Testposition-(Medium). Which engine can find it?

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

Moderator: Ras

User avatar
Ovyron
Posts: 4562
Joined: Tue Jul 03, 2007 4:30 am

Re: Testposition-(Medium). Which engine can find it?

Post by Ovyron »

HIARCS Paderborn 2007 takes 1:09 to find it and 1:45 to realize that black is winning:

CPU0: AuthenticAMD x86 Family 6 Model 10 Stepping 0 2133 MHz
HIARCS Paderborn 2007 SP Copyright (c) 2008 Applied Computer Concepts Ltd. All rights reserved. www.hiarcs.com
Engine: HIARCS Paderborn 2007 SP (64 MB)
by Mark John Uniacke
11/25 0:00 +3.17-- 1...Nc5 (172.321) 289
11/25 0:00 +3.54 1...Nc5 2.Nf4 Kf7 3.d4 e5 4.h4 Rf5 5.dxe5 Rxe5 6.Ke2 Ne4 7.f3 Ng3+ 8.Kf2 (245.069) 293
11/25 0:00 +3.53 1...Kf7 (249.938) 293
11/25 0:00 +3.53 1...Kf7 2.Nf4 Nc7 3.d4 Rf5 4.Nd3 (261.364) 293
11/29 0:00 +3.52 1...e5 (282.680) 293
11/29 0:01 +3.33 1...e5 2.f4 Rf5 3.Ke2 e4 (326.211) 297
12/29 0:01 +3.35 1...e5 2.d4 e4 3.Nf4 Nc7 4.h4 Rg4 5.Ke2 Kf7 6.h5 Rg5 7.f3 exf3+ (467.067) 298
13/30 0:01 +3.42 1...e5 2.d4 e4 3.Nf4 Nc7 4.h4 Rg4 5.h5 Kf7 6.Ng6 Rg5 (613.631) 307
14/30 0:03 +3.39 1...e5 2.d4 e4 3.Nf4 Nc7 4.h4 Rg4 5.h5 Kf7 6.Ke2 Rg5 7.f3 exf3+ 8.Kxf3 Ne8 9.a3 (993.049) 315
14/30 0:04 +3.38 1...Nc7 (1.379.480) 314
14/30 0:06 +3.17++ 1...Nc7 (1.978.108) 311
14/32 0:14 +2.69 1...Nc7 2.Nh4 Ne8 (4.546.651) 309
15/32 0:18 +2.83 1...Nc7 2.e4 e5 3.Nh4 dxe4 4.dxe4 Ne8 5.Nf5 Rg6 6.Ke2 Nd6 7.Nxd6 Rxd6 8.Ke3 Rd8 9.c4 b6 10.b3 (5.643.592) 297
16/32 0:25 +2.79 1...Nc7 2.e4 e5 3.Nh4 dxe4 4.dxe4 Ne8 5.Nf5 Rg6 6.Ke2 Nd6 7.Nxd6 Rxd6 8.h4 Rd8 9.b3 b6 (7.237.784) 285
17/34 0:39 +2.94 1...Nc7 2.e4 e5 3.Nh4 dxe4 4.dxe4 Ne8 5.Nf5 Rg6 6.Ne7 Kxe7 7.Qxg8 Re6 (11.476.679) 293
18/37 1:01 +2.92 1...Nc7 2.e4 e5 3.Nh4 dxe4 4.dxe4 Ne8 5.Nf5 Rg6 6.Ne7 Kxe7 7.Qxg8 Re6 8.Qh7 Rf6 9.Ke2 Rf4 10.c4 (17.781.460) 289
18/39 1:09 +2.92 1...Rxg2 (20.340.826) 290
18/39 1:17 +2.69++ 1...Rxg2 (24.142.119) 309
18/39 1:45 -2.36 1...Rxg2 2.b4 Nb8 3.b5 Rh2 4.Kg1 Rxh3 5.d4 Nd7 6.a3 Kf7 7.Kg2 Rh5 (33.728.369) 318
19/39 1:54 -2.61++ 1...Rxg2 (36.767.791) 319
19/39 2:04 -2.76 1...Rxg2 2.b4 Nb8 3.b5 Rh2 4.Kg1 Nd7 5.d4 Kf7 6.Kxh2 Nf8 7.Kg2 Ng6 (40.406.057) 323
20/41 2:32 -2.96 1...Rxg2 2.b4 Nb8 3.b5 Rh2 4.d4 Rh1+ 5.Ke2 Ra1 6.a4 Rxa4 7.Kd3 Nd7 8.e4 Ra5 (49.614.211) 324
best move: Rg5xg2 time: 3:19.267 min n/s: 333.575 nodes: 66.393.681
Top
Jouni
Posts: 3907
Joined: Wed Mar 08, 2006 8:15 pm
Full name: Jouni Uski

Re: Testposition-(Medium). Which engine can find it?

