Old test position

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

Moderators: hgm, Rebel, chrisw

Jouni
Posts: 3293
Joined: Wed Mar 08, 2006 8:15 pm

Old test position

Post by Jouni »

[D]6k1/7R/6P1/7P/8/6r1/8/K7 w - - 0 1

Here Ka2 is only winning move (why? I have no idea :) ). Surprisingly SF 2.2.2 finds it fast and in 2 best mode after 20 second has:

Analysis by Stockfish 2.2.2 JA:

1. +- (4.24): 1.Ka2 Rh3 2.Kb2 Rg3 3.Kc2 Rh3 4.Kd2 Ra3 5.Rd7 Rh3 6.Rd5 Kg7 7.Ra5 Kh6 8.Rg5 Kg7 9.Kc2 Rh4 10.Rd5 Rh3 11.Kd2 Rh4 12.Ke3 Kh6 13.Kf3 Rh1 14.Rb5 Rf1+ 15.Ke3 Rh1 16.Kf2 Rh3 17.Kg2 Rh4 18.Kf3 Rh2 19.Rf5 Ra2 20.Ke3 Ra3+ 21.Kd4 Ra4+ 22.Kd3 Rh4 23.Rd5 Rh3+ 24.Kd4 Rh4+ 25.Kc5 Rh1 26.Rf5 Ra1 27.Kb4 Rc1 28.Kb3 Rc7 29.Kb4 Rc8 30.Rc5 Rxc5 31.Kxc5 Kg7 32.Kd5 Kh6 33.Ke4 Kg7 34.Kf5 Kh6 35.Kf6
2. = (0.00): 1.Kb2 Rh3 2.Kc1 Rc3+ 3.Kd1 Rd3+ 4.Ke2 Rg3 5.Kf2 Rh3 6.Kg2 Re3 7.h6 Re5 8.Kf3 Rg5 9.Rg7+ Kh8 10.Ke4 Re5+ 11.Kd4 Rd5+ 12.Kc4 Rc5+ 13.Kxc5

After 30 minutes it has (1 best)

56/93- 26:31 4,351,221,742 2,733,373 +8.36 1.Ka1-a2 Rg3-h3 2.Ka2-b2 Rh3-g3 3.Kb2-c2 Rg3-h3 4.Kc2-d2 Rh3-g3 5.Kd2-e2 Rg3-h3 6.Ke2-f2 Rh3-c3 7.Rh7-e7 Rc3-c2+ 8.Re7-e2 Rc2-c3 9.Re2-e5 Kg8-g7 10.Re5-e7+ Kg7-g8 11.Re7-e3 Rc3-c6 12.Kf2-f3 Kg8-f8 13.Re3-e5 Rc6-c1 14.h5-h6 Rc1-h1 15.g6-g7+ Kf8-f7 16.Re5-e7+ Kf7xe7 17.g7-g8Q Rh1-f1+ 18.Kf3-g2 Rf1-d1 19.Qg8-g5+ Ke7-d6 20.Qg5-d8+

Impressive or bug!?

Jouni
Top
User avatar
Guenther
Posts: 4611
Joined: Wed Oct 01, 2008 6:33 am
Location: Regensburg, Germany
Full name: Guenther Simon

Re: Old test position (6men TBS position)

Post by Guenther »

Jouni wrote:
Impressive or bug!?

Jouni
It must be named Impressive or luck!?
(bug makes no sense at all...)

Guenther
Top
Jouni
Posts: 3293
Joined: Wed Mar 08, 2006 8:15 pm

Re: Old test position (6men TBS position)

Post by Jouni »

Sometimes SF displays +9 in draw position. But here's position after main line:

[D]3Q4/8/3k3P/8/8/8/6K1/3r4 b - - 0 20

Sure this is win! I am surprised because other programs with 5 piece TB has difficulties.

Jouni
Top
zullil
Posts: 6442
Joined: Tue Jan 09, 2007 12:31 am
Location: PA USA
Full name: Louis Zulli

Re: Old test position

Post by zullil »

Jouni wrote:[D]6k1/7R/6P1/7P/8/6r1/8/K7 w - - 0 1

56/93- 26:31 4,351,221,742 2,733,373 +8.36 1.Ka1-a2 Rg3-h3 2.Ka2-b2 Rh3-g3 3.Kb2-c2 Rg3-h3 4.Kc2-d2 Rh3-g3 5.Kd2-e2 Rg3-h3 6.Ke2-f2 Rh3-c3 7.Rh7-e7 Rc3-c2+ 8.Re7-e2 Rc2-c3 9.Re2-e5 Kg8-g7 10.Re5-e7+ Kg7-g8 11.Re7-e3 Rc3-c6 12.Kf2-f3 Kg8-f8 13.Re3-e5 Rc6-c1 14.h5-h6 Rc1-h1 15.g6-g7+ Kf8-f7 16.Re5-e7+ Kf7xe7 17.g7-g8Q Rh1-f1+ 18.Kf3-g2 Rf1-d1 19.Qg8-g5+ Ke7-d6 20.Qg5-d8+

Impressive or bug!?

