Writers of Pro Football Prospectus 2008

Most Recent FO Features

LuckAnd12.jpg

» Scramble Over/Unders: The Souths

Mike and Tom's tour of the league gets musical! Listen to the strangely up-tempo Blinded by the Luck, the haunting melancholy of The Schiano Man, and the endless march of the Saints.

28 Nov 2004

2000 DVOA Ratings and Commentary

by Aaron Schatz

In the beginning, the dark ages when matter had not yet formed into planets and Football Outsiders only had five readers a day, four of whom were our fraternity brothers, there were only the 2002 DVOA ratings.  Each week during the 2003 season, we created new ratings, and after Adam Vinatieri's kick sailed through the uprights we had two years worth of ratings.  But why stop there -- we want to do ratings for every year we can get play-by-play for.  Over the summer, I broke down 2000 and 2001 and promised to unveil both on the site early in the season.  I presented the 2001 ratings and commentary in early September, in lieu of a regular commentary on Week 1.  And I presented the 2000 ratings and commentary... um, never.  I got busy writing about the current season, I got the weekly gigs at ESPN and the New York Sun, and the 2000 commentary sat on my "to do" list.  Week after week went by and if you looked at the DVOA pages you still saw "(coming soon)" next to all the 2000 entries.  Until now!  Powered by leftover turkey, Arizona Diet Green Tea, and the new Rufus Wainwright album, Football Outsiders proudly presents:

2000 DVOA RATINGS AND COMMENTARY!

The 2000 NFL season was dominated by two teams, but these teams did not face each other in the Super Bowl.  These two teams were not only in the same conference, they were in the same division.  The 13-3 Tennessee Titans claimed the AFC Central title and the number one spot in the DVOA ratings, but the 12-4 Baltimore Ravens claimed the more important prize, the Super Bowl championship.

All summer long, as I tweaked and upgraded the DVOA formulas, Tennessee and Baltimore traded places atop the 2000 ratings.  I would not be surprised if the next upgrade on the system puts Baltimore back on top.  But no matter how you look at it, these two teams were far ahead of the rest of the NFL.

The pattern of the top DVOA teams in 2000 may look awfully familiar, because it was a year dominated by the AFC just like the current season.  Eight of the top 11 teams in DVOA were AFC teams, and the eventual NFC champion Giants are down at #13.  At the same time, the top teams in 2000 were very different than the top teams this season.  In 2000, the top teams put defense ahead of offense -- not only the Titans and Ravens but also the Dolphins and the top NFC team, Tampa Bay.  In the current season, on the other hand, teams that put defense ahead of offense are mostly struggling (Buffalo, Washington, Chicago) and there are a number of highly ranked teams that put offense ahead of defense (Indianapolis, Philadelphia, San Diego, New York Jets).

Before we discuss the most interesting teams of 2000, let's show you the numbers.  These are the Football Outsiders team efficiency ratings for 2000, 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.)

DVOA represents adjusted statistics.  OFFENSE and DEFENSE DVOA are adjusted for opponent quality and 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 actual wins.  WEIGHTED DVOA gives a stronger consideration to games late in the season.  Remember that, as always, defense is better when it is NEGATIVE. Estimated wins, variance, and other numbers here.


