Author: Robert Hyatt
Date: 07:29:07 03/27/02
Go up one level in this thread
On March 27, 2002 at 10:06:09, Joshua Lee wrote: >The "impotent pair" That's what i was taught to understand this as. Not >Difficult to understand, but what would be interesting is how many lines of code >are needed to make a program understand this and other "Positional" Idea's. This depends on a lot of things. 1. data structure used to represent the chess board; 2. how important is it? IE is it a critical piece of knowledge and is it difficult to compute? If so, it might be done incrementally to save time but expend more code. > >Would someone explain What can be programed and what can't? Can someone tell me >If you have mostly eval, how much of your code in relation to the processor >cycles get's computed how much does it take etc? Compare two programs like a >Little Goliath versus Shredder, What do they do per each clock cycle how can one >search over 1Million nodes while the other under 100Kn? > One is a fast search with a "light" evaluation. The other is based on a "heavier evaluation" which costs NPS. _anything_ can be done in the evaluation code. Anything a human can quantify can be coded. But it is often too slow, too special-purpose, etc... and in many cases the programmer simply doesn't understand the concept and therefore can't possibly code anything for it. IE see past discussions about "pawn levers" and "pawn majorities". >I would like to write a program eventually with the things that haven't been >emplemented or things that are supposed to be difficult.
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.