Team Rankings: Week Fifteen

Despite putting up a mere 14 points on the lowly Cleveland Browns, the Pittsburgh Steelers gained ground on the Texans in the pursuit of our top ranking heading into the playoffs. The Texans dropped a full .03 GWP this week while their opponents' GWP didn't move at all.

If Pittsburgh puts up a good show on the road against the San Francisco 49ers, they will likely take over the top spot here at Advanced NFL Stats. The Texans get to stay home and play the Panthers and their wounded defense while the third place Packers -- also within striking distance these past few weeks -- will be able to destroy the Chiefs.

New England and New Orleans are probably out of the running, as are the Giants and anyone else below them in the rankings. At this point, it's all a crapshoot because teams are going to be resting players and possibly giving opponents games. Barring some unforeseen circumstances, Green Bay is going to be undefeated heading into the final week against the Lions, where a win could mean the top overall ranking heading into the playoffs. If I had to put money on a team finishing at the top of the board, I'd throw some greenbacks on the Green and Yellow.

Notes and Fun Facts
  • For the first time in, well, ever, I feel bad for Tony Romo. The Dallas Cowboys' QB completed over two-thirds of his passes on Sunday night and threw for over 300 yards while racking up four touchdowns. His running game also netted him almost 140 yards on the evening, and his team still lost to the G-Men. The Cowboys are now going to need to sneak in to the NFC Wild Card, as the Giants now own the tiebreaker over them for the NFC East crown.
  • Despite losing to the Texans, the Cincinnati Bengals win the award for "Biggest Mover and Shaker" this week. Sure, the 49ers dropped more spots than Cincy gained, but it's no fun awarding someone for doing something poorly. Actually, it can be fun, but in a more sick and sadistic sort of way.
  • Can the Colts end the season with a GWP below 0.25? I'd like to see it happen to not only them, but the Chiefs, too. It's a train wreck. The carnage is unbelievable, but you just can't look away. And we can't mention a train wreck without at least mentioning Tim Tebow's name, right?
Now, without further ado, here is the Unrated Directors' Cut of your rankings for week fifteen.

2 PIT20.760.4954
3 GB30.730.47225
4 NE50.670.51129
5 NO40.650.48324
9 PHI100.610.511015
14 SD160.520.46928
17 SF110.500.48208
30 TB290.300.521930
31 KC310.280.503021


  • 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

17 Responses to “Team Rankings: Week Fifteen”

  1. Anonymous says:

    I think Cincinnati's jump is largely due to the fact that, unfortunately for an objective ranking system, nothing can be done to account for injuries to key players. We all agree the Houston Texans are not the same caliber team without pieces like Schaub and Johnson (and Leinart...?), and we can mentally adjust their GWP to whatever subjective value we wish.

    But the model sees Cincy as allowing only 20 points to the 7th ranked offense, when clearly the offense is truly ranked somewhere below 7th. I know i'm simplifying it and that the model doesn't look at PA and ranks, etc., but you get the point.

    How do we feel about the ripple effects that, for example, Houston's GWP (which is probably closer to an artifact than an estimate without key players) is having on their opponents?

    I fully understand there are no (effective) objective ways to go about solving this problem...I just want to gauge everyone's opinion on how much of an impact it's having.

  2. Anonymous says:

    Actually, if I understand the basis of this site's numbers, there IS an effective way to take injuries into account. Take the replacement player's stats and sub them in for the injured player.

    In fact, the author of this site did exactly this a couple of weeks ago when he adjusted the Texans' ranking by assuming that the new QB would be a 0. Non-negative, but not the positive play they'd been getting, either. It dropped their ranking to 7, I believe.

    Now take a look at TJ Yates' current rankings. An admittedly small sample size (2.5 games) but he has very nice numbers. You can argue that they are not accurate if you want (team's will adjust with more game film, etc...). A very nice performance against a quality opponent on the road this week in Cincinnati has us Texans fans feeling much more optimistic about our chances in the playoffs.

    However, your statement 'nothing can be done to account for injuries to players' is just not accurate.

  3. Josh Katz says:

    Great post. I'm congenitally incapable of feeling bad for Tony Romo, but I suppose I could understand why someone might.

    Should note though, the Cowboys still have a fairly straight path to the division title since they play the Giants again in Week 17. For the Giants to clinch before then, they'd need to win both of their next two games and Dallas would have to lose to the Eagles. ... Certainly possible, but far from a sure thing.

    All told, there's about a four in five chance that the NFC East will be decided in the NYG/DAL rematch the final week of the season.

  4. Pete says:

    Interesting that 4 of the top 6 teams are awesome offense/lousy defense teams. Maybe this year, some offenses are way out on the end of the bell curve, while there is less variation in defense performance across teams.

  5. Steve says:

    This is a question about the defensive rankings and a basic, fundamental question about the stats that is perhaps answered somewhere else...

    In the defensive rankings, the Packers are ranked 25. On the graph on the top of the page, the Packers are ranked above (about) 15 other teams, so they'd be ranked 17th in defense. I realize that the graph is using dEPA and the rankings here are WPA. Is one of these stats more predicitive and the other more explanatory of the past? What's the difference?

    Thanks for the explanation!

