Author: Joe McCarro
Date: 09:03:16 06/10/02
Go up one level in this thread
As far as playability I thought the best program was power chess.(note though I'm pretty much a patzer) The best featuure it had was that during the game it was also preparing an analysis of the game that was ready to go *right after* the game ended. The game was still fresh in your mind and you could learn from your mistakes right away. How many times have you played a game of chess let program x analyze if for a couple of hours left and never went through it? Since no other program gives immediate analysis after the game, I would want a program that often plays weaker moves but punishes my tactical mistakes immediately. Otherwise I might never see them. The idea of power chess to have a set opening that it steered you into was also kinda nice and bad at the same time. the nice thing wax you ended up learning a new openg prety well the bad part is if you went outside its opening it no londger seemed to play dumbed down at all. It would be nice if a programmer developed a complete opening repetiore in this fasion using more or less main lines. Power Chess did a good job with king pawn openings but didn't address any others. I think what it did was lead you down the opening and then every so often it would make an opeinging mistake which if you spotted it woudl give you a material advantage that should win the game. Once you got that material it played its best moves in order to force you to play accurately. On June 07, 2002 at 09:15:17, Robert Henry Durrett wrote: >We all know that it's just a matter of time before chess computers will be able >to effortlessly whip the strongest GMs. Maybe this year. Maybe next. But >soon. > >So, after that, what would "the users" wish to be the direction of chess >computer development? > >Chess computer users who use their computers for analysis will want their chess >computers to continue to become stronger and stronger, without bound. They will >want their silicon analysis tool to consistently produce moves much stronger >than any human could consistently come up with, and do in "no time flat." These >users will wish to see a continuation of the trend toward exceedingly strong >chess computers. > >But what about those users who wish to play chess against their computers? > >Such users have a very legitimate desirement: To optimize their "silicon chess >buddy" so that they will get the most educational and fun practice available. > >It is a well known fact that you learn most by playing opponents who are >stronger but not so strong as to be discouraging. It is often said that playing >against humans who are 100 rating points above you will give you the best >benefit. > >The obvious way for chess computer designers to provide that capability is to >design "precision handicapping" into their computers. > >Whether the user is an absolute beginner or the likes of Kramnik or Kasparov, >the desirement will be the same: to be able to set the chess computer so that >it will play exactly 100 rating points above the user's skill level. > >It should be easy to chess computer's skill level to within 10 or 20 rating >points of the desired performance level and to be confident that the chess >computer's performance will be at that level. > >Since most users will wish to develop their skill at winning against human >opponents, it is important that the chess computer perform like a human. > >Once such a tool becomes available, users will cheer loudly. > >Bob D.
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.