Post by Jouni »

Easy position, because even Rybka finds it fast on slow PC:

1...e5 2.f4 Rh5 3.fxe5 Rxe5 4.Ke2 Kf7 5.Nf4 Nc7 6.d4 Rf5 7.h4
+- (2.60) Depth: 13 00:00:06 249kN
1...e5 2.f4 Rh5 3.fxe5 Rxe5 4.Ke2 Kf7 5.Nf4 Nc7 6.d4 Rf5 7.h4 Ne8
+- (2.60) Depth: 14 00:00:07 301kN
1...e5 2.f4 Rh5 3.fxe5 Rxe5 4.Ke2 Kf7 5.Nf4 Nc7 6.d4 Rf5 7.h4 Ne8 8.Ne6
+- (2.68) Depth: 15 00:00:10 419kN
1...e5 2.f4 Rh5 3.fxe5 Rxe5 4.Ke2 Kf7 5.Nf4 Nc7 6.d4 Rf5 7.h4 Ne8 8.Ne6
+- (2.60) Depth: 16 00:00:21 896kN
1...Nc7 2.h4 Rg4 3.f3 Rg3 4.Kf2 Rh3 5.a3 Kf7 6.c4 dxc4 7.dxc4 Rh1 8.c5
+- (2.02) Depth: 16 00:01:43 4625kN
1...Nc7 2.e4 dxe4 3.dxe4 Ra5 4.a3 Rb5 5.b4 Rb6 6.Nf4 Ra6 7.Ng6+ Kf7 8.Ne5+
+- (1.82) Depth: 17 00:02:15 6100kN
1...Nc7 2.e4 dxe4 3.dxe4 Ra5 4.a3 Rb5 5.b4 Rb6 6.Nf4 Ra6 7.Ng6+ Kf7 8.Ne5+
+- (1.80) Depth: 18 00:02:22 6451kN
1...Nc7 2.e4 dxe4 3.dxe4 Ra5 4.a3 Rb5 5.b4 Rb6 6.Nf4 Ra6 7.Ng6+ Kf7 8.Ne5+
+- (1.80) Depth: 19 00:02:38 7213kN
1...Nc7 2.e4 dxe4 3.dxe4 Ra5 4.a3 Rb5 5.b4 Rb6 6.Nf4 Ra6 7.Ng6+ Kf7 8.Ne5+
+- (1.80) Depth: 20 00:03:04 8424kN
1...Rxg2 2.Kxg2 Nb8 3.c4 dxc4 4.dxc4 Nc6 5.c5 Nd8 6.c6 bxc6
³ (-0.54) Depth: 20 00:03:22 9590kN
1...Rxg2 2.Kxg2 Nb8 3.c4 Nc6 4.cxd5 exd5 5.e4 dxe4 6.dxe4 Ne5 7.f4 Ng6 8.Qxg8+
-+ (-2.16) Depth: 20 00:05:42 17443kN

Jouni
Top
Vinvin
Posts: 5333
Joined: Thu Mar 09, 2006 9:40 am
Full name: Vincent Lejeune

Re: Testposition-(Medium). Which engine can find it?

Post by Vinvin »

0:13 -2.53 1...Rxg2 2.Kxg2 Nc5 3.f4 Nxd3 4.b3 Nc1
1:38 +2.14 1...Rxg2 2.b4 Rh2 3.Kg1 Rxh3 4.d4 Kf7

Amazing :)
Top
bob
Posts: 20943
Joined: Mon Feb 27, 2006 7:30 pm
Location: Birmingham, AL

Re: Testposition-(Medium). Which engine can find it?

Post by bob »

George Tsavdaris wrote:Black has one choice to win. All other moves lose.
1...Rxg2!! 2.Kxg2 Nb8! (or possibly 2...Nc5) and black wins, despite what engines say for a while in the beginning, until they realize the situation.