  6. bigmouth says:

    Steve, I believe EPA is predictive, while WPA is descriptive.

  7. Dan S says:

    I believe the rankings use DPass (net adjusted yards per attempt), DRushSR (because it's less noisy and more predictive than EPA), and DInt%.

    WPA isn't used in the rankings.

  8. Ben Stuplisberger says:

    I guess I'm just a softie, but I've felt bad for Tony Romo for a long time. He has to be the unluckiest QB currently in the league.

    Brian, I see you're getting clever with the writing there: not sure if you should do that. ;)

    All kidding aside, the two top spots are populated by very balanced teams, while the next several spots are occupied by offensive juggernauts. Any data on how balance effects performance in the playoffs? My intuition is that offensive dominant teams can beat the odds because luck favors the offense through turnovers, and the fact that there is no upper limit on points scored, but obviously a defense cannot give up fewer than zero points. In other words, a few high leverage plays seem to favor teams capable of scoring at will. Any thoughts on this?

  9. Brian Burke says:

    In terms of predictive vs explanatory, it's more of a spectrum than one or the other. From explanatory to predictive of future wins, it would go...WPA, EPA, GWP. I'm not sure where to throw SR. It's predictive for running but not so much for passing. It depends on the style of passing game (deep vs dink).

    EPA is past net point difference generation. Think of it as an improved team point differential.

  10. Anonymous says:

    @ Ben Stuplisberger

    If i understand your post correct; yes "offensive teams" have a slight advantage to prevail at the end of the playoffs. See here:

    Karl, Germany

  11. Anonymous says:

    @ Brian Burke

    I see OAK "leads" by far in PEN-Rate. How many times is that now since the Raiders moved back and forth between LA and OAK in the 80`s? That can´t be coincidence*, small sample size or "splits"!

    It seems the NFL must really hate Mr. Davis beyond death since he sued the NFL.

    (*NFL-Rosters completely overhaul after 5/6 years; and it´s impossible that the Raiders sign the dirtiest players over three decades, and the same time all the different coaches supposedly can´t teach discipline)

    I think it´s time to study this abnormality further :-)

    Karl, Germany

  12. James says:

    Karl, it's uniform color:

    "This study investigated whether or not people would call penalties more on players wearing black jerseys than those wearing a different color. Two identical plays were filmed of a potential penalty occurring with one video having the offense in black and defense in white and the other with offense in white, defense in black. A group of college kids and a group of professional referees were taken and each split into two groups (forming four groups). The first group watched in color, seeing the jerseys were black vs white, and the second group watched in black and white, not being able to tell what color the dark jersey was as it could have been black, blue, red, green, etc.

    The results showed that for the color film, black jerseys were called penalties on more than the white jerseys, even though they were identical plays. For the black and white films, when the dark jersey color was not known, there was no difference in the way the play was called."

  13. Ben Stuplisberger says:

    Thanks Karl... I remember that article, which was mainly refuting the defense wins championships meme, but I am thinking more along the lines of balanced/vs. unbalanced towards offense. The two teams at the top on great but not the best on both sides of the ball, but the next few are elite on offense, weak on defense. I wonder if being very good on both sides is as successful as elite on offense/weak on defense. The data to answer my question is most likely contained in the article you linked, but I am not very good with statistics. :)

  14. James says:

    Btw Zach, you completely misanalyzed the NFC East playoff picture. Not only do the Cowboys not have to sneak into the playoffs via the wild card, but it's very unlikely they'll be able to win the Wild Card at all.

    In fact, the Cowboys still control their own destiny to the division title, in no small part because they play the Giants the final week of the regular season. If the Cowboys win that game, then the Cowboys will hold the ultimate tiebreaker, not the Giants.

  15. Anonymous says:

    Love this stuff but I do question it's usefullness. Relying on stats for the NFL is inherently misleading. Matchups often dictate planning (run vs. pass, blitz vs cover)that can change from week to week, key injuries can change everything so that least week's numbers or the whole seasons's numbers are meaningless, there is plenty of garbage time (and resting players this time of year)that distorts stats, 4th quarter scheming even if not garbage time distorts the true picture (work clock, prevent defense, etc.)and I could go on. Of course the teams with the highest win percentages, particularly those who have played against opponents with better win percentages, will be at the top so it appears that all this other stuff is meaningful. How can Houston and Pittsburgh be above Green Bay? GB would correctly be at least a 6 point favorite against either if they played tomorrow. NYG at 6 and Dallas at 8? C'mon. Denver at 24? That said, I can read this stuff all day.

  16. coldbikemessenger says:

    I am sure there was a typo when they did not make Denver number one!

  17. Arash Sadat says:


    I love your work on this and I use your rankings in tandem with Michael Buoy's 'Betting Market Power Rankings' every week before I make a play.

    I have one suggestion--and you've probably heard this before--that would make this a more useful tool from a betting perspective and more a accurate gauge of teams' relative strength.

    Can you weight more recent results more heavily? It's interesting that Michael uses only a team's last five games and that lets him predict lines fairly accurately. In most cases, a team that is significantly overrated or underrated by these rankings (vis a vis the betting market) has either been playing much better or much worse in their last five. If these rankings could control for that, this would be an (even more) amazing tool.

Leave a Reply

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