Test position: Very tough and just tough

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

Moderator: Ras

Albert Silver
Posts: 3026
Joined: Wed Mar 08, 2006 9:57 pm
Location: Rio de Janeiro, Brazil

Re: Test position: Very tough and just tough

Post by Albert Silver »

Eelco de Groot wrote:
Martin Thoresen wrote:Here's 1 hour and 42 minutes with Houdini on the TCEC computer, 8 GB hash with Large Pages:

[d]

Code: Select all

Houdini 1.5a t6:
  4/17	00:00	       1 870	0	-0,88	19.b3 Rc8 20.bxc4 bxc4 21.Rab1
  5/17	00:00	       2 794	0	-0,88	19.b3 Rc8 20.Nd4 Na6 21.bxc4 bxc4
  5/17	00:00	       4 060	253 000	-0,86	19.a4 a6 20.Nd4 Qe8 21.axb5 axb5
  6/17	00:00	       5 296	331 000	-0,75	19.a4 a6 20.Nd4 Qe8 21.Rf3 Qc8 22.Nc6
  7/17	00:00	       6 611	413 000	-0,70	19.a4 a6 20.Nd4 Qe8 21.Rf3 Nh7 22.Qh4
  8/18+	00:00	       9 858	616 000	-0,61	19.a4
  8/19	00:00	      16 901	1 056 000	-0,51	19.a4 a6 20.Nd4 Be8 21.Nc6 Bxc6 22.dxc6 Qe8 23.axb5 axb5 24.Rxa8 Qxa8
  9/25	00:00	      92 293	2 251 000	-0,54	19.a4 bxa4 20.Nd2 Rb8 21.Ndxc4 Rb7 22.Qg3 Bb5 23.h4 Qd7
 10/27-	00:00	     153 781	3 271 000	-0,61	19.a4 bxa4
 10/27	00:00	     191 501	3 419 000	-0,54	19.a4 b4 20.Nxc4 bxc3 21.bxc3 Rc8 22.Ne3 Na6 23.e5 dxe5 24.Nxe5 Rxc3 25.Rab1
 11/27-	00:00	     218 054	3 634 000	-0,61	19.a4 b4
 11/29	00:00	     691 536	6 649 000	-0,55	19.a4 Qb8 20.Nd4 bxa4 21.Nxc4 Bb5 22.Nxb5 Qxb5 23.Rxa4 Nxe4 24.Qxe7 Qc5+ 25.Kh1 Nf2+ 26.Kg1 Nd3+ 27.Kh1 Nxd5 28.Qxd6 Qxd6 29.Nxd6 Nxb2
 12/29	00:00	     778 424	6 768 000	-0,55	19.a4 Qb8 20.Nd4 bxa4 21.Nxc4 Bb5 22.Nxb5 Qxb5 23.Rxa4 Nxe4 24.Qxe7 Qc5+ 25.Kh1 Nf2+ 26.Kg1 Nd3+ 27.Kh1 Nxd5 28.Qxd6 Qxd6 29.Nxd6 Nxb2
 13/29+	00:00	   1 149 285	7 765 000	-0,48	19.a4
 13/35	00:00	   2 038 986	9 617 000	-0,57	19.a4 b4 20.Nxc4 bxc3 21.bxc3 Qc8 22.Ne3 Na6 23.e5 dxe5 24.Nxe5 Qxc3 25.Nxd7 Nxd7 26.Rab1 Rfe8 27.Rb3 Qd2 28.Bd3
 14/35+	00:00	   2 521 801	9 967 000	-0,51	19.a4
 14/35	00:00	   2 874 462	10 121 000	-0,49	19.a4 b4 20.Nxc4 Qc8 21.Nfd2 bxc3 22.bxc3 Rb8 23.Rf3 Rd8 24.Rg3 Nce8 25.Qe3 Qc7 26.Qd4 Kf8 27.e5
 15/35	00:01	   4 455 096	11 336 000	-0,47	19.a4 b4 20.Nxc4 Qc8 21.Nfd2 bxc3 22.bxc3 Rb8 23.Qe3 a5 24.Rab1 Qa6 25.h3 Rfe8 26.Nb6 Kf8 27.Bd3
 16/35-	00:01	   5 358 110	11 750 000	-0,54	19.a4 b4
 16/41	00:01	   9 659 166	13 359 000	-0,54	19.a4 b4 20.Nxc4 Qc8 21.Nfd2 bxc3 22.bxc3 Rb8 23.Rad1 Rd8 24.h4 Kf8 25.h5 Ng8 26.Qe3 Bh6 27.Qxa7 Nf6 28.Qd4 Qa6 29.Rb1 Bxa4 30.Bxa4 Qxa4
 16/46	00:01	  17 081 335	14 562 000	-0,53	19.e5 dxe5 20.Nxe5 Nce8 21.Rf3 Qc7 22.Nxd7 Qxd7 23.a4 b4 24.cxb4 Qd6 25.Nxc4 Qxb4 26.Ne3 Qxb2 27.Rb1 Qd4 28.Rd1 Qc5 29.Rg3 Rc8 30.Qh4
 17/46+	00:02	  22 688 056	14 770 000	-0,46	19.e5
 17/46	00:02	  24 897 213	14 810 000	-0,45	19.e5 dxe5 20.Nxe5 Nce8 21.Rf3 Qc7 22.Nxd7 Qxd7 23.Rg3 a6 24.Re1 Qc7 25.a4 Qe5 26.Rd1 Rd8 27.Bb1 Qd6 28.Kh1 Qb6 29.axb5 axb5
 18/46+	00:02	  29 234 829	14 953 000	-0,39	19.e5
 18/48+	00:03	  45 345 941	15 609 000	-0,25	19.e5
 18/51	00:06	  92 364 181	16 426 000	-0,10	19.e5 dxe5 20.Nxe5 Nce8 21.Rf3 Nh7 22.Qh4 f6 23.Nc6 Bxc6 24.dxc6 Ng5 25.Rg3 Qb6 26.Qf4 Rd8 27.a4 Rd2 28.h4 Nf7 29.Be4 bxa4 30.Rxa4 Ne5 31.Rb4
 19/51+	00:08	 129 397 958	16 739 000	-0,04	19.e5
 19/51	00:10	 164 684 623	16 838 000	 0,00	19.e5 dxe5 20.Nxe5 Nce8 21.Rf3 Nh7 22.Qh4 f6 23.Nxd7 Qxd7 24.Rg3 Ng5 25.a4 Nd6 26.Qf4 Ngf7 27.Qg4 Ne8 28.Qd4 Ne5 29.Ng4 Nxg4 30.Rxg4 Nd6 31.Rxg7+ Kxg7 32.Qg4+ Kh8 33.Qh3+ Kg8 34.Qg3+ Kh8 35.Qh3+ Kg8
