Week 17 DVOA Ratings

by Aaron Schatz

2009 was a very strange NFL season, at least when it comes to statistics, and it ends with a particularly strange Week 17. Take a season with a lot of good teams packed tightly together in total efficiency, add in a week of unexpected blowouts and teams playing like it was the preseason, and you get some very screwy-looking final 2009 DVOA ratings.

Sitting at the top of our ratings for the entire season are the Baltimore Ravens, who barely made the playoffs with a 9-7 record. The Ravens are the second wild card team to top the ratings, following last year's Eagles; because the Eagles had a tie, the Ravens now set a mark for the worst win-loss record by the team that leads the league in DVOA. As noted in last week's column [1], DVOA says the Ravens were much better than their 9-7 record even though they don't come off as conventionally "unlucky" in any way. Their schedule was actually a little easier than average. They were reasonably consistent and they excelled at the right times -- in fact, they were also the league's best team according to our estimated wins formula, with 12.1 estimated wins. They just lost close games to good teams. It was a very strange season.

The Ravens actually were one of three wild card teams at the top of the DVOA ratings. The highest division champion was New England, fourth in DVOA despite going just 10-6. You may not have noticed, but actually only two teams in the entire AFC were better than 10-6. The Chargers and Colts are of course the favorites to come out of the AFC for Super Bowl XLIII, but they aren't favored by DVOA -- San Diego because they started slow, Indianapolis because they gave up on their last two games, and both teams because they had more regular wins than Pythagorean wins.

Actually, those Pythagorean win totals give a great demonstration of how little separates the top teams in 2009. For those who don't know, Pythagorean wins are a simple projection of how many games a team would be expected to win based only on points scored and allowed. Green Bay led the league with 11.8 Pythagorean wins, making this just the second 16-game season where no team finished with at least 12 Pythagorean wins. (The other was 2003, when New England and Kansas City tied for the league lead with 11.4 projected wins.) However, eight different teams this year finished between 11.1 and 11.8 Pythagorean wins. Before 2009, there had never been a season where more than six teams had at least 11 Pythagorean wins.

That group of eight, by the way, doesn't include Indianapolis. Because of their colossal mail-in job over the last two weeks, the Colts end up with 10.8 projected wins -- yet they still end up with the fifth highest difference between regular and Pythagorean winning percentage since the AFL-NFL merger. (They were on their way to breaking the record until Week 16.)

The "sitting starters" thing has completely wreaked havoc on the weighted DVOA ratings that we use for the playoff odds simulation [2]. Green Bay comes out as the best team in football right now, lowering the strength of earlier games... but the Colts drop to 14th thanks to the Curtis Painter experience. To make the playoff odds simulation a little bit more accurate, we've adjusted things by using weighted DVOA through Week 16, not Week 17, for four teams whose games were affected by sitting starters: Arizona, Green Bay, New Orleans, and San Diego. We're also using the same "halfway through Week 16" rating we used last week for Indianapolis. These adjustments make for a smaller gap between Green Bay and the rest of the league and move Indianapolis up to sixth, which is much more realistic.

The best team in weighted DVOA that won't be playing in the postseason is Carolina. The Panthers have been sensational since midseason and finish the year fourth in weighted DVOA, but believe it or not they don't end up with the all-time biggest difference between total season DVOA and weighted DVOA. That record belongs to the 2001 Washington Redskins [3].

As you can see, these big finishes didn't necessarily guarantee a playoff run the following season. The 1997-1998 Atlanta Falcons are the example the Panthers hope they can follow, but most of these teams actually won fewer games the next year. And yes, that table does include DVOA ratings for the 1993 Houston Oilers and Tampa Bay Buccaneers. That's a little addition to our historical data that I'm saving for the dead week after the conference championship games.

Of course, while the top of the league was well-balanced in 2009, the bottom of the league was anything but. Perhaps the best statement about the current state of the league's worst franchises is that the five teams that finished at the bottom of the 2008 ratings [4] also finished at the bottom of the 2009 ratings in slightly different order. It would have been the bottom six teams, except Cleveland's end-year winning streak moved the Browns into 26th place for the year, slightly ahead of Tampa Bay. Detroit becomes just the second team in DVOA history to finish with a rating below -50%, actually coming out worse than they did during their 0-16 season of 2008, and worse than every team since 1993 except for the 2005 San Francisco 49ers. St. Louis is 31st in DVOA but last in Pythagorean wins, with just 1.6 projected wins. That's the third-lowest Pythagorean total for any team in a 16-game season, ahead of only the 1991 Colts (1.4) and the 2000 Browns (1.5).

Two units this year come close to the best of the DVOA Era, starting with the San Diego pass offense. San Diego ends up with 62.2% pass offense DVOA, which trails only the 2007 Patriots and the 2004 Colts. Astonishingly, the Chargers did this despite finishing dead last in the league in run offense. They completely obliterate the previous record for the biggest gap between a team's pass offense DVOA and run offense DVOA, which belonged to the 2003 Tennessee Titans. That team had a gap of 64.7% DVOA; the Chargers have a gap of 73.4% DVOA.

