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

Re: 2002-03 Predictions, anyone?

Expand Messages
  • harlanzo
    ... similar to ... simply ... way. so ... be one ... would you ... scored by ... who have ... 82 and - ... I think it is just simply a matter of seeing to what
    Message 1 of 31 , Oct 8, 2002
      --- In APBR_analysis@y..., bchaikin@a... wrote:
      > >
      > most players produce more stats with an increase in possessions,
      similar to
      > how someone would increase their stats with more minutes played.
      simply
      > logical as the player is getting more chances/opportunities either
      way. so
      > looking at production per possession to normalize the numbers might
      be one
      > way to solve this...
      >
      > however it also depends what you mean by "efficiency". what stats
      would you
      > use to define efficiency? as an example, if you divide points
      scored by
      > individual player possessions, you'll get high values for players
      who have
      > low touches/min but who shot alot per possession and with a high FG
      > percentage with the greatest "efficiency", like mike mitchell in 81-
      82 and -
      > egads - mel turpin in 85-86....
      >
      > what would be your criteria for efficiency?....
      >

      I think it is just simply a matter of seeing to what extent fg% and
      to/possessions change in between 2 seasons of the same player. Take
      some examples of players who shifted from go-to-guys to supplementary
      guys (or vice versa):

      Hawkins mpg ppg fg% To/min
      90-91 38.9 22.1 47.2 .068
      95-96 34.4 15.6 47.3 .058

      Barkley
      95-96 37.1 23.2 50.0 .082
      96-97 37.9 19.2 48.4 .075

      Pippen
      92-93 38.6 18.6 47.3 .079
      93-94 38.3 22.0 49.1 .084

      Larry Johnson
      95-96 40.4 20.5 47.6 .056
      96-97 34.4 12.8 51.2 .052

      Anthony Mason
      94-95 32.4 9.9 56.6 .049
      95-96 42.2 14.6 56.3 .061

      I know it is a small smaple but it does not look like, as Dean said,
      there is much fg% variation, but turnovers are always up. (i did not
      look at how many extra shots or free throws they generate in each
      situation).
    • Dean Oliver
      ... From the article on my website: http://www.rawbw.com/~deano/articles/JordanvsOlaj.html The nontechnical form of the formula to estimate D stops is
      Message 31 of 31 , Oct 14, 2002
        --- In APBR_analysis@y..., "Dean LaVergne" <deanlav@y...> wrote:
        > Please refresh my memory - how are Defensive Stops calculated?

        From the article on my website:

        http://www.rawbw.com/~deano/articles/JordanvsOlaj.html

        The nontechnical form of the formula to estimate D stops is

        Defensive Stops =
        Min*[(OppFGA-OppFGM-OppOR-TMBLK)/2+(OppTO-TMSTL)]/TMMIN
        + STL + 0.5*(DR+BLK)

        Basically, the point is to estimate how many misses a player forces,
        how many turnovers they force, then augment them with actual stats
        like blocks and turnovers. A stop is a change of possession, of
        course, and a missed shot or block only does part of that (whereas a
        forced turnover does the whole thing). A defensive rebound does the
        other part. This formula is, uh, nontechnical because it doesn't
        weight things by how difficult they are. On some teams a defensive
        rebound deserves more weight than on others (e.g., when the team has
        a hard time getting them). It doesn't make a huge difference at the
        NBA level.

        The big estimate is what is in the square brackets, estimating how
        many forced misses and forced TO's a guy has. I've compared it with
        our Project D Scoresheet stuff and it's definitely only accurate for
        some players. Speaking of that, I really need to finish that work,
        but I first gotta get the manuscript in. And work. And go traveling
        again. Grr.

        DeanO
      Your message has been successfully submitted and would be delivered to recipients shortly.