Here's a suggestion Joe.
I have a "home made" ratings system and a few weeks ago decided to
see if I could verify his accuracy. I made an "offensive only" rating
and applied it to teams rather than individuals. Then compared the
results with the teams Offensive Efficiency (Pts./Poss.).
Well, the rating followed Off.Eff. reasonably closely. Some teams
ratings differed from their Off.Eff.(probably because my ratings are
not very good) but the best teams ranked higher and the worse lower.
I know that your system can't do this (at least with the adjustments
for position), but why don't you try to take away the position
adjustments and give it a try. It should show if the weights are
seriously screwed up.
--- In APBR_analysis@yahoogroups.com, "Joe" <myviewonthis@y...> wrote:
> > Being 4 days behind isn't THAT outdated ;) You'e been tweaking
> > quickly.
> true. i thought i'd been patient enough with several earlier
> but i will try to use this as a lession to keep being patient.
> I do
> > challenge you to show when I have been "just wrong" with any of
> points I
> > make.
> i won't take it because i dont want to slip back into it again.
> posts are there but it is all water under the bridge now.
> > Finally - you mentioned my circular reasoning earlier - how I
> thought your
> > system double hurt smaller guys - THEN I didn't like the method
> how you
> > rewarded those little guys. I know it looks as though I was
> > no matter what, but I think you misinterpreted my point. Here it
> is in a
> > nutshell:
> > I feel as though people need to try to make their systems as
> objective as
> > possible. This means just use the stats - and make it make sense
> in various
> > scenerios. I never have liked when people (including myself -
> > into the same thing) try to do final subjective adjustments to
> their work at
> > the end to make the results come out "better". The results
> > will come out better when they do this - since thats why they did
> it lol.
> > However, that subjectivity (multiplying by some number plucked
> of the
> > air if a player plays a certain position) then clouds the
> So - your
> > adjustments at the end (the position multipliers part) I'll be
> quite suspect
> > of. All things being equal (same team, pace, whatever) - two
> players with
> > the exact stats shouldn't rate differently because one may be
> labeled a PG,
> > the other a SG.
> > So, my original point - I feel if one feels they need to make
> > adjustments at the end to make the ratings work - then they
> look at
> > their original weights first, because they probably aren't
> optimized. That
> > was my original problem with the system that I was critiquing -
> > are improvents IMO. I can tell you are a guy that really will
> crunch, work,
> > rework, scrutinize, etc. to get your work where you feel it need
> go - so
> > I definitely should stand back and let you do this. Maybe, when
> you are all
> > done (so I won't fall prey to that outdated thing) - I'll do my
> > with it - and let you know if I see discrepencies - and you can
> with it
> > what you feel, including ignoring it lol.
> > Good luck with your work - and have fun, don't let criticism from
> > (like me) allow you to get down or pissed. I was only trying to
> help you
> > (believe it or not) - giving you a different take by my
> > OK - that nutshell was big - and not really to the point. My bad
> > Dan
> i hear your point about disagreeing with my method of shooting
> adjustment but i still feel the adjustment is pretty good and fair.
> havent seen better. despite the friction, i appreciate the
> consideration of my work and your response to it. but i feel it
> would be best to move on. enjoy the game.
> > _________________________________________________________________
> > FREE pop-up blocking with the new MSN Toolbar get it now!
> > http://clk.atdmt.com/AVE/go/onm00200415ave/direct/01/