Writers of Pro Football Prospectus 2008

Most Recent FO Features

GurleyTod16.jpg

» OFI: SEC Surprises

In an opening week where even the elite teams in college football looked mortal, the SEC had two big surprises in Texas A&M and Georgia defeating their South Carolinian opponents by big scores.

04 Jan 2005

Final 2004 Team Efficiency Ratings

by Aaron Schatz

Here are the final Football Outsiders team efficiency ratings for the 2004 regular season, measured by our proprietary Value Over Average (VOA) system that breaks down every single play and compares a team's performance to the league averaged based on situation in order to determine value over average.  (Explained further here.)

When all is said and done for 2004, the New England Patriots stand on top of the DVOA ratings, very narrowly surpassing the Pittsburgh Steelers and Indianapolis Colts.  This continues the rise of the Patriots since Bill Belichick took over in 2000, where each season saw the team's total rating increase over the year before.  Just ignore that unexpected Super Bowl championship in the middle. 


YEAR RANK TOTAL W-L OFF RANK DEF RANK
2000 22 -10.0% 5-11 -7.5% 21 5.7% 22
2001 16 -1.4% 11-5 -8.7% 21 -3.4% 19
2002 11 12.5% 9-7 4.0% 17 -3.4% 10
2003 2 22.4% 14-2 -0.8% 13 -22.0% 3
2004 1 35.6% 14-2 26.3% 4 -9.1% 6

The Patriots also top the league in estimated wins for the second straight year (13.2), and they top the league in pythagorean wins (projected from points scored and allowed) for the second straight year (12.4).  If you're a Football Outsiders veteran who has been reading the site since last year, you may remember that the 2003 Patriots became the first team in history to be both the league's best team and its luckiest team -- they had the most pythagorean wins and the biggest difference between their actual wins and their projection.  While this year the Pats once again end up with more wins than their scoring would project, they don't lead the league in that category.  Instead, the "luckiest" team this year was the Pittsburgh Steelers, whose scoring would normally indicate only 11.5 wins.  Also very lucky were Atlanta, which projects to 8.1 wins, and St. Louis, which projects to only 6.1 wins.  On the flipside, the pythagorean projections say Tampa Bay should have gone 8-8, Kansas City 9-7, and Buffalo 11-5.  

Normally a team that wins at least two fewer games than its pythagorean projection will rebound the following season, but this marks the second straight season that Tampa Bay has underperformed its projection by such a large amount.  This is a statistical indication that the Bucs are totally, completely screwed up beyond belief, but Tampa fans already know this.

The one measure that New England does not top is weighted DVOA, which attempts to indicate how teams are playing now by lowering the impact of early season games.  Instead, Indianapolis ranks first, with Buffalo and Pittsburgh right behind.  Like everyone else, I was surprised when Buffalo could not beat a cast of primarily Pittsburgh backups to make it into the playoffs.  They would have been very dangerous.  In fact, the Bills very nearly pass the freefalling Philadelphia Eagles to make the top four teams of the season all AFC squads.  And that, of course, brings us to the problem we first identified last week: what to do about all the teams sitting their starters at the end of the season.

For most of the season, the DVOA ratings have featured the big four and then everyone else.  In fact, back in Week 13, I noted that the four big 2004 teams were on track to rank as four of the top seven teams since my DVOA breakdowns start in 1999, with the Eagles on top.  Since then, Philadelphia's DVOA has dropped almost in half.  And here's a newsflash: it didn't just happen in the last two weeks.  They narrowly beat a bad Washington team and they narrowly beat a bad Dallas team, and they did both those things with their starters in the lineup.  Heck, the first game had Terrell Owens in the lineup.

Then came the two games where the Eagles sat everyone.  They played so badly in those two games that I have to wonder if Paul Tagliabue came to them with some kind of Brewster's Millions-type deal: "If you can get outscored by more than 40 points in the next two games, I'll guarantee you the Super Bowl title."

Last week I agonized over what to do about the problem of so many teams sitting all their starters.  After all, Atlanta and Philadelphia had sat their starters in Week 16 and both teams that week posted their worst one-game DVOA rating of the year.  When seven different teams sat their starters in Week 17, it was going to skew my ratings completely.

But guess what: it didn't.  Most of the teams that had nothing to play for in Week 17 did fine.  The Pittsburgh and San Diego backups actually beat the starters from other teams in the top ten.  Green Bay, playing a mix of starters and backups, annihilated the Bears.  The Patriots didn't play their best, but they didn't play a terrible game either.  The Falcons played most of their starters and played basically an average game, which isn't that surprising since they are really no better than an average team.

