and other Bradley-Terry models).

There’s no need to think about probability

Actually, you are. The first criterion is that a team can't be below a team they've unambiguously established a beatpath to. That's just a description for the likelihood function being a step function - or "the probability of a team with lower beatpath ranking to beat a team with higher beatpath ranking is zero." Then, your next criterion for establishing the power rankings is which method of tiebreakers gives the best retroactive prediction percentage. That's the same thing as saying "out of all possible rankings, given the likelihood function, which ranking requires the least number of unlikely occurances"? That's exactly equivalent to looking for the solution of maximum likelihood.

Wish I could find an archive of weekly power rankings for elo and laplace; I’d love to see how much worse beatpaths is at the weekly picks.

I'm pretty sure there's a program out there to compute Elo rankings given win/loss data, so you could clearly do it. And Laplace's method is pretty straightforward, especially with Colley's method. I might be able to write something for it.

Honestly, if it would outperform Elo or Laplace's method, you should publish it. There are a lot of implications for this kind of stuff, none of which has anything to do with sports.

by Crushinator (not verified) :: Thu, 10/12/2006 - 2:22am

Keep in mind also who the two games Grossman played in last year. One of them was a playoff game, which adds a ton of variables for a young players first high pressure game.

Orton vs Grossman isn't even comparable. Orton was abysmal last year. Even in the end, when Grossman came out in place of Orton during the Atlanta game, you could INSTANTLY see the difference. Grossman was just much, much, much more accurate than Orton.

Now, I haven't seen a ton of bears games this year so I don't know about his decision making or dropped picks or what have you - but accuracy counts a hell of a lot in the NFL and Grossman can make much better throws than Orton can.

by Fnor (not verified) :: Thu, 10/12/2006 - 2:32am

#272: That issue has also been done to death. Football in general has sample-size issues. While he has only played a handful of games, he has been to 3 camps, played multiple preseasons, played in practices (although not full-on, granted), benefitted from being the focus of the QB coach and the coordinator both in attention and drafting plays, watched tape and watched games for the past 3 years. He has essentially been Phillip Rivers, or Steve Young of yore, except he got a lot more focus and attention than those guys did when the top dogs were still in town.

There are, however, statistics that have good sample sizes over a few games or even from game to game. Completion percentage is not an extremely useful stat, but last year we noticed something interesting: low completion percentage + low interception rate = unsustainable situation. It's either luck or scheme or likely both, but there have been an awful lot of interceptions dropped against the Bears this year, some of them at very important junctures of the game. This hints, at least to me, that the interceptions are going to go up rather than the completion percentage, which does not bode well.

As for the anecdotal evidence, that's out there, too. Poor decisions, an insistence to throw downfield no matter what, inaccuracy, an inability to deal with pressure. There are some good things, too, like a very good play action and good delivery on crossing routes, but there are some serious concerns here. He's playing like Favre used to, except he doesn't have Favre: The Wonder Years type skill.

by Fnor (not verified) :: Thu, 10/12/2006 - 2:43am

#276: So, you didn't see many Bears games this year? How about last year? Did you keep in mind that Chicago was near the tops in the league in dropped passes last year? Did you consider that Muhammad simply stopped trying halfway through last season because his buddy Grossman wasn't playing? Did you think about Grossman's completion percentage this year? Or are you just saying what everyone else said about last year's QB situation because, hey, that's what everyone else says?

Why in God's name did I move to Chicago... I should've stayed in Columbus....

by Jake (not verified) :: Thu, 10/12/2006 - 3:00am


Ok, so you're charting 2 WRs and 1 TE split wide as a three WR set? I thought a 3 WR set had, you know, 3 WRs in it. A TE split wide is still a TE.

by DavidP (not verified) :: Thu, 10/12/2006 - 3:10am

When you visit a football information site such as this. And you read that the Patriots have the number 4 ranked passing offense. You simply hit the back key and get the hell out of here. What are you people smoking?

by Fnor (not verified) :: Thu, 10/12/2006 - 3:25am

#279: Is Mike Vrabel lined up as a TE still a LB? If a TE is split wide, he's running WR routes, and doing a WR's job. Counting him as a TE would actually be a hinderance when doing analysis, because he's not doing any of the things a TE would be doing on that play.

by tunesmith (not verified) :: Thu, 10/12/2006 - 7:04am

You listed a 56-4 record: what’s the fourth game? I could look it up, but I’m lazy. :)

I forget what the 4th game was - at the time, it was actually possible to reorder the power rankings to be 57-3, so it actually isn't necessarily true that the beatpath graph will lead to the most accurate retrodictions. The current power rankings are 71-3 now. That 4th game was one of the New Orleans games, I think.