 20/62	00:14	 245 456 740	17 471 000	 0,00	19.e5 dxe5 20.Nxe5 Nh7 21.Qg3 Nf6 22.Qg5 Nh7
 21/62	00:15	 266 496 430	17 561 000	 0,00	19.e5 dxe5 20.Nxe5 Nh7 21.Qg3 Nf6 22.Qg5 Nh7
 22/62-	00:34	 591 085 996	17 726 000	-0,06	19.e5 dxe5
 22/62	00:38	 674 925 152	17 793 000	-0,03	19.e5 dxe5 20.Nxe5 Ncxd5 21.Nxd5 Nxd5 22.f6 Qb6+ 23.Kh1 Nxf6 24.Rf3 Be6 25.Re1 Rad8 26.Rg3 Ne8 27.Qh5 f5 28.Ng6 Rd7 29.Nxf8 Kxf8 30.Bxf5 Qf2 31.Rg1 Bf7 32.Bg6 Bxg6 33.Qxb5 Rd3 34.Rxg6 Nd6 35.Qg5 Bf6 36.Qg4 Kf7 37.Qh5 Rd2 38.Rg3+ Ke6 39.Rf3 Qh4 40.Re3+ Kd7
 23/71-	00:42	 766 026 050	18 141 000	-0,10	19.e5 dxe5
 23/71+	01:31	1 635 812 738	18 105 000	+0,02	19.Qh4
 23/73+	05:16	5 681 258 675	17 966 000	+0,46	19.Qh4
 23/74	11:45	12 666 795 645	17 962 000	+0,52	19.Qh4 Na6 20.g4 Qb6 21.Rae1 Rfb8 22.g5 Ne8 23.g6 Nf6 24.e5 dxe5 25.gxf7+ Kxf7 26.Nxe5+ Ke8 27.Qg5 Bf8 28.d6 Qc5 29.Qg3 Kd8 30.Ng6 Bh6 31.dxe7+ Kc8 32.Rf2 Bxe3 33.Qxe3 Qxe3 34.Rxe3 Kc7 35.Rd2 Rg8 36.a4 Rab8 37.Rd4 Nc5 38.Re5
 24/76-	14:16	15 455 514 675	18 055 000	+0,46	19.Qh4 Na6
 24/76	14:26	15 629 815 018	18 062 000	+0,46	19.Qh4 Nh7 20.f6 Bxf6 21.Qh5 Bg7 22.e5 f5 23.Nxf5 Bxf5 24.Bxf5 Rxf5 25.Qxf5 dxe5 26.Nxe5 Qxd5 27.Ng6 Re8 28.Rae1 Bf6 29.Ne5 Bg7 30.Qg6 Rf8 31.Ng4 Kh8 32.Ne3 Qc5 33.Rf5 Rxf5 34.Qxf5 Be5 35.Qh5 Ne6 36.Kh1 Ng7 37.Qh6 Kg8 38.Rf1
 25/76+	17:25	18 984 154 631	18 165 000	+0,52	19.Qh4
 25/76	21:57	24 034 601 825	18 246 000	+0,56	19.Qh4 Nh7 20.f6 Bxf6 21.Qh5 Bg7 22.e5 f5 23.Nxf5 Bxf5 24.Bxf5 Rxf5 25.Qxf5 dxe5 26.Nxe5 Qxd5 27.Ng6 Re8 28.Rae1 Bf6 29.Ne5 Bg7 30.Qg6 Rf8 31.Ng4 Kh8 32.Ne3 Qc5 33.Rf5 Rxf5 34.Qxf5 Be5 35.Kh1 e6 36.Qh5 Bd6 37.Qxc5 Bxc5
 26/76+	38:47	43 026 742 939	18 487 000	+0,63	19.Qh4
 26/76+	1:04:20	71 719 583 226	18 581 000	+0,81	19.Qh4
 26/77	1:41:49	113 943 779 032	18 651 000	+0,93	19.Qh4 Nh7 20.f6 exf6 21.e5 f5 22.Qg3 Ne8 23.Bxf5 Bxf5 24.Nxf5 Kh8 25.exd6 Qf6 26.N3d4 Qg6 27.Nxb5 Qxg3 28.hxg3 Rd8 29.Nxa7 Nxd6 30.Nc6 Rd7 31.a4 Nf6 32.Ne5 Rdd8 33.Rad1 Nfe4 34.Nc6 Rd7 35.a5 Ra8
 27/77-	1:42:15	114 505 301 894	18 664 000	+0,84	19.Qh4 Nh7
 27/77	1:42:39	114 970 514 766	18 669 000	+0,84	19.Qh4 Nh7 20.f6 exf6 21.e5 f5 22.Qg3 dxe5 23.Nxe5 Bc8 24.Nc6 Qg5 25.Qxg5 Nxg5 26.d6 Ne8 27.Ne7+ Kh8 28.Nxc8 Rxc8 29.Rxf5 Nh7 30.d7 Rd8 31.dxe8Q Rfxe8 32.Nf1 Re2 33.Rf2 Rxf2 34.Kxf2 b4 35.Rd1 Rb8 36.cxb4 Bxb2 37.Ne3 Rxb4 38.Rd8+ Kg7 39.h4 Nf8 40.Nf5+ Kg8 41.Rc8 Bc3 42.Ne3 Bd4 43.a3 Rb2 44.Rxc4 Ne6 45.Kf3 Ra2 46.a4 Ra3 47.Rc8+ Kg7
Best move: Qh4, Value: +0,84, Depth: 27/77, Time: 1:42:39,064, 114 970 514 766 Nodes, 18 669 kN/sec.
Thanks a lot Martin and Louis! Wow, even the TCEC computer has stopped playing games and the TCEC rating lists will now be delayed. The Washington Post can't publish a new article on Houdini because of this, and goes broke. This starts the final collapse of the US economy, looting and arson in the big cities in the North, China invades Taiwan and Japan in the following turmoil, to protect vital interests in their influence zone, uprising in Tibet and Mongolia ensues, all this because of a single chess position. It better be worth it...
Just shows the human brain is still not to be underestimated. :wink:
"Tactics are the bricks and sticks that make up a game, but positional play is the architectural blueprint."
Albert Silver
Posts: 3026
Joined: Wed Mar 08, 2006 9:57 pm
Location: Rio de Janeiro, Brazil