Jouni
I doubt what you are seeing is Stockfish finding a win.

What is the score at the completion of your depth 57 iteration? I ask because here's what I see:

Code: Select all

info depth 51 seldepth 81 score cp 606 nodes 1091235501 nps 7372166 time 148021 multipv 1 pv a1a2 g3d3 h7f7 d3g3 f7f5 g8g7 a2b2 g3h3 f5e5 h3g3 e5b5 g7h6 b2c2 h6g7 b5c5 g7g8 c2d1 g3g2 d1e1 g2h2 e1f1 g8g7 c5f5 h2h4 f1g1 g7h6 g1f2 h4a4 f5d5 a4a7 f2e2 a7a1 e2e3 a1a4 d5e5 a4a8 e3d2 a8a3 e5f5 a3a8 d2e2 a8a3 e2d2
info depth 52 currmove a1a2 currmovenumber 1
info depth 52 currmove h7b7 currmovenumber 4
info depth 52 seldepth 81 score cp 597 upperbound nodes 1159622305 nps 7317660 time 158469 multipv 1 pv a1a2 g3h3 a2b2 h3g3 b2a2
info depth 52 currmove a1a2 currmovenumber 1
info depth 52 currmove a1b2 currmovenumber 5
info depth 52 currmove h7a7 currmovenumber 9
info depth 52 currmove g6g7 currmovenumber 10
info depth 52 currmove h7h6 currmovenumber 14
info depth 52 seldepth 81 score cp 589 upperbound nodes 1160633782 nps 7314304 time 158680 multipv 1 pv a1a2 g3h3 a2b2 h3g3 b2a2
info depth 52 currmove a1a2 currmovenumber 1
info depth 52 currmove h7c7 currmovenumber 3
info depth 52 currmove h7g7 currmovenumber 9
info depth 52 currmove h7h6 currmovenumber 14
info depth 52 seldepth 81 score cp 577 upperbound nodes 1161896252 nps 7310696 time 158931 multipv 1 pv a1a2 g3h3 a2b2 h3g3 b2a2
info depth 52 currmove a1a2 currmovenumber 1
info depth 52 currmove h7f7 currmovenumber 6
info depth 52 currmove h7a7 currmovenumber 9
info depth 52 currmove h7b7 currmovenumber 10
info depth 52 currmove h7c7 currmovenumber 11
info depth 52 currmove h7e7 currmovenumber 12
info depth 52 seldepth 81 score cp 559 upperbound nodes 1162567038 nps 7308157 time 159078 multipv 1 pv a1a2 g3h3 a2b2 h3g3 b2a2
info depth 52 currmove a1a2 currmovenumber 1
info depth 52 currmove g6g7 currmovenumber 6
info depth 52 currmove h7c7 currmovenumber 9
info depth 52 currmove h7g7 currmovenumber 10
info depth 52 currmove h7h6 currmovenumber 11
info depth 52 currmove h7d7 currmovenumber 12
info depth 52 seldepth 81 score cp 532 upperbound nodes 1164282722 nps 7304524 time 159392 multipv 1 pv a1a2 g3h3 a2b2 h3g3 b2a2
info depth 52 currmove a1a2 currmovenumber 1
info depth 52 currmove g6g7 currmovenumber 4
info depth 52 currmove a1b2 currmovenumber 11
info depth 52 currmove h7b7 currmovenumber 12
info depth 52 seldepth 81 score cp 491 upperbound nodes 1166169538 nps 7304584 time 159649 multipv 1 pv a1a2 g3h3 a2b2 h3g3 b2a2
info depth 52 currmove a1a2 currmovenumber 1
info depth 52 currmove h7h8 currmovenumber 6
info depth 52 currmove h7f7 currmovenumber 7
info depth 52 currmove h7b7 currmovenumber 8
info depth 52 currmove h7h6 currmovenumber 9
info depth 52 currmove g6g7 currmovenumber 10
info depth 52 currmove a1b2 currmovenumber 11
info depth 52 currmove h5h6 currmovenumber 12
info depth 52 currmove h7c7 currmovenumber 13
info depth 52 currmove h7g7 currmovenumber 14
info depth 52 seldepth 81 score cp 430 upperbound nodes 1168731048 nps 7302332 time 160049 multipv 1 pv a1a2 g3h3 a2b2 h3g3 b2a2
info depth 52 currmove a1a2 currmovenumber 1
info depth 52 currmove g6g7 currmovenumber 4
info depth 52 currmove h7c7 currmovenumber 13
info depth 52 currmove h7e7 currmovenumber 14
info depth 52 seldepth 81 score cp 338 upperbound nodes 1171797617 nps 7300010 time 160520 multipv 1 pv a1a2 g3h3 a2b2 h3g3 h7d7 g3g5 d7d8 g8g7 b2c3 g5h5 d8d7 g7g6 c3b2 h5h2 b2c3 h2h5
info depth 52 currmove a1a2 currmovenumber 1
info depth 52 currmove h7h6 currmovenumber 6
info depth 52 currmove h7f7 currmovenumber 7
info depth 52 currmove h7b7 currmovenumber 8
info depth 52 seldepth 81 score cp 202 upperbound nodes 1175703936 nps 7297704 time 161106 multipv 1 pv a1a2 g3h3 a2b2 h3g3 b2a2
info depth 52 currmove a1a2 currmovenumber 1
info depth 52 currmove h7g7 currmovenumber 5
info depth 52 currmove h7a7 currmovenumber 9
info depth 52 currmove h7c7 currmovenumber 10
info depth 52 currmove h7e7 currmovenumber 11
info depth 52 seldepth 81 score cp 0 nodes 1210506659 nps 7340407 time 164910 multipv 1 pv a1a2 g3h3 a2b2 h3g3 b2a2
and later

