Author: Paulo Soares
Date: 14:59:22 05/27/99
Go up one level in this thread
On May 27, 1999 at 08:48:18, Harald Faber wrote:
>On May 27, 1999 at 08:18:59, blass uri wrote:
>
>>>>>Right, I don't know how strong a chess player with ELO1800 is, I just have a
>>>>>guess. I am at 1800DWZ, our best club player has 2000DWZ and an ELO2200.
>>>>>I see a lot in analysis but when this guy starts to play/analyze I am astonished
>>>>>again and again how much more he sees and realizes.
>>>>
>>>>Harald,
>>>>
>>>>It makes the following experience: ask for somebody to choose 5
>>>>positions, of preference closed, and tries to make the analysis of
>>>>those positions with the aid of some program of which you likes, for
>>>>one hour. After that ask for the best player of his club making the
>>>>analysis of the same positions, in the same time, without using a
>>>>program. I bet with you a beer(man talk that Germany beer is
>>>>excellent), as his analyses will be very better.
>>>>Paulo Soares
>>>
>>>What do you bet? That the program's analysis is better?
>>>Then I can say you already lost, I did it with OPEN (!) positions, see the game
>>>Gambitsoft-Forum vs IM Hans Klarenbeek, game still to go on.
>>>HUMANS (in my club) came up with GOOD, JUSTIFIED and REASONABLE move suggestions
>>>and IDEAS. Programs only had stupid move suggestions (or even worse ideas), e.g.
>>>16.Nf5 and 16.g5 were suggested. 16.g5 might have been a considerable move but
>>>16.Nf5 is not.
>>
>>I think that he does not say that the program's analysis is better but that the
>>analysis of the team of human with 1800 rating + program is better.
>>Uri
>
>It is still wrong. :-)
>It took 1900+ (DWZ) to show the right way(s). Programs+me failed. We had some
>interesting ideas of defence but not for the attack.
Harald, I have the certainty that you are missed in the considerations numbers
2 and 3 that I made in reply to Robert Yatt:
"2. Which is the experience that has that person in this type of analysis?
Also it is important, and I say on that due to my experience playing
postal chess, assisted for computer, for some years.
3. Which the time useed to make the analysis?
Better is much time, and I am speaking of hours."
I win the bet, and I invite you and Uri(because he is helping me)
to drink a good Germany beer! (it's only a joke).
About the truth, I think that this subject " human+machine, in analysis ", must
be better evaluated, therefore I (Estimated ELO=1800) have gotten satisfactory
resulteds, and I think that Uri (ELO=2020), also. The test considered for Uri
must decide the question.
I also find that the analyses of a GM can and must be contested by anyone
that finds in conditions for that, without no demerit for GM,
therefore today we have powerful tools for that: a good program and
a good hardware, and, very important, time.
Only a throught, about the Rebel10.5xRohde game, why GM Rohde did not
analysed 23.Rxb4, this is not a good move for white? And was his analyses
for 26. Bg2, better then this: 26....Qc2 27. Qe1 Qd2 28. Qb1(not Qxf1)
Re8...etc?
I want to leave clearifyly that I am not wanting to diminish GM Rohde, or
any GM, I am only placing my thoughts regarding the possible strong in
analysis of team "human+machine+program+time" in analyses for low
ELO of human.
Paulo Soares.
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.