Looking back at FEI's preseason projections, we find that most teams did about what they were supposed to do -- but not in the Big Ten, where things got screwy.
16 Jul 2007
by Doug Farrar
Last year's NFC West was not a good collection of teams. Seattle won the division at 9-7 despite ranking 25th in DVOA, and it seems that this division title came at the expense of a schedule that can politely be called "less than stellar." While writing the Seattle chapter for Pro Football Prospectus 2007 (which comes out on July 23, plug plug plug) I added up the final DVOA numbers for each division for the 2006 season, to see how much worse the NFC West was compared to the other seven divisions. (For those who may be coming to this article from another site, DVOA is Defense-adjusted Value over Average, our primary advanced metric, and it is explained here.)
Using the combined DVOA for each division, I came up with the following totals:
Zoinks. As you can see, the Seahawks weren't exactly facing a bunch of dynasties on their way to their third straight division championship (especially since they also faced every team in the NFC North and got obliterated by Chicago, THAT division's one respectable representative), which is how a team with a -13.5% total DVOA wins a division in the first place. The 8-8 Rams actually "led" the West with a -5.2% DVOA, but they were swept by Seattle in two games that were each decided by last-second Josh Brown field goals. In the second game, the Seahawks would not have kept it close without a 90-yard Nate Burleson punt return, taking advantage of the Rams' 31st-ranked special teams.
The Seahawks' luck against the division rival Rams wasn't duplicated throughout the division; they were swept by the 49ers (DVOA: -20.3%) and split the season series with Arizona (DVOA: -18.1%). Seattle barely eked out the title, one year after becoming the third team in NFL history to win its division by seven games (the others being the 1975 Los Angeles Rams and 2004 Philadelphia Eagles). This unimpressive title in a bad division illustrated Seattle's overall downturn in performance after 2006. Moral: a division title can hide a multitude of sins.
As we'll soon see, the NFC West has been far more "sinful" than any other division, with five of the ten worst DVOA seasons. Nobody knows this better than the Rams. In 2004, St. Louis became the worst team in the 11-year history of DVOA to win a playoff game when they beat the division champion Seahawks, 27-20, at Seattle's Qwest Field. The Rams sported an 8-8 record and a -23.0% DVOA, ranked 30th ahead of only the 5-11 Bears and the 2-14 49ers. The Seahawks (-5.0% DVOA) didn't have much room to complain.
Remembering this playoff win by St. Louis got me to thinking about the strength of each division through DVOA's history and the fact that DVOA would be a pretty decent barometer of said strength. Because DVOA is accrued on a play-by-play basis, is adjusted for opponent strength, and is situation-oriented (measuring both scoring and the ability to move the chains), it's probably the best way we have to reveal which divisions since 1996 have been the best-to-worst. So, I simply added the total DVOA for each team per division per year, and came up with the numbers discussed below.
This method may be too simple -- there probably should be some sort of adjustment for divisions like the 2006 AFC West, whose 0.0% DVOA (Blutarsky!) was more reflective of the abysmal Oakland Raiders (-30.2% DVOA) then it was of the Kansas City Chiefs and the Denver Broncos, whose 4.0% and -3.7% DVOA pretty much canceled each other out. And should the San Diego Chargers, whose 29.9% DVOA was the second-best in the league next to Baltimore, be penalized by being tied to a league-average division when they only got to play the Raiders twice, even though the Raiders played the Chargers tight in a 21-14 loss in the second game?
It's possible that strength of schedule (as measured by Weighted DVOA), Estimated Wins (which calculates "wins" based on adjusted situational criteria) and Pythagorean Wins (which calculates "wins" based on points scores and points allowed) should be added to the math for a more accurate portrayal of division strength. However, the situational and opponent-adjusted aspects of DVOA keep us out of the woods to a point.
I should also mention Doug Drinen's excellent blog at pro-football-reference.com. His recent post about the best and worst divisions since 1970, based on what he calls the "Simple Rating System", sent me further on my way when thinking about division strength, what it means, how hard it is to accurately weigh, and that DVOA might tell us a great deal.
Here are the five best divisions since 1996:
This division featured the best of New England's three Super Bowl teams of the new millennium (2001 DVOA: 6.2%; 2003 DVOA: 22.8%; 2004 DVOA: 35.7%), and this feat was made more impressive by the strength of the New York Jets and Buffalo Bills in 2004. Buffalo may have finished 22nd in Offensive DVOA, but the Bills ended the season first in DVOA for both defense and special teams and third in overall DVOA at 29.8% (the Colts finished second with 33.8%). This was the first year of the Mike Mularkey era, and the last of three seasons for Drew Bledsoe in Buffalo. The Bills were 6-2 in the second half of the year, but one of those losses came against Pittsburgh's second string in Week 17, keeping them out of the playoffs despite a 9-7 record.
