Re: [computer-go] SlugGo approach: GNU vs.Goliath

[David G Doshay <ddoshay@xxxxxxx>]

On 30, Dec 2004, at 12:09 PM, Jay Scott wrote:

> David G Doshay <ddoshay@xxxxxxxxxxxxxxxxx>:
> > disclaimer warning: ALL OF IT IS PURE SPECULATION!
> I have a different speculative theory, based on the nature of playout
> search. ...
> This theory predicts that search depth is key and branching factor is
> less important.
I do not think anything we have seen from SlugGo could support or
refute such a theory, mostly because we have not tried forward
branching yet.

> The playout search works because it is making available
> information about the far future--perhaps poor information, but better
> than was available before.
Yes. The only surprise from SlugGo is that this information is better
than many people's expectation.

IMHO, SlugGo proves only that the moves in GNUGo's top ten list
are slightly better in a different order than GNU Go originally
evaluated them, and that the US-THEM trick can help identify those
that should move higher up in the list.

Oh, and that having lots of CPUs lets you try things that otherwise
would take too long.	;^)

Cheers,
David


_______________________________________________
computer-go mailing list
computer-go@xxxxxxxxxxxxxxxxx
http://www.computer-go.org/mailman/listinfo/computer-go/