One thing - I don't see the power rankings as equidistantly spaced at all. Some sequences of teams are crunched together in quality, some are far apart (like Oakland in last place - right now I imagine they're last by a LOT).

Beyond that, I think we're talking orthogonally. I don't contest that beatpaths can be phrased mathematically. I'm saying that it doesn't have to be. It's implemented without math, it can be described without math. ELO and Laplace can't be described so simply. Sure, Laplace is "logical" but that's semantics. It requires a fraction bar and decimal points and stuff to describe it and implement it.

The tiebreakers have nothing to do with the 56-4 or the 71-3. It's the beatpath graph that guarantees that record (or better). Every possible power rankings (meaning, any possible tiebreaker) will be consistent with the beatpath graph. The way I've tested tiebreakers is through backtesting - its ability to predict the next week's games (running through the entire 2005 season, week by week, at the end of the season). I forget the records back then but I believe it did slightly worse than the aggregate picks of the Yahoo users.

by Travis (not verified) :: Thu, 10/12/2006 - 8:55am

Re: 273

Chicago extends at least to Gary, Valparaiso and S Bend, and Indy extends up to the lake? The TV markets might even touch?

The Indianapolis TV market doesn't extend to the lake, but does extend to White County, IN, which borders Jasper County, IN, the southeast extent of the Chicago TV market. Map

Champaign Ill looks closer to Indy than Chicago–what TV market is that considered to be in?

Champaign is its own TV market, the 82nd biggest in the country. Map

by Fnor (not verified) :: Thu, 10/12/2006 - 10:42am

#283: It doesn't actually take over Valparaiso, because that town very specifically has Indiana's only Spanish restaurant. If they let Chicago take over, they wouldn't be the only Spanish restaurant in the Chicagoland area, and they'd have to buy new billboards or something.

by dbt (not verified) :: Thu, 10/12/2006 - 10:53am

Muhsin was dropping balls from the beginning of the season, and kept dropping them once Rex came back. He simply worked harder this year and has been much better on that score this year.

I don't think Orton is a terrible QB, but I'm pretty sure that for the next couple years a [healthy] Rex is better.

("Ewwww, Columbus")

by Pat (not verified) :: Thu, 10/12/2006 - 10:56am

It’s the beatpath graph that guarantees that record (or better).

That's true. It's your choice of fluke win elimination that gives a high agreement, but that's somewhat implicit in the idea of contradicting as few games as possible.

Sure, Laplace is “logical� but that’s semantics. It requires a fraction bar and decimal points and stuff to describe it and implement it.

No, it doesn't! It require math to find an ideal solution in one exact step. But you can implement it solely with logic. Heck, I could do it in the NFL using playing cards, or a table and a bunch of markers.

You could absolutely do Laplace's method graphically.

One thing - I don’t see the power rankings as equidistantly spaced at all.

When you're treating predictions as perfect (i.e. if ranking of team X is higher, team X is the pick) you're implicitly declaring the rankings to be equidistantly spaced - or the likelihood function to be a step function. Either or - the solution's identical either way.

by Travis (not verified) :: Thu, 10/12/2006 - 11:35am

Re: 284

Couldn't it be in both Indiana AND Chicagoland? It's not like they bill themselves as "Indianapolis's only Spanish restaurant." :)

Re: 273, again

South Bend, in St. Joseph County, IN, is its own TV market.

by mediator12 (not verified) :: Thu, 10/12/2006 - 11:45am

Thanks for trying to explain that Aaron.
Correct me if I am wrong, but the idea was to explain the deviance between a bend but do not break defense and instead I felt you were rather justifying the system with the number of games played.

I do not care how poor an offense is in the red zone going 0-5 is an excellent accomplishment considering they were handed the ball twice in six plays on the second and third possession getting three points. Therefore, the Linehan explanation seems very shaky due to its subjective nature of past performance with another team.

Two of Denver's young players made huge mistakes in those first three games. DJ Williams missed several cut backs and got stiff armed from Stephen Jackson or he gets 65 rushing yards in that game.

Darrent Williams was picked on to the extent of 33 completions in 46 opportunities including 9 first down catches on those seven scoring drives that started outside FG range. That was through 3 games. Sure Bailey is awesome, but if Darrent was just average to replacement value the defense would have been even more stingy in Yards and more importantly points allowed.

