Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Gelfand Position from World Cup

Author: Dave Gomboc

Date: 07:43:47 09/08/00

Go up one level in this thread


On September 08, 2000 at 09:46:36, Stephen A. Boak wrote:

>On September 08, 2000 at 01:05:22, Howard Exner wrote:
>
>>A nice win by Gelfand.
>>[D]3k4/2pP2b1/3p2P1/1K1P1BP1/8/8/8/8 w - -
>>
>>
>> id Gelfand,B - Ye Jiangchuan; bm Kc6;
>>
>>In the game Gelfand played Kc6 and Ye Jiangchuan resigned.
>>The winning technique as explained on Kasparov's site explains
>>that white's bishop must go to either a4 or e8, allowing the white king to
>>penetrate via the f5 square. Both work nicely, yet it made me wonder if computer
>>programs will discover this winning technique. How do programs
>>fair here? The theme of first protecting the d7 pawn with the king, then
>>with the bishop(on a4 or e8), and finally marching the king around to the
>>kingside may prove difficult for programs.
>
>Somebody me help me analyze a clear win here.  I sure can't find it, even with
>Fritz6's help.
>
>I tried the case of Be8 (suggested by Shipov as winning).
>
>I tried Fritz6a and it couldn't find the win, even after I played out many of
>the moves in the supposed winning strategy (specifically, positioning B on e8,
>and then bringing the white king back to f4 or g4 area).
>
>It looks to me like black can set up a dynamic (changing) fortress in that event
>(Be8) which holds the draw.
>
>Until Be8 is achieved by white, Black shuffles his Bishop on the long diagonal.
>
>After Be8 is achieved by white (with white K on c6):
>
>STEP 1: Black plays his king to e7 (keeping white king from f6 entry) after
>white king leaves c6 to go to b5 to return to kingside; then
>
>A. While both g5 & g6-pawns still exist:
>
>Corresponding squares for black Bishop (BB) to go to, for each white king (WK)
>square:
>WK BB
>f5 g7
>h5 g7
>[Thus white can't sac while his king is on f5 or h5, since BB will block the
>g-pawns.  White's king must be *behind* the g5 pawn (not beside it), prior to
>sac of g6-pawn.]
>
>e4 e5 or h8 (not g7)
>f4 e5+ or h8
>g4 e5 or h8
>h4 e5 or h8
>
>B. White sac's the g6-pawn with g6-g7 (while white king starts *behind*
>g5-pawn):
>
>WK BB
>g7 xg7
>f5 b2 or c3 (long diagonal)
>h5 b2 or c3 (long diagonal)
>g6 c1 or d2 (attack g5)
>h6 c1 or d2 (pin g5)
>
>after any g5-g6 push, Black easily gets to g7, keeping out the white king from
>h6 or f6.
>
>So, where is the forced win, with Be8?
>
>--Steve

I don't know about Be8, but I think Ba4 might work... you get the White king to
f4/g4 (to move to f5 next), place the white bishop on c6, by now the Black king
sits on e7 (to prevent Kf5), so sacrifice the d-pawn by promoting it, thereby
allowing Kf5-g6 and winning?

That is a 15-second analysis, it could be completely incorrect. :)

Dave



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.