[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [pbmserv-dev] ratings
Ahh, as I understand it the current behavior was meant
to prevent people "locking in" high provisional
ratings by achieving one win and then playing a series
of weak opponents. A quick fix would be to say that
whenever the current formula would have caused you to
lose rating points after a win, instead your rating
would remain fixed and the game would not "count"
towards achieving established status (i.e., it would
be for you as if the game had not been played).
Cheers,
Lyman
--- Richard Rognlie <rrognlie@gamerz.net> wrote:
> The biggest issue is...
>
> while you are provisional, you can *lose* point when
> you beat a
> weak opponent. Many users a dismayed by this
> behaviour.
>
> The FIBS sysem apparently does not suffer from that
> behaviour.
> (If I understand it correctly).
>
> On Sun, Sep 22, 2002 at 01:40:17PM -0700, Lyman Hurd
> wrote:
> > I have discovered that when you get them old
> enough
> > you can get the kids to PLAY abstract games with
> you
> > (voila two endeavors for the price of one :-)).
> >
> > I had to ask, what specific deficiencies of the
> old
> > ratings system is the new system supposed to
> address?
> >
> > I know that there has been an open thread on team
> > games and multiplayer games and how to rate them
> > fairly. Is there concern that there are specific
> > instabilities in the current system?
> >
> > Of course all ordering systems assume transitivity
> > which does not always hold. Also ones performance
> > against near and far rankings may be markedly
> > different (when I first learned go I learned from
> a 6
> > dan. For quite a while I played much stronger
> against
> > really strong players (with appropriate handicap)
> than
> > against ones nearer my level (not knowing how to
> > capitalize on mistakes my teacher never made :-)).
> >
> > I have found the page that describes the new
> formulas
> > and was just curious (I have a longstanding
> interest
> > in dynamical systems that makes this sort of
> > investigation irresistible!).
> >
> > Cheers,
> >
> > Lyman
> >
> > --- Richard Rognlie <rrognlie@gamerz.net> wrote:
> > > On Sun, Sep 22, 2002 at 01:05:19PM -0700, Scott
> > > Huddleston wrote:
> > > > Richard,
> > > >
> > > > Thanks for the update. I'm just curious, not
> > > trying to dun.
> > > >
> > > > Can you give some general ideas about what the
> new
> > > rating system
> > > > might eventually be? Would it mainly entail
> > > changes to the rating
> > > > system, changes in code architecture, or both?
> > >
> > > The new system is FIBS (first internet
> backgammon
> > > server) based.
> > > it's "real" (vs. integer). And the issue is
> > > serious changes to
> > > the code architecture. Essentially, I need to
> move
> > > all the rating
> > > from it's own class to a "part" of the Game
> class,
> > > so it can be
> > > inherited by each game engine.
> > >
> > >
> > > > The "raise the kids" game has its share of fun
> too
> > > :-)
> > >
> > > I've got them today (while my wife is at a baby
> > > shower... we're having
> > > fun laughing at the dog... (he's being silly
> for
> > > some reason...)
> > >
> > > http://www.gamerz.net/rrognlie/images/Hairy/
> > >
> > > He's a big hairy dog.. hence the name. a little
> > > over 4.. acting like
> > > a pupppy today.
> > >
> > > > Thanks,
> > > > - Scott
> > > >
> > > > >The answer is... I've never figured out a
> good
> > > way to convert the
> > > > >existing pbmserv rating code to a per game
> > > mechanism that is inheritable
> > > > >by class.
> > > > >
> > > > >It needs work on my part... yep yep yep. I
> > > dropped the ball...
> > > > >
> > > > >*years* ago.
> > > > >
> > > > >I'd been hoping to work on it during my
> summary
> > > vacation this year
> > > > >but it seems my wife thinks I should help
> with
> > > the kids... go figure.
> > > > >
> > > > >8^)
> > > > >
> > > > >On Sun, Sep 22, 2002 at 12:32:25PM -0700,
> Scott
> > > Huddleston wrote:
> > > > >>
> > > > >> On Tue, 17 Sep 2002, John Williams wrote:
> > > > >> >I finally found out what happened to the
> > > fabled "new rating system" which
> > > > >> >was going to be integrated in at one
> point,
> > > but I have not looked at it to
> > > > >> >determine whether it supports joint wins,
> or
> > > how much work is left to make
> > > > >> >it work at all.
> > > > >>
> > > > >> John,
> > > > >>
> > > > >> What IS the story with the fabled "new
> rating
> > > system" that never appeared?
> > > > >>
> > > > >> Thanks,
> > > > >> - Scott
> > > >
> > > >
> > > > To unsubscribe, send a message to
> > > esquire@gamerz.net with
> > > > unsubscribe pbmserv-dev@gamerz.net
> > > > as the BODY of the message. The SUBJECT is
> > > ignored.
> > > >
> > >
> > > --
> > > / \__ | Richard Rognlie / Sendmail Ninja /
> > > Gamerz.NET Lackey
> > > \__/ \ | http://www.gamerz.net/rrognlie/
> > > <rrognlie@gamerz.net>
> > > / \__/ | No trees were killed in the sending
> of
> > > this message.
> > > \__/ | However, a great many instructions
> were
> > > executed.
> > >
> > >
> > > To unsubscribe, send a message to
> esquire@gamerz.net
> > > with
> > > unsubscribe pbmserv-dev@gamerz.net
> > > as the BODY of the message. The SUBJECT is
> ignored.
> > >
> >
> >
> > __________________________________________________
> > Do you Yahoo!?
> > New DSL Internet Access from SBC & Yahoo!
> > http://sbc.yahoo.com
> >
> >
> > To unsubscribe, send a message to
> esquire@gamerz.net with
> > unsubscribe pbmserv-dev@gamerz.net
> > as the BODY of the message. The SUBJECT is
> ignored.
> >
>
> --
> / \__ | Richard Rognlie / Sendmail Ninja /
> Gamerz.NET Lackey
> \__/ \ | http://www.gamerz.net/rrognlie/
> <rrognlie@gamerz.net>
> / \__/ | No trees were killed in the sending of
> this message.
> \__/ | However, a great many instructions were
> executed.
>
>
> To unsubscribe, send a message to esquire@gamerz.net
> with
>
=== message truncated ===
__________________________________________________
Do you Yahoo!?
New DSL Internet Access from SBC & Yahoo!
http://sbc.yahoo.com