Author: Albert Silver
Date: 06:57:18 10/06/04
Go up one level in this thread
On October 06, 2004 at 01:28:01, Andrei P wrote:
>On October 06, 2004 at 00:05:28, Albert Silver wrote:
>
>>On October 05, 2004 at 23:50:37, Andrei P wrote:
>>
>>>I found this on Ossimtz' site. apparently, a ~2000 strength player used
>>>fritz5.32 to direct all his moves during a match, of course his opponets never
>>>suspecting anythng. He got 2600+ performance out of 9 games. My best guess is
>>>that the cpu was PII 450 Mhz.
>>>
>>>So, today's top programs clearly should be 2800+ on a decent hardware, PROVIDED
>>>the opponets is NOT suspecting.
>>
>>And provided they think you're rated 1900 as the famous cheater was.
>>
>>
>
>agreed, Albert, this should have made a difference in the first 4-5 rounds or
>so, but in the second half of the rounds the 1900 player must have been looked
>at like an equal by the players with similar performance. so, the last games
>can't be attributed to underestimating the opponent.
Yes, they can. No matter how well a 55-year-old 1900 player may be performing,
there is simply no way a GM is going to presume he is facing another GM-strength
player. At best (worse) they may assume he is playing at 2300 strength. The fact
that he was in fact a computer makes it even worse, since the best way to blow
away a player who is assumed considerably weaker (and a 2300 would still be much
weaker) is to play on their weakness. In this case, that overperforming 1900
might be playing a 2400 positional game (meaning that would be the assumption),
as tactics are the domain of the young, so they might veer for a sharper game.
That would mean dealing with a 2600+ tactical middlegame from Fritz.
Albert
>
>>>
>>>the dude who performed the 1998 "test" seems to haver retired from chess in
>>>1999, but his test was a really nice one in support of computer superiority.
>>>
>>>http://www.uni-klu.ac.at/~gossimit/c/curious.htm the file is alwer.zip
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.