Re: computer-go: Most simple Go rules

[Nick Wedd <nick@xxxxxxxxxxxxx>]

In message <200106251611.MAA19384@xxxxxxxxxxxxxxxxx>, Don Dailey
<drd@xxxxxxxxxxxxxxxxx> writes

>But you have  to realize that without rules,  you have a big  problem.
>Suppose that you have  decided on Tromp-Taylor,  but by this  rule one
>player wins,  but the other  player wins if  you accomodate  him.  You
>have given away the integrity of the whole tournament unless you stick
>with the rules you choose at the beginning.
>
>You cannot cover every possibility, but a very  good strategy is to be
>absolutely clear in advance  about  the BIG things  like how  you will
>score the  end  game.   Tell   people they   cannot  play until   they
>understand the rules and stick with this.   
>
>When people are involved,  you  will always have  trouble with  one or
>two.  There is no solution to this, so you might as well maintain some
>dignity and integrity and strictly enforce rules that determine winner
>and loser.

If I have understood Tromp-Taylor correctly, then all stones on the
board at the end of the game are alive.  A lot of people aren't going to
be happy with this.  I doubt that any of the programs that have entered
EGC01 will carry on playing until they have captured all their
opponent's dead stones.  Even if I ask them to, some of them won't do
it, and some may drop out.

Here is what I intend to enforce:
    After two passes, the two programs will state the score.  If they
agree, then that is the result.
    If they don't agree, or are not both able to state the score, then
their programmers will state the score.  If they agree, then that is the
result.
    If the programmers also don't agree, I will use my judgement to
assess the status of groups.  If the position I posted in a previous
email had reached this stage, I would have inferred from Black's failure
to fill the dame around its unsettled group, that it believed that this
group was dead.  I would therefore have scored this group as dead.

Nick
-- 
Nick Wedd