The brilliant chess move (49)

[Glarean]

.

The weekly series of chezz puzzles "The brilliant chess move" (Glarean Magazine) includes
tactical and strategic examples of great human chess art, and simultaneously
it is composed of very hard positions for most chess software.

The youngest member of this Hall of Fame is a Endgame Study by A. Mouterde.

Regards: Walter

White to move




Solution:
http://glareanverlag.wordpress.com/cate ... schachzug/

.

[AdminX]

Intel Core 2 Quad Q9000 @ 2GHz
Analysis by Rybka 3:

1.Rf5 Bd4+ 2.Kd3 Kxh7 3.Rxd5 Rxd5 4.Bxd5 Kh6 5.Bf3 Kg5
= (0.20) Depth: 6 00:00:00 4kN
= (0.16) Depth: 9 00:00:00 27kN

1.Bxd5 Bg5+ 2.Ke4[] Kxh7 3.Rf7+ Kh6[] 4.Rxa7 Kxh5 5.a4 Bd2 6.a5 Bxa5 7.Rxa5 Re8+ 8.Kf4 Rf8+ 9.Ke5 Kg4
+/= (0.30) Depth: 9 00:00:00 67kN
+/= (0.34) Depth: 23 00:04:09 30054kN, tb=542

1.h6+ Kg6[] 2.Rxf6+ Kxf6[] 3.Bxd5[] Re8+[] 4.Kd3[] Rh8 5.Bg8[] Ke5 6.a3[] a6 7.a4[] Kd6 8.Ke4[] a5 9.Kf5 c4 10.Bxc4 Rxh7 11.Kg6[] Rh8 12.Kg7[] Rc8 13.h7[] Ke5 14.Bg8[] Rc7+ 15.Kg6[]
+/= (0.54 !) Depth: 23 00:04:50 35666kN, tb=646
+- (3.23) Depth: 23 00:08:28 58447kN, tb=2706

[AdminX]

Intel Core 2 Quad Q9000 @ 2GHz
Analysis by Naum 4:

1.Rf5 Bd4+ 2.Kf4 Kxh7 3.Bxd5 Kh6 4.Kg4 Rd7 5.a4 a5 6.Kf4 Re7 7.Be4 Re8 8.Bc6 Re1
+/= (0.27) Depth: 6/14 00:00:00 6kN
= (0.16) Depth: 16/28 00:00:01 2501kN, tb=21

1.Bxd5 Rxd5 2.Rxf6 Rxh5 3.Ra6 Kxh7 4.Kd3 Rd5+ 5.Kc4 Rd7 6.Kxc5 Rf7 7.Rc6 Rf5+ 8.Kc4 Kg7 9.Ra6 Rf4+ 10.Kd5 Rf5+ 11.Kd4 Rf4+ 12.Ke5 Rf7 13.a4 Rb7
= (0.25) Depth: 16/35 00:00:04 9693kN, tb=4526
= (0.16) Depth: 26/48 00:03:22 544mN, tb=684240

1.h6+ Kg6 2.Rxf6+ Kxf6 3.Bxd5 Re8+ 4.Kd3 Rh8 5.Bg8 Ke5 6.Kc4 Kd6 7.Kb5 a6+ 8.Kc4 Kc6 9.a3 Kb6 10.a4 Kc6 11.Kd3
+- (1.76) Depth: 26/62 00:06:28 1061mN, tb=862616
+- (1.76) Depth: 27/49 00:07:34 1249mN, tb=912449

[Eelco de Groot]

Nice study Walter!