[d]5kbQ/pp4p1/n3p2p/3p2r1/8/2PPP2P/PP3PN1/5K2 b - - 0 1


One very interesting position that occurs in the same theme is the following too difficult endgame to judge:
[d]5kbQ/pp4p1/n3p2p/3p4/3P1P2/2P4P/PP3PK1/8 b - - 0 1
Black has a free Knight that can be a killer for the extra reason it can win some tempo in zugzwang positions that will occur, but white has a free King as also a Pawn more. Any verdict?
This does not look to be _that_ hard:

Code: Select all

               17    12.59   3.02   1. ... e5 2. d4 Kf7 3. Ne1 e4 4. Ng2
                                    Nc7 5. Nf4 Ne8 6. h4 Rf5 7. Ke2 Nf6
                                    8. Kd2 Ng4 9. Ke2 Rf6
               17    20.82   2.67   1. ... Nc7 2. Nh4 Nb5 3. Nf3 Rf5 4.
                                    Nd4 Nxd4 5. cxd4 Rf7 6. Ke2 Rc7 7.
                                    Kd2 Kf7 8. f4 Rc8 9. h4
               17    23.21     -1   1. ... Rxg2!!
               17    24.52     -3   1. ... Rxg2!!
               17    25.76     -M   1. ... Rxg2!!
               17    28.99  -2.17   1. ... Rxg2 2. Kxg2 Nb8 3. c4 Nc6 4.
                                    cxd5 exd5 5. f4 Nd8 6. e4 dxe4 7. dxe4
                                    Nf7 8. Qxg8+ Kxg8 9. e5 g6 10. e6 <HT>
               17->  29.09  -2.17   1. ... Rxg2 2. Kxg2 Nb8 3. c4 Nc6 4.
                                    cxd5 exd5 5. f4 Nd8 6. e4 dxe4 7. dxe4
                                    Nf7 8. Qxg8+ Kxg8 9. e5 g6 10. e6 <HT>
                                    (s=3)
               18    30.98  -2.11   1. ... Rxg2 2. Kxg2 Nb8 3. c4 Nc6 4.
                                    cxd5 exd5 5. f4 Nd8 6. e4 dxe4 7. dxe4
                                    Nf7 8. Qxg8+ Kxg8 9. e5 g6 10. e6 Nd6
                                    (s=2)
               18->  31.17  -2.11   1. ... Rxg2 2. Kxg2 Nb8 3. c4 Nc6 4.
                                    cxd5 exd5 5. f4 Nd8 6. e4 dxe4 7. dxe4
                                    Nf7 8. Qxg8+ Kxg8 9. e5 g6 10. e6 Nd6
               19    39.51     -1   1. ... Rxg2!!
               19     1:19  -2.50   1. ... Rxg2 2. Kxg2 Nb8 3. f4 Nc6 4.
                                    e4 dxe4 5. dxe4 Nd8 6. Kf3 Nf7 7. Qxg8+
                                    Kxg8 8. e5 Kh7 9. Ke3 Kg6 10. Ke4 Kh5
                                    11. a3 Kg6 <HT>
               19->   1:19  -2.50   1. ... Rxg2 2. Kxg2 Nb8 3. f4 Nc6 4.
                                    e4 dxe4 5. dxe4 Nd8 6. Kf3 Nf7 7. Qxg8+
                                    Kxg8 8. e5 Kh7 9. Ke3 Kg6 10. Ke4 Kh5
                                    11. a3 Kg6 <HT>
               20     1:25  -2.50   1. ... Rxg2 2. Kxg2 Nb8 3. f4 Nc6 4.
                                    e4 dxe4 5. dxe4 Nd8 6. Kf3 Nf7 7. Qxg8+
                                    Kxg8 8. e5 Kh7 9. Ke3 Kg6 10. Ke4 Kh5
                                    11. a3 Kg6 <HT>
               20->   1:26  -2.50   1. ... Rxg2 2. Kxg2 Nb8 3. f4 Nc6 4.
                                    e4 dxe4 5. dxe4 Nd8 6. Kf3 Nf7 7. Qxg8+
                                    Kxg8 8. e5 Kh7 9. Ke3 Kg6 10. Ke4 Kh5
                                    11. a3 Kg6 <HT>
               21     2:08  -2.69   1. ... Rxg2 2. Kxg2 Nb8 3. f4 Nc6 4.
                                    h4 Nd8 5. e4 Nf7 6. Qxg8+ Kxg8 7. Kf3
                                    Nd6 8. a4 g6 9. b4 Kg7 10. a5 dxe4+
                                    11. dxe4 Nc4 12. e5 b5
               21->   2:09  -2.69   1. ... Rxg2 2. Kxg2 Nb8 3. f4 Nc6 4.
                                    h4 Nd8 5. e4 Nf7 6. Qxg8+ Kxg8 7. Kf3
                                    Nd6 8. a4 g6 9. b4 Kg7 10. a5 dxe4+
                                    11. dxe4 Nc4 12. e5 b5
              time=2:14  mat=2  n=243735798  fh=94%  nps=1.8M
              ext-> check=7.1M 1rep=136K mate=2K pp=0 reduce=151.0M/33.9M
              predicted=0  evals=161.3M  50move=0  EGTBprobes=0  hits=0
              SMP->  splits=2578  aborts=229  data=6/128  elap=2:14
My core-2 2.0ghz laptop gets this in just over 5 seconds. I ran it on a dual-quad-core box and found it in under one second...
Top
F. Bluemers
Posts: 880
Joined: Thu Mar 09, 2006 11:21 pm
Location: Nederland

