Position to be solved first by head and then by your engine

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

Moderators: hgm, Dann Corbit, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
Post Reply
User avatar
cdani
Posts: 2189
Joined: Sat Jan 18, 2014 9:24 am
Location: Andorra
Contact:

Position to be solved first by head and then by your engine

Post by cdani » Sat May 07, 2016 6:01 pm

Difficult but interesting.
Some good engines cannot solve it.
White plays and ??.

[d]Kb4rr/2PP4/R1p5/1kp5/p7/P7/8/8 w - - 0 1

User avatar
Ajedrecista
Posts: 1535
Joined: Wed Jul 13, 2011 7:04 pm
Location: Madrid, Spain.
Contact:

Position to be solved first by head and then by your engine.

Post by Ajedrecista » Sat May 07, 2016 6:37 pm

Hello:
cdani wrote:Difficult but interesting.
Some good engines cannot solve it.
White plays and ??.

[d]Kb4rr/2PP4/R1p5/1kp5/p7/P7/8/8 w - - 0 1
I expected a draw with some kind of stalemate involving the rock sacrifice and underpromotions. I almost got it but my underpromotions were not right. Here I go with my busted solution (I used engine later for check my line):

1.- Rb6+ (correct).
1.- ..., Kxb6 (correct). Other moves lead to 2.- Rxb8 and white wins (correct).
2.- c8=N+ (incorrect).

I was looking for something like 2.- ..., Rxc8; dxc8=N+, Rxc8? Draw by stalemate.

Well, I am a patzer so I am glad that I found the first move by myself, without engines, and guess the result and the way to get it (by stalemate).

A tip for people who want to try to solve the problem is that they must look for underpromotions (c pawn, d pawn or both) that pin/s in some way the promoted pawn/s.

Spoiler alert! This study was composed by Yochanan Afek and was published in 1981. You can find the solution here.

Regards from Spain.

Ajedrecista.

User avatar
vittyvirus
Posts: 646
Joined: Wed Jun 18, 2014 12:30 pm
Full name: Fahad Syed

Re: Position to be solved first by head and then by your eng

Post by vittyvirus » Sun May 08, 2016 8:34 am

Yaka seems to have an idea.

