Team Rankings: Week Seventeen

After a convincing run on top, the Houston Texans are no longer the best team on the board. The Charlie Batch-led Steelers have taken over the top ranking thanks to a Texans loss to the now 30th ranked Colts. Unless the Ravens lose to the Bengals this week, the Steelers will be the fifth seed in the AFC Playoffs. While they may have to convince themselves that they don't need home field advantage to win games in the postseason, it certainly won't help their chances against the pesky Ravens and the powerful Patriots.



Other Changes of Note
  • Philadelphia is now a top-3 team according to these rankings, and they will need a win on Sunday just to get to 8-8 this year. If most of the team comes back intact next season, the Eagles will certainly be a team to watch if they can make a few improvements along their offensive line and give Vick more opportunities to complete passes.
  • The Jets have lost all respect from the rankings, falling to 20th from their previous perch of 16th. The Jets defense is still highly thought of, but they could still use another pass rusher opposite of Calvin Pace. The offense is the Jets Achilles heel, and their coordinator is likely going to pay the price for it. They will have to ask themselves if Mark Sanchez deserves another season to prove himself under center during the offseason.
  • Being ranked below the Colts and the Rams is no easy feat, but the Tampa Bay Bucs have managed to do it. Congrats!
Without further ado, here is your uncut and unrated director's edition of the rankings for week seventeen.




RANKTEAMLAST WKGWPOpp GWPO RANKD RANK
1 PIT20.750.4744
2HOU10.740.4681
3 PHI60.680.5269
4 NO30.680.47323
5 GB50.670.46228
6 NE40.650.50130
7NYG80.650.53520
8DET100.610.48132
9BAL90.600.48156
10DAL70.600.50722
11ATL120.580.501211
12 SD110.540.471027
13 SF130.520.50177
14CHI170.510.51253
15CIN150.500.481418
16MIA180.490.532012
17OAK140.490.481116
18TEN210.490.471814
19BUF200.460.511624
20NYJ160.460.52285
21WAS190.450.512213
22CAR230.430.48932
23ARI220.430.491919
24SEA240.410.522410
25DEN250.360.503017
26 KC270.350.512621
27CLE260.350.502325
28JAC280.320.53328
29STL290.310.543115
30IND310.310.532929
31MIN320.300.512726
32 TB300.290.522131


TEAMOPASSORUNSR%OINT%OFUM%DPASSDRUNSR%DINT%PENRATE
ARI5.9434.31.26.0551.70.48
ATL6.7402.10.96.5613.10.39
BAL5.9412.31.65.4613.10.36
BUF6.2443.90.76.8544.20.38
CAR7.0463.20.77.3532.90.55
CHI5.9394.21.06.2652.90.41
CIN6.0412.90.95.8581.90.44
CLE5.2412.30.86.0512.10.42
DAL7.0442.11.06.5562.90.41
DEN5.4423.22.06.5581.80.41
DET6.7382.30.75.6593.60.50
GB8.2401.40.77.1515.00.27
HOU7.0432.11.25.1603.40.42
IND5.2382.42.17.1541.70.27
JAC4.3383.31.06.2623.20.36
KC5.8373.81.26.9594.40.47
MIA6.1402.51.66.3552.30.39
MIN5.3453.01.27.0551.40.46
NE7.9461.90.67.2533.30.37
NO7.6472.10.66.4571.30.39
NYG7.6382.91.16.4503.40.39
NYJ5.5412.92.26.1593.60.43
OAK7.0394.60.96.0543.00.66
PHI7.0494.71.66.0593.00.41
PIT7.2443.01.34.9602.00.44
SD7.0413.41.06.8583.70.36
SF5.9391.20.76.0653.90.51
SEA5.6352.71.26.1624.20.52
STL4.7371.61.86.3562.60.44
TB5.9443.92.27.5532.80.53
TEN6.4352.61.05.8582.00.47
WAS6.0434.21.56.4582.60.44
Avg6.3412.91.26.3572.90.43

  • 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

21 Responses to “Team Rankings: Week Seventeen”

  1. Anonymous says:

    Just goes to show you how statistics can hide the truth.....

  2. Anonymous says:

    How does Det end up with dRank = 2?

  3. Keith Goldner says:

    Det's pass defense has been extremely efficient this year and important in victories. They lead the league with a -1.71 WPA.

  4. Anonymous says:

    Brian,

    Had a question which I don't know has been covered yet or not..

    What is more variable; offensive efficiency or defensive efficiency? That is, should we expect offenses to have more games where they stray from their true efficiency or defenses?

