Author: blass uri
Date: 21:27:15 08/15/98
Go up one level in this thread
On August 15, 1998 at 11:12:37, Kevin Mulloy wrote: >Dear Friends at CCC, > I was wondering if I might pose a question: Does anyone believe that it >would be possible for one of the current crop of top computer programs to post a >Master rating in correspondence chess? I can lend a little insite, but I would >like the opinion of the good people that work with computers and the computer >programs every day. I am currently a USCF correspondence chess Candidate Master >(2064). I have played more than 30 games and -- to date -- I have no losses and >only 2 draws. I cannot trust your USCF correspondence chess rating. If you have 28 wins and 2 draws I think that maybe you should have a master rating in correspondence chess so the fact that you have only 2 out of 3 against programs may prove that they are close to master rating. I can trust your correspondence chess rating only if your result is close to 50% against at least 6 players with correspondence rating similiar to your's. Uri I have played against players that I believed used computers. >Since my own style of play with both the white and black pieces tends to be >closed, I have noticed that computer players tend to attack well, but against >the closed positions I play they have trouble making any advances in the middle >game. I am able to come out with a better game positionally and this has worked >very well for me. I have went as far as playing Hiarcs6, Fritz5 and MCP7 in >simulated "postal games" and I am 1win, 0 losses and 2 draws. I can not beat >these programs in a standard 40 moves in 120 min game. Most of the time I can >not even draw against these programs at these time controls. Does anyone have >an opinion on this? Is there anyone that has played several games against the >top programs at "correspondence time controls" that may have simular results? > With my limited experience, I don't think that -- against closed positions >-- the current software could achieve a master rating at postal. If anyone >would like to simulate a game against their program, the only three rules that >you need to know are: 1. Any opening library or research is legal. 2. You are >allowed 3 days per move. 3. You can not use any computer programs to generate >a move for you. In an actual postal game you are not allowed to ask for advise >from anyone else about a move. > I would be very interested to hear from anyone that plays such a match or >has noticed some of the same things that I have while playing closed positions. > Thank You, > Trapper
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.