Re: Picking the Difference-Makers for the All-NBA Teams
- Equation (1) has nothing to do with league/team averages. Variables
are included in that equation for each of the players and that is
where the adjustments for specific opponents are made. I added some
language yesterday that further explains what is going on in that
--- In APBR_analysis@yahoogroups.com, tajallie@h... wrote:
> I saw the regression part which basically adjusts the value of each
> recorded activity against league/team averages. . .but I didn't see
> anything that adjusts for specific opp. Is that correct?
> --- In APBR_analysis@yahoogroups.com, "dan_t_rosenbaum"
> <rosenbaum@u...> wrote:
> > There is nothing in the All-NBA piece that describes precisely
> > am doing.
> > The piece that I link to in article discusses that in gory
> > do this accounting through a fairly complex regression approach.
> > http://www.uncg.edu/eco/rosenbaum/NBA/winval2.htm
> > --- In APBR_analysis@yahoogroups.com, tajallie@h... wrote:
> > > Dan,
> > >
> > > I am unclear how you adjust for player who typically compete
> > against
> > > stronger players . . .in the article you say you adjust for who
> > > player play against but I didn't see any calculation that would
> > > consider if say Bruce Bowen always played when
> > Kobe/Carter/Peja/Melo
> > > and sat down against when the banch players came in. Did I miss
> > or
> > > is their some consideration of the hisorical performce of the
> > > opposite players on the court. (ie anyone who is on the court
> > > relatively against KG would see his rating decrease bcs of it,
> > right)
> > >
> > >
> > > --- In APBR_analysis@yahoogroups.com, "dan_t_rosenbaum"
> > > <rosenbaum@u...> wrote:
> > > > Here is a piece using my WINVAL-stuff that is more fun and
> > > to
> > > > digest. In it I discuss which players I would have selected
> > > All-
> > > > NBA, along with what players just don't do well in my
> > > > plus/minus system.
> > > >
> > > > http://www.uncg.edu/bae/people/rosenbaum/NBA/stats1.htm