Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Can your program find Rxg2?

Author: Uri Blass

Date: 11:56:56 04/28/02

Go up one level in this thread


On April 28, 2002 at 14:35:41, Roy Eassa wrote:

>On April 28, 2002 at 14:17:54, John Merlino wrote:
>
>>On April 28, 2002 at 01:19:27, Uri Blass wrote:
>>
>>>On April 27, 2002 at 18:41:28, John Merlino wrote:
>>>
>>>>On April 27, 2002 at 18:12:27, Uri Blass wrote:
>>>>
>>>>>The following position is from my last game
>>>>>
>>>>>[D]6k1/RR4p1/5p1p/4p3/8/P7/4rrPP/3K4 b - - 0 31
>>>>>
>>>>>I played 31...Rxg2 and after 32.a4 Ra2 my opponent resigned.
>>>>>
>>>>>
>>>>>
>>>>>After the game he said that he needed one more tempo.
>>>>>
>>>>>I said that I was lucky because I saw the idea of Rxg7+ that almost win the game
>>>>>only after I played 31...Rxg2.
>>>>>
>>>>>I played 31...Rxg2 fast because the move seemed to be almost forced(I did not
>>>>>like the idea of Rd2+ Kc1 and the idea of Ra2 with the double mate threat is
>>>>>gone)
>>>>>
>>>>>I also said in the anlaysis that the best was still 32.Rxg7+ Rxg7 33.Ra8+
>>>>>inspite of the fact that white is losing after 33...Kf7 34.Kxe2 Rg2+ because in
>>>>>the worst case black will need to sacrifice the rook for the a pawn when the
>>>>>black passed pawns are going to be srtronger than the rook.
>>>>>
>>>>>
>>>>>I find that some top programs cannot find 31...Rxg2
>>>>>for black because they do not evaluate correctly the pawn endgame.
>>>>>
>>>>>
>>>>>Uri
>>>>
>>>>It takes Chessmaster 9000 (pre-Beta) 2:05 on a PIII-600 to see Rxg2, but about a
>>>>minute later it discards it for Rd2+ (which turns out to be two checking moves
>>>>to force the White King farther away from the Black pawns BEFORE taking on g2):
>>>
>>>The problem is that the line that chessmaster suggests is not forced.
>>>
>>>31...Rd2+ 32.Kc1 Rc2+ 33.Kb1 Rxg2 34.Rxg7+ is not a forced line from white point
>>>of view.
>>>
>>>White can choose 32.Ke1
>>>
>>>Uri
>>
>>Well, maybe not "forced", but, according to CM, 31...Rd2+ 32.Ke1 leads to this
>>(with an eval that is instantly about 1.50 worse for White than if 32.Kc1):
>>
>>Time	Depth	Score	Positions	Moves
>>0:00	1/3	-2.51	2176		32...Rfe2+ 33.Kf1 Rxg2 34.a4 Rgf2+
>>					35.Kg1
>>0:00	1/4	-2.49	6886		32...Rfe2+ 33.Kf1 Rxg2 34.a4 Rdf2+
>>					35.Ke1 Re2+ 36.Kd1
>>0:00	2/5	-2.42	32451		32...Rfe2+ 33.Kf1 Rxg2 34.a4 Rdf2+
>>					35.Ke1 Ra2 36.Kf1 f5 37.a5 Rgf2+
>>					38.Ke1
>>0:01	3/6	-2.46	79710		32...Rfe2+ 33.Kf1 Rxg2 34.a4 Rdf2+
>>					35.Ke1 Ra2 36.Kf1 Kh8 37.h4 Rgf2+
>>					38.Ke1 Rfe2+ 39.Kf1
>>0:03	4/7	-2.50	269671		32...Rfe2+ 33.Kf1 Rxg2 34.a4 Rdf2+
>>					35.Ke1 Ra2 36.Kf1 Rg4 37.h3 Rf4+
>>					38.Ke1 Re4+ 39.Kf1
>>0:08	5/8	-2.54	756383		32...Rfe2+ 33.Kf1 Rxg2 34.a4 Rdf2+
>>					35.Ke1 Ra2 36.Kf1 Kh8 37.h4 Rgc2
>>					38.Rb1 Kh7 39.h5
>>0:27	6/9	-2.44	2437336		32...Rfe2+ 33.Kf1 Rxg2 34.a4 Rdf2+
>>					35.Ke1 Ra2 36.Kf1 Kh8 37.Ra8+ Kh7
>>					38.Raa7 Raf2+ 39.Ke1 Rc2 40.Kf1
>>					Rg4 41.h3 Rf4+ 42.Ke1
>>1:26	7/10	-2.71	8085064		32...Rfe2+ 33.Kf1 Rxg2 34.a4 Rdf2+
>>					35.Ke1 Ra2 36.Kf1 Kh8 37.Ra8+ Kh7
>>					38.Raa7 e4 39.Re7 e3 40.Rxe3 Rxh2
>>					41.Kg1 Kg6
>>4:38	8/11	-2.81	26452346	32...Rfe2+ 33.Kf1 Rxg2 34.a4 Rdf2+
>>					35.Ke1 Ra2 36.Kf1 Kh8 37.Ra8+ Kh7
>>					38.Raa7 h5 39.Rd7 Rgc2 40.Ke1 Kg6
>>					41.a5 Kf5
>>
>>It probably all leads to the same loss, anyway....
>>
>>jm
>
>
>It DOES matter, though, because it affects what is the "best" move.  It seems
>that it is indeed best to do the two checking moves first (pushing the king
>away), THEN take the pawn.  I.e., Rxg2 is NOT the best first move.

I disagree.

I do not see why pusing the king away is better.
It only delay the win.

Rxg2 immediatly force Rxg7+ and after Rxg7 a lot of programs  do not know to
evaluate correctly the pawn endgame.

Their main line before changing their mind from Rxg2 is not Rxg2 Rxg7+ Rxg7 Ra8+
but Rxg2 Rxg7+ Rxg7 Rxg7+??

Uri



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.