Writers of Pro Football Prospectus 2008

Most Recent FO Features

KhanSha1.jpg

» Futures: My Expansion Franchise

You've just been awarded an NFL expansion team and must build your personnel department. How would you do it? Matt Waldman takes on the exercise.

16 Sep 2003

Week 2 Team Efficiency Ratings

by Aaron Schatz

It looks like both Ian Dembsky of Scramble for the Ball and the VOA prediction system knew what they were doing when they listed Seattle as the surprise team of the year.  It is still too early for these ratings to have any serious significance, but after two weeks the Seattle Seahawks rank as the best team in the NFL using our proprietary 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.)

Despite an overtime loss to St. Louis, San Francisco drops only one spot in the rankings to #2.  The VOA system says they totally dominated St. Louis and should never have even come close to going to overtime.  Looking over the play by play, it looks like the 49ers lost this one on penalties (which admittedly I need to make stronger in the formula), and to lose an important division game like this despite outplaying your opponent is the kind of thing that returns to bite you in the ass at season's end.

Speaking of which, Tampa Bay actually moves up to #5 this week despite the loss to Carolina.  Every year we talk about "the worst undefeated team in football" and this year there is no question it is the Panthers.  There they are, down at #29, below Detroit.  Detroit!  Now, this is clearly a place where adding special teams ratings will make up some of the difference, but not all the difference.  Carolina has very good special treams.  My in-progress work on special teams ratings, for example, shows that Todd Sauerbrun was the best punter in football last year by a huge margin, almost twice as valuable as any other punter.  Signing him as the league's best-paid punter was a great move by the Panthers.  But the kind of special teams plays that have won Carolina these two games aren't the kind of plays that happen week after week.  You don't block two field goals and the game-winning extra point every week.  Week 1 was similar, as Carolina blocked a punt and got some ridiculously long kickoff and punt returns.  Carolina's special teams are good, but not that good.  They get this week off, but if they are favored over Atlanta in Week 4 by more than the three-point home field advantage, I'll take the Falcons.

Buffalo has enjoyed two huge wins this year, but they earned each win in a completely different fashion.  The Week 1 victory over the Patriots was driven by defense.  The Bills scored a ton of points, but the offense didn't play particularly well -- they thrived off great field position as the defense forced the Pats to either turn the ball over or punt from back in their own end.  After Week 1, the Bills offense actually ranked as below average.  Week 2 switched things around.  The Bills offense had the highest efficiency rating of Week 2, +51.4%, but the defense was below average, allowing the Jags offense +10.6% VOA.  Put the two games together, and Buffalo ranks #4 overall.

Part of the reason for Buffalo's schizophrenic start is that they played the two teams with the biggest discrepancy between offense and defense this year.  New England really does rank #4 in defense this year.  Commentators cried about how losing Lawyer Milloy decimated the Pats defense, but it was the offense that lost the first game.  In Week 2, the offense improved to being average (literally, they were 0% VOA) and the defense was superb against the struggling Eagles.  The Pats are #4 in defense but #24 in offense, which makes them #13 overall.  Still, since they were #30 after Week 1, they take the biggest jump in this week's rankings.

The flip side of New England is Jacksonville, who completely outplayed Carolina in Week 1 and lost based on horrible special teams play, and then had a reasonable offensive game ruined by a miserable defensive performance in Week 2.  Jacksonville ranks #5 in offense but #25 in defense, which puts them at #14, one spot below the Pats.  The two teams play each other in Week 15, and if they keep playing in this same manner it will be one of the least-hyped, most interesting games of the season.  By the way, Jacksonville's #14 ranking also makes them the flipside of Carolina as well -- they are "the best winless team in football."

If there is anything we learned from Week 2, it is the importance of taking strength of opponent into account when looking at a team's performance.  Remember how last week Arizona and Detroit ranked #1 and #3 in offense after their shootout at Ford Field?  Well, this week the Arizona and Detroit offenses had to play actual professional football defenses, and now they are ranked #26 and #21 in offense.

Still, we can't yet create adjustments based on strength of schedule for 2003.  For example, the Philadelphia offense has gotten clobbered by two of the league's best defenses.  If we did opponent adjustments this early, Tampa Bay's defensive ranking would be penalized because New England also dominated Philadelphia.  If every team keeps the Philly offense down, it can't be that hard, so the system adjusts accordingly.  Once Philly's offense has a good game or two, the adjustment will be more reasonable.  Of course, with the upcoming schedule featuring Buffalo, Washington, and Dallas, it's gonna be a long year for anyone who drafted Donovan McNabb in their fantasy league.