That leaves four games in the past two weeks where the "sitting starters" effect may be distorting the DVOA ratings: Indianapolis-Denver in Week 17 (Jim Sorgi actually was fine, but Dominic Rhodes was terrible), Atlanta-New Orleans in Week 16, Philadelphia-St. Louis in Week 16, and Philadelphia-Cincinnati in Week 17.  In reality, it isn't that many more games than in previous seasons.  The effect isn't so great that we have to re-do all the ratings in order to correct for it.  We simply have to acknowledge the issue when we look at these teams.  For example, Denver, San Diego, and Baltimore ended up close to each other, with Denver in the lead ranked #7.  If the Colts had played the starters more in this week's game, maybe Denver is actually slightly lower and ends up ranked #9.  It doesn't really change the fact that Denver, San Diego, and Baltimore are all at about the same level of quality this season, and it doesn't really change the fact that Indianapolis, Pittsburgh, and New England are all at about the same level of quality this season.  (Come on, snow... come on, snow... come on, snow...)

The team that gets most affected here is Philadelphia, because they played two of these rotten games.  But after some thought, I've come to the opinion that Philadelphia's poor performance in the last two weeks tells us just as much about the Eagles as their big wins earlier in the year.  It tells us that their depth is a problem, just as San Diego's win tells us that their depth is a strength and their backups are hungry (Discuss: Doug Flutie in Arizona would have won the NFC West).  Two straight losses with nothing to gain from playing hard might not be a big deal had the Eagles been the hottest team in football before those games, but when you combine these losses with the two close wins over bad teams, I start to wonder if the wheels are, in fact, coming off.

And yet (I think this is the 5437th time I've typed this so far this season) they still might be the best team in the terrible NFC.  According to DVOA, only one of the NFC playoff teams has a positive rating over the last four games: Minnesota.  And they lost three of those games.  For crying out loud, the team that ranks 30th in the NFL made the playoffs.  Heck, the Rams might even win their first game, since they seem to have Seattle's number.

The season ends with the AFC so dominant over the NFC that the AFC features 10 of the top 11 teams and 12 of the top 15.  On top of the NFC being so bad, DVOA thinks the wrong teams made the playoffs, although the NFC teams are so grouped together in mediocrity that it probably doesn't really matter.  The NFC playoff field, if chosen by the estimated wins formula, would feature Carolina as NFC South champion, Minnesota as NFC North champion, either Detroit or Tampa Bay as the second wild card alongside Atlanta, and the Rams and Packers sitting at home.

* * * * *

Last year, in the final DVOA ratings for the season, I included a rank of the following year's schedule, based on the average 2003 rating of 2004 opponents.  I decided not to do that this year because we all know that a number of teams will be much better next year, and a number of teams will be much worse.  Better to wait and do a schedule strength ranking when we do statistical projections that can predict some of that.

LAST WEEK rating has been replaced by LAST YEAR, representing final 2003 rank.

Final numbers for individual positions will be up as soon as I can lay them out Tuesday.  Same goes for offensive and defensive line numbers.  Lots more analysis is coming later this week with another year of extensive playoff preview articles.  We've also got our Playoff Fantasy Draft and Roundtable coming later this week, with special guest Will Carroll from our partner Baseball Prospectus, and a Loser League Season Wrap-Up.

Speaking of BP, you can ask me all your playoff-related questions when I do another online chat at BaseballProspectus.com, Wednesday afternoon at 2pm Eastern.  Click this link to ask questions ahead of time.

* * * * *

OFFENSE and DEFENSE DVOA are adjusted based on strength of opponent as well as to consider all fumbles, kept or lost, as equal value.  SPECIAL DVOA is adjusted for type of stadium (warm, cold, dome, Denver) and week of season.  NON-ADJ TOTAL VOA does not include these adjustments.  DVOA is a better indicator of team quality.  VOA is a better indicator of wins so far.

As always, positive numbers represent more points so DEFENSE is better when it is NEGATIVE.  LAST YEAR represents 2003 rank.