Code: Select all

info depth 64 seldepth 83 score cp 0 nodes 3812913798 nps 7773318 time 490513 multipv 1 pv a1a2 g3h3 a2b2 h3g3 b2a2
and later still

Code: Select all

info depth 66 seldepth 98 score cp 0 nodes 10029017918 nps 7608681 time 1318102 multipv 1 pv a1b2 g3h3 b2a2 h3g3 a2b1 g3g5 h7d7 g5h5 d7d8 g8g7 d8a8 g7g6 a8g8 g6h6 g8b8 h5h1 b1b2 h6h7 b8c8 h1h3 c8a8 h7g7 a8c8 g7f7 c8a8 h3h1 a8d8 h1h5 d8c8 h5h2 b2b3 h2h1 c8d8 f7g7 b3c3 h1h3 c3d4 h3h4 d4d3 g7h7 d8a8 h4g4 a8a6 g4g5 d3e4 h7g8 a6a8 g8g7 a8a6 g7g8
and one more. Note that SF is back to Kb2

Code: Select all

info depth 70 seldepth 98 score cp 0 nodes 15451666610 nps 8069636 time 1914791 multipv 1 pv a1b2 g3h3 b2c1 h3c3 c1d2 c3h3 d2c2 h3g3 h7a7 g3g5 c2d3 g5h5 a7a8 g8g7 d3d4 g7g6 a8d8 g6h7 d4d3 h7h6 d3c3 h6g5 d8c8 g5h6 c8d8
Last edited by zullil on Fri Feb 10, 2012 3:08 pm, edited 1 time in total.
Top
Michel
Posts: 2272
Joined: Mon Sep 29, 2008 1:50 am

Re: Old test position

Post by Michel »

GnuChess also switches quickly to Ka2 so there must be some obvious advantage in this move.

