RE: [computer-go] Learning : was Chess programs versus go programs

[Arend Bayer <arend.bayer@xxxxxx>]

On Mon, 13 Dec 2004, Mark Boon wrote:

> > -----Original Message-----
> > From: computer-go-bounces@xxxxxxxxxxxxxxxxx
> > [mailto:computer-go-bounces@xxxxxxxxxxxxxxxxx]On Behalf Of Arend Bayer
> > Sent: Monday, December 13, 2004 14:36
> > To: computer-go
> > Subject: RE: [computer-go] Learning : was Chess programs versus go
> > programs
> >
> > This depth cutoff is relative to the full-board position we are
> > analyzing, i.e. if we are at depth 4 in life-and-death reading and call
> > out to the tactical reading, it will only read 3 plys with 5 liberties
> > at the safety cutoff. I didn't get around to making the same experiment
> > with GNU Go but that's why I don't expect it to be a drastic speed
> > improvement.
> 
> I don't understand this. What use is reading 4-liberty tactics in 3 ply?
> Even 3-liberty tactics won't be possible unless you don't count pass as a
> ply. You might just as well save the time and declare it alive in those
> cases. Am I missing something?
> 
> Anyway this depth restriction seems rather arbitrary. I wouldn't want
> something to think it's dead because it can't read a 60-ply ladder to
> capture a stone on the outside.

Sorry, you misunderstood me. In the tactical reading, a string is
considered tactically stable if
a) it has 5 liberties, or
b) it has 4 liberties, and we are beyond FOURLIB_DEPTH (default 7), or
c) it has 3 liberties, and we are beyond DEPTH (default 16).

In other words, the attacker gives up beyond depth 16 unless the string
can be captured in a ladder, etc.

Arend

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