Author: Mike S.
Date: 14:18:02 07/12/00
Go up one level in this thread
On July 12, 2000 at 10:44:11, Daniel Chancey wrote: >The main concern is the # of times the comptuer lets itself into getting >hammered by a raging attack. So many times the computer ignores it, eats a pawn >and dies. (Like the trap used to kill termites) >Here are my suggestions that hopefully could make a much better computer and put >them in near Super GM status: >1.Count # of squares near own king, that the opponent is attacking into the >evaluation. >2.Create software that makes the computer aware of attacking possibilities and >stop it. >3.The computer should avoid exchanging good for bad 99% of the time. >4.The computer should have software that allows it to have an idea, go for it, >and not switch horses unless the idea is bad. >5.Never play GM's until 2001! (lol) I think that chess programs have had ever since, what you write in No. 1,2 and 3. It's more a problem of increasing the quality of those algorithms, to not only recognise 90% or 95%, but 99,5% of dangerous situations, caused by traps and sacrifices... While during the eighties, a "trojan horse" situation was almost everytime a win for the human opponent, you will see that today it's much harder to construct a situation where the computer will take the poisoned piece (and if he takes, the attack might not succeed). The trouble with programs having ideas is IMO, that too many fans or customers look at rating list and computer tournament results. If a program is let's say 100 elo weaker against it's top software competitor's, but has an attractive style with ideas and brilliant attacks... who does really care for the product? Only a minority I'm afraid. Regards, M.Scheidl
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.