Code: Select all

 28	+3.71	200.9M	2:09.13	Ka2 Rh3 Kb2 Rg3 Kc2 Rf3 Rc7 Rg3 Rc3 Rg2+ Kd3 Rg5 Rc8+ Kg7 Rc7+ Kf6 Rf7+ Ke6 Rf8 Ke7 Rh8 Kf6 Ke3 Re5+ Kf4 Re7 Rc8 Rd7 Rc5 Rb7 Rc4 Rd7 Ke4 
 27	+3.64	164.7M	1:45.23	Ka2 Rh3 Kb2 Rg3 Kc2 Rf3 Rc7 Rg3 Rc3 Rg2+ Kb3 Kg7 Rc7+ Kf6 Rc5 Kg7 Rf5 Rg3+ Kb4 Rg4+ Kc3 Rg3+ Kd4 Kh6 Ke4 Rg4+ Kd3 Rb4 Kc3 Rg4 Re5 Ra4 Kd3 
 26	+3.64	149.7M	1:35.13	Ka2 Rh3 Kb2 Rg3 Kc2 Rf3 Rc7 Rg3 Rc3 Rg2+ Kb3 Kg7 Rc7+ Kf6 Rc5 Kg7 Kc4 Kh6 Kd3 Rg3+ Ke4 Rg4+ Kf3 Rb4 Ra5 Rb3+ Ke4 Rb4+ Kd3 Rg4 Ke3 Rc4 
 25	+3.63	123.5M	1:17.88	Ka2 Rh3 Kb2 Rg3 Kc2 Rf3 Rc7 Rg3 Rc3 Rg2+ Kb3 Kg7 Rc5 Kh6 Kc3 Kg7 Kd4 Rg4+ Ke3 Kf6 Kf3 Rg1 Ra5 Rf1+ Ke4 Re1+ Kd3 Rd1+ Kc3 Rd8 Kc4 Kg7 
 24	+3.64	101.3M	1:03.49	Ka2 Rh3 Kb2 Rg3 Kc2 Rf3 Rc7 Rg3 Rc3 Rg2+ Kb3 Kg7 Rc5 Kh6 Kc3 Rg3+ Kd4 Rg1 Kd3 Rd1+ Kc4 Rd8 Re5 Rf8 Kd3 Kg7 Kd4 
 24	+3.39	43.3M	0:27.36	:-( 
 23	+3.64	24.1M	0:15.42	Ka2 Rg5 Kb3 Rc5 Kb2 Rc4 Rd7 Rh4 Rd8+ Kg7 Rd5 Rh2+ Kc3 Rg2 Kd4 Kh6 Rf5 Rg4+ Kd3 Rb4 Ke3 Ra4 Re5 Rb4 Ra5 Rb8 Kd3 Rb4 
 22	+3.65	18.2M	0:11.80	Ka2 Rg5 Kb3 Rc5 Kb2 Rc4 Rd7 Rh4 Rd8+ Kg7 Rd5 Rh2+ Kc3 Rg2 Kd4 Kh6 Rf5 Rg4+ Kd3 Rb4 Ke3 Ra4 Re5 Ra3+ Ke4 Ra4+ Kd5 Rf4 Ke6 
 21	+3.65	14.8M	0:09.65	Ka2 Rg5 Kb3 Rc5 Kb2 Rc4 Rd7 Rh4 Rd8+ Kg7 Rd5 Rh2+ Kc3 Rg2 Kd4 Kh6 Rf5 Rg4+ Kd3 Rb4 Ke3 Ra4 Rc5 Rb4 Kd3 Ra4 
 20	+3.66	11.1M	0:07.41	Ka2 Rg5 Kb3 Rc5 Kb2 Rc4 Rd7 Rh4 Rd8+ Kg7 Rd5 Rh3 Rg5 Rh2+ Kc3 Rh1 Kd3 Rd1+ Kc4 Rd8 Rd5 Re8 Kd4 
 19	+3.64	5.9M	0:04.04	Ka2 Rg5 Kb3 Rc5 Kb2 Rc4 Ra7 Rh4 Ra5 Kg7 Kc3 Kh6 Rf5 Rg4 Kd3 Rb4 Ra5 Rf4 Ke3 