TEAM TOTAL
DVOA
W-L NON-ADJ
TOTAL VOA
WEIGHTED
DVOA
RANK OFFENSE
DVOA
OFF.
RANK
DEFENSE
DVOA
DEF.
RANK
SPECIAL
DVOA
S.T.
RANK
1 TEN 36.1% 13-3 31.1% 37.5% 1 3.6% 12 -27.2% 2 5.4% 5
2 BAL 32.6% 12-4 51.8% 24.2% 4 -6.5% 20 -31.5% 1 7.6% 2
3 MIA 24.7% 11-5 21.1% 23.0% 5 -5.1% 19 -21.1% 4 8.6% 1
4 PIT 24.6% 9-7 26.9% 26.3% 3 11.2% 8 -12.1% 9 1.4% 15
5 OAK 24.5% 12-4 29.4% 29.0% 2 14.8% 5 -4.8% 15 4.9% 7
6 TB 21.5% 10-6 26.9% 20.6% 7 0.8% 14 -16.7% 6 4.1% 10
7 DEN 20.0% 11-5 35.1% 18.1% 10 14.6% 6 -7.6% 13 -2.2% 22
8 STL 18.0% 10-6 18.8% 18.9% 9 31.4% 1 11.9% 26 -1.5% 17
9 IND 16.8% 10-6 15.0% 11.2% 13 27.5% 2 9.0% 24 -1.6% 19
10 WAS 13.7% 8-8 9.7% 5.8% 16 5.7% 11 -14.2% 7 -6.2% 29
11 NYJ 13.3% 9-7 0.0% 13.2% 11 -4.4% 17 -21.6% 3 -3.9% 25
12 PHI 12.5% 11-5 19.3% 4.0% 17 -1.8% 16 -9.1% 11 5.1% 6
13 NYG 9.1% 12-4 21.7% 19.2% 8 7.7% 10 -7.2% 14 -5.8% 28
14 DET 6.6% 9-7 5.7% 3.3% 18 -17.1% 25 -18.2% 5 5.5% 4
15 JAC 6.5% 7-9 11.1% 21.0% 6 12.0% 7 7.0% 23 1.6% 14
16 GB 6.3% 9-7 2.6% 11.7% 12 0.4% 15 -1.0% 17 4.8% 8
TEAM TOTAL
DVOA
W-L NON-ADJ
TOTAL VOA
WEIGHTED
DVOA
RANK OFFENSE
DVOA
OFF.
RANK
DEFENSE
DVOA
DEF.
RANK
SPECIAL
DVOA
S.T.
RANK
17 BUF 0.1% 8-8 3.1% -1.1% 19 3.4% 13 -11.0% 10 -14.4% 31
18 KC -0.1% 7-9 2.2% -10.2% 21 8.1% 9 4.1% 19 -4.1% 27
19 NO -0.5% 10-6 -13.3% 8.9% 14 -4.8% 18 -7.9% 12 -3.5% 23
20 SF -4.3% 6-10 4.3% 7.9% 15 14.8% 4 15.2% 28 -3.9% 26
21 MIN -5.4% 11-5 -1.7% -5.9% 20 20.4% 3 26.7% 31 0.8% 16
22 NE -9.7% 5-11 11.5% -23.6% 27 -7.5% 21 5.7% 22 3.5% 12
23 CAR -16.0% 7-9 -14.6% -16.6% 23 -20.9% 27 -1.3% 16 3.7% 11
24 DAL -16.0% 5-11 -24.0% -17.4% 24 -13.9% 24 5.2% 21 3.0% 13
25 CHI -18.8% 5-11 -29.3% -12.9% 22 -12.5% 23 4.4% 20 -1.9% 21
26 SEA -21.5% 6-10 -25.4% -19.9% 26 -9.8% 22 17.3% 29 5.7% 3
27 ATL -26.0% 4-12 -32.0% -32.8% 29 -28.4% 29 1.9% 18 4.3% 9
28 SD -29.9% 1-15 -48.8% -17.9% 25 -39.0% 31 -12.9% 8 -3.9% 24
29 CIN -40.4% 4-12 -46.2% -31.7% 28 -18.9% 26 13.6% 27 -7.9% 30
30 CLE -45.5% 3-13 -51.8% -51.6% 31 -31.9% 30 11.8% 25 -1.7% 20
31 ARI -46.9% 3-13 -51.6% -45.6% 30 -21.4% 28 24.0% 30 -1.5% 18

The 2000 AFC Central was a pretty remarkable division.  It not only had the two best teams in football, it had three of the top four teams in TOTAL DVOA and four of the top six teams in WEIGHTED DVOA because Jacksonville went 5-3 over the second half of the year and lost their final two games by only three points apiece.  Meanwhile, down at the bottom of the division were two of the three worst teams in the league, Cincinnati and Cleveland.  But if you think that the AFC Central teams only had high ratings because they got to beat up on the terrible twins of Ohio, remember that the Ravens, Steelers, and Titans all had to play each other twice as well.