Code: Select all

Yaka 0.0 x64 by Syed Fahad
position fen Kb4rr/2PP4/R1p5/1kp5/p7/P7/8/8 w - - 0 1
search 15 20
info depth 1 score cp 974 nodes 1 nps 1000 tthits 0 pv a6a4
info depth 1 score cp 1353 nodes 7 nps 7000 tthits 0 pv c7b8q
info depth 2 score cp -616 nodes 22 nps 22000 tthits 0 pv c7b8q g8b8
info depth 2 score cp 140 nodes 176 nps 176000 tthits 0 pv a6b6 b5b6
info depth 2 score cp 170 nodes 180 nps 180000 tthits 0 pv a6a5 b5a5
info depth 3 score cp 711 nodes 409 nps 409000 tthits 0 pv a6a5 b5a5 c7b8q
info depth 4 score cp -1251 nodes 1256 nps 1256000 tthits 13 pv a6a5 b5a5 c7b8q
g8b8
info depth 4 score cp -1227 nodes 1331 nps 1331000 tthits 14 pv a6b6 b5b6 c7b8q
g8b8
info depth 4 score cp -809 nodes 1765 nps 1765000 tthits 14 pv d7d8q g8d8 c7d8q
h8d8
info depth 4 score cp -761 nodes 2383 nps 2383000 tthits 14 pv c7c8q g8c8 d7c8q
h8c8
info depth 5 score cp -136 nodes 11759 nps 3919000 tthits 396 pv c7c8q g8c8 d7c8
q h8c8 a6a4
info depth 5 score cp 0 nodes 21233 nps 1179000 tthits 2037 pv a6b6 b5b6 c7b8q g
8b8
info depth 6 score cp -737 nodes 27782 nps 5556000 tthits 2918 pv a6b6 b5b6 c7b8
q g8b8
info depth 6 score cp -669 nodes 31439 nps 5239000 tthits 3426 pv c7c8q g8c8 d7c
8q h8c8 a6a4 b5a4
info depth 7 score cp -268 nodes 98021 nps 4455000 tthits 14730 pv c7c8q g8c8 d7
c8q h8c8 a6a4 b5a4 a8b7
info depth 7 score cp -139 nodes 117552 nps 6186000 tthits 20613 pv a6b6 b5b6 c7
b8q g8b8
info depth 8 score cp -1195 nodes 287678 nps 26152000 tthits 55487 pv a6b6 b5b6
c7b8q g8b8
info depth 8 score cp -733 nodes 315231 nps 13705000 tthits 66407 pv c7c8q g8c8
d7c8q h8c8 a6a4 b5a4 a8b7 c5c4
info depth 9 score cp -281 nodes 826360 nps 7945000 tthits 216857 pv c7c8q g8c8
d7c8q h8c8 a6a4 b5a4 a8b7 c5c4 b7c8
info depth 9 score cp -267 nodes 981930 nps 6771000 tthits 251094 pv c7c8r g8c8
d7c8q h8c8 a6a4 b5a4 a8b7 c5c4 b7c8
info depth 9 score cp 0 nodes 1504683 nps 3292000 tthits 428040 pv a6b6 b5b6 c7b
8q g8b8
info depth 10 score cp 0 nodes 2789550 nps 24686000 tthits 746241 pv a6b6 b5b6 c
7b8q g8b8
info depth 11 score cp 0 nodes 8379435 nps 19578000 tthits 2302995 pv a6b6 b5b6
c7b8q g8b8
info depth 12 score cp 0 nodes 30386116 nps 38317000 tthits 8458686 pv a6b6 b5b6
 c7b8q g8b8
info depth 13 score cp 0 nodes 88499128 nps 20857000 tthits 22326408 pv a6b6 b5b
6 c7b8q g8b8
info depth 14 score cp 0 nodes 307219877 nps 14745000 tthits 59926824 pv a6b6 b5
b6 c7b8q g8b8
info depth 15 score cp 0 nodes 942651713 nps 46961000 tthits 162186500 pv a6b6 b
5b6 c7b8q g8b8
bestmove a6b6 nodes 3425660855
PS the nps might be buggy.

F. Bluemers
Posts: 867
Joined: Thu Mar 09, 2006 10:21 pm
Location: Nederland
Contact:

Re: Position to be solved first by head and then by your eng

Post by F. Bluemers » Sun May 08, 2016 8:56 am

cdani wrote:Difficult but interesting.
Some good engines cannot solve it.
White plays and ??.

[d]Kb4rr/2PP4/R1p5/1kp5/p7/P7/8/8 w - - 0 1
My head did not solve it :lol:

Dirty:

Code: Select all

 7	  0.00 	11193  	0:00.00	Rb6+ Kxb6 d8=B Bxc7 
  7	  0.00 	10615  	0:00.00	Rb6+ Kxb6 d8=B Bxc7 
  6	  0.00 	9868    	0:00.00	Rb6+ Kxb6 d8=B Bxc7 
  6	  0.00 	9002    	0:00.00	Rb6+ Kxb6 d8=B Bxc7 
  6	 -5.40!	5139    	0:00.00	Rb6+! 
  6	 -7.90?	5139    	0:00.00	Rb6+? 
  5	 -7.90 	5139    	0:00.00	Rb6+ 
  5	 -6.62!	3206    	0:00.00	Rb6+! 
  5	 -7.90 	2964    	0:00.00	c8=Q Rxc8 dxc8=Q Rxc8 Kb7 
  4	 -8.96 	2659    	0:00.00	c8=Q Rxc8 dxc8=Q Rxc8 
  4	 -8.96 	2073    	0:00.00	c8=Q Rxc8 dxc8=Q Rxc8 
  4	 -7.52!	1216    	0:00.00	c8=Q! 
  3	 -8.96 	1216    	0:00.00	c8=Q Rxc8 dxc8=Q 
  3	 -8.96 	737      	0:00.00	c8=Q Rxc8 dxc8=Q 
  2	 -8.96 	580      	0:00.00	c8=Q Rxc8 
  2	 -8.96 	325      	0:00.00	c8=Q Rxc8 
  2	 -11.43 	291      	0:00.00	d8=B Bxc7 
  1	 -6.47 	163      	0:00.00	d8=B 
Best
Fonzy

