Author: Sune Larsson
Date: 03:31:18 06/22/01
Go up one level in this thread
On June 22, 2001 at 06:20:21, Uri Blass wrote: >On June 21, 2001 at 23:45:21, Dann Corbit wrote: > >>On June 21, 2001 at 17:15:59, Sune Larsson wrote: >> >>> >>> [D]4k2r/rp5p/pR4p1/4qp2/3pp1PQ/8/2P4P/R5K1 w k - 0 28 >>> >>> >>> This position is from IM Berg-GM Hector, Sigeman tournament - Malmo. >>> These two very tactical skilled Swedes had battled it out just like >>> old masters did in the 19th century. Young Emanuel Berg sacked one >>> pawn after another but, as it looked, in vain. Now, all good stories >>> give some glory to the brave one and that is also what happened here. >>> In the above position white, 4 pawns less, has the resource 28.Ra5!! - >>> with the point 28.-Qxa5 29.Qf6! (not 29.Re6+ Kd7). At the present >>> moment it looks like this 28.Ra5!! is good enough to draw the game >>> for white. >>> >>> >>> Test 1: Can your program find the move 28.Ra5!! - evals? >> >>It's a terrible [definitely losing] blunder, throwing away the rook for no >>compensation: > >> >>[D] 4k2r/rp5p/pR4p1/R3qp2/3pp1PQ/8/2P4P/6K1 b k - acd 15; acn 532177870; acs >>1000; ce 1126; pv Qxa5 Rxg6 hxg6 Qxh8+ Kf7 Qxd4 Qe1+ Kg2 Qe2+ Kg1 Qxg4+ Kf1 Qf4+ >>Kg2 Ra8 Qd5+ Kf8 c4 Qf3+ Kg1 Qe3+ Kg2 Qe2+ Kg1 Ra7; > >The big blunder is in the way that chess programs work. >I am not surprised that they cannot see nothing. > >The reason is simply the fact that they do not use a good algorithm. > >Programs except gustav cannot see a mate in 70 that was posted here and again >the reason is simply the fact that all programs except gustav do not use the >right algorithm to find mate in 70. > >After 1...Qxa5 2.Qf6 black must play 2...Qe1+ to get big advantage by the >evaluation of the programs. > >after 3.Kg2 black is again forced to play a check. >The number of checks is very high and after enough checks programs do not >extend enough and cannot see that black has to play checks. > >Uri That's totally correct. 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.