Once again, since the variables used to measure league average performance are from 2002, the average VOA is not 0%.  Last week the offenses came back a little bit, so the league VOA moved from -11.5% to -9.8%.  Still, it has been a defensive year so far.

Finally, let's talk about Jamal Lewis and his record-setting 295 yards against Cleveland.  Believe it or not, the VOA system gives Baltimore only +1.7% VOA for rushing offense in Week 2.  Huh?  Well, there are two issues at play here.  First, that number includes runs by Chester Taylor and a couple other players.  Strip those away to leave just Lewis, and you get a +33% VOA.*  Better, I suppose, but you would still expect a record-setting day to be a bit higher.  VOA says that Oakland, San Diego, and Minnesota all had better weeks rushing.  That's because Lewis got 193 of these yards on three runs.  The system judges all runs over 40 yards the same, which makes some sense when you think about it.  Occasionally a defensive player will catch up to a back on one of these huge breakaway runs -- as one did on Lewis' 48-yard scamper in the second quarter -- but it is rare.  Usually the difference between a 45-yard TD and a 75-yard TD is nothing more than field position.  The rest of the game, Lewis had the usual mix of good and bad runs.  Despite his record, he had 13 different runs of two yards or less on either first-and-10 or second-and-8 or more.  Contrast that with Moe Williams and Onterrio Smith, who together had only four different runs of two yards or less on either first-and-10 or second-and-8 or more.  Everyone loves an 82-yard touchdown, but one yard on first down is not a successful rushing play.

As always, positive numbers represent more points so DEFENSE is better when it is NEGATIVE.  Here are the ratings through Week 2:
 


TEAM
TOTAL
VOA
LAST WEEK
RANK
OFFENSE
VOA
OFFENSE
RANK
DEFENSE
VOA
DEFENSE
RANK
1 SEA 85.4% 2 27.6% 2 -57.7% 1
2 SFO 76.7% 1 19.8% 4 -56.9% 2
3 MIN 63.1% 4 30.9% 1 -32.2% 5
4 BUF 45.7% 3 23.7% 3 -22.0% 11
5 TAM 42.6% 9 10.6% 7 -32.0% 6
6 KAN 37.7% 8 1.5% 12 -36.2% 3
7 IND 28.9% 16 7.2% 8 -21.7% 12
8 WAS 26.8% 10 5.2% 9 -21.6% 13
9 DEN 21.3% 12 4.5% 10 -16.8% 15
10 OAK 12.8% 11 14.8% 6 2.0% 21
11 DAL 12.6% 19 -0.4% 13 -12.9% 17
12 PIT 10.5% 6 -13.4% 17 -23.9% 8
13 NWE 6.6% 30 -25.8% 24 -32.4% 4
14 JAC 3.9% 7 18.0% 5 14.1% 25
15 HOU 2.0% 5 -11.7% 16 -13.6% 16
16 GNB 1.8% 29 -20.3% 23 -22.2% 10
TEAM
TOTAL
VOA
LAST WEEK
RANK
OFFENSE
VOA
OFFENSE
RANK
DEFENSE
VOA
DEFENSE
RANK
17 CIN -7.0% 21 -15.8% 20 -8.9% 19
18 BAL -7.8% 27 -39.4% 29 -31.5% 7
19 NYJ -8.6% 23 4.0% 11 12.6% 24
20 NYG -9.3% 13 -30.2% 27 -20.9% 14
21 NOR -9.3% 31 -20.1% 22 -10.8% 18
22 ATL -10.2% 14 -6.3% 15 3.9% 22
23 CLE -23.6% 17 -45.9% 31 -22.3% 9
24 STL -25.9% 20 -28.2% 25 -2.3% 20
25 MIA -26.5% 28 -4.6% 14 21.9% 30
26 TEN -33.8% 22 -14.5% 19 19.3% 28
27 SDG -34.5% 25 -13.4% 18 21.0% 29
28 DET -35.8% 18 -19.7% 21 16.1% 27
29 CAR -46.8% 26 -30.8% 28 16.0% 26
30 PHI -48.8% 24 -42.5% 30 6.2% 23
31 ARI -60.0% 15 -29.0% 26 31.0% 31
32 CHI -102.2% 32 -69.2% 32 33.0% 32

*Note that this is not Lewis' individual rushing VOA rating.  Individual VOA ratings use additional variables -- for example, to compare RB to other RB and QB to other QB -- that are not used in these team ratings.  We won't have individual ratings until a few weeks from now.

Posted by: Aaron Schatz on 16 Sep 2003

comments