The other historically great unit in 2009 is Cleveland's special teams. Cleveland's rating of 8.8% is twice as high as any other team [6], and currently ranks as the third-highest rating of the DVOA Era behind the 2002 Saints and the 2007 Bears. The Vikings were higher than Cleveland as of a few weeks ago, but slipped in the last few weeks, and no longer have the best year-to-year improvement of any special teams in the DVOA Era. Still, the Vikings go from 32nd last year to an impressive third this year.

And yes, the Pittsburgh Steelers finish with the worst net kickoff value of any team in our database. They aren't dead last in special teams because they actually were positive in three of the other four areas of special teams that make up our ratings.

When it comes to individual stats, 2009 was emphatically a year for quarterbacks. Passing stats have never been better around the NFL. Tom Brady [7] ends up leading the league [8] with 2,170 passing DYAR, the fourth-highest total of the DVOA Era. Peyton Manning [9] is second with 1,932 passing DYAR... and that's the fifth-highest total of the DVOA Era. In total, six quarterbacks from 2009 end up in the all-time top 15 for passing DYAR.

I know a number of readers disagree with the opponent adjustments that give Brady the season lead in passing value. As I noted in an Extra Points post a couple weeks ago, Brady has played against the hardest schedule of opposing pass defenses of any quarterback in the past 17 years. Six of his games came against the top four pass defenses according to DVOA, with four others against the Ravens (7), Saints (9), and Dolphins (11). Against all of these defenses, except for New Orleans, Brady outplayed most or all of his contemporaries. He had the second and fifth-highest DYAR totals allowed by the Jets this year. He had the highest DYAR total allowed by the Broncos, the highest DYAR total allowed by the Dolphins (in the Week 9 game), and the third-highest DYAR allowed by the Panthers. His 250 DYAR against the Bills in Week 1 was double what any other quarterback did against them for the rest of the season.

Some other individual stat notes from 2009, both good and bad:

• JaMarcus Russell finishes the season with -755 passing DYAR, which is the fourth-lowest season of all-time. He ends up ahead of David Carr [10] in 2002, Bobby Hoying [11] in 1998, and Alex Smith [12] in 2005.
• Chris Johnson easily leads the league with 343 rushing DYAR [13], nearly 100 ahead of every other running back, but his awesome 2,000-yard season doesn't even come close to the top rushing DYAR seasons of all-time because it was so boom-and-bust. (Priest Holmes in 2002 and Terrell Davis in 1998 each had over 500 rushing DYAR.) Johnson finishes with a 45 percent Success Rate, 32nd in the league among backs with at least 100 carries. Ryan Grant passed Ray Rice in the final week to rank second in rushing DYAR and quietly had a very good season. Do you remember anyone even mentioning Ryan Grant this year?
• Sidney Rice finishes up as the league leader in receiving DYAR for wide receivers [14], although Vincent Jackson [15] leads qualifying receivers in DVOA and is second in DYAR. Mike Thomas [16] of Jacksonville actually edges out Wes Welker [17] for the league lead in catch rate... albeit on 100 fewer passes. (Thomas caught 77 percent of 62 passes, Welker 76 percent of 162 passes.)
• Although Darrius Heyward-Bey [18] didn't end up with enough pass targets to make our main ranking list for wide receivers, there has to be some sort of recognition for his mind-blowing 23 percent catch rate on 40 passes. No other player since 1994 has put up a catch rate lower than that with at least 25 passes.
• Antonio Gates sets a new record for the best DYAR season by a tight end [19] with 358 DYAR. Here are your all-time single-season leaders for DYAR by a tight end, which goes back to 1994 since I haven't had the time to finish the 1993 receiving numbers yet. Gates has four of the top ten seasons, and one other 2009 tight end sneaks onto the list at the bottom:

Fun end-season housekeeping: All team and individual stats pages should now be updated with final 2009 numbers. FO Premium will be updated with final 2009 ratings sometime tonight. We'll get all the 2009 stats onto the player pages sometime in the next week or two. Loser League results will be announced in Scramble for the Ball on Wednesday.

* * * * *

These are the Football Outsiders team efficiency ratings through the end of the 2009 regular season, 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 [20].)

OFFENSE and DEFENSE DVOA are adjusted to consider all fumbles, kept or lost, as equal value. SPECIAL TEAMS DVOA is adjusted for type of stadium (warm, cold, dome, Denver) and week of season. WEIGHTED DVOA is adjusted so that earlier games in the season become gradually less important. It better reflects how well the team is playing right now. LAST WEEK represents rank after Week 16, while LAST YEAR represents rank in 2008.

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

To save people some time, please use the following format 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>

• ESTIMATED WINS [21] 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.
• NON-ADJ TOTAL VOA lists performance without adjustments for fumble recovery, weather/altitude, or opponent strength.
• 2009 SCHEDULE lists average DVOA of opponents played this season, ranked from hardest schedule (#1, most positive) to easiest schedule (#32, most negative).
• PYTHAGOREAN WINS [22] represent a projection of the team's expected wins based solely on points scored and allowed.
• VARIANCE [23] measures the statistical variance of the team's weekly DVOA performance. Teams are ranked from most consistent (#1, lowest variance) to least consistent (#32, highest variance).