Re: computer-go: Searches

["P.J.Leonard" <P.J.Leonard@xxxxxxxxxx>]

Joan Pons Semelis wrote:

> > - Contact fight reader: Not necessarily very deep, but reading has to be
> > stopped somewhere (enough liberties? stable string (what ever that
> > means), or something else). hard to define clear goals or to evaluate
> > end positions. (if I hane at the end of 2 or 3 stones, is the cutting
> > points how much more serious risk?)
> I tried it, but the tree exploded and couldn't (yet :-) find a way to
> control it.


 I found incrementally expanding the nodes helps a bit. 

 Adjusting the objective evaluation from optimistic to pessimistic (or
visa versa)
should also help. That is initially prune with a heuristic that helps
thin the tree
out. Only when you have proven the pruned tree should you adjust the
heuristic and start
adding more variations that are less likely to change the status.
(Haven't tried this yet).


  cheers Paul.