It took me a while but I think Ancalagon can now solve it! But the version that can do it I more or less specially made for this study, so I don't know if I did it right. The Rook that ends up on h8 gets zero mobility and you can use that of course to speed up the solution. But to do that I had to change the Stockfish mobility routines as well, and they are probably not in balance anymore. I suspected the NPS to go down a lot maybe the slowdown is still within limits but not enough data. At least the solution time is better now

[D]3r4/p5kP/2B2b2/2pp3P/8/4K3/P7/5R2 w - -

Engine: Ancalagon 1.3 Weak Squares 180 Board Control middlegame 50 endgame 50
Build 364 (Athlon 2009 MHz, 256 MB) by Romstad, Costalba, Kiiski, de Groot

2.00 0:01 +0.23 1.Bxd5 Rxd5 2.Rxf6 Rxh5 3.Ra6 Kxh7
4.Rxa7+ Kg6 5.a4 Rh3+ 6.Kd2 Kf5
7.Ra5 Ke5 8.Rxc5+ Kd4 9.Rb5 Rh2+
10.Kd1 (965.446) 506

2.07 0:02 +0.24 1.Rf5 Bd4+ 2.Kd3 Kxh7 3.Bxd5 Kh6
4.Kc4 Re8 5.a4 (1.278.928) 517

3.01 0:05 +0.12 1.Rf5 Bd4+ 2.Kf3 Kxh7 3.Bxd5 Rb8
4.Kg4 Kh6 5.Rg5 Rb6 6.Rg8 (3.318.374) 570

3.02 0:05 +0.13 1.Bxd5 Rxd5 2.Rxf6 Rxh5 3.Ra6 Kxh7
4.Rxa7+ Kg6 5.a4 Rh3+ 6.Kd2 Kf5 7.a5 c4 (3.357.964) 573

4.01 0:09 +0.43 1.Bxd5 Rxd5 2.Rxf6 Rxh5 3.Ra6 Kxh7
4.Rxa7+ Kg6 5.Kd3 Rh4 6.Ra5 Rh2 7.a4 Rh4
8.Ra8 Kf5 9.a5 Rd4+ 10.Kc3 Ke6 (6.085.787) 644

5.01 0:36 +0.50 1.Bxd5 Rxd5 2.Rxf6 Rxh5 3.Ra6 Kxh7
4.Rxa7+ Kg6 5.Kd3 Rh4 6.Ra5 Rh2 7.a4 Rh4
8.Ra8 Kf5 9.a5 Rd4+ 10.Kc3 (23.410.704) 644

6.01 0:38 +0.30 1.Bxd5 Rxd5 2.Rxf6 Rxh5 3.Ra6 Kxh7
4.Rxa7+ Kg6 5.Kd3 Rh4 6.a4 Kf6 7.a5 Ke5
8.a6 Rd4+ 9.Kc2 Ra4 10.Re7+ Kd5 (25.152.728) 650

7.01 0:50 +0.37 1.Bxd5 Rxd5 2.Rxf6 Rxh5 3.Ra6 Kxh7
4.Rxa7+ Kg6 5.Kd3 Rh4 6.Kc3 Kf6
7.Ra5 Rh2 8.a3 Rh5 9.Ra6+ Ke7 10.a4 Rh4
11.a5 Ra4 (34.146.070) 670

8.01 1:00 +0.24 1.Bxd5 Rxd5 2.Rxf6 Rxh5 3.Ra6 Kxh7
4.Kd3 Rd5+ 5.Kc4 Rd7 6.Kxc5 Kg7
7.Kc6 Re7 8.Kd5 Rd7+ 9.Ke6 Rf7
10.Rc6 Rf6+ 11.Kd5 Rxc6 12.Kxc6 (42.146.667) 691

8.02 1:20 +0.29 1.h6+ Kg6 2.Rxf6+ Kxf6 3.Bxd5 Rh8
4.Bg8 Kg6 5.Kd3 Kxh6 6.Kc4 Kg7
7.Kxc5 Kg6 8.a4 Kg7 9.a5 Kg6 10.Kd4 Kg7 (54.452.911) 676

9.01 1:45 +0.51 1.h6+ (73.220.513) 692

10.01 2:40 +1.50 1.h6+ (108.841.053) 677

11.01 3:10 +1.77 1.h6+ Kg6 2.Rxf6+ Kxf6 3.Bxd5 Rh8
4.Bg8 Kg6 5.Ke4 Kf6 6.a4 Kg5 7.Kd3 Kg4
8.Kc4 Kg5 9.Kxc5 Kg6 10.a5 Kg5
11.Kb5 Kxh6 12.Ka6 Kg5 13.Kb7 Kf4
14.Kxa7 (130.645.417) 684

