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

Re: Let's discuss scouting (How would you do a better job?) (long)

Expand Messages
  • Kevin Pelton
    ... I think this task you describe is one of the most important aspects of using college stats. To use them, first you have to understand them, and how well
    Message 1 of 55 , Jun 1, 2004
    • 0 Attachment
      > To figure out which stats are meaningful pts and effeciency
      > probably are not as important as competition and athletic stats -
      > steals, blocks, shot attempts / team attemtps (i.e. making your
      > own shot), and rebounds.

      I think this task you describe is one of the most important aspects
      of using college stats. To use them, first you have to understand
      them, and how well they translate to the NBA.

      A fellow poster at Hoopsworld did this analysis on one draft class,
      I believe it was 1999 (min. 800 NBA minutes). Here's what he found:

      0.933 AST/48
      0.902 REB/48
      0.747 AST/TO
      0.739 BLK/48
      0.734 FT%
      0.699 2FG%
      0.653 TO/48
      0.601 FGA/48
      0.596 FG%
      0.583 TS%
      0.548 3FG%
      0.523 STL/48
      0.457 PTS/48
      0.449 PF/48
      0.364 FTA/48

      Here's what I get from my database from 2000-02:

      0.884 R48
      0.867 B48
      0.837 A48
      0.833 3PA%
      0.734 3P%
      0.733 FTA%
      0.731 S48
      0.643 FT%
      0.545 2P%
      0.502 Ps/min
      0.358 Eff

      It's pretty clear that rebounding translates the best, which
      shouldn't be a major surprise. I think we have to evaluate these
      numbers in the context of the year-to-year consistency of these
      stats for NBA regulars. Here's what I found for these in a study
      last summer:

      0.938 R48
      0.926 B48
      0.718 Eff
      0.521 2PT%

      (I grouped FT% and 3P% into a "shoot" rating and A/TO and A48 into
      a "pass" rating; since they're not strictly the same, I won't
      include them.)

      The point of including these is that while the two-point percentage
      transition may be awfully variable from the NCAA to the pros, the
      same is true for NBA players year-to-year. So I'm not sure the issue
      is really that of athleticism so much as the inherent inconsistency.
      There is probably a tie to possession usage.

      Stil, the bigger point is that betting on a guy who's an outstanding
      rebounder and shot-blocker is a safer bet than betting on guys with
      really outstanding efficiencies in college.

      JohnH in particular has theorized in his books that steals are a
      proxy for athleticism and the ability to make the transition to the
      NBA -- along the same lines as the blocks and rebounds thinking here.

      I decided for the heck of it to find the correlation between a given
      statistic and the NBA translation (using my per-minute efficiency):

      0.522 TO%
      0.294 A48
      0.156 S48
      0.045 3PA%
      -0.091 3P%
      -0.135 Ps/min
      -0.204 FTA%
      -0.235 2P%
      -0.245 R48
      -0.250 FT%
      -0.288 B48
      -0.580 Eff

      These results are a little odd. I think they mostly show that small
      players' efficiency translates better than big players' -- an
      important reason for the factor analysis I've previously mentioned.
      That a high TO% is actually a good thing is not a huge surprise --
      DeanO, amongst others, has mentioned that trying to do a lot and
      failling as a rookie is a good sign.

      Steals are actually a pretty good proxy. Rebounds and blocks,
      because of the big man issue, fare poorly.
    • Mike G
      ... The ... have to ... much all top ... mediocre ... Take it further and just state that the vast majority of college stars do not succeed at the NBA level --
      Message 55 of 55 , Jun 3, 2004
      • 0 Attachment
        --- <danthestatman@h...> wrote:

        > >From: "John Hollinger" <alleyoop2@y...>
        > >One other thing: High assist guys almost never succeed as pros.
        The
        > >reason is that to be good enough to be in the NBA, you almost
        have to
        > >be the best player on your college team. And if you're the best
        > >player, you're probably leading the team in points rather than
        > >assists.
        > >
        > I assume you mean high assist, lower scoring guys. Pretty
        much all top
        > tier PGs were high assist guys in college - but very few were
        mediocre
        > scorers in college ...

        Take it further and just state that the vast majority of college
        stars do not succeed at the NBA level -- be they assist guys,
        scorers, or whatever.

        Is the best player on a bad team more likely to succeed in the NBA
        than the sidekick on a championship team? I doubt it.

        Bobby Hurley was often likened to Bob Cousy. I imagined we might
        find out how Cousy would do in the '90s NBA. Did we ?
      Your message has been successfully submitted and would be delivered to recipients shortly.