Tough test position

[maxchgr]

[d]2r2r1k/pp1b2pp/3b3q/5p2/1P1Pn3/P2BN1P1/1B3PQP/R3R1K1 b - - 0 21

How long for your engine to find
21...Nxf2! 22.Qxf2 f4 23.gxf4 Rxf4 24.Qg2 Rcf8 -/+
(Gelfand's analysis of a variation in Grischuk-Gelfand, Wijk aan Zee, 2002)

This seems very hard for most engines.

[Terry McCracken]

[d]2r2r1k/pp1b2pp/3b3q/5p2/1P1Pn3/P2BN1P1/1B3PQP/R3R1K1 b - - 0 21

How long for your engine to find
21...Nxf2! 22.Qxf2 f4 23.gxf4 Rxf4 24.Qg2 Rcf8 -/+
(Gelfand's analysis of a variation in Grischuk-Gelfand, Wijk aan Zee, 2002)

This seems very hard for most engines.

Very nice! Do you have a link to the game?

[Terry McCracken]

[d]2r2r1k/pp1b2pp/3b3q/5p2/1P1Pn3/P2BN1P1/1B3PQP/R3R1K1 b - - 0 21

How long for your engine to find
21...Nxf2! 22.Qxf2 f4 23.gxf4 Rxf4 24.Qg2 Rcf8 -/+
(Gelfand's analysis of a variation in Grischuk-Gelfand, Wijk aan Zee, 2002)

This seems very hard for most engines.

Very nice! Do you have a link to the game?

It appears to be a draw with an edge for Black. It's unlikely an engine will find this move, even though it is best and takes away White's advantage.

Computer Chess in fact has a long way to go, it's far from finished.

Terry

[BBauer]

Hallo,

Glaurung 2.0.1 took 2:05 min
Toga II 1.4.2JD took 2:07 min
Strelka 2.0 B took 3:04 min
Rybka 1.0 Beta took 7:55 min

to find Ne4xg2. Score went up to +1.09 for Toga and Strelka.
kind regards
Bernhard

[Marc MP]

[d]2r2r1k/pp1b2pp/3b3q/5p2/1P1Pn3/P2BN1P1/1B3PQP/R3R1K1 b - - 0 21

How long for your engine to find
21...Nxf2! 22.Qxf2 f4 23.gxf4 Rxf4 24.Qg2 Rcf8 -/+
(Gelfand's analysis of a variation in Grischuk-Gelfand, Wijk aan Zee, 2002)

This seems very hard for most engines.

Very nice! Do you have a link to the game?

It appears to be a draw with an edge for Black. It's unlikely an engine will find this move, even though it is best and takes away White's advantage.

Computer Chess in fact has a long way to go, it's far from finished.

Terry

Hi Terry,

Chess Tiger 2007.1 finds it after some time:

Code: Select all

...

00:07:33.9	-1,32	19	255936584	Rce8 Rad1 Qh5 f4 Re7 Nc4 Bc7 Ne5 Ba4 Rc1 Rfe8 Bxe4 fxe4 Bc3 Bc6 Nxc6 bxc6 Re3 Qd5 

00:10:15.8	-1,24	19	336910906	Ng5 h4 Ne4 Nxf5 Bxf5 Bxe4 Qg6 Bxf5 Qxf5 Rac1 Rxc1 Rxc1 b6 f4 Re8 Qc6 Rd8 

00:13:02.3	-1,24	20	421299699	Ng5 h4 Ne4 Nxf5 Bxf5 Bxe4 Qg6 Bxf5 Qxf5 Rac1 Rxc1 Rxc1 b6 Rc6 Qb1+ Bc1 Re8 

00:19:16.3	-0,35	20	630435263	Nxf2 Qxf2 f4 gxf4 Rxf4 Qg2 Rh4 Nf1 Bh3 Qc2 Qf4 Qf2 Qxf2+ Kxf2 Rf8+ Ke3 Bf4+ Kf3 Bd6+ 

00:22:44.0	-0,35	21	742568726	Nxf2 Qxf2 f4 gxf4 Rxf4 Qg2 Rh4 Nf1 Bh3 Qc2 Qf4 Qf2 Qxf2+ Kxf2 Rf8+ Ke3 Bf4+ Kf3 Bd6+ Ke3 Bf4+ Kf3 Bd6+ Ke3 Bf4+ Kf3 Bd6+ Ke3 

The game is here: http://www.chessgames.com/perl/chessgame?gid=1241400 , but recall the test position is from Gelfand's analysis, not the actual game.

[maxchgr]

thats correct game, and as Marc said it is position of gelfand's analysis though.
Rybka winfinder found it 4 seconds for me though! and of course rybka 2.3.2a found it in something around 30 seconds and Fritz 10 took a while but found it. Many engines i tried just take very long and dont find it.