Although Tennessee had a very good defense in 2000, I'm sure most people are a little stunned that it comes so close to Baltimore's defense atop the ratings -- after all, the 2000 Baltimore defense is considered by many to be the greatest defense of all time.  But you may notice that while Baltimore and Tennessee are very close to each other in DVOA, Baltimore is far superior to Tennessee in non-adjusted VOA.  The adjustment is almost entirely on defense, and it is a byproduct of the confused schedule caused by a six-team division in the three years between the Cleveland expansion in 1999 and the Houston expansion in 2002.  Each team in the AFC Central played ten division games, and then either two or three games with other AFC teams and either three of four games against the NFC East, scheduled by throwing darts at a map of the United States.

As a result, Tennessee's non-division schedule included five offenses ranked in the top half of the league according to our ratings (KC, NYG, WAS, BUF, PHI) while Baltimore played only one offense ranked in the top half of the league (WAS) and three offenses in the bottom ten (DAL, SD, ARI).  Pittsburgh, by the way, also has their defensive rating penalized by an easy slate of opponents, although not quite as easy because Arizona wasn't on the schedule.

One more interesting note on the Baltimore defense: it was the only dominant defense of the past few years to be based on stopping the run instead of controlling the pass.  For all the cliches about the need to stop the run, in general the best defenses are ranked the highest because they have the best ratings against the pass, not against the run -- just like the best offenses are ranked the highest because they are the best through the air, not on the ground.  The 2000 Ravens only rank seventh in defense against the pass, but their rushing defense DVOA of -41.9% is far and away the best of the past six seasons.  Only one other team of the past six seasons had a rushing defense DVOA of better than -30%, the 1999 San Diego Chargers.

* * * * *

Despite all the fascinating goings-on in the AFC Central, the most interesting team of 2000 is over in the AFC East.  Let's look at the remarkably screwed up world of the 2000 Buffalo Bills, who hold the NFL record for worst special teams of all time and dumbest coaching decision ever made.

When all the plays are edited and all the numbers are run, the first numbers I get are offense and defense.  And when I added together offense and defense, Buffalo came out with a DVOA of 14.4%.  Now, that wasn't one of the best ratings for the year by any means, but it really stood out when compared to Buffalo's record.  How could a team that went 8-8 and was outscored by its opponents 350 to 315 end up with a rating that positive?  This wasn't a fluke result as astonishing as the current Kansas City Chiefs, but it was close.

What made it even stranger was the fact that DVOA is a better indicator of how a team will do the following year than wins or points.  In general, a team that finishes with more wins than its DVOA would otherwise indicate will regress the following year, and a team that finishes with fewer wins than its DVOA would otherwise indicate will rebound.  Look at the 2000 ratings above and compare to 2001 records.  Pittsburgh, with a DVOA higher than nine teams with better won-loss records, went 13-3 the next year.  Minnesota and New Orleans, each with double digit wins despite negative DVOA ratings, declined to 5-11 and 7-9 the following year.

Except the 2001 Bills weren't a good team, as might be indicated by their 2000 performance (or at least, my initial computation of it).  They were a terrible team, the worst team in the AFC, 3-13 with the second-worst DVOA in the league.  What the heck was going on here?  Did DVOA just not work for 2000? There was one step remaining.  I had not run special teams yet.  Could special teams possibly be that negative?

