Efficiency Rankings - Week 9

The team rankings below are in terms of generic win probability. The GWP is the probability a team would beat the league average team at a neutral site. Each team's opponent's average GWP is also listed, which can be considered to-date strength of schedule, and all ratings include adjustments for opponent strength.

Offensive rank (ORANK) is offensive generic win probability, which is based on each team's offensive efficiency stats only. In other words, it's the team's GWP assuming it had a league-average defense. DRANK is is a team's generic win probability rank assuming it had a league-average offense.

GWP is based on a logistic regression model applied to current team stats. The model includes offensive and defensive passing and running efficiency, offensive turnover rates, defensive interception rates, and team penalty rates. If you're scratching your head wondering why a team is ranked where it is, just scroll down to the second table to see the stats of all 32 teams.

Click on the table headers to sort.


RANKTEAMLAST WKGWPOpp GWPO RANKD RANK
1 SD20.810.4314
2 NYG30.790.5321
3 PIT10.790.5143
4 TEN60.710.5852
5 IND40.710.56710
6 KC50.680.51157
7 GB70.640.50614
8 MIA90.630.541711
9 HOU100.620.61325
10 PHI80.610.48118
11 NE110.560.54824
12 NYJ180.540.51266
13 BAL130.530.491913
14 CHI160.520.48279
15 MIN210.500.532012
16 DAL110.500.57928
17 WAS120.490.531817
18 NO200.490.421418
19 ATL150.450.471227
20 CIN210.430.462226
21 CLE190.410.532519
22 OAK300.410.451616
23 DEN220.380.481031
24 SF250.360.452120
25 DET270.360.552422
26 JAC310.350.592330
27 TB230.340.421332
28 BUF280.330.562829
29 CAR260.310.46325
30 STL290.290.403015
31 SEA240.230.432923
32 ARI320.210.433121

And here are team efficiency stats through week 8.


TEAMOPASSORUNOINT%OFUM%DPASSDRUNDINT%PENRATE
ARI4.74.65.21.46.24.42.90.44
ATL6.24.22.00.37.04.14.50.36
BAL6.43.62.50.86.04.22.20.39
BUF5.34.53.10.96.75.00.50.33
CAR4.73.35.72.55.53.64.50.38
CHI5.74.05.60.35.53.63.40.36
CIN6.13.82.51.46.24.53.70.38
CLE5.93.93.81.86.93.93.40.42
DAL6.83.73.70.06.94.52.60.54
DEN7.02.91.61.77.14.62.30.52
DET5.63.63.00.86.04.93.40.54
GB6.74.23.30.55.64.64.20.39
HOU6.45.32.60.37.24.11.40.34
IND7.03.70.71.05.84.92.60.35
JAC5.94.24.91.07.94.43.10.37
KC6.25.21.70.35.83.82.20.35
MIA6.33.93.20.86.13.82.30.25
MIN5.94.75.11.26.33.92.60.44
NE6.44.21.80.06.93.93.30.39
NO6.53.63.30.95.84.12.20.41
NYG6.84.84.61.85.03.53.10.45
NYJ5.74.81.81.75.83.42.00.53
OAK6.04.93.30.95.74.71.70.65
PHI6.34.91.60.85.64.04.60.58
PIT7.14.03.71.26.02.63.80.32
SD7.74.22.31.65.23.52.80.39
SF6.03.93.61.16.63.63.00.56
SEA5.03.63.00.06.43.92.20.46
STL5.13.62.70.25.44.32.70.44
TB6.04.21.30.96.65.26.40.42
TEN6.74.22.41.15.64.14.10.55
WAS6.04.12.81.16.14.62.60.38
Avg6.14.13.11.06.24.13.00.43

  • Spread The Love
  • Digg This Post
  • Tweet This Post
  • Stumble This Post
  • Submit This Post To Delicious
  • Submit This Post To Reddit
  • Submit This Post To Mixx

8 Responses to “Efficiency Rankings - Week 9”

  1. Matt says:

    Given SD's OPASS efficiency number (7.7) is so much higher than the rest of the league, or even the second best (colts at 7.1) and that Rivers is working with a hobbled group of receivers, should Rivers be the frontrunner for MVP?

    I realize WPA is probably a better metric for this, but looking at the efficiencies the 7.7 stands out pretty strongly.

  2. Anonymous says:

    Matt, you should also factor in the Colts' (i.e. Manning's) far superior OINT rate. Those two combined make up passing and, in fact, Rivers and Manning are tied at 6.7 AYPA.

  3. Probable Picks says:

    I know, Brian, that your stance is that special teams' performances are uncorrelated from week to week. Which stats have you looked at, exactly?

    I haven't looked at this specifically, but I wouldn't be surprised to see some consistency in return yards (opponent adjusted, of course). Even with a quick glance at the kickoff return avg and punt return avg, I see some similar teams near the top and bottom. And of course it's easy to remember a couple of examples of returners who had consistently good seasons: Hester, Hall come to mind.

    Maybe something to look into, especially since return yards influence starting field position and can account for a significant number of yards over the course of a game.

  4. Brett says:

    What stands out to me: AFC South teams have all played an extremely tough schedule thus far. HOU, JAX, and TEN are the top 3 in Opp GWP, and Indy is not far behind (They definitely benefit from not having to play themselves).

  5. James says:

    Prob Picks, Hester and Hall had good *seasons*, but did any of them have consistently good careers? If Hester and the Bears special teams unit was consistently good at returning kicks then you'd expect them to keep doing it with some regularity like how quarterbacks put up consistent Y/P.

    However, I remember seeing a stat from a Bears game earlier in the season that said like Hester had 8 TD returns in his first 30 games and 0 in his last 30. That seems wildly unpredictable to me.

  6. Probable Picks says:

    James,

    It's true that the performances of Hester and Hall varied from season to season (and I believe that the Bears had Hester on returns less often for awhile), but the point isn't to measure correlation between seasons, but between games within seasons. Like I said, I don't know if such a correlation exists; maybe those examples are the exception, not the rule. TDs/return are not the unit of measurement that I would advocate anyway.

    It could be that a small number of returns per game makes the average yards per return too random to produce a strong correlation across games. But speaking from observational memory, there seem to be some teams each season that have consistently better return games than the rest.

    The attempts are still somewhat low (and perhaps that invalidates any conclusions), but there seems to be a relatively large range in avg yards per return:
    http://espn.go.com/nfl/statistics/team/_/stat/returning/sort/yardsPerKickReturn
    http://espn.go.com/nfl/statistics/team/_/stat/returning/sort/yardsPerPuntReturn

    Anyway, just something to consider.

  7. Michael says:

    Return yards can be misleading. I think the appropriate metric would be "Average Starting Field Position After Kickoff" and "Average Starting Opponent Field Position After Kickoff". Unfortunately, I don't think any of the standard websites have this kind of stat. You'd probably have to derive it from the play-by-play data.

    And I think punt statistics would be absolutely impossible to use in a predictive model. Even if you used one of Brian's advanced stats like EPA vs. Expected EPA per punt, it would end up being explanatory but not predictive (too dependent on lucky bounces, etc.).

  8. Jonathan says:

    Wow, the NFC West really is pathetic.

Leave a Reply

Note: Only a member of this blog may post a comment.