zullil
Posts: 6442
Joined: Mon Jan 08, 2007 11:31 pm
Location: PA USA
Full name: Louis Zulli

Re: Position to be solved first by head and then by your eng

Post by zullil » Sun May 08, 2016 10:22 am

vittyvirus wrote:Yaka seems to have an idea.

Code: Select all

Yaka 0.0 x64 by Syed Fahad
position fen Kb4rr/2PP4/R1p5/1kp5/p7/P7/8/8 w - - 0 1

info depth 15 score cp 0 nodes 942651713 nps 46961000 tthits 162186500 pv a6b6 b
5b6 c7b8q g8b8
bestmove a6b6 nodes 3425660855
PS the nps might be buggy.
Something else seems buggy, since your engine is evaluating as 0 a PV that ends in mate.

[D] Kr5r/3P4/1kp5/2p5/p7/P7/8/8 w - - 0 3

User avatar
vittyvirus
Posts: 646
Joined: Wed Jun 18, 2014 12:30 pm
Full name: Fahad Syed

Re: Position to be solved first by head and then by your eng

Post by vittyvirus » Sun May 08, 2016 12:43 pm

zullil wrote:
vittyvirus wrote:Yaka seems to have an idea.

Code: Select all

Yaka 0.0 x64 by Syed Fahad
position fen Kb4rr/2PP4/R1p5/1kp5/p7/P7/8/8 w - - 0 1

info depth 15 score cp 0 nodes 942651713 nps 46961000 tthits 162186500 pv a6b6 b
5b6 c7b8q g8b8
bestmove a6b6 nodes 3425660855
PS the nps might be buggy.
Something else seems buggy, since your engine is evaluating as 0 a PV that ends in mate.

[D] Kr5r/3P4/1kp5/2p5/p7/P7/8/8 w - - 0 3
Hi Louis!
Hope you're doing fine.
Well, I think that you're right on this one. I'd look into it.

rabbits23
Posts: 144
Joined: Fri Sep 12, 2014 2:57 am
Location: Randwick Australia

Re: Position to be solved first by head and then by your eng

Post by rabbits23 » Mon May 09, 2016 12:25 am

Definitely a challenging problem Daniel.I was unable to sort it out even with the help of Komodo 9.42. Stockfish 7.32 gave me the solution immediately though.
1Rb6* KxR 2 Pd8 Ba7 3Pc8 stalemate.
Allan

User avatar
vittyvirus
Posts: 646
Joined: Wed Jun 18, 2014 12:30 pm
Full name: Fahad Syed

Re: Position to be solved first by head and then by your eng

Post by vittyvirus » Mon May 09, 2016 3:01 pm

zullil wrote:
vittyvirus wrote:Yaka seems to have an idea.

Code: Select all

Yaka 0.0 x64 by Syed Fahad
position fen Kb4rr/2PP4/R1p5/1kp5/p7/P7/8/8 w - - 0 1

info depth 15 score cp 0 nodes 942651713 nps 46961000 tthits 162186500 pv a6b6 b
5b6 c7b8q g8b8
bestmove a6b6 nodes 3425660855
PS the nps might be buggy.
Something else seems buggy, since your engine is evaluating as 0 a PV that ends in mate.

