Author: Albert Silver
Date: 11:51:16 12/31/99
Go up one level in this thread
On December 31, 1999 at 08:11:57, Graham Laight wrote:
>On December 30, 1999 at 15:47:03, Albert Silver wrote:
>
>>Sorry to say this, but that has been my point from the beginning. You are saying
>>that these 2500-2600 computers are only well equipped to handle one style of
>>play, and will be at a disadvantage against opponents whose style do not fit it;
>>yet their 2500-2600 human peers do not (of course there are always exceptions).
>>Could it be that there IS indeed a difference? Otherwise their ratings are a
>>fiction.
>>
>> Albert Silver
>
>I have to disagree.
>
>For this to be true, the existence of "anti-computer" tactics would have to be
>something which only applies in the 2500+ Elo range.
>
>In reality, "anti-computer" tactics exist all through the Elo ranges - probably
>even more so at the lower end of the range.
>
>Speaking for myself, I used to be able to beat lots of chess computers, using
>tactics like swapping off material, playing for the endgame, carrying on
>regardless how poor the position (because the computer would fall over in the
>endgame), doing nothing (computers could be trusted to compromise their position
>sooner or later), and things like this.
>
>However, once I acquired computers (and programs) over 2000 Elo, none of this
>would work any more. Even under the most favourable conditions, I simply could
>not beat them anymore (without taking the time to learn a lot of new chess
>skills).
>
>Some humans have not experienced this yet, and so are still in a position to
>talk about the great weaknesses of computers. Hopefully, the new batch of
>programs will help them experience what for them will be a new perspective.
>
>-g
I understand but I wasn't talking about anti-computer tactics. Uri said that
positional players were tougher opponents for computers than tactical players,
and this was unfair as it would lower the computer's rating. This is nonsense.
Should I also declare that I am in fact rated 2400 FIDE _except_ against
positional players? If the program is at 2600 (or whatever rating) strength,
this should not be dependent on the style of the opponent's play.
Albert Silver
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.