Re: computer-go: Why Ko

[Erik van der Werf <E.vanderWerf@xxxxxxxxxxxxx>]

I think that if a player plays more than N^2 moves in one game on an NxN
board, you can safely conclude that he plays badly. 

Therefore you could limit the number of moves to (possibly even a
fraction of) N^2. If a player makes more moves he should then
automatically loose the game. (I believe something like this is called
T-truncated Go)

Ofcourse you still have to check for direct ko's, if your program plays
an illegal move it might be best to change it into a pass and hand over
one prisoner to the opponent (i.e. a direct negative reinforcement of
one stone).

Erik



Joshua Berryman wrote:
> 
> I know ko is supposed to be one of the most subtle and important aspects
> of the game, but its a funny rule in that no-one can gain anything by
> breaking it:
> 
> If you take the rule as `never repeat a previous board state' then if a
> program does do this then wahey! its back to square one. :)
> 
>  I was just wandering if I should trouble to include a ko-checker in my RL
> based program: hopefully, it should prefer a quick reward to a delayed one
> and learn to follow the ko rule of its own accord.
> 
> Sanity check?
> 
> Josh