RE: [computer-go] Modern brute force search in go

[Arend Bayer <arend.bayer@xxxxxx>]

> > At 13:51 7-11-2004 -0500, Don Dailey wrote:
> > >
> > >
> > >Just my thoughts.  All of this  is even more true of Go.  There exists
> > >(in  principle) a  future GO  player than  can make  the very  best Go
> > >player (of today or yesterday) look like a baby.
> > >
> > >- Don
> 
> To be honest, I consider all of this pure speculation again. Top pro's
> estimate 'god's level' to be anywhere between 3 and 5 stones stronger than a
> top pro. I do think top pro's are biased in this respect, but they're not
> totally ignorant. And they're definitely not stupid either, so I'd say they
> are better qualified to make such estimates than anyone else. Until we have
> some hard data or good theories that say otherwise, why make such claims?
> Unless 5 stones qualifies as 'making look like a baby', which in a pro's eye
> it very well might do.

It's not pure speculation. Youcan  try to determine god's rankd by looking
at winning probabilities in even games between players of different
strength -- at god's level, even a 0.1 rank difference amounts
to a winning percentage of 100%. I think someone did this with the data
in the EGF ranks, and concluded it would be about 2-3 stones above
today's top pros.

> I think for a top pro, 30 moves deep is not unusual. But most moves are
> played based on a much shallower search.

Even I can read 30 moves occasionally (I am EGF-4d). Some top pro
claimed he would read ahead 100 moves sometimes, and I find this claim
entirely credible. 

But I doubt the relevance of this question with regards to go, it seems
very much a chess programmers point of view (no disrespect intended).
Since a program will never be able to do 30 ply lookahead on a 19x19
board using standard search techniques, you will need a lot of forward
pruning, and the strength of the go program will depend much on the
quality of this pruning (even if you assume an excellent position
evaluation).

Further, if a human does 20 ply fullboard lookahead, this is backed up
with a lot of local reading that is not included in the 20 plies. The
same will IMO hold true for any strong go program. This reduces the
relevance of this "20" even more.

Arend


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