Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: How many years it takes before ...

Author: Ron Langeveld

Date: 09:41:05 06/19/01

Go up one level in this thread


On June 18, 2001 at 13:37:05, Joshua Lee wrote:

>I agree with what you said , seeing how tactics flow out of a superior position
>The More positional Progam will get to a point where it will see the tactics and
>keep tactics from happening .... Has anyone figured out how deep some tactics
>have been , what ply depth is needed?

Well, I don't think this can be figured out, unless someone thinks solving chess
is practically possible. The relevant question though would be a statistical
one. Since some tactics will remain beyond the horizon for even the fastest
computers, the question should be what percentage of the tactics you want to
see. Overlooking a deep combination in one game of a match of a 1000 games does
not hurt, as long as you see the large population of 'short' tactics. Let's
assume that knowledge-driven engines are currently able to spot 90% of the
tactics (short term). If they lose 10% of the games as a result, will it be
compensates by enough positional wins ? I don't have any hard data, just a
hunch. Maybe they need to see at least 99% of the tactics that is seen by faster
engines.



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.