Re: Test position: Very tough and just tough

Post by Albert Silver »

Martin Thoresen wrote:Here's 1 hour and 42 minutes with Houdini on the TCEC computer, 8 GB hash with Large Pages:

[d]

Code: Select all

Houdini 1.5a t6:

27/77-	1:42:15	114 505 301 894	18 664 000	+0,84	19.Qh4 Nh7

27/77	1:42:39	114 970 514 766	18 669 000	+0,84	19.Qh4 Nh7 20.f6 exf6 21.e5 f5 22.Qg3 dxe5 23.Nxe5 Bc8 24.Nc6 Qg5 25.Qxg5 Nxg5 26.d6 Ne8 27.Ne7+ Kh8 28.Nxc8 Rxc8 29.Rxf5 Nh7 30.d7 Rd8 31.dxe8Q Rfxe8 32.Nf1 Re2 33.Rf2 Rxf2 34.Kxf2 b4 35.Rd1 Rb8 36.cxb4 Bxb2 37.Ne3 Rxb4 38.Rd8+ Kg7 39.h4 Nf8 40.Nf5+ Kg8 41.Rc8 Bc3 42.Ne3 Bd4 43.a3 Rb2 44.Rxc4 Ne6 45.Kf3 Ra2 46.a4 Ra3 47.Rc8+ Kg7
Best move: Qh4, Value: +0,84, Depth: 27/77, Time: 1:42:39,064, 114 970 514 766 Nodes, 18 669 kN/sec.
It is rather remarkable that even with so much computing power it goes for the much weaker 19.Qh4 Nh7 20.f6?

20.e5 is stronger. Ex: 20...f6 21.exd6 exd6 22.Qg3 Qb8 23.Nd4 (preparing Ne6) Qb6 24.Kh1 Rae8 25.Ne6 etc.
"Tactics are the bricks and sticks that make up a game, but positional play is the architectural blueprint."