Re: Testposition-(Medium). Which engine can find it?

Post by F. Bluemers »

chesstiger 2007 update 3

Fonzy
Top
OliverBr
Posts: 865
Joined: Tue Dec 18, 2007 9:38 pm
Location: Munich, Germany
Full name: Dr. Oliver Brausch

Re: Testposition-(Medium). Which engine can find it?

Post by OliverBr »

OliThink found it in 16 seconds (on a 64bit OS it would last about 11 seconds)

Code: Select all

olithink -sd 13 "5kbQ/pp4p1/n3p2p/3p2r1/8/2PPP2P/PP3PN1/5K2 b - - 0 1"
 1  -204      0        49  a6c5
 2  -220      0       341  a6c5 f2f4
 3  -224      0      1002  a6c5 f2f4 g5g3
 4  -240      1      3527  a6c5 f2f4 g5g3 d3d4
 5  -220      3      9532  a6c5 h3h4 g5g4 f2f3 g4a4
 6  -224      4     31171  a6c5 g2f4 e6e5 f4d5 c5d3 b2b4
 7  -244     12    144511  a6c7 b2b4 b7b6 h3h4 g5g4 f2f3 g4g3
 8  -248     29    348267  a6c5 g2f4 e6e5 b2b4 e5f4 e3f4 g5f5 b4c5
 9  -248     59    722030  a6c5 g2f4 e6e5 b2b4 e5f4 e3f4 g5f5 b4c5 f5f4
10  -252    139   1802615  a6c5 g2f4 e6e5 d3d4 e5f4 e3f4 g5f5 d4c5 f8f7 b2b4
11  -240    640   9887131  a6c7 a2a4 a7a5 d3d4 b7b5 a4b5 c7b5 h3h4 g5g4 f2f3 g4g3
12    62   1635  24586894  g5g2 f1g2 a6b8 f2f4 b8c6 c3c4 c6d8 c4d5 e6d5 e3e4 d8f7 h8g8 f8g8
13    62   1826  27986970  g5g2 f1g2 a6b8 f2f4 b8c6 c3c4 c6d8 c4d5 e6d5 e3e4 d8f7 h8g8 f8g8 e4d5
1. ... g5g2

kibitz W: 62 Nodes: 21420294 QNodes: 6566676 Evals: 6293748 cs: 1826 knps: 1532
Top
sorcerers_apprentice

Re: Testposition-(Medium). Which engine can find it?

Post by sorcerers_apprentice »

Hi all,
Very interesting position. Black's King, Bishop and two pawns are rendered immobile, as is the white Queen. So after the exchange sacrifice it's King and 6 pawns vs. Knight and 4 pawns!!. I remember reading somewhere that the estimated value of the King as an attacking piece in the endgame is about 4 pawns; that certainly shows in this position.