As for the Jets, they finished fourth in DVOA at 26.4% behind a solid season from Chad Pennington and a rushing title from Curtis Martin. The 10-6 Jets beat the Chargers in the wild card round and lost to the Steelers in the Divisional frame. New England took the league by storm with a franchise-best 14-2 record, playoff wins over the Colts and Steelers, and a tight 24-21 win over the Eagles in Super Bowl XXXIX. Winning a third Super Bowl in four seasons is impressive enough, but when you do so by going 5-1 in a division like this (losing only to the 4-12 Dolphins late in the year), that really is a season for the ages.
Another division with three of the top teams in overall DVOA for the season. Denver finished second in DVOA with 30.7%, behind only Indianapolis' 33.7%. The Chiefs (27.5%) and Chargers (21.5%) were slotted fourth and sixth, respectively, behind eventual Super Bowl entrants Pittsburgh and Seattle. Denver's 13-3 season ended in a 34-17 AFC Championship game loss to the Steelers. Neither Dick Vermeil's Chiefs nor Marty Schottenheimer's Chargers made the playoffs. The Bengals (18.9%) and Jaguars (17.0%) had better records, which makes their DVOA inferiority small consolation, to be sure. Filling out the division, as has become customary, was a 4-12 Oakland Raiders team.
What, these guys again? Not really, because Oakland and San Diego were switched. The Raiders finished second in overall DVOA (30.9%) behind the Tampa Bay team that would maul them in the Super Bowl. Oakland had a balanced attack, finishing second in offense and seventh in defense. The team that out-dueled Oakland for the Offensive DVOA crown was the Chiefs -- of course, Kansas City could finish no better than 8-8 after ranking 29th in Defensive DVOA, one of the biggest dichotomies between units in recent memory. Denver (16.2%) finished 9-7, and the Chargers matched the Chiefs' 8-8 record despite a -4.4% DVOA. To paraphrase Joaquin Andujar, this was "Onetoughdivision".
The 12-4 Jets took the division and their 26.4% DVOA ranked fourth behind Denver, Minnesota, and San Francisco. The Parcells-led J-E-T-S made it as far as the AFC Championship, where they succumbed to John Elway's final comeback. All three AFC wild cards were also taken by AFC East teams. The Dolphins (21.6%) and Bills (16.0%) both finished 10-6, and Miami beat Buffalo in the wild card round before being trounced by Denver. The 9-7 Patriots (8.7%) completed the best of the five-team divisions. The only team in this division to miss the postseason was Indianapolis. Peyton Manning would certainly have better seasons than this ... but in his rookie season, he'd have to be content with a 3-13 record (IND: -15.3% DVOA, 26th overall) and 28 interceptions. The Colts would reverse that record in 1999, but that's a story for another day.
Super Bowl Champion Tampa Bay led this division with their 12-4 record and league-best 34.0% DVOA; had the Bucs put together even a league-average offense (they finished 21st in Offensive DVOA), they could have been terrifying, since they finished first in Defensive DVOA and 8th in special teams. Their -33.6% Defensive DVOA (remember, Defensive DVOA is better when it's negative) was more than twice the total for any other team that year.
The 9-6-1 Falcons (15.9%), led by Michael Vick in his only truly great season (10th in DPAR for quarterbacks, 825 yards rushing) beat the Packers in that famous Lambeau wild card game. They lost to the Eagles in the following week, as Vick fell to Donovan McNabb, the only QB with a better rushing DPAR that season. The 9-7 Saints (8.3%) of Jim Haslett and Aaron Brooks finished third. Yes, there as a time when Brooks was fairly productive, as well -- he finished right behind Vick in DPAR that year. The 7-9 Carolina Panthers had a -9.2% DVOA in John Fox's rookie year as a head coach, just one season removed from their 1-15 nightmare under George Seifert (-23.7% DVOA).
And now, the FO YukBurgers: the worst Divisional DVOA seasons in recorded history (i.e., back to 1996).