 18	+3.72	2.5M	0:01.66	Ka2 Re3 Kb2 Rg3 Kc2 Re3 Kd2 Ra3 Ke2 Rg3 Kf2 Ra3 Rf7 Rh3 Rf5 Ra3 Ke2 Ra2+ Ke3 Ra4 
 17	+3.67	1.8M	0:01.20	Ka2 Re3 Kb2 Rg3 Kc2 Re3 Kd2 Ra3 Ke2 Rg3 Kf2 Ra3 Rf7 Rh3 Rf5 Rc3 Ke2 Rc7 h6 
 16	+3.81	1.1M	0:00.80	Ka2! 
 16	+3.63	1.4M	0:00.99	Ka2 Re3 Kb2 Rg3 Kc2 Re3 Kd2 Ra3 Rf7 Rg3 Rf5 Rh3 Ke2 Rg3 Rd5 Kg7 Rd3 Rg4 
 15	+3.56	807387	0:00.59	Ka2 Re3 Kb2 Rg3 Kc2 Re3 Ra7 Rg3 Kd2 Rg5 Rh7 Rg3 Ke2 Rg5 Ke3 Rd5 
 14	+3.68	655418	0:00.48	Ka2 Re3 Kb2 Rg3 Kc2 Re3 Kd2 Rh3 Ke2 Rg3 Kf2 Rd3 Ra7 Rh3 Ra5 
 13	+3.55	550193	0:00.41	Ka2 Re3 Kb2 Rg3 Kc2 Re3 Rc7 Rg3 Rc8+ Kg7 Rc5 Rg4 Kd2 Re4 
 13	+3.47	309707	0:00.23	:-( 
 12	+3.72	169547	0:00.13	Kb2 Re3 Kc2 Rh3 h6 Rg3 Rg7+ Kh8 Kd2 Rd3+ Ke1 Re3+ Kf2 Rf3+ Kg2 Rc3 
 11	+3.83	105291	0:00.08	Kb2 Re3 Kc2 Rh3 h6 Rg3 Rg7+ Kh8 Kd2 Rd3+ Ke1 Re3+ Kf2 Re5 
 10	+3.87	79157	0:00.07	Kb2 Re3 Kc2 Rg3 Kd2 Rg4 Kc3 Rg3+ Kc4 Rg4+ Kd5 Rg2 
  9	+3.89	56314	0:00.05	Kb2 Re3 h6 Rg3 Rg7+ Kh8 Kc2 Rc3+ Kd2 Rd3+ Ke2 Rc3 
  8	+3.89	31407	0:00.03	Kb2 Kf8 Kc2 Kg8 Kd2 Kf8 Rc7 Ke8 
  7	+3.75	14679	0:00.02	Kb2 Rg1 Kb3 Rg4 Kc2 Rg3 Rc7 
  6	+3.67	6547	0:00.00	Kb2 Rg1 Kb3 Rg5 Kc4 Re5 
  5	+3.66	5422	0:00.00	Kb2 Rg1 Kb3 Rb1+ Kc4 Re1 
  5	+3.59	3800	0:00.00	Kb2! 
  5	+3.26	2767	0:00.00	Kb1 Rg2 Rc7 Rh2 Rc8+ Kg7 
  4	+3.34	790	0:00.00	Rc7 Rg2 Re7 Rf2 
  3	+3.41	417	0:00.00	Rc7 Rg2 Re7 
  3	+3.40	287	0:00.00	Re7 Rg5 Re8+ Kg7 
  2	+3.46	67	0:00.00	Re7 Rg2 
  1	+3.52	20	0:00.00	Re7
Top
User avatar
JuLieN
Posts: 2949
Joined: Mon May 05, 2008 12:16 pm
Location: Bordeaux (France)
Full name: Julien Marcel

