bug-gnubg
[Top][All Lists]
Advanced

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

Re: [Bug-gnubg] Double Decisions


From: Christian Anthon
Subject: Re: [Bug-gnubg] Double Decisions
Date: Mon, 20 Aug 2007 22:32:44 +0200

Hi Peter,

GNU backgammon is a world class opponent and assumes, in a sense at
least, that your and its strengths are equal. GNU backgammon's doubles
are based on this assumption. So while it in a given position would
win e.g. 60% of the games when playing against itself, it is likely to
win e.g. 80% (or even 99%) of the games when playing against you.
Especially in a complicated position like the one you report.

The bottom line is keep studying, for example by enabling the build in
tutor function.

Christian.

On 8/19/07, Peter Carlson <address@hidden> wrote:
> Why does gnu *almost* always recommend to take a double.  I have noticed
> after playing literally hundreds of games against gnu that 99% of the
> time it recommends to take a double it is just plain WRONG!  besides why
> would gnu offer a double if the odds weren't greatly in its favor to
> win???  here is just one example:
>
> Match MBlgARAAAAAA
> Position when 2x offered: bMdCgClwt4MBGA
> Position when I decided to resign: 7x8AAGB2NwiAAQ
>
> Peter
>




reply via email to

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