TEAM TOTAL
DVOA
LAST
YEAR
NON-ADJ
TOTAL VOA
W-L OFFENSE
DVOA
OFF.
RANK
DEFENSE
DVOA
DEF.
RANK
SPECIAL
DVOA
S.T.
RANK
1 NE 35.6% 2 32.1% 14-2 26.3% 4 -9.1% 6 0.2% 16
2 PIT 35.4% 16 36.3% 15-1 16.6% 7 -15.0% 4 3.8% 7
3 IND 34.7% 4 37.0% 12-4 38.9% 1 2.3% 18 -1.8% 22
4 PHI 28.7% 9 24.0% 13-3 17.7% 6 -3.5% 13 7.4% 3
5 BUF 28.6% 22 30.4% 9-7 -5.1% 21 -24.5% 1 9.2% 1
6 NYJ 23.8% 18 26.7% 10-6 23.4% 5 2.8% 19 3.1% 11
7 DEN 19.6% 12 18.7% 10-6 8.2% 10 -14.2% 5 -2.8% 23
8 SD 19.2% 25 22.0% 12-4 16.4% 8 -6.6% 11 -3.8% 29
9 BAL 18.9% 6 12.7% 9-7 -2.5% 16 -16.6% 2 4.7% 5
10 KC 12.1% 1 3.8% 7-9 29.0% 2 16.2% 28 -0.8% 19
11 CIN 9.8% 23 -0.1% 8-8 7.6% 11 1.1% 17 3.2% 10
12 CAR 4.2% 20 8.9% 7-9 0.1% 15 -7.1% 10 -3.0% 25
13 MIN 1.5% 10 4.6% 8-8 28.0% 3 22.8% 31 -3.7% 28
14 JAC -0.1% 19 -3.9% 9-7 -7.5% 25 -8.5% 8 -1.0% 20
15 HOU -1.2% 29 -5.2% 7-9 0.2% 14 -2.4% 14 -3.7% 27
16 TB -2.0% 8 -0.5% 5-11 -6.1% 23 -8.8% 7 -4.7% 31
TEAM TOTAL
DVOA
LAST
YEAR
NON-ADJ
TOTAL VOA
W-L OFFENSE
DVOA
OFF.
RANK
DEFENSE
DVOA
DEF.
RANK
SPECIAL
DVOA
S.T.
RANK
17 WAS -2.4% 24 -0.7% 6-10 -15.3% 28 -16.0% 3 -3.2% 26
18 DET -2.9% 31 0.5% 6-10 -3.3% 18 3.3% 20 3.8% 8
19 ATL -2.9% 30 1.7% 11-5 -6.2% 24 0.6% 16 3.9% 6
20 SEA -3.3% 7 6.0% 9-7 4.0% 13 4.5% 21 -2.8% 24
21 GB -5.0% 5 -2.5% 10-6 13.3% 9 20.0% 29 1.7% 13
22 NYG -10.0% 28 -4.7% 6-10 -5.4% 22 6.3% 22 1.6% 14
23 NO -10.6% 17 -2.0% 8-8 -4.2% 20 14.1% 26 7.7% 2
24 DAL -14.9% 15 -17.3% 6-10 -3.3% 19 12.4% 25 0.8% 15
25 MIA -15.5% 11 -22.3% 4-12 -28.5% 31 -7.5% 9 5.4% 4
26 OAK -19.1% 27 -28.1% 5-11 -3.2% 17 14.3% 27 -1.6% 21
27 ARI -19.4% 32 -10.3% 6-10 -20.5% 30 -1.3% 15 -0.2% 18
28 CLE -19.4% 21 -28.5% 4-12 -11.9% 27 9.7% 24 2.3% 12
29 TEN -22.5% 3 -20.5% 5-11 -10.4% 26 7.9% 23 -4.2% 30
30 STL -25.9% 13 -24.2% 8-8 4.8% 12 21.0% 30 -9.7% 32
31 CHI -32.0% 26 -31.2% 5-11 -39.4% 32 -4.1% 12 3.3% 9
32 SF -46.5% 14 -47.5% 2-14 -19.9% 29 26.7% 32 0.1% 17

  • 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.  It is based on the number of games a team has played so far, so teams which have not had their bye will appear higher.
  • PYTHAGOREAN WINS represent the number of wins projected from the team's points scored and allowed, as described in this article.
  • WEIGHTED DVOA represents an attempt to figure out how a team is playing right now, as opposed to over the season as a whole, by making recent games more important than earlier games.
  • 2004 SCHEDULE lists average DVOA of opponents played this season, ranked from hardest schedule (#1, most positive) to easiest schedule (#32, most negative).
  • 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).