Those two young players were getting abused by not executing their role in the system, but you put up something about Linehan's offense trending from MIA as proof? I expected better analysis than that.

Also, you failed to mention WHY denver does not have more takeaway's this year. They have dropped 8 potential INT's including 4 against Brady and the PI against Caldwell on the lone TD drive where Bailey was poised to INT that second play throw.

As for the Red Zone, they have not allowed thirty rushing yards either in 11 possessions and ZERO rushing TD's. With the DL being the biggest question mark, they have been superior in performance in the red Zone. The majority of Red Zone TD's last year were Runs and through four games they have zero.

I believe the scheme change has allowed some more plays to be made between the 20's as they try to surrender fewer big plays and lock in their zone coverages in the intermediate range. They have allowed a ton of passing yards underneath and almost every play that has been completed was completed within 15 yards of the LOS. The only 40+ yard play was the result of Blitzing into a screen pass and Missing the tackle until 40 yards later. (DJ Williams was one of those too.)

This is my frustration with DVOA. It only measures results without explanation. Sure, DEN has surrendered a lot more Yardage than other good defenses but the numbers only tell part of the story. As well as the defense has played overall, two young players have surrendered some key third down's and a bunch of missed play yardage and that has brought the performance overall down.

by Sean (not verified) :: Thu, 10/12/2006 - 12:22pm

I suppose it is inevitable that whatever team is doing well at the moment is going to have insufferable fans coming out of the woodworks and setting up strawman arguments to prove that they are getting the Rodney Harrison treatment. It's the only explanation I can come up with for how Kyle Orton has made it into the discussion. Look, as someone who is not drinking the Bears Kool-Aid, let me make a few basic points:

1) No one is disputing that the Bears have performed at a higher level than any other team so far this season. The DVOA is pretty clear on that matter. It does not inherently follow, however, that the Bears are going to continue to play at that level. If you want to assume that the team is going to continue to field the #3 offense, the #2 defense and the #1 special teams unit in the league, you are welcome to do so, and you can point to the first five games as evidence. That doesn’t mean that the people who think the team will not continue to perform at such a high level on offense and/or special teams are living in denial. It just means that they are looking at the personnel and recognizing that very few of the players have demonstrated that they can play elite football over the course of a season. It’s on those players to prove themselves. Anyone who was watching the Rams in 1999 and thought that Kurt Warner needed to show it for more than five games was being pretty reasonable. Maybe the Bears offense will hold up and maybe it won’t.
2) There’s really no need to bring Kyle Orton into the conversation. Everyone is perfectly willing to acknowledge that Grossman is a better player and that his presence should make the offense better as a result. However, there are entirely legitimate reasons to be concerned about him. He has not showed anything in his limited time in the league to suggest he can sustain this level of play. He was consistently terrible throughout the preseason. (You can trot out the argument that the preseason means nothing and you can’t draw any conclusions from it, but you’re talking to someone who watched every preseason game and now has Alex Smith, Frank Gore, Antonio Bryant, Jerricho Cotchery and Greg Jennings on his fantasy team, among others. The preseason means a lot if you know what you are looking at. When players or teams struggle with their execution in the preseason, it usually carries over.) He was a streaky player in college. And last but certainly not least, his statistics are not in line with the game notes of people who have been watching Grossman play. Put that all together and there are legitimate reasons for expecting a drop-off in Grossman’s effectiveness, and a corresponding drop in the overall effectiveness of the offense and the team.
3) The defense is tremendous, no doubt. The defense was also tremendous last year. They then went and got destroyed in the playoffs, at home, on a day when the offense put up more than enough points. (They also got ripped in 2001, which is probably ancient history, but certainly doesn’t help the perception that Chicago is not the most intimidating of playoff venues.) The defense does not get to walk away from that performance against Carolina until they expunge it in the playoffs.

Bears fans have every reason to be ecstatic about their team. Non-Bears fans have reason to be suspicious of the team’s ability to sustain their current level of play, and of their ability to beat talented and playoff-tested teams come January. And there it is.

by Jake (not verified) :: Thu, 10/12/2006 - 12:29pm


No, Vrabel is a LB lined up at TE. Clark is a TE split wide, but not a Wide Reciever.

by Starshatterer (not verified) :: Thu, 10/12/2006 - 12:55pm

I suppose it is inevitable that whatever team is doing well at the moment is going to have insufferable fans coming out of the woodworks and setting up strawman arguments to prove that they are getting the Rodney Harrison treatment.
Either that, or the insufferable fans are setting up those strawmen to give 'em the Rodney Harrison treatment.