Anyway. Here's what old Shredder Classic thinks of it; it took it some bit of time to come up with a winning line for Black:

Code: Select all


5kbQ/pp4p1/n3p2p/3p2r1/8/2PPP2P/PP3PN1/5K2 b - -

Engine: Shredder Classic 1.3 (512 MB)
by Stefan Meyer-Kahlen

11/22  0:01   -4.72++  1...e5 (1.076.361) 683 

11/22  0:02   -4.71    1...e5 2.f4 Rf5 3.Ke2 e4 4.d4 (1.458.300) 682 

11/22  0:02   -4.70++  1...Kf7 (1.641.234) 683 

11/22  0:02   -4.59    1...Kf7 2.Nf4 Nc7 3.d4 e5 4.Nd3 e4 
                       5.Ne5+ Kf8 6.c4 dxc4 7.Nxc4 (1.977.237) 677 

12/24  0:04   -4.69    1...Kf7 2.d4 e5 3.h4 Rf5 4.Ke2 exd4 
                       5.cxd4 Nb4 6.a3 Nc2 7.f3 b6 (2.949.971) 690 

13/25  0:07   -4.71    1...Kf7 2.d4 e5 3.h4 Rf5 4.Ke2 Nc7 
                       5.f3 exd4 6.cxd4 Ne6 7.b4 b5 8.Ne1 (5.094.485) 696 

13/25  0:09   -4.70++  1...e5 (6.436.814) 694 

13/26  0:12   -4.61    1...e5 2.f4 Rf5 3.Ke2 e4 4.d4 Rf6 
                       5.Nh4 Nc7 6.b3 Nb5 7.Kd2 Kf7 (8.720.742) 692 

13/26  0:13   -4.60++  1...Nc5 (9.576.377) 691 

13/26  0:14   -4.52    1...Nc5 2.Nf4 Kf7 3.Ke2 Rg1 4.d4 Na4 
                       5.Nxe6 Nxb2 6.Nd8+ Kf8 7.Nxb7 Nd1 
                       8.a3 (10.350.621) 696 

14/28  0:19   -4.67    1...Nc5 2.Nf4 Kf7 3.h4 Rg4 4.d4 Rxf4 
                       5.exf4 Nd3 6.f5 exf5 7.b3 Nf4 8.a4 (13.751.213) 697 

14/29  0:24   -4.66++  1...e5 (17.161.511) 701 

14/29  0:26   -4.65    1...e5 2.f4 Rf5 3.Ke2 e4 4.d4 Nb8 
                       5.Nh4 Rf6 6.a4 (19.067.760) 713 

14/29  0:31   -4.64++  1...Nc7 (22.304.872) 718 

14/29  0:45   -4.29++  1...Nc7 (32.056.454) 710 

14/29  0:56   -4.09    1...Nc7 2.Nh4 Nb5 3.f4 Rg3 4.f5 (40.298.785) 710 

14/29  0:58   -4.08++  1...Rxg2 (41.720.837) 712 

14/29  0:58   -3.99    1...Rxg2 2.Kxg2 Nc5 3.f4 Nxd3 4.b3 Nc1 
                       5.c4 Nxa2 6.cxd5 exd5 7.Kf3 Nc3 8.f5 (42.045.063) 713 

15/24  1:00   -3.75    1...Rxg2 2.Kxg2 Nc5 3.f4 Nxd3 4.b4 Nc1 
                       5.a3 Na2 6.Kf1 (43.199.694) 716 

15/30  1:20   -3.74++  1...Nc7 (57.484.039) 710 

15/30  1:24   -3.73    1...Nc7 2.Nh4 e5 3.d4 exd4 4.cxd4 Ne8 
                       5.Nf3 Rh5 6.Ne5 Rxh3 7.Ng6+ Kf7 
                       8.Kg2 Rh5 9.Ne5+ (61.013.826) 720 

16/30  1:56   -3.84    1...Nc7 2.Nh4 e5 3.e4 dxe4 4.dxe4 Ne6 
                       5.Nf5 Nd8 6.Ne7 Kxe7 7.Qxg8 g6 (83.666.071) 719 

16/30  2:06   -3.83++  1...Rxg2 (90.641.956) 719 

16/30  2:09   -3.48++  1...Rxg2 (93.652.013) 722 

16/30  2:40   +2.01    1...Rxg2 2.b4 Rh2 3.Kg1 Nc7 4.c4 Rxh3 
                       5.Kg2 Rh4 6.cxd5 Nxd5 7.a3 Kf7 8.Kg3 Rh1 
                       9.d4 (116.839.306) 728 