    I guess there's also two ways to look at this; does offense or defense stray more often and does offense or defense stray more dramatically. One doesn't necessarily predispose the other.

  5. Ken R. says:

    Brian,

    One thing that I question is the value of early season numbers vs current numbers. For example, 8 weeks ago the Texans with Matt Schaub as quarterback were peaking statistically and were, of course, number 1 in the rankings. Since Schaub has been injured, the Texans numbers (particularly passing and turnovers) have been off. It seems clear that the Texans aren't the team that they were with Schaub at QB, but it also seems that their early season numbers (generated with a pro-bowl QB) are creating a bias for the current team ranking given that the current QB is a 3rd string rookie that has struggled in the passing game. Do the rankings give equal ranking to current and early season performance? Are the Texans really the current number 2?

  6. Brian Burke says:

    Ken--equal strength. Not to say that that's the true ranking, it's just that adjustments beyond that are up to the reader.

    Anon-Great question. First cut, I would guess that defensive performance is more variable. The reason is because offenses tend to set the numbers, and defenses (except the very best and worst) are more or less at their mercy. So defensive performance will be more of a reflection of the week-to-week variability of the offenses they face.

  7. Anonymous says:

    Is there any way to favor new data over old data without throwing out all the old stuff? I can't help but think that some teams (most notably the Texans) are being propped up by their past success, whereas others may be dragged down by their past failures. When I look at a chart that says the Texans are the second best team in the NFL if you include all of Matt Schaub's early season performance, I am also thinking that perhaps a team like Miami that started poorly but has improved in the second half of the season is being underrated.

    It just makes me question the intended use of this chart. It seems like it's meant more as a way to measure a team's whole season, not how good they are right now.

  8. Anonymous says:

    Anon - from what I understand dramatic mid-season improvements or team shattering injuries happen so infrequently that by excluding early season games or by giving more weight to more recent games you're going to make more inaccuracies than accuracies.

    Brian would know better though. Could also be a good project for the community site if someone's up for the task.

  9. Anonymous says:

    Gotta really rethink your methodology.
    Put it this way - you get your top three (Pitt, Philly, Houston). I'll take your next three (NO, GB and NE). Whose group has the next Super Bowl winner?
    And as far as offense vs. defense variability - wouldn't it be the same, as it's a zero-sum game?

  10. Brian Burke says:

    No, don't have to rethink the methodology at all, thanks. You just need to understand the purpose of the model.

    Regarding zero-sum and offense vs. defense variability: Offenses vary more widely than defenses. Yes, zero-sum considerations dictate their means are equal, but that doesn't mean their variances must be equal at all. Year in, year out, offenses vary more widely.

  11. Matthew says:

    Brian,
    As someone who frequently reads posts and follows updates on this site, I have learned to adjust to the fact that these rankings are neither intended to be seen as an orderd listing of the current "best" teams in the NFL, nor as a prognostication for future success. They are simply an accumulation of relevant statistics that can be used as one tool amongst many to draw conclusions from. That being said, it is hard to fault new readers for making the assumption that the implicit purpose of a ranking system is to determine the current best team in the league and provide some predictive power for future outcomes. To respond dismissively, as I have often seen you do, when people question your methodology in earnest (even if they manage to be blunt and/or rude in the process of doing so) strikes me as counter-productive to this website's efforts. If your goal is to inform and enlighten your readers as to the value of looking at NFL statistics in a relevant, rigorous, and objective manner, it may behoove you to be the bigger person in comment threads and focus on instruction and constructive criticism. Some people have personal agendas and are not open to listening, but that is not true of everyone. I would neither underestimate nor dismiss the effects your comments have on potential readers.

  12. Anonymous says:

    Yeah, I got a bit confused by phrases like "a convincing run on top," "best team on the board," and "top ranking" and the rather persistent allusions that higher ranked teams were better than lower ranked teams.
    I dug around on your site a bit but couldn't find an explanation of what the rankings mean. Could you provide a link?
    Also, I don't disagree with you that offenses are more variable than defenses. But every play deviates from the average by a certain amount, positively for the offense and negatively for the defense (or vice versa). Wouldn't that mean the variance of all offensive plays equals the variance of all defensive plays?

  13. Anonymous says:

    Matthew,

    What the fuck are you talking about? How are these rankings not a "prognostication for future success"?? That's *exactly* what they are... Why does this site attract so many idiots that don't even bother to read any of the relevant information. So many people here look at the rankings and decide that they don't match with their intuition so they write moronic comments. I'm not sure how Brian deals with it.