Borderline knockdown for the win!

by Wanker79 (not verified) :: Thu, 10/12/2006 - 1:12pm

Re: 288

So just because a weakness in the Denver defense has been exploited we're supposed to just ignore it?!? That makes absolutely no sense.

by Fnor (not verified) :: Thu, 10/12/2006 - 1:23pm

#290: Okay, so when he's split wide and the QB gets by an OLB edge blitz, do we blame him for poor blocking? When the RB is lined up as a TE, do we blame him for not picking up a MLB blitz?

While saying Vrabel's a LB or Heap's a TE is completely true, it's also semantic. For those plays, they're playing at the role of another position. The play was drawn up for the person to perform like the position he has lined up in. To insist that a TE split wide must be classified in a chart of the formation as a TE is not only pendantic, but seriously mischaracterizes what actually happened on the field.

by Jason (not verified) :: Thu, 10/12/2006 - 1:34pm

Fnor, as a Steelers fan, your denouncing of another team's young (successful) QB is misplaced and laughable.

by Fnor (not verified) :: Thu, 10/12/2006 - 1:52pm

#285: That's a perfectly fine argument. People could make a good case that Grossman is better than Orton. That's a discussion we could have. What I generally refer to with Muhammad last year (aside from dropped passes, that's just his style) is his repeated complaining about the QB situation and the offence in general, the fact that he went very public with it, and sloppy route-running and a couple alligator arms as last season went on.

What bothers me is people (a lot of them) just stating "Orton was terrible, Grossman was so clearly better that a chior of angels appeared over soldier field and heralded him with praise" when Grossman came in and pretty much played at the same level. A lot of this is was the excitement of a more open playbook than was allowed for Orton. It wasn't effective at all last year, but hey, it was more interesting and therefore better.

What I most strenuously object to, however, is how people are (and were) completely stoning an unprepared rookie thrown into a situation with subpar weapons and a crippled playbook. It's not particularly accurate or fair.

by Fnor (not verified) :: Thu, 10/12/2006 - 1:53pm

#294: So, by that logic, am I not allowed to comment on linebackers or safties either?

by Jason (not verified) :: Thu, 10/12/2006 - 1:57pm

Re 296
Why? Are your LBs and safeties sucking up the joint, too?

by dbt (not verified) :: Thu, 10/12/2006 - 2:49pm

#295: What was irritating was the perception that orton shouldn't "lose his job" because "they're winning with him". I think Orton, given time to develop, will be a fine QB.

#289: 2001 is ancient history. That team had the 6th best DVOA in the NFC, a projected wins of 9.1 but a record of 13-3 off a series of fluky wins. I believe the only regulars still on the team are the two centers, urlacher and mike brown.

Last year Carolina beat the Bears injury-depleted secondary and outscored 'em. It still concerns me. Mike Brown has a history of injury problems and Tillman's dropoff was huge. That being said, I think Chris Harris is being groomed as Mike Brown's backup/replacement, and the unit overall should not be quite so weak.

As a point of reference, I didn't recognize several names of guys who logged serious minutes in that playoff game in the secondary at all when they were brought up a couple weeks ago in relation to their new team.

Regarding this year -- we'll just have to wait and see. But if the over-under is 13.5 wins, I'll happily take the over. [I'm making up that O/U]

by mediator12 (not verified) :: Thu, 10/12/2006 - 3:05pm


No. You can not ignore, you have to account for them. Missed tackles and Blown coverages are part of the game. However, there was no mention of that in the argument Aaron presented.

The weakness was not making the plays that were there, guys who can play very well were underperforming. If you saw DJ Williams against the Pats or Ravens, you missed the one who was horrendous versus KC and STL.

The same thing applies to to Darrent Williams. He was very good versus BAL, but horrendous versus KC, STL, and NE.

The point being, players have different performances from game to game. And, As Aaron said earlier DVOA accounts for the Overall performance over a large sample size. It can not easily explain the Matchups from game to game and it also does not take into effect playcalling execution. There is no way to account for yardage due to missed tackles and blown assignments easily.

by mediator12 (not verified) :: Thu, 10/12/2006 - 3:35pm

Plus, Aaron brought playcalling and execution into the mix when he he commented on the ineffectiveness of Linehan as a Red Zone playcaller.