Author: Russell Reagan
Date: 15:26:52 06/06/04
Go up one level in this thread
On June 05, 2004 at 15:09:32, Anthony Cozzie wrote: >1. A quick glance at Go and I think a endgame tablebase for go would be >impossible. I could be wrong. Well, yes. It was more of a theoretical point. For instance, imagine if we changed the three-fold repetition rule in chess, or that the FIDE used a different rule from USCF, and there were a dozen different countries that handled the repetition rule differently. It would either be more difficult, more time consuming, or the tablebases would simply not work under some rule systems. The point is, something like tablebases are based on theoretical exactness. If the three-fold repetition rule suddenly becomes the four-fold repetition rule, or becomes non-existant, the tablebases aren't accurate. That's a lot of work invested for nothing. Maybe I'm building this up to be something bigger than it really is, similar to how someone new to computer chess programming might be afraid of using a transposition table because it is theoretically unsound, but not so practically. >2. I thought the rules for Go were standard (and very simple) except for the >komi. I think the repetition rules differ from country to country. Ko, and all that (I think, I haven't looked into Go in a while). One thing I was wanting to do was write a program to play on very small Go boards. I figured I would start with a 2x2 board. That would be easy to solve and ensure that I had implemented all of the rules correctly. Come to find out, there are different outcomes depending upon what rule set you are using. There are numerous discussions of the outcome of a 2x2 game on rec.games.go. If that's the case with a 2x2 board, why not bigger boards? Again, maybe not a big deal in reality, but coupled with the fact that the computer go community is not as big as the computer chess community, it is not enough to overcome intertia, yet.
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.