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

Re: Assisted/unassisted field goals

Expand Messages
  • HoopStudies
    ... layup), ... for ... shot, ... you ... Make ... the ... My logic provides a mechanism for assigning credit. It encourages passes to players in good scoring
    Message 1 of 5 , Jan 25, 2002
    • 0 Attachment
      --- In APBR_analysis@y..., "mikel_ind" <msg_53@h...> wrote:
      > --- In APBR_analysis@y..., "HoopStudies" <deano@r...> wrote:
      > >..... Basically, if shooting the ball is a 50% proposition,
      > > I give the assistant 1/2 of that 50% credit = 25% credit. If
      > > shooting the ball is an 80% proposition (a big man making a
      layup),
      > >I
      > > give the assistant 1/2 of 80% = 40%. This rewards an assistant
      for
      > > making passes to good shooters and penalizes an assistant who
      > >passes
      > > to poor shooters ....
      >
      > I still don't get this logic. If you pass to a guy with an 80%
      shot,
      > you have 80% of an assist. If you pass to a guy with a 40% shot,
      you
      > have 40% of an assist. The efficiency factor is built-in.
      >
      > Put it another way: Make 5 passes to Shaq, you get 4 assists.
      Make
      > 5 passes to Fox, you get 2 assists. Why factor by anything, when
      the
      > shooting percentages take care of that for you?
      >

      My logic provides a mechanism for assigning credit. It encourages
      passes to players in good scoring position and encourages good
      shooters to shoot relatively more.

      Your example is actually a good one to illustrate why I do what I
      do. Why risk 5 passes and possible turnovers to Fox when you should
      be risking 5 passes to Shaq? Until the defense adjusts to double
      team Shaq (to force turnover on the pass or prevent his shot or force
      him to turn it over), passing to Shaq is a dominant strategy. But
      teams, of course, do defense Shaq more heavily than Fox to try to
      balance the risks. It's relatively easier to pass to Fox, who could
      shoot every time if he wanted, but defenses don't respect that threat
      as much. This logic accounts for this.

      Part of the logic for this mechanism came when someone asked me the
      question almost 10 years ago: "Why don't teams just shoot all 3
      pointers? The expected points per 3 pt shot is higher." Game
      theory. Teams that shoot 35% from 3pt land and 45% from 2pt land
      still shouldn't shoot 3 point shots every time. Defenses will adjust
      knowing that that is their strategy, lowering the 3pt%. The natural
      equilibrium is for 3pt% to be relatively high because good shots are
      set up by threats from 2 pt shots. Accounting for the feedback of
      defensive adjustments is what I try to build in to counting Scoring
      Possessions. This distribution of credit for assists is an example
      of how I do it.

      >
      > My observation is that players who don't pass much tend to also not
      > look for the good pass. Passing is where it's at. That's why you
      > see whole teams of low-percentage shooters, and other teams of all
      > high-percentage shooters. People look for each other -- or they
      > don't.

      Hmm, lots of observations. Not sure what the contrary would be.
      Teams that shoot high percentage overall but have only one or two
      guys with high shooting percentages? Those would be teams that don't
      pass much but don't need to? They would be the exception rather than
      the rule. Is that what we would test?

      No one doubts that passing is important. It's making sure that we
      don't double-count for it as most linear weights systems do (a score
      with an assist and a fgm is given more credit than one without the
      assist, which just doesn't add up to the team number). If passing
      makes teams better shooters, isn't that already reflected in the
      stats? Sounding a lot like MikeG's example, if assists make for
      better shooting, say 60fgm/100fga on 40 assists vs. 50fgm/100fga on
      20 assists, doesn't

      linear weight credit = pts - (fga - fgm)

      already account for the benefit of assists (credit = 80 vs 50)? Why
      do you need assists? (Of course assists ARE important, but they are
      part of how you got those points, not a bonus in themselves.)

      The teamwork involved in basketball is so fundamentally different
      than anything in baseball. That's what makes basketball analysis so
      much harder. (And makes football essentially impossible.) Passing
      is the basic part of this. My model of how passing is reflected in
      the success of an offense is pretty different and pretty
      complicated.

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