Author: Russell Reagan
Date: 08:49:37 05/18/02
Go up one level in this thread
On May 18, 2002 at 08:26:50, Marc van Hal wrote: >To my point of vieuw program 1 does not have enough positional knowledge neither >does program 2 >If the positional learning should work corectly program1 should in time easely >win against program2 but also against other programs. >Otherwise the statement of positional leaning is false! >Regards Marc van Hal Sorry, I think you're wrong. For example, if my dad teaches me how to play chess when I'm 10 years old, and he's not very good himself, eventually I will develop strategies (learning!) and reach a level of skill where I can beat him regularly, even if I only play him and no one else for say, 3 years. After that, by your statements above, you are asserting that I should be able to go out and beat anyone else that I play, even though I myself am really only at a novice level myself after becomming proficient at beating my dad. So your statement "If the positional learning should work corectly program1 should in time easely win against program2 but also against other programs." is false, and therefore so is your conclusion. If program2 has a playing strength of about 800 ELO, and program1 eventually "learns" and is able to beat program2 100% of the time, that IN NO WAY means that program1 should be able to beat other programs. All it means is that program1 can beat program2 now. You didn't consider the case where program2 is the worst program in the entire world. After program1 gets better than program2, it's only the second worst program in the world, and it could still lose 100% of it's games to every other computer program it ever played against. Russell
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.