Re: Old test position

Post by JuLieN »

Prédateur switches nearly instantly to Ka2 too.... (although in my case that certainly must be a bug! :lol: )

But even if I let it search for an hour and a half its score won't increase: still +2.07, and its mainline doesn't looks so bright to me...

Code: Select all

FEN: 6k1/7R/6P1/7P/8/6r1/8/K7 w - - 0 1 

Predateur Bureau:
  3/4	00:00	         398	0	+2,05	Th6 Tg2 Th7
  3/6	00:00	         711	44 000	+2,06	Td7 Tg2 Rb1
  4/6	00:00	       1 478	92 000	+2,06	Td7 Tg2 Rb1 Rh8
  5/8	00:00	       2 913	93 000	+2,06	Td7 Tg2 Rb1 Rh8 Ta7
  5/10	00:00	       4 696	151 000	+2,07	Ra2 Tg2+ Rb3 Rf8 Tf7+ Rg8 Rc3
  6/8	00:00	       5 458	176 000	+2,07	Ra2 Tg2+ Rb3 Rf8 Tf7+ Rg8 Rc3 Rh8
  7/11	00:00	       9 480	305 000	+2,07	Ra2 Tg2+ Rb3 Rf8 Tf7+ Rg8 Rc3 Rh8 Tc7
  8/13	00:00	      19 087	615 000	+2,07	Ra2 Tg2+ Rb3 Rf8 Tf7+ Rg8 Rc3 Rh8 Tf8+ Rg7 Tf7+ Rh6
  9/16	00:00	      31 876	678 000	+2,07	Ra2 Tg2+ Rb3 Rf8 Tf7+ Rg8 Rc3 Rh8 Tf8+ Rg7 Tf7+ Rg8 Rb4 Td2
 10/18	00:00	      51 006	1 085 000	+2,07	Ra2 Tg2+ Rb3 Rf8 Tf7+ Rg8 Rc3 Rh8 Tf8+ Rg7 Tf7+ Rg8 Rd4 Th2 Th7
 11/19	00:00	     100 437	712 000	+2,06	Ra2 Tg2+ Rb3 Rf8 Tf7+ Re8 Th7 Tg3+ Ra2 Tg2+ Ra1 Tg1+ Ra2 Tg2+ Ra1 Tg1+ Ra2 Tg2+ Rb1
 12/21	00:00	     361 132	1 101 000	+2,06	Ra2 Rf8 Th8+ Rg7 Th7+ Rf6 Tf7+ Rg5 Th7 Tg2+ Rb3 Rf6 Rc4 Tg4+ Rc3 Tg2 Tf7+ Rg5
 13/21	00:00	     420 652	1 121 000	+2,06	Ra2 Rf8 Th8+ Rg7 Th7+ Rf6 Tf7+ Rg5 Th7 Tg2+ Rb3 Rf6 Rc4 Tg4+ Rc3 Tg2 Tf7+ Rg5 Th7 Rh6
 14/26	00:00	     727 520	1 258 000	+2,07	Ra2 Rf8 Rb2 Tg2+ Rc3 Tg3+ Rd4 Tg2 Tf7+ Rg8 Rc3 Tg3+ Rc2 Tg2+ Rc3 Tg3+ Rc2 Tg2+ Rc3 Te2
 15/25	00:00	     889 840	1 265 000	+2,07	Ra2 Rf8 Rb2 Tg2+ Rc3 Tg3+ Rd4 Tg2 Tf7+ Rg8 Rc3 Tg3+ Rc2 Tg2+ Rc3 Tg3+ Rc2 Tg2+ Rc3 Tg3+ Rc2 Tg2+ Rc3 Ta2 Rc2
 16/28	00:01	   1 217 192	1 277 000	+2,07	Ra2 Rf8 Rb2 Tg2+ Rc3 Tg3+ Rd4 Tg2 Tf7+ Re8 Th7 Rf8 Rc4 Tg4+ Rc3 Rg8 Tf7 Tg2 Th7
 17/28	00:01	   1 620 499	1 296 000	+2,07	Ra2 Rf8 Rb2 Tg2+ Rc3 Tg3+ Rd4 Tg2 Tf7+ Re8 Th7 Rf8 Tf7+ Re8 Th7 Rf8 Tf7+ Re8 Th7 Rf8 Tf7+ Re8 Re5
 18/32	00:02	   2 543 959	1 356 000	+2,07	Ra2 Rf8 Rb2 Tg2+ Rc3 Rg8 Tf7 Tg3+ Rc2 Tg2+ Rc3 Tg3+ Rc2 Tg2+ Rc3 Tg3+ Rc2 Tg2+ Rd3 Th2 Th7 Tg2 Tf7 Rh8 Th7+ Rg8 Tf7
