Author: Odd Gunnar Malin
Date: 23:28:09 09/02/04
Go up one level in this thread
On September 02, 2004 at 17:20:36, Mike S. wrote: >On September 02, 2004 at 16:47:25, Odd Gunnar Malin wrote: > >>On September 02, 2004 at 16:12:28, Mike S. wrote: >> >>>But these are just mainlines (pv). For such tests, you'd have to actually >>>**play** a game (fragment), to see what an engine really can achieve. Displaying >>>a high evaluation does not mean by far yet that, an engine can get a clear win. > >>Yes, this was what I meant. I'm not sure it is too interesting. Maybe one (as >>human) shoud stydy this first and then see if one can win with white and draw >>with black most of the time against a top engine. > >How do GMs estimate that position? AFAIK a general rule is, if you have only >*one* weakness you can defend in most of the cases. Here, White has only *one* >strong point (rook on the 7th rank), rest is even symmetrical currently. > I guess this is one of the theme to learn from the startup position. How to use this bind to create more weakness, Kopec mention three elements to be taken care of here. There is another saying too here, this time from Dvoretsky. The wording is mine since I take it from the memory. - The most important factor is rook activity. - For the defending part, it is even worth a pawn or setting the king in an inferior position if you manage to break lose with your rook. Danny kopec on the video say somthing similar. Eg. Don't give up your dominating rook placement for one pawn, two pawn should be ok. >(Obviously it will be easier for White to activate his King...) > >Should White "theoretically" be able to win this? I mean, if Black can be >considered to be able to defend with/against best play, we cannot really >critizise engines which can't win this with White... :-) Nobody to critizise here, but such a position should be good as a startposition to learn this type of endgames. And since engines can't planning, they could be good sparingpartner where you would see your learningprogress (lost->draw-win(?)) when you learn more about the endgame. For adults it isn't easy to get human sparingpartners ready when you yourself have time. This remines me, what is a good engine? One you can learn from or one who do it good on a computer ratinglist. (This can of course be the same engine.) In the above diagram it is maybe just as 'good' if an engine do a thematic blunder as it is if it do a briliance. Odd Gunnar > >Regards, >Mike Scheidl
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.