Hi Erik,
From your words, I think the main reason that Steenvreter defeated Mogo and
Crazy Stone is that Steenvreter handled life and death, seki and nakade much
better than them. Still, many people will be interested if you publish
something about "I had a bit different philosophy on what the playouts were
supposed to do than most other people at that time." :)
Aja
----- Original Message -----
From: "Erik van der Werf" <erikvanderw...@gmail.com>
To: "Aja" <ajahu...@gmail.com>; <computer-go@dvandva.org>
Sent: Monday, January 03, 2011 9:16 AM
Subject: Re: [Computer-go] Fwd: News on Tromp-Cook ?
Hi Aja,
On Mon, Jan 3, 2011 at 1:28 AM, Aja <ajahu...@gmail.com> wrote:
Many people will be very interested if you publish something about
Steenvreter. Beating Mogo without learning Mogo's paper at that time?
That's
incredible. :)
Well, that's not exactly true either. I did of course read some of
their reports before the 2007 olympiad and some of it even directly
contributed. One example is UCB-tuned. Another is how they
parallellized the search. Also indirectly, discussions with various
people and ideas that came by on this list must have influenced me
somehow. The Mogo team definitely did some great work.
The playouts just happened to be something for which I already had my
own code. Perhaps it was more important that I had a bit different
philosophy on what the playouts were supposed to do than most other
people at that time. I wanted my playouts to evaluate positions well
(and not necessarily play strong moves). So that's what I optimized
them for, focusing especially on errors in human final positions,
which was easy because I already had a large collection of scored 9x9
games from my time in Maastricht. As a consequence Steenvreter got
most of the nakade and seki issues right long before others figured
those things out...
Best,
Erik
_______________________________________________
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