Green Bay finished 12-4, but their DVOA in doing so (7.9% overall, 2.1% offense, -9.3% defense) would imply that their record came at the expense of more than a few patsies. That the Pack was the only team in the NFC Norris to finish with a positive DVOA in 2002 goes a long way toward confirming this. The collective DVOA of their opponents (-2.1%, 24th in the NFL) and Pythagorean projection of 9.8 wins would seem to do the rest. This was also the team that lost the aforementioned wild card playoff game to the Falcons, the only time Green Bay has ever lost a home playoff game, after going 8-0 at home in the regular season.
The 6-10 Vikings (-12.4%) actually posted the division's best Offensive DVOA at 8.5%, but giving up 442 points (30th in the league) and ranking 28th in Defensive DVOA at 15.4% sank them. Minnesota finished the season at a .500 clip after losing their first four, and they'd post three straight winning (though unimpressive DVOA) seasons from 2003-2005 under Mike Tice as the offense exploded and the defense fought for respectability. The 4-12 Bears (-16.3%) won their first two games but then reeled off eight straight losses behind a singularly unimpressive offense by committee and a defense that fell down one year after leading the Bears to a 13-3 record.
As for those poor, poor Detroit Lions? They finished 3-13 with a -35.4% DVOA. It's a copy/paste proposition for the Matt Millen era, current 10-win predictions notwithstanding.
Neither the Seahawks, who actually did so, nor the 49ers, who challenged them at the end of the season, deserved to win the NFC West, but rules are rules. Each division has to have a "champion." Seattle's -13.5% total DVOA, a huge drop-off from the 26.2% they posted in their 2005 Super Bowl season, came about through depleted lines on both sides of the ball and injuries to Matt Hasselbeck and Shaun Alexander. They had to field most of their starters against a horrid Tampa Bay team in the season finale just to avoid the indignity of an 8-8 division title. The 49ers, who had posted the worst team DVOA in history in 2005 (-56.6%), would find their -20.3% DVOA in 2006 to be the sort of improvement that their 7-9 record would seem to imply after going 4-12 the year before. After setting a pace to give up more points than any team in NFL history through the first half of the season, Team Nolan tightened things up a bit. A super-aggressive off-season makes them everyone's division darlings for 2007.
Forgotten in all this statistical drama were the 8-8 Rams (-5.2%), who posted the division's best DVOA. Extreme early difficulty in the red zone gave way to offensive effectiveness with Marc Bulger and Steven Jackson as the year progressed. If the defense had been league-average instead of 29th at 13.1%, St. Louis might have had something to talk about in Scott Linehan's first season. In Dennis Green's last season, the 5-11 Cardinals crowned the asses of their opponents on a disturbingly regular basis. They had a -18.1% DVOA, including a -7.3% Offensive DVOA, to the surprise of many who speak of the powerful Arizona passing game.
We're getting near the basement now ... when the team atop the division represents the NFC in the Super Bowl and there's still this sort of total number, there are issues. Seattle's 23.2% DVOA (13-3 record) was fifth in the league and the best in team history. San Francisco's historically bad DVOA has already been discussed, and this was the 49ers' second straight season at the bottom (a record the Raiders have a very good chance of tying in 2007). St. Louis' -16.0% DVOA, a season after finishing 30th, reflected the competitive ambiguity of Mike Martz' last days. Risk/reward offenses, an increased inattention to special teams, and a defense that was springing leaks just about everywhere told the story.
The 5-11 Cardinals (-12.3% DVOA) were who we thought they were (6.0 Pythagorean wins, and yet another negative Offensive DVOA), which would, improbably, lead to yet another offseason of "Arizona will win the division next year!" predictions. Are we over those yet?
This is an interesting case. Three teams at .500 or better, but we're down to second-to-last in DVOA. Mathematical error? Not quite. The worst of the five-team divisions is here because the 10-6 Cowboys were the only team to post a positive DVOA (14.6%), and even they couldn't keep all sides better than league average (4.5% defense). Dallas won the division but lost in the wild card round to the 9-7 Cardinals of Vince Tobin and Jake Plummer (-18.7%). That's the lowest DVOA of any team to win a playoff game until those 2004 Rams came around. The third-place Giants stumbled to an 8-8 mark with Jim Fassel at the helm and the QB tandem of Kent Graham and the "immortal" Danny Kanell (whose jersey our own Bill Barnwell proudly wears while watching NFL games in the Port Authority Bus Terminal). Rounding up this cavalcade of world-beaters was the 6-10 Redskins (-20.2%) of Norv Turner (really!) and Trent Green (yes, indeed), and a 3-13 Eagles team which can be explained as follows: Bobby Hoying, Koy Detmer, and Rodney Peete. Those were the quarterbacks who each took 129 or more snaps from center.
