[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Not quite {was: Re: [Bug-gnubg] Double Decisions

From: Joseph Heled
Subject: Re: Not quite {was: Re: [Bug-gnubg] Double Decisions
Date: Wed, 22 Aug 2007 18:40:16 +1200

On 8/22/07, Jim Segrave <address@hidden> wrote:
> On Tue 21 Aug 2007 (13:31 -0400), Roy A. Crabtree wrote:
> > Some responses; most of what you said I concur with, but NOT most
> > of your conclusions.
> >
> > To reiterate: gnubg is a great result.  Keep on keepin' on.
> Thanks.
> I have a great deal of trouble actually reading your responses, they
> often seem to have sections missing or almost appear to be two
> different blocks of text interleaved.

As far as I know Roy (or whatever his true name is) is a known troll
and we had a brush with him in the past.I suggest we stop feeding him.

> A match equity table (MET), which gives the match winning chances for a
> player at a given match score. All of the METs provided with gnubg
> assume that both players are equally strong. An interested player
> could attempt to make a MET for unequal players - I believe Walter
> Trice has explored what such a table would be like. I am not aware of
> any such tables being available, but if one is, it could be put into a
> format such that gnubg could use it.

I played with those ideas at the time but could not find anything
which worked. Even in the most simple case I took - e.g. two nets
whose relative strength is known a-priori (via self play) - a biased
MET did not perform better, perhaps even worse than an "even" one.

I remember asking several strong players (can't recall who, sorry,
perhaps they were only strong relative to me) if they would double
weak opponents sooner or later, and answers have been inconclusive.
When I tried changing the doubling algorithm to account for the
difference in strength it always came worse. That probably means I was
not smart enough.

It seem that there should be a way to take this information into
account, but I have not seen any concrete suggestion and was unable to
conjure one myself.


reply via email to

[Prev in Thread] Current Thread [Next in Thread]