TEAM TOTAL
DVOA
W-L ESTIM.
WINS
RANK PYTH
WINS
RANK WEIGHTED
DVOA
RANK PAST
SCHED
RANK VARIANCE RANK
1 NE 35.6% 14-2 13.2 1 12.4 1 33.0% 4 4.4% 9 28.4% 3
2 PIT 35.4% 15-1 11.9 4 11.5 4 41.4% 3 4.6% 8 17.2% 17
3 IND 34.7% 12-4 12.3 3 11.5 2 43.1% 1 0.0% 18 14.4% 25
4 PHI 28.7% 13-3 12.3 2 11.5 3 23.9% 5 -4.4% 22 28.2% 5
5 BUF 28.6% 9-7 11.6 5 11.0 6 43.0% 2 1.1% 16 31.8% 1
6 NYJ 23.8% 10-6 11.1 6 10.2 7 19.2% 9 4.1% 10 16.0% 19
7 DEN 19.6% 10-6 10.2 8 10.1 8 19.4% 8 1.1% 15 26.1% 7
8 SD 19.2% 12-4 10.7 7 11.2 5 20.7% 6 1.8% 12 10.2% 29
9 BAL 18.9% 9-7 10.1 9 9.6 9 15.1% 10 10.8% 2 22.0% 10
10 KC 12.1% 7-9 9.5 10 9.0 11 4.2% 14 5.8% 6 15.7% 22
11 CIN 9.8% 8-8 8.8 12 8.1 15 20.2% 7 8.8% 4 19.4% 14
12 CAR 4.2% 7-9 8.8 13 8.4 12 10.5% 11 -4.4% 24 15.6% 23
13 MIN 1.5% 8-8 9.0 11 8.2 13 4.8% 13 -5.1% 28 15.0% 24
14 JAC -0.1% 9-7 8.4 14 7.3 18 -1.7% 18 5.0% 7 8.4% 31
15 HOU -1.2% 7-9 8.0 17 7.1 19 -1.6% 17 1.4% 14 13.6% 26
16 TB -2.0% 5-11 8.0 16 7.9 17 4.2% 15 -7.3% 31 12.9% 27
TEAM TOTAL
DVOA
W-L ESTIM.
WINS
RANK PYTH
WINS
RANK WEIGHTED
DVOA
RANK PAST
SCHED
RANK VARIANCE RANK
17 WAS -2.4% 6-10 7.0 21 7.1 20 8.2% 12 -2.2% 19 17.6% 16
18 DET -2.9% 6-10 8.0 18 6.4 24 -7.0% 21 -5.1% 27 19.8% 13
19 ATL -2.9% 11-5 8.1 15 8.1 14 -4.1% 19 -5.8% 29 28.3% 4
20 SEA -3.3% 9-7 7.9 19 7.9 16 -13.2% 23 -8.5% 32 10.3% 28
21 GB -5.0% 10-6 7.6 20 9.0 10 3.0% 16 -4.8% 25 26.6% 6
22 NYG -10.0% 6-10 6.6 24 6.7 22 -24.2% 28 0.4% 17 29.8% 2
23 NO -10.6% 8-8 6.9 22 6.6 23 -5.9% 20 -4.9% 26 7.5% 32
24 DAL -14.9% 6-10 6.6 23 5.1 29 -18.3% 24 1.5% 13 9.9% 30
25 MIA -15.5% 4-12 6.3 25 5.7 27 -7.3% 22 7.7% 5 18.8% 15
26 OAK -19.1% 5-11 5.9 26 5.1 28 -21.0% 26 10.3% 3 16.0% 20
27 ARI -19.4% 6-10 5.3 29 6.8 21 -20.4% 25 -6.4% 30 22.4% 9
28 CLE -19.4% 4-12 5.6 27 4.9 30 -22.3% 27 13.7% 1 21.6% 11
29 TEN -22.5% 5-11 5.2 30 5.8 26 -29.8% 30 3.4% 11 21.4% 12
30 STL -25.9% 8-8 5.4 28 6.1 25 -27.3% 29 -3.4% 21 22.9% 8
31 CHI -32.0% 5-11 4.3 31 4.8 31 -42.3% 31 -3.1% 20 16.8% 18
32 SF -46.5% 2-14 2.3 32 3.4 32 -50.3% 32 -4.4% 23 16.0% 21

Posted by: Aaron Schatz on 04 Jan 2005

1 comment, Last at 10 Apr 2006, 2:33pm by cari arneson

Comments

1
by cari arneson (not verified) :: Mon, 04/10/2006 - 2:33pm

brett favre is the all time greatest. nobody compares to him. so all you brett favre haters can just get lost and fall off the earth