Hard Test

Discussion of anything and everything relating to chess playing software and machines.

Moderator: Ras

Patatov

Hard Test

Post by Patatov »

[d] 8/8/3k2p1/1p2p2p/pNp1P2P/P1Pn4/1P2K1P1/2bN4 b - - 0 47

This position comes from a game Keres-Portisch, Moscow 1967
Here Portisch played 47...Nc5! and went on winning brillantly on move 90.
The trap was 47... Nxb2 ? 48. Nxb2 Bxb2 49. Kd2 Bxa3 50. Kc2 Kc5 (50... Bxb4 51. cxb4 = a3 52. Kc3 Kc6; 50... Bxb4 51. cxb4) 51. Na6+ Kb6 52. Nb4 =
Is there any engine that avoid Nxb2? I think this is a very difficult position for them
BBauer
Posts: 658
Joined: Wed Mar 08, 2006 8:58 pm

Re: Hard Test

Post by BBauer »

Great position and great trap!

The trap leads to a fortress and it is well known that computers have
difficulties to solve such positions. C. Donninger once saw such a problem and decided to do nothing.

Thank you
Bernhard
AGove

Re: Hard Test

Post by AGove »

[d]8/8/3k2p1/1p2p2p/1Pp1P2P/p1K5/6P1/8 b - - 0 52

Never mind the start position from the game, how many engines understand this position from one of the variations?
Tony Thomas

Re: Hard Test

Post by Tony Thomas »

Mr.Governator, I had the same position against queen 3.09 while playing over the position that was posted. I had no clue what to do.
BBauer
Posts: 658
Joined: Wed Mar 08, 2006 8:58 pm

Re: Hard Test

Post by BBauer »

Yace gives a score of +5.4 fot black.
Don't know what others do.
So a program may have an Elo value of 3000 but has no idea about
this position. :lol:
Bernhard