Author: Christophe Theron
Date: 10:51:06 04/09/01
Go up one level in this thread
On April 08, 2001 at 02:39:26, Uri Blass wrote:
>On April 08, 2001 at 01:28:20, George Humprey wrote:
>
>>On April 08, 2001 at 01:08:20, Uri Blass wrote:
>>
>>>On April 08, 2001 at 00:51:41, Christophe Theron wrote:
>>>
>>>>On April 07, 2001 at 23:31:18, Uri Blass wrote:
>>>>
>>>>>On April 07, 2001 at 18:25:44, robert flesher wrote:
>>>>>
>>>>>>Hello everyone i downloaded Chess Genius 2 last night.
>>>>>
>>>>>Do you mean to the Dos version or to 2.016?
>>>>>
>>>>> I must say that anyone
>>>>>>who likes genius this is a must. I find the interface that comes with Shredder
>>>>>>is rather poor with a slow mouse response. However this interface is fast and
>>>>>>simple. For strength i tested it Against good old hiarcs 7.01 30 min each. I
>>>>>>expected Hiarcs to destroy genius as this is a old program. After Three games 1
>>>>>>draw two wins for genius. You can imagine my shock.
>>>>>
>>>>>Did you use one computer for the test.
>>>>>
>>>>>I ask because I have doubts about tests on one computer because one program may
>>>>>be slowed down by a significant factor.
>>>>>
>>>>>
>>>>> I know everyone This is not
>>>>>>enough to draw hasty conclusions, although it shows that even old GENIUS2 can
>>>>>>hold its own to newer commercial programs. Maybe this shouldnt come as a
>>>>>>surprise as it was this same program that beat Kasparov himself in in 1994 at an
>>>>>>intel speed tournement.
>>>>>
>>>>>No
>>>>>Genius2.9 that is almost identical to Genius3 won kasparov in 1994.
>>>>>
>>>>>Kasparov was wrong to think that he is playing Genius2 and I guess that this is
>>>>>the reason that he lost.
>>>>
>>>>
>>>>
>>>>This kind of reasonning is a mystery to me.
>>>>
>>>>Can you explain to me why Kasparov (or somebody else) is supposed to have no
>>>>problem if it is a known program, and be totally disturbed if it is a slightly
>>>>improved version?
>>>
>>>The question is if it was only slightly improved version.
>>>It is possible that kasparov could win Genius2 easily at home because of some
>>>positional weakness that was corrected in Genius3.
>>>>
>>>>Do you think that a human can prepare against all the possible chess positions
>>>>if he has the opportunity to test the program?
>>>>
>>>>Are you going to say that it is a matter of opening preparation? In this case,
>>>>do you believe that Lang is so stupid that he would have played with the same
>>>>book as the commercial version?
>>>
>>>I do not say it but it is possible that kasparov knew Ganius2 and expected it to
>>>play some kind of mistake that Genius3 knew not to do.
>>>
>>>It is possible that kasparov did not play the best move in a position that is
>>>not in theory because he expected Genius to blunder.
>>>
>>>Uri
>>
>>
>> I disagree, there is an infinite amount of possiblities in chess, even in the
>>opening phase, it is impossible for kasparov to prepare for every opening move.
>>The only way that grandmasters are able to prepare is if they have a fair amount
>>of games of their opponent. If a programmer simply switches books, it would be a
>>very lucky chance that a program will play a specific move, anticipated by the
>>grandmaster.
>
>I disagree.
>Humans can know about specific weaknesses of a program.
>
>An example:
>It is possible to predict that some programs are going to do the mistake Bxh7
>and lose the bishop after g6 that helps to trap the bishop.
>
>It is possible that kasparov played a move that is not the best because he knew
>about specific weakness of Genius2 that was not a weakness of Genius3.
>
>We need to analyze the games with Genius2 in order to know.
I have never heard of such positional mistake that would have existed in Genius2
and that would have been solved in Genius3.
And you are also assuming an incredible amount of preparation for Kasparov
against a microcomputer program.
Come on, your assumptions belong to science fiction. Come back on earth.
Christophe
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.