Game Probabilities - Week 12

Game probabilities for week 12 are available at the New York Times. This week I look at the Jets' strange Pythagorean record.

...This season the Jets are doing their best to defy James, Pythagoras and their formulas. They have scored 183 points and allowed 268 on their way to a 5-5 record. That’s a net point difference of -85, which is second worst in the league, topping only the Jaguars. According to the Pythagorean expectation, the Jets should have about a .265 winning percentage, good for a 3-7 record or perhaps 2-8...

  • 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

9 Responses to “Game Probabilities - Week 12”

  1. Anonymous says:

    Is there a way to see these probabilities if we don't have access to the NY Times?
    You only get 10 free articles a month from them, and if that's reached, we can't see the Advanced NFL stats probabilities until the next month.

  2. Anonymous says:

    As a Bill james devotee, I'm not sure his theories transfer to NFL games as the teams play 1/10 the number of games and statistical theory is thus less reliable. For example, in the case of the Jets, they have been blown out three times distorting the stats. The occasional blow-out in baseball gets subsumed in 162 games; the high high low low nature of this year's NYJ makes the stats less reliable in my view.

  3. Jim Glass says:

    Small sample size. (Especially 10 games v 162).

    the high high low low nature of this year's NYJ makes the stats less reliable in my view.

    To compensate specifically for this within the inherent small sample size of the NFL compared to baseball and other sports I worked the Pythagorean method a new way (at least I've never seen anyone else do it) -- figuring a Pythagorean result for each game and averaging the results for each team. This effectively leads to more observations for each team, and knocks down the value of outlier scores.

    The result correlated measurably more strongly with actual W-L% than did the standard Pythagorean method. I put the result in the Community. (Gee, is it near two years ago now?)

    I haven't run these numbers for this year, but don't think it would make my Jets look any better since their outliers are their wins. Sometimes you just can't get around small sample size.

  4. Chase Stuart says:

    I tried to get around the "outlier" issue with the Jets blowouts by noting where the Jets performance ranked compared to every other team against that opponent. So instead of -40 against the Bengals, the Jets just get graded as having the worst performance against Cincinnati.

    The Jets are still in the bottom five. I'm trying to figure out why FO and ANS both like the Jets more than points differential (or this method).

  5. NateTG says:

    > I'm trying to figure out why FO and ANS both like the Jets more than
    > points differential (or this method).

    The Jets have a spectacularly negative interception differential at -11. Those will affect the point differential, but that rate is not really sustainable. My gut feeling is that the ANS model tends to over-discount high variance events like turnovers a bit.

  6. LamKram says:

    Does your probability model take into account injuries and backup QBs? Your model gives Green Bay an 83% chance of winning. Seems a bit high given that Scott Tolzien will be under center (even against the lowly Vikings). The Vegas moneyline has the game at +170/-200, implying a 65% win probability for the Packers. Should I bet the farm on the Packers?

  7. Anonymous says:

    @ LamKram

    Bet, if you need to, but no guarantees (b/c of random occurrences). Since 1980 true starting QBs have a .524 Wng.-Pct., ALL backups .476.
    It´s now up to you to grade Tolzien. W/o the Ints he has the same stats as Rodgers. He now has more reps, thus it should make the quality difference even smaller...

    Karl, Germany

  8. Brian Burke says:

    Historically bad week for Vegas, but even worse for the efficiency model.

  9. bigmouth says:

    Anonymous #1: You should have unlimited access when you click on links to NYTimes from other sites like this one or twitter. If not, just open up the articles in an incognito/private tab.

Leave a Reply

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