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

1748Re: Cross Generational Simulating/Comparisons (Enough)

Expand Messages
  • Dean Oliver <deano@rawbw.com>
    Feb 3, 2003
    • 0 Attachment
      All right! For whatever reason there is a disconnect here on this
      topic. Let's hold off a while on this train of thought and converse
      on some other things. It's getting to some rather uninformative long-
      winded counterproductive stuff.

      So let's not talk about this thing for a month or so. Hopefully some
      of these other tracks will address other issues that may lead to
      productive discussion...LATER!!!

      DeanO
      Moderator

      --- In APBR_analysis@yahoogroups.com, "schtevie2003 <schtevie@h...>"
      <schtevie@h...> wrote:
      > Bob, I read your posting, shaking my head in disbelief. What
      > could possibly be your motivation for not responding to any of the
      > substantive points in the previous posting? They were not hard
      > to identify in the text; I pointed them out as they arose, inviting
      a
      > response, noting that for an alternative interpretation of the data
      > to prevail that these were the relevant points at issue. But a
      > complete lack of response is what you offered. In fact, all you
      > posted was a numerical illustration of the problem that I initially
      > presented (way back when)! And that problem is that one cannot
      > identify progress or regress based on movements in average
      > offensive (equivalently defensive) productivity alone.
      > Congratulations on your having mastered the first sentence of
      > paragraph one, now proceed onwards!
      >
      > I am left with the belief that the only plausible answer for your
      > lack of responsiveness (if not the rudeness) is that you are
      > entirely cynical. (After all, your stated profession demands in
      > training that you have the intellect to understand the type of
      > arguments presented - based as they are on simplest math and
      > logic. And you have shown yourself to be very dogged pursuing
      > evidence - all that S.J. Gould stuff - when you think you have a
      > case.) I can only think that you have seen the error of your ways
      > and are hoping no one else has (?) and are blustering to the
      > end, hoping that I will tire of repeating the argument - and
      > imploring you to respond - thereby taking some odd comfort in
      > being the last one to opine. Such an approach has its
      > consequences however and ones I would be very concerned
      > about them if I were you. In particular, your credibility erodes.
      At
      > best (?) one is left with the misimpression that you lack a certain
      > acuity. At worst is the inference that you lack trustworthiness,
      > being willing to obfuscate for the purposes of protecting a vested
      > interest.
      >
      > So here we go again - this time with much greater brevity.
      >
      > To be interpreted: the rise (to the mid-80s) and fall (thereafter)
      of
      > offensive productivity (well actually just the rise, but the fall -
      I
      > presume - is at least finally admitted to as being fact?)
      >
      > Interpretation offered: falling game pace identifies observed
      > increase is offensive productivity as a net (that is lower bound)
      > improvement in offensive "technology" (that is distinct from player
      > height, strength, or athleticism) as opposed to a decrease in
      > defensive "technology".
      >
      > Basis of inference: Possessions by definition consist exclusively
      > of fast-breaks and non-fast-breaks (that is half-court set
      > offenses). By definition fast-breaks take less time and yield
      > higher points per possession. Thus, all else equal (READ: IF
      > THERE HAS BEEN NO CHANGE IN OFFENSIVE
      > "TECHNOLOGY") if the game slows, this means that by
      > definition there are either fewer fast-breaks or longer half-court
      > possessions. In either of these two instances, the implication is
      > that average offensive productivity falls. THUS, THERE MUST
      > HAVE BEEN A CHANGE IN OFFENSIVE TECHNOLOGY.
      >
      > Bob, there is the argument. Deal with it (and all the better if no
      > gratuitous insults are offered) or just sit down.
      >
      > Oh yes, one final thing. If behind your inability to deal with the
      > arguments is your inquietude with the jarring statistical
      > conclusion (12 to 14 points, blah, blah, blah) perhaps it is
      > comforting to know that I share it to a certain degree, but absent
      > ameliorating explanations (or some error of my ways) I feel
      > obliged to accept the calculation as is.
      >
      > That said, it would be nice if there were some other evidence of a
      > time machine sort that could to some degree corroborate the
      > logical inference above. And perhaps (only perhaps) there is.
      > Consider two cases:
      >
      > 1. Teams that went retro to their detriment (thanks to Mike G. for
      > the idea). As a possible (and I only say possible) case in point,
      > the '90 Nuggets became the '91 Nutties and increased their
      > possessions per game by 8.4 and dropped their points per
      > possession by .033. Now, if one looks at these changes and
      > links them to movements in the league averages these are
      > equivalent to going back in time from 1990 to 1978 in the former
      > case and to between '78 and '79 in the latter. Now, whether any
      > ultimately persuasive evidence can ever be gleaned from looking
      > at the basketcases on the NBA over time is not clear to me, but
      > it's a thought in progress.
      >
      > 2. More persuasive to me now is a reasoned interpretation of
      > the performance of NBA All-Stars in Olympic/World Competition.
      > Witness how over a mere decade, the rest of the world went from
      > being whipping boys to being serious competitors. (Sorry, no
      > hard data on average victory margins of European and South
