[Top][All Lists]

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

Re: [gnugo-devel] problem with score feature of 3.2

From: bump
Subject: Re: [gnugo-devel] problem with score feature of 3.2
Date: Thu, 20 May 2004 21:51:19 -0700

Adam Burry wrote:

> Ok. I stopped being lazy and tested it with 3.4. 10 of the games now seem to 
> work properly! But 2 of them still have trouble. I've enclosed them.
> Once again the issue is that gnugo seems to assume a non-integer komi and
> yet ends up with an integer result.
>       gnugo --score finish -l file.sgf --quiet
> B[hd];W[ce];B[eb];W[dc];B[db];W[bc];B[bb];W[cd];B[cb];W[fg];B[gf];W[eg];B[gg]
> W[fb];B[ec];W[fc];B[dd];W[ee];B[bd];W[be];B[ac];W[hb];B[ae];W[ic];B[he];W[af]
> B[bf];W[ad];B[ag];W[bc];B[cg];W[ab];B[dg];W[ef];B[ae];W[bh];B[cf];W[dh];B[ch]
> ;W[eh];B[aa];W[ba];B[ca];W[de];B[aa];W[bd])

I cannot confirm your statement that you get an integer score
with GNU Go 3.4. But there is a problem with the current
CVS that is fixed by Gunnar's patch. Generally we are more
interested in problems with the current version than with 3.4.

You can see better what is going on by adding the debug flag DEBUG_SCORE:

$ gnugo -l game.sgf --quiet --score finish -d0x010000

Turning to another game, a serious problem is contained in one of the first
set of games Adam gave:


Running --score finish turns up a blunder leading to this position:

GN[GNU Go 3.5.6 load and print]
KM[4.5]RU[Japanese]AP[GNU Go:3.5.6]AW[fa][bb][cb][db][fb][gb]

W wants to play at G1, which is a blunder, turning
the bottom into seki.


reply via email to

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