On Thu, Sep 10, 2009 at 01:54:49PM +0200, Magnus Persson wrote:
> This is very interesting. Here is a crazy idea, maybe it the same as
> Marks but I want to take it to its extreme.
>
> Since AMAF values are so helpful, perhaps one can let go of the idea
> of sequential play following the rules of go, and basically play
> moves in parallel on all empty intersection. Compute new state (do
> captures) and repeat a fixed number of times and evaluate.
>
> Since I do not understand GPU programming at all I cannot judge
> whether this would make things easier to implement.
>
> Also a lot of weird thing will happens when you do things in
> parallel. What happens when two adjacent blocks are captured at the
> same time. Are both removed? Are there tricks to remove the one that
> was most likely to captured to begin with.
Well, after one iteration, _all_ blocks should be captured since there
will be no empty intersections, right? (Even if you don't play in
eye-like spaces, you will have this situation until some of these
actually form.)
The easiest heuristic would seem to be to capture the smallest group,
but I have hard time imagining this working well, _and_ how to choose
the smallest group efficiently (other than that, modifying my source to
do this should be fairly easy).
--
Petr "Pasky" Baudis
A lot of people have my books on their bookshelves.
That's the problem, they need to read them. -- Don Knuth
_______________________________________________
computer-go mailing list
[email protected]
http://www.computer-go.org/mailman/listinfo/computer-go/
