Author: Sune Larsson
Date: 10:05:20 05/19/01
Go up one level in this thread
On May 19, 2001 at 12:09:01, Uri Blass wrote: >On May 19, 2001 at 11:49:48, Sune Larsson wrote: > >>On May 19, 2001 at 10:26:32, Uri Blass wrote: >> >>>On May 19, 2001 at 09:36:00, Sune Larsson wrote: >>> >>>>On May 18, 2001 at 15:47:05, Dan Andersson wrote: >>>> >>>>>The King plays the Old-Indian defence, and gets to exchange pieces thus getting >>>>>a solid and passive position that seems playable or at least without any serious >>>>>weakness. Whites bishop is by no means weak as it has the ability to >>>>>singelhandedly defend the Queen side while the heavy pieces regroup and prepare >>>>>the opening pawn break e4-e5, the bishop could make blacks defence against this >>>>>hard by the threat of Bb5. Seeking a no existing initiative (or maybe kings >>>>>position bonuses) The King erroneously goes on a rampage on the kings side, >>>>>helping white advance his pawns. Passive defence is the only way in Old-Indian, >>>>>black should defend passively making white pay for the e4-e5 advance. Clearly >>>>>not a good opening for The King. >>>>> >>>>>Regards Dan Andersson >>>> >>>> >>>> Nice summary - still the Old-Indian carries some poison... >>>> The famous game Averbakh-Kotov, Zuerich 1953 springs to mind. >>>> >>>> [D]1r4nk/1p1qb2p/3p1r2/p1pPp3/2P1Pp2/5P1P/PP1QNBRK/5R2 b - - 0 30 >>>> >>>> Averbakh has just played 30.Ne2 when Kotov unleashes his shocking 30.-Qxh3+!! >>>> The white king is forced to a heavy walk right into the mine field at f5. >>>> >>>> The whole Old-Indian game: >>>> >>>> >>>>Averbakh,Y - Kotov,A [A55] >>>>Candidats Tournament Zuerich (14), 23.09.1953 >>>> >>>>1.d4 Nf6 2.c4 d6 3.Nf3 Nbd7 4.Nc3 e5 5.e4 Be7 6.Be2 0-0 7.0-0 c6 8.Qc2 Re8 9.Rd1 >>>>Bf8 10.Rb1 a5 11.d5 Nc5 12.Be3 Qc7 13.h3 Bd7 14.Rbc1 g6 15.Nd2 Rab8 16.Nb3 Nxb3 >>>>17.Qxb3 c5 18.Kh2 Kh8 19.Qc2 Ng8 20.Bg4 Nh6 21.Bxd7 Qxd7 22.Qd2 Ng8 23.g4 f5 >>>>24.f3 Be7 25.Rg1 Rf8 26.Rcf1 Rf7 27.gxf5 gxf5 28.Rg2 f4 29.Bf2 Rf6 30.Ne2 Qxh3+ >>>>31.Kxh3 Rh6+ 32.Kg4 Nf6+ 33.Kf5 Nd7 34.Rg5 Rf8+ 35.Kg4 Nf6+ 36.Kf5 Ng8+ 37.Kg4 >>>>Nf6+ 38.Kf5 Nxd5+ 39.Kg4 Nf6+ 40.Kf5 Ng8+ 41.Kg4 Nf6+ 42.Kf5 Ng8+ 43.Kg4 Bxg5 >>>>44.Kxg5 Rf7 45.Bh4 Rg6+ 46.Kh5 Rfg7 47.Bg5 Rxg5+ 48.Kh4 Nf6 49.Ng3 Rxg3 50.Qxd6 >>>>R3g6 51.Qb8+ Rg8 0-1 >>>> >>>>According to Bronstein, after 33.Kf5 Kotov had very little time left until >>>>the time control. Otherwise he would surely have found Stahlberg's postmortem >>>>suggestion 33.-Ng4!, as a cleaner and quicker way to victory. >>> >>>I doubt it. >>>I am also not convinced that 33...Ng4 is a cleaner and quicker way to win. >>> >>>I think that Ng4 is a typical computer move and not a human move. >>> >>>Humans usually are going to think about sacrificing pieces only if they cannot >>>find a good idea without sacrificing pieces and in this case there is a good >>>idea without sacrificing pieces. >>> >>>stahlberg's analysis was done after the game and when people analyze after the >>>game they are less afraid to sacrifice pieces because it is not their game. >>> >>>White can also try to defend after 33...Ng4 by 34.Nxf4 Rg8 35.Nh5 