Author: Robert Hyatt
Date: 12:46:32 08/16/99
Go up one level in this thread
On August 16, 1999 at 15:13:39, Robert Hyatt wrote: >On August 16, 1999 at 14:57:44, Rafael Vasquez wrote: > >>On August 16, 1999 at 13:55:29, Robert Hyatt wrote: >> >>>The only thing I have seen "guarapuava" try on ICC is the trojan horse attack. >>>Nothing else looks particularly interesting or effective. I haven't seen >>>anything else from him, and now that most every crafty has him 'tagged' he is >>>losing 100 games for every game he draws. I don't call that an 'effective >>>strategy against computers'. >>> >>>IMHO of course... >> >>Bob, Nowadays is normal that we have to play thousands of games in order to have >>some games for a database. Imagine the lots of game I had to filter. But >>Guarapuava is stuborn enough to have almost 100 won games.You have to recognize >>that a player has to be STRONG to have more than 70 won games against Crafty in >>1999. I hope more players will let know their "single method" to us. That way in >>the end we will have an arsenal. > >70 is not 'hard'... because he plays a _lot_ of different crafty clones on ICC. >However, I think he is 'done' as news spreads fast. :) > > > > >>Due to the philosofy of the website I am not posting the hundreds of thousands >>games won by Crafty. > > >that was my point of course... that once a crafty operator 'catches on', >Guarapuava is finished with that one... > > > > >> >>My Very Best Regards to you Bob,(I admire you're work and you as a person) >> >>Rafael Vasquez Sorry... thumb hit the send button before brain was finished. :) In any case, let's define "anti-computer" strategy. (And yes, I think that they _definitely_ exist). I define this as a 'overall plan' to exploit some inherent weakness in computer chess programs. Almost all are horrible at detecting long-term kingside attacks. Almost all are horrible at playing in positions where the pawns are locked up. So any sort of plan to reach those types of positions and then using some 'knowledge' to dictate where to break through in the least risky way, I call "anti-computer". I don't call 'anti-computer' any strategy where you have to lose thousands of games before you find a way to win once. My status... The _biggest_ problem I have today (in Crafty) is not losing vs GM players, rather it is avoiding draws against them. It has reached the point where many GM players _only_ play for draws against Crafty on ICC. And this is even harder to defend against than kingside attacks... I'm having some success in combating this, but it is _very_ difficult. If a GM wants a draw, it is _very_ difficult to prevent him from getting one... And this is made more difficult by a large opening book where I haven't made any attempt to define openings by class (a good project for someone) like drawish, sharp, hard to lose, but maybe hard to win also, etc...
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.