Oh, yes.  They could.  The 2000 Buffalo Bills had the worst special teams of any team in any season for which I have data.  I would not be shocked if they had the worst special teams of all time, except maybe for some expansion teams in the sixties and seventies.  Add Buffalo's -14.4% DVOA to the 14.4% offense and defense and you get: 0%, perfect for an 8-8 team.  Here's a look at the five worst special teams of 1999-2003, plus one team which would make that list if it continues at its 2004 pace.  The numbers represent points above or below average for each aspect of special teams, adjusted for weather and altitude, like our usual special teams ratings:


YEAR TEAM FG/XP KICKOFFS KICK
RETURNS
PUNTS PUNT
RETURNS
S.T.
DVOA
2000 BUF -6.8 -33.6 -9.6 -14.5 -15.2 -14.4%
2002 CIN 2.2 -13.8 7.0 -35.7 -23.7 -11.5%
2004* STL 1.7 -6.3 -13.0 -12.0 -7.5 -10.7%
2001 CIN -13.5 -11.0 -8.5 -4.5 -13.3 -9.2%
2001 NYG 4.2 -17.2 -17.4 -6.1 -10.2 -8.4%
2002 WAS -11.1 -3.0 -4.8 -23.0 -4.4 -8.3%

*(Through Week 11 of 2004)

Everything about special teams was horrible for the Bills that year, but Steve Christie was the biggest black hole among a galaxy of sucking black holes.  He was bad at kicking field goals, but that's nothing compared to his kickoffs.  Christie's average kickoff went only 55.6 yards, 7.5 yards less than the league average.  And the bad kickoffs didn't just come late in the season in Buffalo's usual snow and wind.  Week 8 against Minnesota -- in a dome! -- Christie had five kickoffs of 56 yards or less.  On the season, 18 of his kickoffs went for 50 yards or less.  No other kicker had more than 10 kicks that short.  Once Christie's kick had come down, Buffalo allowed opponents kick returns worth 22 points more than the league average -- no other team in the NFL gave up kick returns worth half that.  So the average opposing drive after a Buffalo kickoff started at the 37-yard line.  Wow.

But special teams alone were not enough to take an 11-5 team that had been knocked out of the playoffs by a fluke kickoff return the year before and turn it into a shining 8-8 beacon of mediocrity.  No, the coach of this team also had to make the wrong decision on one of the most mismatched quarterback controversies of all time.  Somehow, despite all logical evidence screaming the name "Doug Flutie" in a voice louder than any ever heard among the dormitories of Chestnut Hill, Wade Phillips handed the keys to Rob Johnson.  Cue sound of plane crash.

In 1998, Flutie took over a 1-3 team in Week 6 and finally proved he could play quarterback in the NFL, going 8-3 over the next eleven games before sitting out the final contest with a playoff spot assured.  In 1999, he led the Bills to an 11-4 record and once again sat out the final game with a playoff spot assured.  But this time, coach Wade Phillips shockingly decided that a 31-6 victory over the defensively challenged Colts was enough proof to start Johnson over Flutie in the AFC Wild Card game in Tennessee.  Johnson had a terrible game, throwing for only 131 yards, but Buffalo nearly pulled off the upset win until the Music City Miracle sent them home.

So in 2000, did Phillips go back to Flutie?  No, of course not, he started Johnson for the first six games and only gave Flutie the ball when Johnson had his inevitable injuries.  Flutie went 3-1 from Weeks 8-11, and when Johnson was healthy it was back to Johnson and back to losing.

The difference between the 2000 Bills with Johnson at quarterback and the 2000 Bills with Flutie at quarterback is astonishing.  Not only was Flutie far superior to Johnson as a passer, the running game performed better when Flutie was starting, and even the defense performed far better when Flutie was starting.  You would think the quality of the quarterback should have no effect on defensive performance -- read this article about last year's Falcons by Al Bogdan -- but the 2000 Bills are at least one data point to suggest that there is an effect.  Here is the DVOA for each Buffalo unit with Johnson starting and Flutie starting, plus the DVOA for just the passing game:


Buffalo DVOA by Starting Quarterback, 2000
Starting QB W-L PASSING
OFFENSE
TOTAL
OFFENSE
DEFENSE SPECIAL
TEAMS
TOTAL
DVOA
Rob Johnson (Weeks 1-7, 12-15) 4-6 1.2% -1.3% -5.1% -11.7% -7.9%
Doug Flutie (Weeks 8-11, 16-17) 4-2 20.0% 10.7% -21.4% -18.1% 14.1%

And then in 2001, Flutie went off to San Diego and had a good season with a team that was terrible otherwise, and Rob Johnson was his usual crappy self and the Bills went 3-13, which caused the Bills to trade for Drew Bledsoe, who of course was on the decline, and Buffalo has been a mess for years even though they have a good defense now.  Coming soon: Gregg Williams as Count Olaf in the new film, Tom Donahoe's A Series of Unfortunate Events.

* * * * *

Meanwhile, over in the NFC, all three divisions (remember when there were only three?) were won by a team with a lower DVOA than the team in second place.

In the Central division, close wins gave Minnesota the division title over the team with the NFC's top DVOA rating, Tampa Bay.  Minnesota's points scored and allowed project to a team with only 8.6 wins, not 11, and Tampa Bay's points scored and allowed project to a team with 11.3 wins, not 10.

That's not the case in the West, though, where the Rams, coming off a Super Bowl win, battled the Saints, who were 3-13 the year before.  Given that DVOA is a better measure of year-to-year consistency than wins and losses, it is no big shock that the Rams were #8 in DVOA (sandwiched between two Super Bowls) and the Saints were #19 in DVOA (sandwiched between 3-13 and 7-9).  So how did they end up with the same 10-6 record -- and virtually identical Pythagorean projections based on points scored and allowed?  To be honest, I'm not quite sure.  There wasn't much difference between the two teams in any of the usual "luck" measurements: schedule difficulty, week-to-week inconsistency, or the percentage of recovered fumbles.  The Rams won about as many games as you would expect from looking at their play-by-play; the Saints were the team that didn't match up.  I looked at each game for New Orleans that year and there didn't seem to be any where they won despite being outplayed, although they did have a habit of only narrowly beating some really bad teams.  They beat 1-15 San Diego by a single point, 4-12 Atlanta by two points, 6-10 San Francisco by four points, and 3-13 Arizona by eleven points in a game that was much closer than the final score (The Cardinals outgained the Saints 394-247, but four turnovers in the second half cost Arizona a chance at the win).

In the East, although the Giants were only the #13 team in DVOA over the entire year, they were the second-strongest team in the NFC going into the postseason, and when Philadelphia upset Tampa Bay in the first round of the playoffs it opened the door for the Giants to march all the way to the Super Bowl.  If we had Football Outsiders back in January 2001, New York's 41-0 cremation of the Vikings would not have been such a surprise.

New York and Philadelphia both went to the playoffs, but DVOA says that Washington was the better NFC East team over the course of the season.  The Redskins started 6-2, but blew it by losing six of their next seven games.  The following year, they went the exact opposite direction, starting terribly and finishing strong.  Washington has just been a generally weird team since Daniel Snyder bought them.  To the right you'll find one of my neat week-to-week DVOA graphs, only this one has both the Giants and Redskins on it to show the difference between the two seasons.  The heavier curved line represents the trend of performance, and Weeks 4 and 14 are the two games the teams played against each other.

* * * * *

A look at the 2000 quarterback numbers is interesting four years later.  Peyton Manning was the top quarterback of 2000 in DPAR, what a shock, but it is interesting to note that Trent Green was the top quarterback in DVOA.  For those few who do not remember, Green was injured at the start of the previous year, and Kurt Warner took over and had his improbable MVP season as the Rams won the Super Bowl.  In the middle of 2000, Warner got injured and it was Green's turn to take over -- and he was better, over those six games, than Warner was over the rest of 2000.  The team went 2-4 because they gave up a zillion points, and because those six games were against more difficult opponents.  But Green had 42.8% DVOA and 59.7 DPAR compared to Warner's 17.2% DVOA and 48.0 DPAR.

