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

Re: games won and lost

Expand Messages
  • Dean Oliver
    ... lost by ... I have done all but the 3rd option before. I ve never put together much of a study other than one section of this article:
    Message 1 of 2 , Feb 22, 2001
    • 0 Attachment
      --- In APBR_analysis@y..., "John Grasso" <johng@s...> wrote:
      > Has anyone ever calculated a very basic simple stat - games won and
      lost by
      > player?
      >
      > This can be done in one of several ways -
      > - games won by player's team whether or not he played in the game
      >
      > - games won by player's team when he played in the game
      >
      > - games won by player's team when he played x minutes in the game
      >
      > It might lead to some interesting results.

      I have done all but the 3rd option before. I've never put together
      much of a study other than one section of this article:

      http://www.tsoft.com/~deano/articles/iwldef.html

      For the '97-98 season, Adam Keefe was the best looking at your 2nd
      bullet. You end up with very strong team correlations within one
      season. Over the course of a career, you would hopefully lose some
      of that, but it wouldn't distinguish between Stockton and Malone,
      Jordan and Pippen, for example.

      The third bullet is a little weird. What I have found is that
      minutes played don't correlate very well with winning for a lot of
      players. For instance, you take out your starters if you are winning
      by 30 in the 3rd quarter, so their minutes are low in this case.
      Further, we know that if the good players don't get much playing
      time, the team generally won't win much. The two factors tend to
      cancel each other out. Perhaps you could look at non-blowouts. I've
      never done that.

      Another thing that was once suggested to me was to look for similar
      boxscores, where teammates all played similar minutes. That should,
      by their thinking, predict how well the team did. So if Seattle
      played LA twice and all their guys played about the same number of
      minutes, they should score the same, allow the same. I've never
      tested it.

      Dean Oliver
      Journal of Basketball Studies
    Your message has been successfully submitted and would be delivered to recipients shortly.