  14. Matthew says:

    Anonymous,
    Perhaps I have misunderstood, but I made the assumption that a ranking system that does not take into account current injuries, is not weighted more significantly towards the latter part of the season, and disregards special teams is not going to be a perfect prognostic tool. Brian has mentioned in posts, to cite but one example, that If it took injuries into account then Houston would not be ranked as high. I am not saying that it is not useful for making predictions, but that it is a system that does not understand all aspects of the game and should therefore be used as a tool amongst many from which to derive conclusions.
    Also, you have disregarded the point of my post. Whether or not I misconstrued the ranking system (it is entirely possible that I did), my sentiment stands that I, nor anyone else, should be attacked for commenting in earnest. I have spent days reading the fine print on this website and my thoughts, whether they be slightly amiss as to the intentions of the site, are not moronic and should not be treated as such. It is easy to be dismissive of others, and sometimes they deserve it, but more often than not it is toxic to comment threads and on a larger scale destroys what sense of common humanity we may share.

  15. Matthew says:

    Anonymous,
    Perhaps I have misunderstood, but I made the assumption that a ranking system that does not take into account current injuries, is not weighted more significantly towards the latter part of the season, and disregards special teams is not going to be a perfect prognostic tool. Brian has mentioned in posts, to cite but one example, that If it took injuries into account then Houston would not be ranked as high. I am not saying that it is not useful for making predictions, but that it is a system that does not understand all aspects of the game and should therefore be used as a tool amongst many from which to derive conclusions.
    Also, you have disregarded the point of my post. Whether or not I misconstrued the ranking system (it is entirely possible that I did), my sentiment stands that I, nor anyone else, should be attacked for commenting in earnest. I have spent days reading the fine print on this website and my thoughts, whether they be slightly amiss as to the intentions of the site, are not moronic and should not be treated as such. It is easy to be dismissive of others, and sometimes they deserve it, but more often than not it is toxic to comment threads and on a larger scale destroys what sense of common humanity we may share.

  16. bigmouth says:

    The Niners have beaten 2 of top-3 teams, and 4 of the top-10 teams, but are still mired in 13th place. At this point, I think you have to wonder if the failure to include special teams in these rankings is seriously skewing the results.

  17. Anonymous says:

    bigmouth: As a niners fan myself you have to admit a good number of those wins were due to turnover luck which is obviously not repeatable in any significant way. Once the fortunes regress to the mean we're likely to see an Atlanta-style beatdown.

  18. Anonymous says:

    Referring to Atlanta in the playoffs last year in above post btw.

  19. Anonymous says:

    Matthew:
    Am I the anonymous you're talking about? I wrote the first follow-up to your post, not the second. [The other guys are imposters(!)] I have no beef against you. If I read correctly, you stood up for me. I'm genuinely sorry if anything I wrote came off in a hurtful way.
    I am a bit puzzled by a web site that purports to create statistics that value players and teams more accurately than, say, yards per carry, then pivots to say that its ranking of teams is not a measure of the relative quality of those teams.
    Even if one ignores current injuries and takes these rankings as the cumulative accomplishments of teams to date, I'm hard put to say that Pittsburgh, Houston and Philly have had better seasons than Green Bay, New Orleans and New England.
    I could be wrong, of course. But any system that puts, say, Philadelphia, ahead of, say, Green Bay, really should explain what the system sees that others do not - or at least make clear what it is trying to measure. I have a lot of experience explaining mathematical models to executives, and the most important skill one must have is the ability to explain the counterintuitive.
    Next, you have to show how the model is actionable. What can you do with it? How is the information valuable? Here, it sounds like these rankings are neither actionable nor valuable.
    But my questions and rather barbed comments are about the site itself and this post in particular. I apologize if they came off as as slam against you.

  20. Matthew says:

    Anonymous (the one who didn't imply that I was a moronic idiot):
    No worries, was definitely referring to the other anonymous and not you. Agree strongly with your point that great pains must be taken to explain the counter-intuitive. Not all of us are long time readers who implicitly understand every algorithm and statistic on this site and it would be nice to consistently provide explanations for that which doesn't make sense at face value.

  21. Anonymous says:

    "I'm hard put to say that Pittsburgh, Houston and Philly have had better seasons than Green Bay, New Orleans and New England"

    Yeah too bad thats not what the rankings are trying to say...

    All they are trying to say is that if Pitt/Hou/Phil play at the efficiency they have throughout the season they are better than GB/NO/NE. Its a forward looking ranking using efficiency statistics that are repeatable. If you want a ranking that just matches your intuition of what you think the teams should be ranked why not just go to the ESPN rankings?

Leave a Reply

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