 19/35	00:02	   3 290 238	1 376 000	+2,07	Ra2 Rf8 Rb2 Tg2+ Rc3 Rg8 Tf7 Tg3+ Rc2 Tg2+ Rc3 Tg3+ Rc2 Tg2+ Rc3 Tg3+ Rc2 Tg2+ Rd3 Th2 Th7 Tg2 Tf7 Tg3+ Rd2 Tg2+ Rd3 Th2
 20/34	00:03	   4 548 603	1 386 000	+2,07	Ra2 Rf8 Rb2 Tg2+ Rc3 Rg8 Tf7 Tg3+ Rc2 Tg2+ Rc3 Tg3+ Rc2 Tg2+ Rc3 Tg3+ Rc2 Tg2+ Rb3 Th2 Th7 Tg2 Rc4 Tg4+ Rb3 Tg2 Rc4 Th2
 21/38	00:04	   6 272 529	1 408 000	+2,07	Ra2 Rf8 Rb2 Tg2+ Rc3 Rg8 Tf7 Tg3+ Rc2 Tg2+ Rc3 Tg3+ Rc2 Tg2+ Rc3 Tg3+ Rc2 Tg2+ Rb3 Rh8 Tf8+ Rg7 Tf7+ Rg8 Th7 Rf8 Tf7+ Rg8 Th7 Tg3+ Rb2 Tg2+ Rc3
 22/44	00:06	   8 772 228	1 417 000	+2,07	Ra2 Rf8 Rb2 Tg2+ Rb3 Rg8 Tf7 Tg3+ Rb4 Tg2 Td7 Tg1 Th7 Tg2 Tf7 Rh8 Rc3 Tg3+ Rc2 Tg1 Tf8+ Rg7 Tf7+ Rg8 Th7 Tg2+ Rc3 Tg3+ Rc2 Tg2+ Rc3
 23/47	00:36	  51 402 120	1 408 000	+2,07	Ra2 Rf8 Td7 Tg2+ Rb3 Tg3+ Rb2 Rg8 Th7 Tg1 Tf7 Tg2+ Rb3 Tg3+ Rc2 Tg2+ Rc3 Rh8 Rc4 Tg4+ Rb3 Tg5 Tf8+ Rg7 Tf7+ Rg8 Th7 Tg3+ Rb2 Tg2+ Rc3 Rf8 Tf7+ Rg8
 24/45	00:50	  69 609 920	1 380 000	+2,07	Ra2 Rf8 Td7 Tg2+ Rb3 Tg3+ Rb2 Rg8 Th7 Tg1 Rb3 Tg3+ Rc4 Tg4+ Rc3 Tg2 Tc7 Tg4 Tc8+ Rg7 Tc7+ Rf8 Tf7+ Rg8 Th7 Tg2 Ta7 Rh8 Tf7 Rg8 Rc4 Rh8
 25/46	01:31	 126 297 239	1 381 000	+2,07	Ra2 Rf8 Td7 Tg2+ Rb3 Tg3+ Rc2 Tg2+ Rc3 Tg3+ Rb2 Rg8 Th7 Tg2+ Rc3 Rf8 Tc7 Tg5 Tf7+ Rg8 Th7 Rf8 Tf7+ Rg8 Th7 Rf8 Tf7+ Rg8 Th7 Rf8 Tf7+ Rg8 Th7 Tg2
 26/54	06:49	 583 583 283	1 425 000	+2,07	Ra2 Tg2+ Rb3 Rf8 Tf7+ Rg8 Th7 Tg3+ Rb2 Tg2+ Rc3 Rf8 Tb7 Tg5 Th7 Tg2 Tb7 Tg5 Th7 Tg2 Tb7 Tg5 Th7 Tg2 Tb7 Tg5 Tb8+ Rg7
 27/49	07:19	 629 200 939	1 431 000	+2,07	Ra2 Tg1 Rb2 Tg2+ Rb3 Tg1 Tf7 Tg2 Th7 Tg3+ Rc4 Tg2 Tb7 Tc2+ Rb3 Tg2 Tf7 Tg4 Rb2 Tg2+ Rc3 Tg4 Ta7 Tg3+ Rd4 Tg2 Tf7 Td2+ Rc4 Tg2 Rc3 Th2 Th7
 28/56	17:40	1 517 857 212	1 430 000	+2,07	Ra2 Tg2+ Rb3 Td2 Tf7 Tg2 Th7 Tg1 Tf7 Tg2 Rb4 Tg4+ Rb3 Tg2 Rb4 Tg4+ Rb3 Tg2 Rb4 Tg4+ Rb3 Tg2 Rb4 Tg4+ Rb3 Tg2 Rb4 Tg4+ Rb3 Tg2 Rb4 Tg4+ Rb3 Tg2 Rb4
 29/60	1:25:08	7 099 095 626	1 389 000	+2,07	Ra2 Tg2+ Ra3 Rf8 Tf7+ Rg8 Th7 Tg3+ Rb2 Tg2+ Rb3 Tg3+ Rb4 Tg1 Rc3 Tg2 Tc7 Rh8 Rb3 Tg4 Tc3 Tg1 Tc2 Rg7 Tc7+ Rf8 Tf7+ Rg8 Rc2 Tg2+ Rd3 Tg4 Rc2 Tg2+ Rd3 Th2 Th7
"The only good bug is a dead bug." (Don Dailey)
[Blog: http://tinyurl.com/predateur ] [Facebook: http://tinyurl.com/fbpredateur ] [MacEngines: http://tinyurl.com/macengines ]
Top
Jouni
Posts: 3293
Joined: Wed Mar 08, 2006 8:15 pm