[Jouni]

Analysis by bright-0.3a:

1...Ng5 2.h4 Ne4 3.Nc4 Bb8 4.Rac1 Be6 5.Ne3 Bd6 6.Rxc8 Rxc8 7.Bxe4 fxe4 8.Qxe4
² (0.60) Depth: 9/35 00:00:04 2287kN
1...Ng5 2.h4 Ne4 3.Rac1 Rxc1 4.Bxc1 Qf6 5.Bb2 Be6 6.Rc1 Qg6
² (0.65) Depth: 10/35 00:00:09 5601kN
1...Nxf2 2.Qxf2 f4 3.gxf4 Rxf4 4.Qg2 Rcf8 5.Re2 Rf3 6.Nf5 R3xf5 7.Bxf5 Rxf5 8.Qxb7
² (0.60) Depth: 10/35 00:00:14 9090kN
1...Nxf2 2.Qxf2 f4 3.gxf4 Rxf4 4.Qg2 Rcf8 5.Re2 Rf3 6.Nf5 R3xf5 7.Bxf5 Rxf5 8.Qxb7
² (0.60) Depth: 11/36 00:00:17 11038kN
1...Nxf2 2.Qxf2 f4 3.gxf4 Rxf4 4.Qg2 Rcf8 5.Re2 Rf3 6.Nf5 R3xf5 7.Bxf5 Rxf5 8.Qxb7 Bb5
² (0.64) Depth: 12/41 00:00:29 19009kN
1...Nxf2 2.Qxf2 f4 3.gxf4 Rxf4 4.Qg2 Rcf8 5.Re2 Rf3 6.Nf5 R3xf5 7.Bxf5 Bxf5 8.Qxb7 Bd3
² (0.60) Depth: 13/41 00:01:11 46224kN
1...Nxf2 2.Qxf2 f4 3.gxf4 Rxf4 4.Qg2 Rh4 5.Nf1 Bh3 6.Qc2 Rf4 7.Qe2 Rcf8 8.Bc1 Rxf1+ 9.Rxf1 Bxf1 10.Bxh6 Bxe2 11.Bxe2 gxh6
= (0.21) Depth: 14/42 00:02:04 78905kN

Jouni

[Terry McCracken]

Hallo,

Glaurung 2.0.1 took 2:05 min
Toga II 1.4.2JD took 2:07 min
Strelka 2.0 B took 3:04 min
Rybka 1.0 Beta took 7:55 min

to find Ne4xg2. Score went up to +1.09 for Toga and Strelka.
kind regards
Bernhard

Toga didn't have a clue! Maybe you ran this on a Octi-Core? You should give the hardware.

Terry

[maxchgr]

wow! impressive by bright. Is bright good at these things?

[kgburcham]

[d] 2r2r1k/pp1b2pp/3b3q/5p2/1P1Pn3/P2BN1P1/1B3PQP/R3R1K1 b - -

Zappa Mexico II x64
Hash 1024

11/41 0:01 +0.55 1...Nxf2 2.Qxf2 f4 3.gxf4 Rxf4 4.Qg2 Rcf8 5.Re2 Rf3 6.Bf5 R8xf5 7.Nxf5 Rxf5 8.Qxb7 (2.935.061) 1662
12/43 0:06 -0.46 1...Nxf2 2.Qxf2 f4 3.gxf4 Rxf4 4.Qg2 Bh3 5.Qe2 Rcf8 6.Rf1 Re8 7.Rf2 Rxe3 8.Qxe3 (12.931.601) 1872
13/43 0:07 -0.46 1...Nxf2 2.Qxf2 f4 3.gxf4 Rxf4 4.Qg2 Bh3 5.Qe2 Rcf8 6.Rf1 Re8 7.Rf2 Rxe3 8.Qxe3 (14.665.178) 1899
14/47 0:45 -1.01 1...Nxf2 2.Qxf2 f4 3.Nf5 Bxf5 4.Bxf5 Rxf5 5.Rac1 Rcf8 6.g4 Rg5 7.Qf3 Qh4 8.h3 h5 9.Re2 hxg4 10.hxg4 Qxg4+ 11.Qxg4 Rxg4+ (92.077.291) 2009
15/47 0:49 -1.01 1...Nxf2 2.Qxf2 f4 3.Nf5 Bxf5 4.Bxf5 Rxf5 5.Rac1 Rcf8 6.g4 Rg5 7.Qf3 Qh4 8.h3 h5 9.Re2 hxg4 10.hxg4 Qxg4+ 11.Qxg4 Rxg4+ (100.170.518) 2018
16/51 1:40 -1.07 1...Nxf2 2.Qxf2 f4 3.h4 fxg3 4.Qg2 Rf2 5.Qe4 Re8 6.Ng4 Rxe4 7.Nxh6 Rxe1+ 8.Rxe1 gxh6 9.Bc1 Bc6 10.b5 Bd7 11.Bxh6 Bxa3 (207.406.664) 2063
17/53 6:20 -1.14 1...Nxf2 2.Qxf2 f4 3.Qd2 fxg3 4.Re2 Rf2 5.Rxf2 gxf2+ 6.Qxf2 Bxh2+ 7.Qxh2 Qxe3+ 8.Qf2 Qxd3 9.Rf1 Be6 10.Qf3 Qb5 11.Re1 Qg5+ 12.Qg2 Qxg2+ 13.Kxg2 Bd5+ 14.Kg3 Kg8 (781.184.286) 2052
18/55 10:36 -1.14 1...Nxf2 2.Qxf2 f4 3.Qd2 fxg3 4.Re2 Rf2 5.Rxf2 gxf2+ 6.Qxf2 Bxh2+ 7.Qxh2 Qxe3+ 8.Qf2 Qxd3 9.Rf1 Be6 10.Qf3 Qb5 11.Re1 Qg5+ 12.Qg2 Qxg2+ 13.Kxg2 Bd5+ 14.Kg3 Kg8 (1.298.608.645) 2039
best move: Ne4xf2 time: 16:29.250 min n/s: 1.890.592 nodes: 1.992.678.891