      > American elite vs. NBA, but it must show a movement of at least
      > 10 points per game on average.) So what would the analogy be?
      >
      > In the case of the NBA history in question, the presumption is
      > that greatness is greatness no matter when, and it is
      > unreasonable to assume improvement, especially dramatic
      > improvement, over relatively short periods of time. As in "are you
      > trying to tell me that the average NBA team in the early 90s was
      > as good or better than the 67 Warriors, clearly one of the greatest
      > teams ever?"
      >
      > Against this we have witnessed just such dramatic improvement
      > against relatively constant NBA talent in a not-too-distantly
      > related theater - though arguably it can in large measure be
      > attributed to improvements in foreign talent as well as in
      > organization (though wasn't the presumption that talent
      > increases weren't that plausible either?)
      >
      > The tentative upshot: massive changes can happen in relatively
      > short periods of time, to the surprise and sadness of many.
      >
      > ***************
      >
      > --- In APBR_analysis@yahoogroups.com, bchaikin@a... wrote:
      > >
      > > Given that I started off recalling data over the 60s and 70s, let
      > me take '60
      > > and '79 as endpoints to begin the discussion. Note a drop in
      > game pace from
      > > 131.2 to 109.7
      > > possessions per game, a decrease of 15.6%. And during
      > which time offensive
      > > productivity increased 11% (.879 to 1.005). Put another way,
      > the average team
      > > in 1979, playing at its slower pace would be expected to beat
      > the average
      > > team of 1960 by 13.82 points per game (.126 P/CP * 109.7
      > CP/game).
      > >
      > > ...so this is your magical brain storm as to why an average
      > 1978-79 team is
      > > almost 14 points better than an average 1959-60 team???
      > you're blatantly
      > > stating, based on this one lonely calculation, that:
      > >
      > > "....the superiority of the average team in the early 80 (in this
      > case
      > > 1978-79) to its 60s counterpart was on the order of 10 to 12
      > points per game
      > > (at a modern game pace). This is equivalent to how much
      > better the best Bulls
      > > team was to the league
      > > average....."
      > >
      > > I say again: yes, yes, and yes, certainly to a first
      approximation
      > and absent
      > > any introduced counter-argument.
      > >
      > > your 1959-60 pts/poss is .879. your 1978-79 pts/poss is 1.005.
      > you are saying
      > > that:
      > >
      > > 1 - (.879)/(1.005) = 0.126 difference in pts/poss between 59-60
      > and 78-79...
      > >
      > > you are then multiplying this pts/poss difference by the 78-79
      > game pace:
      > >
      > > 0.126 x 109.7 = 13.82 pts/g
      > >
      > > and you are then stating that because of this percent difference
      > that since
      > > these pts/poss numbers are the average for their respective
      > season, that an
      > > average team in 78-79 would beat an average team in 59-60
      > by almost 14
      > > points....
      > >
      > > but like professor morbius in the forbidden planet, you are
      > clueless...
      > >
      > > this calculation of yours is absolutely meaningless, because
      > you fail to
      > > realize that pts/poss, while a measure of "offensive
      > productivity", is at the
      > > same time a measure of what the defense allows per
      > possession. you are
      > > measuring simply one side of the ball, and looking only at
      > offense, and your
      > > calculation assumes that the defensive pts/poss remained the
      > same over that
      > > time...
      > >
      > > pts/poss scored is a relative number, what the offense scores
      > against the
      > > defense on average, or what the defense allows the offense to
      > score on
      > > average. it is a relative number stating the "balance" between
      > what the
      > > offense produces and what the defense gives up...
      > >
      > > in 59-60 defenses allowed only 0.879 pts/poss, as compared
      > to in 78-79
      > > defenses allowing 1.005 pts/poss. no matter how you want to
      > look at it, a
      > > defense giving up just 0.879 pts/poss is "better" than a
      > defense that gives
      > > up 1.005 pts/poss - in any game where two teams each have
      > the same number of
      > > possessions, the team giving up the least amount of pts/poss
      > will win. thus
      > > the avg team in 59-60, based on your meaningless logic,
      > states that 59-60 had
      > > the better avg defensive teams...
      > >
      > > stating that an offensive productivity of 1.005 in 78-79 is
      better
      > than 0.879
      > > in 59-60 is exactly the same as saying a defense allowing
      > 1.005 pts/poss in
      > > 78-79 is worse than a defense allowing just 0.879 pts/poss in
      > 59-60...
      > >
      > > this is exactly the same arguement you are using to say teams
      > of 78-79 are
      > > better than teams of 59-60, simply because the avg team
      > pts/poss is greater,
      > > 1.005 to 0.879. it has no basis in fact. using your formula one
      > can say the
      > > teams of 59-60 are 13.82 better on defense than teams of
      > 78-79, and thus if
      > > they played the avg 59-60 team would defeat the avg 78-79
      > team by that 13.82
      > > pts/g, actually even more because the game pace was greater
      > in 59-60...
      > >
      > > 0.126 x 131.2 = 16.5 pts/g
      > >
      > > thus an avg 59-60 team, because of its "better" defense, would
      > outscore an
      > > avg 78-79 team by 16.5 pts/g, normalized for the 59-60 game
      > pace. this
      > > manipulation of the numbers is worthless...
      > >
      > > a completely meaningless calculation that has no basis in
      > fact....
      > >
      > > bob chaikin
      > > bchaikin@b...
    • Show all 29 messages in this topic