Author: Dirk Frickenschmidt
Date: 09:01:34 01/09/98
Go up one level in this thread
On January 08, 1998 at 17:00:28, Amir Ban wrote: Hi Amir, ...some snipped >I watched the game on ICC and there public opinion was not impressed. >One GM described what was going on as "kindergarten". Well I still don't agree with that. Both left theory in move 7. The resulting game was quite interesting and entertaining until some blunders happened. >It seems that both sides played tactically weakly here. Black should >play simply 28... Rxg6 and if 29. Rf8+? Rxf8 30. Qxf8+ Nc8 31. Bxg6 Qg5 >and black wins. Better is 29. Bxg6 Bxf3 30. Qxf3 dxe5 31. dxe5 Qxb4 32. >Qf6 white will regain the pawn but this position looks better for black. Agreed. During the game I thought Karpov was avoiding some king attack he had seen in lines like 28...Rxg6 29.Bxg6 Bxf3 30.Qxf3 dxe5 31.dxe5 Qxb4 32.Rb1. But after for example 32...Qc5+ 33.Kh1 Nc6 34.Be4 Rf8 35.Qe2 nothing will happen to black after playing simply Ka7. >Black's 28... Qd8 is an elementary blunder which you don't expect at >this level, or even at lower levels. Karpov has now lost two games >through tactical blunders ! He must be very rusty. See above: he perhaps was trying to avoid an attack which did not exist. >Going back one move, in view of this, 27. e5 is a mistake since after >27... Bc6 28. Rg3 white's advantage is gone. A fair annotation would be >27. e5? Bc6 28. Ng6? Qd8?? Agreed again. But I still regard the white position as considerably stronger before 27.e5 anyway, so I think the white win is ok after faults on both sides. I think Qb2 would have been a good attacking move. White can sacrifice a pawn and gain a strong attack in some variations. Of course I'm not shure how GMs or IMs would judge this, only being a hobby player... 27.Qb2 A) 27...Nc6 28.b5 axb5 29.R3f2 Na7 30.Rb1 Qg5 31.a4 Rf8 32.axb5 Nc8 (32...Rxf4? 33.b6) 33.Qa3 Nb6 34.Ra2; B) 27...h4 28.h3 Nc6 29.R3f2 e5 30.Nd5 Qd8 31.b5 axb5 32.Bxb5 Nxd4 33.Bxd7 Qxd7 34.Rb1 c6 35.Nf6 Qg7 36.Nxg8 Rxg8 These are only examples, of course, and only an expert analysis could show the truth about the line. >>>The computer programs I used all missed Anand's strong plan for direct >>>king threats with 33. Rd1 Qxb4 34. Rb1! followed by the elegant queen >>>moves to h5 and back to e2 (aiming at a6). >> It seems to me that this is a key point concerning computer analysis: When I tried to use different computer programs for analysis I was surprised how late *all* of them still recognize problems concerning slowly growing kingside attacks. If I remember it correct, none of the program would have chosen the above Qb2 variation, though they slowly(!) began to recognize white had some attack after(!) moves like sacrificing a white pawn (which all or most of them would not have sacrificed within minutes on my P133) and getting attack on the open(ing) a and b files. >I did not understand why 35. Qxh5 was not answered simply by Qxa2. Who >knows, maybe Anand would overlook the mate on g2 :) Maybe Karpov did not like the resulting endgame after 36. Qe2 Qxe2 too much? The four connected black pawns look impressive, but black is a piece down with only two pawns ahead, and the white connected pawns on g/h are as fast as the black ones. Only a guess again... >White can play 39. >Nd6 instead of h3 with a quick win. Correct. But meanwhile Anand did not have to be in a hurry any more. >Amir Thanks for your comments. On the "Lost Boys" server the GM comments and evaluations were much less concrete than it obviously was the case on ICC... Kind regards from Dirk
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.