Diamonds In The Rough

Discussion in 'Nuttin' but Net' started by InstiGATOR1, Aug 13, 2013.

  1. InstiGATOR1
    Offline

    InstiGATOR1 Active Member

    Joined:
    Jan 9, 2011
    Messages:
    2,892
    Likes Received:
    16
    Trophy Points:
    38
    Ratings Received:
    +16 / 0 / -0
    This is by Hanner. Here is an excerpt:

    and here is a link to the rest:

    http://basketball.realgm.com/article/229362/Diamonds-In-The-Rough
  2. themistocles
    Offline

    themistocles Well-Known Member

    Joined:
    Apr 8, 2007
    Messages:
    12,436
    Likes Received:
    65
    Trophy Points:
    48
    Ratings Received:
    +235 / 78 / -20
    Thanks Insti, interesting read.

    A couple of points:

    1 - As a professional statistician, I can tell you that his t-tests are "absurd." His samples are so large that no difference at all would be judged "statistically significant" at what he likely used, the p < .05 level. For that type of analysis involving so much multiplicity he should have used p < .001. But it is not a meaningful bit of information anyway. The question is: "Are the differences 'Meaningful?'"

    2 - Regarding that, if the ratings he use have any validity, and I really question the last analysis where he defines a star as a player who hogs the ball all the time, nonetheless his 2nd and 3rd ratings are quite interesting and really do show a highly "meaningful" difference between even 4 stars and those above 4 stars.
  3. REM08
    Online

    REM08 Well-Known Member

    Joined:
    Jun 1, 2010
    Messages:
    4,526
    Likes Received:
    167
    Trophy Points:
    63
    Location:
    Arkansas
    Ratings Received:
    +511 / 26 / -28
    Despite bringing back feelings of panic I experienced in the watered down 'social sciences' versions of stats classes I've had to take, you make some interesting points. The article was fun to read and seemed scientific to me - but I certainly don't have the ability to do anything other than take his stats as being golden truth. Seems every message board could use a resident statistician...
  4. InstiGATOR1
    Offline

    InstiGATOR1 Active Member

    Joined:
    Jan 9, 2011
    Messages:
    2,892
    Likes Received:
    16
    Trophy Points:
    38
    Ratings Received:
    +16 / 0 / -0
    In fairness to Hanner he only mentions his t-tests in passing. He seems more to be saying the differences in offensive rating are meaningful. Well last year in the SEC the average 5 star JR/SR offensive rating of about 109.0 was about Pat Young, while the average2 star JR offensive rating of between 98.0 to 100.0 was R.J. Slawson of South Carolina. I would say that is a meaningful difference on the basketball court.

Share This Page