[D] Kr5r/3P4/1kp5/2p5/p7/P7/8/8 w - - 0 3
Oh, I did look into the matter. It turns out the search is not buggy, but since I reconstruct the PV from transposition tables, that seems to be the source of problems.
Here's the PV after I played the recommended moves.
[d]Kb4rr/2PP4/1kp5/2p5/p7/P7/8/8 w - - 0 2[/d]

Code: Select all

Yaka 0.0 x64 by Syed Fahad
position fen Kb4rr/2PP4/1kp5/2p5/p7/P7/8/8 w - - 0 2
search 13 23
info depth 1 score cp 700 nodes 1 nps 1000 tthits 0 pv c7b8q
info depth 2 score cp -1227 nodes 16 nps 16000 tthits 0 pv c7b8q g8b8
info depth 3 score cp -107374182 nodes 269 nps 269000 tthits 0 pv c7b8q g8b8
info depth 3 score cp 0 nodes 277 nps 277000 tthits 1 pv c7b8r g8b8
info depth 4 score cp -107374182 nodes 782 nps 782000 tthits 14 pv c7b8r g8b8
info depth 4 score cp -1400 nodes 1509 nps 754500 tthits 14 pv c7c8q g8c8 d7c8q
h8c8
info depth 4 score cp -1294 nodes 2877 nps 1438500 tthits 86 pv c7c8r g8c8 d7c8q
 h8c8
info depth 4 score cp -737 nodes 4601 nps 1533666 tthits 191 pv d7d8b g8d8 c7d8q
 h8d8
info depth 5 score cp -86 nodes 5974 nps 1991333 tthits 241 pv d7d8b g8d8 c7d8q
h8d8
info depth 6 score cp -1195 nodes 16042 nps 1002625 tthits 1813 pv d7d8b g8d8 c7
d8q h8d8
info depth 6 score cp -733 nodes 23488 nps 1957333 tthits 3146 pv c7c8q g8c8 d7c
8q h8c8
info depth 7 score cp -140 nodes 45847 nps 1478935 tthits 8235 pv c7c8q g8c8 d7c
8q h8c8
info depth 7 score cp 0 nodes 47346 nps 23673000 tthits 8768 pv d7d8b g8d8 c7d8q
 h8d8
info depth 8 score cp 0 nodes 120324 nps 1168194 tthits 25096 pv d7d8b g8d8 c7d8
q h8d8
info depth 9 score cp 0 nodes 306965 nps 1340458 tthits 85453 pv d7d8b g8d8 c7d8
q h8d8
info depth 10 score cp 0 nodes 889838 nps 1815995 tthits 258227 pv d7d8b g8d8 c7
d8q h8d8
info depth 11 score cp 0 nodes 2338180 nps 1741012 tthits 864968 pv d7d8b g8d8 c
7d8q h8d8
info depth 12 score cp 0 nodes 5816605 nps 1805277 tthits 2566415 pv d7d8b g8d8
c7d8q h8d8
info depth 13 score cp 0 nodes 16038442 nps 1625627 tthits 6686545 pv d7d8b g8d8
 c7d8q h8d8
bestmove d7d8b nodes 38381358

viditchess
Posts: 20
Joined: Mon Feb 16, 2015 12:58 pm

Re: Position to be solved first by head and then by your eng

Post by viditchess » Tue May 17, 2016 5:53 pm

Rb6 Kb6 d8=B Ba7 c8 B! Nice problem :)
Grandmaster 2617

User avatar
cdani
Posts: 2189
Joined: Sat Jan 18, 2014 9:24 am
Location: Andorra
Contact:

Re: Position to be solved first by head and then by your eng

Post by cdani » Tue May 17, 2016 6:27 pm

viditchess wrote:Rb6 Kb6 d8=B Ba7 c8 B! Nice problem :)
Sure! Glad to have such strong player here at the forum :-)

Post Reply