Rhg6 36.Qg5 >>>Bxg5 37.Kxg4(this is Junior5.9's main line at depth 18 with score of 1.94 for >>>black and the score was more optimistic for black at previous iterations) >>> >>>You can see that there are quiet moves like Rg8 and Rhg6 and humans need to find >>>that white has no good defence by checking all the legal moves of white at every >>>ply(remember that black is a full queen down so even moves that sacrifice a full >>>queen like 36.Qg5 should be checked). >>> >>> >>>Uri >> >> Scattering your doubts might be a huge task - but here are Bronstein's own >> words from the tournament book: >> >> (After 33.Kf5) "...For an understanding of the next phase of the game, bear >> in mind that Kotov had very little time left until the time control, and >> naturally did not wish to spoil such a beautiful and unusual game with some >> hasty move. Therefore, he decides to give a few checks, in order to get the >> game past the 40th move and adjourn it. No doubt, there has to be mate in this >> position; most probably, Kotov saw its basic outlines as far back as his 30th >> move." >> >> (After 33.-Nd7) " Here's the proof: had the queen sacrifice been "accurately >> calculated", Kotov would instead have chosen Stahlberg's postmortem suggestion, >> 33.-Ng4, depriving white of the reply 34.Rg5. After 33.-Ng4, white would have >> had to suffer colossal material losses in order to avert the mate threats." >> >> >> 33.-Ng4 a typical computer move and not a human move?? Then we just disjoin >> grossly in our chess views... >> >> In your line given above - 33.-Ng4 34.Nxf4 Rg8 35.Nh5 Rhg6 36.Qg5 Bxg5 >> 37.Kxg4 white has to suffer playing with an exchange down. Fritz 6 gives >> here -3.34 and lost for white. Junior 6 continues this variation with >> 37.-Bf4+ 38.Kh3 Rxg2 39.Nxf4 exf4 40.Be1 R8g6 41.Bh4 Rxb2 and climbs to -2.74 >> >> Now if doubt is today's trend - I doubt if Averbakh really would have tested >> Kotov any further in this endgame... >> >> Sune > >The evaluation after the forced line >33...Nd7 34.Rg5 Rf8+ 35.Kg4 Nf6+ 36.Kf5 Nxd5+ 37.Kg4 Nf6+ 38.Kf5 Ng8+ 39.Kg4 >Bxg5 is also very optimistic for black and in the game white resigned few moves >later. > >The game was longer but black had no reason to play for the fastest win when he >was in time trouble. Yes, this was exactly what Bronstein described in the tournament book. > >The main problem of programs is to see from the root position that after >39...Bxg5 white is losing but if you give them the position after 39...Bxg5 they >can only see an evaluation of about +3 for black so I doubt if 33...Ng4 is the >fastest win. > >Uri Okey Uri, Stahlberg and Kotov are dead and I *doubt* ;) if either of us would care to contact Bronstein in this issue... So I prefer to trust this man's words from the tournament book. By the way, this game was played in the 14th round where Bronstein had a Bye. Guess the queen sac is easy for present programs, but the trick is to play for the possibility. Deep Fritz changed from 30.-b5 -0.03 to 30.-Qxh3+! -2.06 after about 15 sec on a PIII 800. Sune
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.