12.01 5:31 +2.14 1.h6+ Kg6 2.Rxf6+ Kxf6 3.Bxd5 Rh8
4.Bg8 Kf5 5.Kd2 Kf6 6.Kc3 Ke7 7.Kd3 Kd7
8.Ke4 Kd6 9.a4 c4 10.Kd4 Kd7 11.Kxc4 Kc6
12.Kd3 Kd6 13.Kd4 a6 14.Ke4 (228.200.598) 687

13.01 9:31 +2.67 1.h6+ Kg6 2.Rxf6+ Kxf6 3.Bxd5 Rh8
4.Bg8 Ke5 5.Kd3 Kd6 6.Ke4 Kd7 7.Kd5 c4
8.Kxc4 Kc8 9.Kd4 Kd8 10.Kd5 Kd7
11.a4 Kd8 12.Kd6 Ke8 13.Kc6 Kd8
14.Kb7 (403.462.510) 705

14.01 27:44 +3.17 1.h6+ Kg6 2.Rxf6+ Kxf6 3.Bxd5 Rh8
4.Bg8 Ke5 5.Kd3 Kd6 6.Ke4 c4 7.Kd4 Kc7
8.Kxc4 Kd6 9.Kb5 Kc7 10.Kc5 Kd7
11.Kd5 Ke7 12.Kc6 Kd8 13.a4 Kc8
14.a5 (1.119.043.793) 672

15.01 73:02 +3.45 1.h6+ Kg6 2.Rxf6+ Kxf6 3.Bxd5 Rh8
4.Bg8 Ke5 5.Kd3 Kf6 6.Kc4 Ke7 7.Kxc5 Kd7
8.Kd5 Ke7 9.Kc6 Kd8 10.Kb7 a5 11.a4 Kd7
12.Kb6 Kd6 13.Kxa5 Kc5 14.Ka6 (2.766.834.419) 631

[peter]

Just for the discussion of actuality:


3r4/p5kP/2B2b2/2pp3P/8/4K3/P7/5R2 w - - 0 1

Analysis by IPPOLIT 0.080a:

1.Lxd5 Ld4+ 2.Ke4 Kxh7 3.Tf7+ Kh6 4.Txa7 Kxh5
= (0.16) Tiefe: 6/17 00:00:00 8kN
...
1.Lxd5 Txd5 2.Txf6 Txh5 3.Ta6 Te5+ 4.Kd3 Te7 5.Th6 Kh8 6.Kc4 Te2 7.a3 Te3 8.Ta6 Te7 9.Kxc5 Kxh7 10.Kd6 Tf7 11.Ke5 Kg7 12.a4 Tb7 13.Te6 Kf7 14.Tc6 Td7 15.Tf6+ Ke7 16.Te6+ Kf7 17.Tf6+
= (0.18) Tiefe: 26/49 00:03:13 247mN
1.h6+ Kg6 2.Lxd5 Ld4+ 3.Ke4 Kxh7 4.Tf7+ Kxh6 5.Txa7 Te8+ 6.Kd3 Te3+ 7.Kc4 Tc3+ 8.Kb5 Tc2 9.Th7+ Kxh7 10.Le4+ Kg7 11.Lxc2 c4 12.Kxc4 Lb6 13.a4 Kf6
Tiefe: 26/50 00:03:32 274mN
...
1.h6+ Kg6 2.Txf6+ Kxf6 3.Lxd5 Te8+ 4.Kf4 Th8 5.Lg8 Ke7 6.Ke4 Kf6 7.Kd5 c4 8.Kxc4 Kg6 9.Kb5 Kf6 10.Ka6 Ke7 11.Kxa7 Kd6 12.Kb8 Ke5 13.Kc7 Kf6 14.a4
Tiefe: 28/62 00:09:19 1008mN

But that's single core of course and on the same 2,5Ghz-Intel 32 bit R3 did not find it, thinking over an hour.