17/30  3:05   +2.26++  1...Rxg2 2.b4 (135.329.270) 730 

17/33  3:30   +2.49    1...Rxg2 2.Kxg2 Nb8 3.f4 Nc6 4.c4 dxc4 
                       5.dxc4 Nd8 6.Kf3 Nf7 7.Qxg8+ Kxg8 
                       8.b3 Nd6 9.e4 (154.468.886) 733 

best move: Rg5xg2 time: 3:40.008 min  n/s: 739.219  CPU 99.2%   n/s(1CPU): 745.180  nodes: 162.622.395
And this is what HIARCS 12 thinks of it (all default settings: combinations=off, sharpen PV=off, etc.):

Code: Select all


5kbQ/pp4p1/n3p2p/3p2r1/8/2PPP2P/PP3PN1/5K2 b - -

Engine: HIARCS 12 MP (512 MB)
by Mark John Uniacke

12/29  0:01   -3.31    1...Nc5 2.Nf4 Rf5 3.Ke2 Rf6 4.b3 Kf7 
                       5.f3 b5 6.h4 Rf5 7.h5 (902.905) 460 

13/30  0:02   -3.34    1...Nc5 2.Nf4 Rf5 3.Ke2 Rf6 4.b3 Kf7 
                       5.f3 b5 6.h4 a5 7.d4 Nd7 (1.172.227) 475 

14/30  0:02   -3.33    1...Nc5 2.Nf4 Rf5 3.Ke2 Rf6 4.b3 Kf7 
                       5.f3 b5 6.h4 a5 7.d4 Nd7 8.b4 a4 (1.543.018) 520 

15/31  0:03   -3.39    1...Nc5 2.Nf4 Rf5 3.Ke2 Rf6 4.b3 b5 
                       5.f3 Nd7 6.d4 g5 7.Nh5 (2.176.746) 557 

15/31  0:06   -3.38    1...e5 (3.304.282) 534 

15/31  0:07   -3.35    1...e5 2.d4 Nc7 3.dxe5 Rxe5 4.Nf4 Kf7 
                       5.h4 Ne6 6.Ng6 Rf5 7.f4 Kxg6 8.Qxg8 (3.735.956) 517 

16/32  0:11   -3.22    1...e5 2.d4 e4 3.h4 Rg4 4.h5 Nc7 
                       5.Nf4 Kf7 6.Ke2 Rh4 (6.385.298) 558 

16/32  0:12   -3.21    1...Rxg2 (6.891.601) 570 

16/32  0:12   -3.10++  1...Rxg2 (7.267.081) 579 

16/39  0:23   +1.82    1...Rxg2 2.d4 Rh2 3.Kg1 Rxh3 4.Kg2 Rh4 
                       5.a4 Rg4+ 6.Kf1 Nc7 7.f3 (16.896.698) 733 

17/39  0:26   +2.07++  1...Rxg2 (19.755.534) 747 

17/39  0:37   +2.75    1...Rxg2 2.b4 Rh2 3.a4 Nb8 4.b5 Nd7 
                       5.d4 Nb6 6.a5 Na4 7.b6 axb6 (29.420.854) 777 

18/39  0:40   +2.78    1...Rxg2 2.b4 Rh2 3.a4 Nb8 4.b5 Nd7 
                       5.d4 Nb6 6.a5 Na4 7.Kg1 Rxh3 (31.873.969) 784 

19/39  0:47   +3.03++  1...Rxg2 (38.188.928) 804 

19/39  0:52   +3.03    1...Rxg2 2.Kxg2 Nb8 3.c4 dxc4 (42.868.842) 809 

20/41  0:59   +3.03    1...Rxg2 2.Kxg2 Nb8 3.c4 dxc4 4.dxc4 Nc6 (49.526.151) 831 

best move: Rg5xg2 time: 1:10.825 min  n/s: 854.560  nodes: 60.336.000
Cheers !
Top
Guetti

Re: Testposition-(Medium). Which engine can find it?

Post by Guetti »

No idea why the "professional" engines need depth 16 to find Rxg2. With only material evaluation my toy selects Rxg2 at depth 10; it wants to get on a pawn hunt before it switches to go after the queen.
Probably advanced search and eval stuff is a drawback in this position.