Of course the following year, Green went to Kansas City and threw a hundred interceptions, while Warner topped the QB ratings and took the Rams back to the Super Bowl, so St. Louis did keep the right guy.  But it is interesting that for 2000, Green had more value.

2000's number two quarterback in DVOA behind Green was not Manning or Garcia or Culpepper, it was Brian Griese.  I really think that Griese got a bum rap in Denver, and his career wasn't helped when he got stuck playing a few games behind that horrible Miami offensive line last year.  As I noted in my AFC midseason review, Jake Plummer is proving to be the exact same quarterback as Griese: sometimes impressive, sometimes frustrating, always inconsistent.  I'm not really that surprised that Griese is having success in Tampa Bay this season -- nobody should ever expect him to be Brett Favre, but he'll win if you put him in the right situation.

The other interesting quarterback in the 2000 ratings is Elvis Grbac, who ranks sixth with 72.3 DPAR.  I've never quite understood what happened with this guy.  He was a good backup in San Francisco, was very good in Kansas City, was stuck having a mediocre year with a Baltimore team that was trying to run a ground-based offense with Terry Allen instead of Jamal Lewis, and then what... He quit?  Nobody wanted him?  Some combination?  It's not clear what made Grbac stop playing at the age of 31 when he clearly had a few years left.

* * * * *

At running back, Marshall Faulk put a Barry Bonds-like gap between his performance and everyone else in football.  Despite playing only 14 games, he led the league in DPAR both rushing (60.0, Edge was second at 45.5) and receiving (8.7, Barber was second at 5.8).  His rushing DVOA was 43.6%.  That was more than twice the second-best rushing DVOA by a running back with at least 75 carries that season, which belonged to Tony Richardson of Kansas City at 18.1%.  It was the best rushing DVOA by any running back over the past five seasons -- second was Priest Holmes with 33.4% in 2002.  Faulk also had a success rate of 61%, which was significantly better than the running back who finished second in that statistic, Mike Alstott at 54%.

Meanwhile, down at the bottom of the rushing numbers, Jamal Anderson never should have come back from his ACL injury.

The top wide receiver in DPAR (total value) was Isaac Bruce; the top wide receiver in DVOA (value per play) was Derrick Mason yet again.  I believe he has been the most underrated wide receiver in football since man first walked upright.  The top tight end was Tony Gonzalez, no surprise, but check out the nice year by Kyle Brady of Jacksonville.

2000 numbers are now on the site for individual positions as well as team offense, defense, and special teams.  (I'm still trying to get the offensive and defensive lines to spit out correctly for 2000 and 2001.)  Feel free to ask questions about 2000 in the discussion thread or in email, just like I did with 2001 I'll likely collect the questions into a special 2000 DVOA mailbag -- if I can find time between articles on the current season!

Posted by: Aaron Schatz on 28 Nov 2004

6 comments, Last at 27 Mar 2007, 8:03am by ruth van sickle ford

Comments

1
by dine out for america (not verified) :: Thu, 02/22/2007 - 8:33pm

dine out for america

Modern view of dine out for america.

2
by European Pharmacy Pills Prescriptions (not verified) :: Mon, 03/26/2007 - 4:03am

European Pharmacy Pills Prescriptions

eighty seen?string waiving auditorium perjury

3
by ford stock stereo (not verified) :: Mon, 03/26/2007 - 6:16pm

ford stock stereo

Pros and cons of ford stock stereo.

4
by propellerhead reason refill (not verified) :: Mon, 03/26/2007 - 8:46pm

propellerhead reason refill

Modern view of propellerhead reason refill.

5
by porsche cayenne cd changer (not verified) :: Mon, 03/26/2007 - 10:59pm

porsche cayenne cd changer

porsche cayenne cd changer company

6
by ruth van sickle ford (not verified) :: Tue, 03/27/2007 - 8:03am

ruth van sickle ford

Relevant ruth van sickle ford