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

1720Re: Cross Generational Simulating/Comparisons

Expand Messages
  • Dean Oliver <deano@rawbw.com>
    Jan 28, 2003
    • 0 Attachment
      One last try.

      If a team of 6'7" guys wins in a league where the average is 6'5",
      would that team do as well in a league where the average is 6'7"? My
      sense is No. Coaches/scouts say No. The stats say No. Bob says
      Yes, it's already accounted for in their basic stats.

      Other questions for me to deal with.

      --- In APBR_analysis@yahoogroups.com, bchaikin@a... wrote:
      > The point was that you seem to claim that height differences don't
      > Compare one season to the next and your simulator does about right
      > though average heights change.
      > no - i am not claiming anything. it is you claiming/pointing out
      that the sim
      > does not take into account height differences from year to year,
      and that it
      > should. my point is the sim does just fine reproducing real life
      > within a single season where there are players ranging in height
      > from 5'3" to 7'7" - based on their and their teams real life
      > statistics in a season are the results of what happened when
      players of
      > different heights played each other, no? players play multiple
      seasons, span
      > over not just years but decades...
      > if one was to truly believe your arguments for why you could not
      play one
      > season versus another, whether they be close to each other in time
      or from
      > different eras, simply because the teams don't in real life play
      each other,
      > then there would never have been a world series in baseball between
      AL and NL
      > teams when the leagues only played games within their leagues.
      since the AL
      > only played their AL counterparts, and the NL teams only played
      their NL
      > counterparts, how could anyone have possibly believed or trusted
      the result
      > of a match between the AL and NL champions in a world series?? i
      mean - after
      > all - they never played against one another...this is your logic...
      > same for ABA teams versus NBA teams. they never played one another
      during the
      > seasons - how could anyone trust the results between an ABA team
      and an NBA
      > team, when they have never played each other before? based on your
      logic they
      > couldn't.....
      > but i seem to remember alot of world series and ABA/NBA
      > Coaches and scouts certainly use height as a tool for evaluation.
      > helps shoot over people, pass over people, and catch balls. There
      is a
      > logical basis for height meaning something.
      > height is merely one of a number of facets to a player's game that
      > look at. are you saying height is the only tool coaches/scouts use
      > evaluation? the major, dominant tool? for every coach you can quote
      me that
      > says height is a dominant factor, i can quote you one that says
      speed is more
      > important - for shooting over/thru people, for passing around
      people, for
      > catching balls. there is a logical basis for speed meaning
      something also...
      > And there is a statistical basis, too. An inch over the league
      > height improves both an offense and a defense. Statistically, it's
      true, at
      > least. Not every tall team wins and not every short team loses.
      But we're
      > talking statistics, right?
      > fine, take every team, multiply the players heights (on each of the
      teams) by
      > their minutes played to get the avg height on the court for each
      team, and
      > run that thru history by charting it against W-L record, and come
      up with
      > some kind of correlation factor...
      > but don't forget to also do this for speed, and dribling ability,
      and passing
      > ability, and other facets of players......otherwise, based on your
      logic, you
      > would be leaving out important facets that a sim also does, no?...
      > I can give you the numbers, but if you don't think it's true,
      what's the
      > point?
      > so what if it is true? what is your point? are you suggesting that
      teams from
      > certain eras need to get a 1 or 2 or some odd point spot in each
      game played
      > between teams of different eras to account for some statistical
      height bias?
      > or that rebounding or some other stat should be adjusted because
      one era is
      > an average of one inch lower than another? what about within the
      > season?...
      > what about a statistical bias for speed? are you going to ignore
      speed but
      > somehow compensate teams for height differences? how can you make
      > adjustment for height but not speed?...
      > i could probably do a study that would show that there is a
      statistical bias
      > between NBA blue eyed players and brown eyed players, as if that
      really is
      > applicable to the playing of the game, and if there was indeed a
      > bias, such that say brown eyed players scored more (and scoring
      > who wins or loses) that some statistical bias should be awarded to
      teams with
      > more brown eyed players...
      > when you decide to develop a simulation, if you believe you should
      award or
      > take away from a team's totals merely because of height
      differences, or speed
      > differences, or brown/blue eyed differences, be my guest..
      > if you are going to look at and/or compensate statistically for
      > height, then you probably should also for player weight, whether a
      player is
      > R or L handed, whether they are fast or slow, because i'm sure all
      of these
      > separate facets can be shown to have a statistical bias. how can
      you pick and
      > choose which to use and which not to use?....
      > but since players of different heights, different speeds, and
      different eye
      > color all produce the statistics, i'll just use the actual
      > bob chaikin
      > bchaikin@b...
    • Show all 29 messages in this topic