If ever there was a case for the dissolution of the old five-team divisions, it would be the 1998 NFC East. And if ever there were a case for the removal on an entire division from the NFL altogether...
At last, the La Brea Tar Pits of competitive professional football. Not only did this division provide the worst division champ and playoff winner in DVOA history (two different teams!), but this occurred in a season that saw an entire conference of below-average divisions. That's right -- the NFC East (-9.3%), South (-20.0%) and North (-34.1%) all fell repeatedly in the mud, and only the Eagles (22.9%) and Packers (1.6%) posted positive DVOA marks. But for pure, unadulterated suckitude, the 2004 NFC West will most likely stand as the model for many years to come.
The 9-7 Alleged Division Champion Seahawks (-5.0%) were swept by the 8-8 Rams (-23.0%) and faced the weakest schedule that season (-8.6% opponent DVOA). Wildly inconsistent defense and special teams combined with a combustible offense to bring forth one very exasperating squad. Mike Holmgren, beaten down by on-field drama and off-field power struggles, came very close to quitting after this season. Those aforementioned Rams got past the Seahawks again in the playoffs despite a radically imbalanced offense in which Marshall Faulk's wheels had pretty much fallen off entirely, and Mike Martz called 200 more pass plays than rushing plays. The offense had dipped to league average, ending the "Greatest Show on Turf" era once and for all.
The 6-10 Cardinals (-21.6%) were in the throes of great excitement, as the Dennis Green Era had just begun. At one time, it was possible to say that with a straight face. And the 2-14 49ers (-46.1%), in the last gasps of the uber-dysfunctional Terry Donahue/Dennis Erickson era, were just hoping for better days ahead.
And there you have it -- 11 DVOA seasons in the books, and the best and worst divisions coming out of the same year. Ironically, the teams from the two divisions played each other one time each in 2004, and the AFC East emerged with a 13-3 record over the NFC West. Little surprise, really -- with all the variables and crazy things that can happen in a season, there's very little left untold by this sort of swing from best to worst.
Here's the rest of the list:
|2000 AFC East||46.4%||1996 AFC West||20.2%||2006 AFC West||0.0%||2004 NFC South||-20.0%|
|2004 AFC North||44.1%||1997 NFC Central||17.9%||2000 AFC West||-0.5%||1996 NFC West||-20.1%|
|2002 AFC East||39.9%||2000 NFC Central||15.4%||2005 AFC South||-2.9%||1999 NFC East||-21.8%|
|2005 NFC East||38.2%||2001 AFC West||13.1%||1996 NFC Central||-4.3%||2002 AFC North||-23.6%|
|1997 AFC Central||38.1%||2001 NFC West||12.7%||2006 NFC North||-5.6%||1997 NFC East||-24.2%|
|2004 AFC West||37.6%||2006 AFC East||12.3%||2003 NFC South||-7.0%||1997 NFC West||-24.8%|
|1999 AFC West||34.1%||1999 AFC Central||12.3%||2002 NFC East||-7.6%||2000 NFC West||-28.4%|
|2006 NFC East||22.0%||1997 AFC West||9.5%||2004 NFC East||-9.3%||1997 AFC East||-28.7%|
|1999 AFC East||30.5%||2004 AFC South||.1%||2001 AFC Central||-9.6%||2000 NFC East||-30.2%|
|2006 AFC North||27.5%||2006 AFC South||8.7%||2005 NFC South||-11.2%||1996 AFC East||-30.8%|
|1996 AFC Central||26.3%||2003 AF C West||7.7%||2003 NFC North||-12.9%||2005 AFC East||-33.8%|
|1998 NFC Central||25.1%||2000 AFC Central||4.3%||2001 AFC East||-13.1%||2004 NFC North||-34.1%|
|2005 AFC North||24.7%||1996 NFC East||3.4%||1999 AFC Central||-13.3%||2002 NFC West||-34.3%|
|2003 AFC South||23.2%||1998 AFC West||2.8%||2001 NFC East||-16.3%||1999 NFC West||-34.8%|
|2003 AFC East||22.0%||2001 NFC Central||1.2%||2003 NFC East||-16.9%||1998 AFC Central||-36.3%|
|1998 NFC West||21.6%||2003 AFC North||1.0%||2006 NFC South||-17.4%||2005 NFC North||-37.1%|
|2003 NFC West||-17.4%||2002 AFC South||-37.2%|
46 comments, Last at 09 Oct 2008, 1:50pm by Joshua