Loading ...
Sorry, an error occurred while loading the content.

62Re: Top Players of 2001

Expand Messages
  • Dean Oliver
    Apr 3, 2001
    • 0 Attachment
      --- In APBR_analysis@y..., msg_53@h... wrote:

      > I hate to think I am stuck with my own rating system, without
      > anything to gain from others'. Maybe I am just in a bad mood today.

      With respect to your own rating system, how do you evaluate rating
      systems? What do you hope to gain by looking at others? You're
      not just saying, "Well, I don't agree that Mutombo is better than
      Ratliff, so it couldn't be right," right? What is "accuracy" for a
      rating system? I usually say accuracy is roughly determined by
      applying the system to teams and seeing whether it reproduces the team
      ranks by winning percentages. Most linear weights methods don't do a
      very good job. I'm not sure we can do this with your system and its
      scaling. Without some measure of accuracy (how well does it predict
      winning?), it just becomes a basic argument over whether an overall
      rating is useful. Or I suppose it could be an argument over whether
      the theory behind the formulas are better. With 4th roots in your
      method, I don't think you want to be arguing theory.

      Dean Oliver
      Journal of Basketball Studies
    • Show all 7 messages in this topic