Code: Select all

Juhui, searching!
Trying 120s
 ply   score     time       nodes    pv
   1    -214        0          30    a6c5 
   2    -224        0         128    a6c5 g2f4 
   3    -213        0         926    a6c5 d3d4 c5e4 
   4    -223        2        3656    a6c5 d3d4 c5e4 g2f4 
   5    -225       10       22950    a6c5 g2f4 g5f5 f4g6 f8f7 d3d4 
   6    -234       18       76381    e6e5 f2f4 e5f4 g2f4 a6c5 d3d4 
   7    -227       35      341695    g5f5 g2f4 f5f6 d3d4 a6b8 
   8    -223       73      890608    a6c5 g2f4 e6e5 d3d4 e5f4 e3f4 g5f5 d4c5 f5f4 
   9    -223      141     1965744    a6c5 g2f4 e6e5 d3d4 e5f4 e3f4 g5f5 d4c5 f5f4 
  10    -206      420     6049780    g5g2 f1g2 a6c5 f2f4 c5d3 b2b3 d3c1 c3c4 c1a2 g2g1 
  11    -205      903    13456390    g5g2 f1g2 a6c5 f2f4 c5d3 b2b3 d3c1 c3c4 c1a2 c4d5 e6d5 
>>12    -155     1741    26409480    g5g2 *
  12      96     3745    55365076    g5g2 f1g2 a6b8 e3e4 b8c6 e4d5 e6d5 f2f4 c6d8 c3c4 d8f7 h8g8 f8g8 c4d5 
  13      96     8396   125381313    g5g2 f1g2 a6b8 e3e4 b8c6 e4d5 e6d5 f2f4 c6d8 c3c4 d8f7 h8g8 f8g8 c4d5 

13 (20/20) 0.96 120.01  180246277  g5g2 f1g2 a6b8 e3e4 b8c6 e4d5 e6d5 f2f4 c6d8 c3c4 d8f7 h8g8 f8g8 c4d5 
Score: 0.96  Time: 120.01  Ply: 13/29
Nodes: 180246277  NPS: 1501k
Qnodes:  33045666  18.33% of Nodes
Top
OliverBr
Posts: 865
Joined: Tue Dec 18, 2007 9:38 pm
Location: Munich, Germany
Full name: Dr. Oliver Brausch

Re: Testposition-(Medium). Which engine can find it?

Post by OliverBr »

Guetti wrote:No idea why the "professional" engines need depth 16 to find Rxg2. With only material evaluation my toy selects Rxg2 at depth 10; it wants to get on a pawn hunt before it switches to go after the queen.
Probably advanced search and eval stuff is a drawback in this position.
It's because they prune, razor and nullmove like maniacs. In standard positions this might work well but in extraordinary position like a trapped queen (!) it's a flaw.
Pruning always has the danger that something is overlooked but it's is generally no problem, because such positions are rare and things will be seen in greater depth nontheless.

OliThink does not prune at all in standard search, so ply 12 already delivers the right move.

BTW, what "tool" are you using?
Top
User avatar
George Tsavdaris
Posts: 1627
Joined: Thu Mar 09, 2006 12:35 pm

Re: Testposition- 2 more.

Post by George Tsavdaris »

bob wrote: My core-2 2.0ghz laptop gets this in just over 5 seconds. I ran it on a dual-quad-core box and found it in under one second...
Crafty is too clever for this position.

Perhaps the following 2 positions, always in the same theme, can give it some trouble, although i doubt. Crafty is too clever. :D

1...Rxg2! (or Nxg2 but this has some problems) wins again with 2.Nxg2 Nxg2! 3.Kxg2 Nb8!
After 1...Nxd3?? white is much better with 2.f4.
[d]5kbQ/pp4p1/n3p2p/3p2r1/7N/2PPP2P/PP3PP1/4n2K b - - 0 1

Only 1...a5! **draws(the line that draws for white is just a single one between hundreds of possible replies). All other moves lose.
[d]5kbQ/pp3qp1/4p2p/3p4/8/4P1RP/1P3P2/6K1 b - - 0 1

*The source i had must be wrong. As i see 1...b6! or 1...a6! must draw too. But which engines will play any of these moves?
After his son's birth they've asked him:
"Is it a boy or girl?"
YES! He replied.....
Top

Return to “Computer Chess Club: General Topics”