Re: Old test position

Post by Jouni »

Yes Louis SF evalution goes between 0,00 and +x,xx. That's common with that engine. So You have to stop at correct point!

Jouni
Top
JohnnyZoom3H

Re: Old test position

Post by JohnnyZoom3H »

Tablebases do show Ka2 as winning, with all other reasonable moves only drawing. I have no idea what makes it work, some outmaneuvering of the king with a key-square duel with the rook, perhaps.

I had SF 2.2.2-ja running with 1GB hash and 3 PVs, it thought all K moves equal at about +5 until about ply 48, where the other two went to 0.00. But up around ply 60 even Ka2 went to 0. But now at ply 63 it is coming back positive (0.28 and climbing), it may take a while to settle down though.

I can't even use the PV to figure out the subtleties of this position. Most of the concepts involve what happens with moves not in the PV.

Ugh, as I write ply 63 resolved with score=0.00. This may be a case where one needs 64 bit to use >4GB hashes. I don't have that capability.
Top
zullil
Posts: 6442
Joined: Tue Jan 09, 2007 12:31 am
Location: PA USA
Full name: Louis Zulli

Re: Old test position

Post by zullil »

Jouni wrote:Yes Louis SF evalution goes between 0,00 and +x,xx. That's common with that engine. So You have to stop at correct point!

Jouni
Well, here's how my long search ended. Certainly SF does not solve this:

Code: Select all

info depth 80 currmove a1b2 currmovenumber 1
info depth 80 seldepth 101 score cp 16 lowerbound nodes 71182567366 nps 6827948 time 10425177 multipv 1 pv a1b2 g3h3 b2b1 h3h1 b1a2 h1h2 a2b3 h2g2 b3a4 g2d2 h7b7 d2h2 b7b5 h2a2 a4b3 a2h2 b3c4 g8g7 b5e5 h2h3 c4d4 h3h1 d4d5 h1d1 d5e6 d1g1 e5c5 g1e1 e6d7 e1d1 d7e8 d1e1 e8d8 e1h1 c5g5 h1a1 g5e5 a1a8 d8e7 a8a4 e7e6 a4a6 e6f5 g7h6 e5e8 a6a5 f5f6 a5a6 e8e6
info depth 80 currmove a1b2 currmovenumber 1
info depth 80 seldepth 101 score cp 28 lowerbound nodes 71492673997 nps 6815539 time 10489658 multipv 1 pv a1b2 g3g2 b2b3 g2h2 b3a3 h2h4 a3a2 h4h1 a2b2 h1h2 b2c1 h2h1 c1c2 h1h2 c2d3 h2h4 d3e3 h4h1 e3d4 h1h4 d4d5 h4b4 d5e5 b4b1 e5d4 b1g1 d4c5 g1d1 h7e7 d1h1 e7e8 g8g7 e8e5 h1h2 e5f5 h2c2 c5d5 c2d2 d5e4 d2e2 e4d4 e2d2 d4e5 d2e2 e5d6 e2h2 d6e5 h2e2
info depth 80 currmove a1b2 currmovenumber 1
Bus error
Top
lech
Posts: 1136
Joined: Sun Feb 14, 2010 10:02 pm

Re: Old test position

Post by lech »

zullil wrote:
Jouni wrote:Yes Louis SF evalution goes between 0,00 and +x,xx. That's common with that engine. So You have to stop at correct point!

Jouni
Well, here's how my long search ended. Certainly SF does not solve this:

Code: Select all

info depth 80 currmove a1b2 currmovenumber 1
info depth 80 seldepth 101 score cp 16 lowerbound nodes 71182567366 nps 6827948 time 10425177 multipv 1 pv a1b2 g3h3 b2b1 h3h1 b1a2 h1h2 a2b3 h2g2 b3a4 g2d2 h7b7 d2h2 b7b5 h2a2 a4b3 a2h2 b3c4 g8g7 b5e5 h2h3 c4d4 h3h1 d4d5 h1d1 d5e6 d1g1 e5c5 g1e1 e6d7 e1d1 d7e8 d1e1 e8d8 e1h1 c5g5 h1a1 g5e5 a1a8 d8e7 a8a4 e7e6 a4a6 e6f5 g7h6 e5e8 a6a5 f5f6 a5a6 e8e6
info depth 80 currmove a1b2 currmovenumber 1
info depth 80 seldepth 101 score cp 28 lowerbound nodes 71492673997 nps 6815539 time 10489658 multipv 1 pv a1b2 g3g2 b2b3 g2h2 b3a3 h2h4 a3a2 h4h1 a2b2 h1h2 b2c1 h2h1 c1c2 h1h2 c2d3 h2h4 d3e3 h4h1 e3d4 h1h4 d4d5 h4b4 d5e5 b4b1 e5d4 b1g1 d4c5 g1d1 h7e7 d1h1 e7e8 g8g7 e8e5 h1h2 e5f5 h2c2 c5d5 c2d2 d5e4 d2e2 e4d4 e2d2 d4e5 d2e2 e5d6 e2h2 d6e5 h2e2
info depth 80 currmove a1b2 currmovenumber 1
Bus error
I described it: http://talkchess.com/forum/viewtopic.php?t=42244 :?
Top

Return to “Computer Chess Club: General Topics”