Author: Paul Clarke
Date: 08:51:31 05/25/05
Go up one level in this thread
On May 25, 2005 at 06:17:02, Tord Romstad wrote: >On May 25, 2005 at 05:56:22, Vasik Rajlich wrote: >> >>So maybe you could say that if you want to be an engineer, write a chess >>engine, while if you want to be an innovator, write a go engine. > >That's indeed an excellent summary. I would like to add that if you want >to combine both types of skills, shogi is somewhere in the middle (though >of course much closer to chess). Shogi does indeed look interesting. I've written a toy shogi program using alpha-beta search with the usual refinements; it plays fairly badly, particularly when there are a few pieces available to drop and the branching factor jumps to 200 or more. It could be improved quite a lot without abandoning alpha-beta: the evaluation function is cobbled together from my limited shogi knowledge, and I've payed very little attention to performance. However, I'm tempted to switch to something similar to B*, mainly because it looks like an interesting thing to play with. There's an interesting description of the 1997 computer shogi champion YSS at http://www32.ocn.ne.jp/~yss/book_e.html (the level of detail is similar to Ed Schroeder's pages on Rebel).
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.