TESTPOSITION!!

[GARGAMEL]

hello

[Event "Studie, Ta2"]
[Site "?"]
[Date "1978.??.??"]
[Round "?"]
[White "Eisert,S & Rehm,H"]
[Black "_"]
[Result "1-0"]

[d]q1b4n/1p1pKp2/1p1p1Pp1/1P1P2Pp/2P4P/8/7R/1N1k4 w - - 0 1[d]
1. Ra2!! Qb8 2. Kd8 Kc1 3. Nd2 Kd1 4. Nb3 Ke1 5. Nc1 Kd1 6. Ne2 Ke1 7. Nc3 Kf1 8. Nd1 Ke1 9. Nf2 Kf1 10. Nd3 Kg1 11. Ne1 1-0

Which engine can solve it?

regards Anton

[Steve Maughan]

None!!

Way to deep for today's engines.

[JuLieN]

None!!

Way to deep for today's engines.

I beg to desagree... providing your engine is sufficiently clueless, it will actually only consider the correct move

Code: Select all

FEN: q1b4n/1p1pKp2/1p1p1Pp1/1P1P2Pp/2P4P/8/7R/1N1k4 w - - 0 1 

Predateur Bureau:
  2/2	00:00	         118	0	-7,76	Rxd6 Da4
  2/3	00:00	         218	0	-7,56	Cc3+ Rc1 Rxd6
  3/6	00:00	         910	0	-7,59	Cc3+ Rc1 Ca2+ Rd1 Rxd6
  4/8	00:00	       1 933	128 000	-7,58	Cc3+ Rc1 Ce2+ Rc2 Rxd6
  5/10	00:00	       6 073	404 000	-7,58	Cc3+ Rc1 Ca2+ Rd1 Cc3+
  6/16	00:00	      23 807	767 000	-7,63	Cc3+ Re1 Ce4 Da3 Th1+ Re2 Th2+ Re1 Cxd6
  7/16	00:00	      47 637	610 000	-7,60	Cc3+ Re1 Ta2 Db8 Ta1+ Rd2 Ce4+ Rc2 Cxd6 Dc7
  7/12	00:00	      61 479	661 000	-6,63	Rf8 Rc1 Th1+ Rb2 Rg7 Da4 Rxh8 Dxc4
  9/24	00:00	     379 555	1 282 000	-6,52	Ta2 Db8 Cc3+ Re1 Te2+ Rf1 Rd8 Da8 Te8 Da1 Rxc8 Dxc3
 10/19	00:00	     479 852	1 230 000	-5,51	Ta2 Db8 Cc3+ Re1 Te2+ Rf1 Rd8 Da8 Te8 Rg2 Txh8 Da3 Rxc8
 11/20	00:00	     577 375	1 233 000	-5,53	Ta2 Db8 Cc3+ Re1 Te2+ Rf1 Rd8 Da8 Te8 Rg2 Txh8 Da3 Rxc8 Dxc3
 12/26	00:00	     806 142	1 323 000	-5,42	Ta2 Db8 Cc3+ Re1 Te2+ Rf1 Rd8 Da8 Te8 Rg2 Txh8 Rh3 Rc7 Da3 Rxb6
 13/26	00:01	   1 170 111	1 337 000	-5,45	Ta2 Db8 Cc3+ Re1 Te2+ Rf1 Rd8 Rg1 Te8 Rg2 Txh8 Rh3 Ca4 Da7 Cxb6 Dxb6+ Rxc8
 14/28	00:01	   1 804 847	1 392 000	-5,45	Ta2 Db8 Cc3+ Re1 Te2+ Rf1 Rd8 Rg1 Te8 Rg2 Txh8 Rh3 Ca4 Da7 Cxb6 Dxb6+ Rxc8 Rxh4
 15/32	00:02	   3 739 969	1 477 000	-5,41	Ta2 Db8 Rd8 Re1 Cc3
 16/32	00:03	   5 443 783	1 457 000	-5,41	Ta2 Db8 Rd8 Re1 Cc3 Rf1 Te2 Rg1 Te8 Rg2 Txh8 Rh3 Ca4
 17/41	00:06	   9 842 794	1 468 000	-4,39	Ta2 Db8 Cc3+ Re1 Rd8 Rf1 Ca4 Da8 Rc7 Da5 Rxc8 De1 Rxb7 Dxh4 Ta1+ Rg2 Ta2+ Rg3 Cxb6
 18/38	00:10	  14 740 920	1 440 000	-4,39	Ta2 Db8 Cc3+ Rc1 Ce2+ Rb1 Cc3+ Rc1 Ce2+ Rd1 Cc3+ Rc1 Ce2+ Rb1 Cc3+ Rc1 Ce2+ Rb1 Cc3+ Rc1 Ce2+ Rd1 Cc3+ Re1
 19/43	00:21	  29 438 865	1 378 000	-4,39	Ta2 Db8 Cc3+ Re1 Rd8 Rf1 Ca4 Da8 Cc3 Db8

[lech]

My computer is busy, but it looks like another example of the use VALUE_DRAW = 0 in the engines. It is impossible to deny zero. The remaining problem (cascade of zugzwangs) is not too deep for some engines rather.
There are no positions too deep but imperfect programs.
God forbid from artificial intelligence!

[Steve Maughan]

LOL!!

[Steve Maughan]

It's so tough for a computer to find the true solution. The problem is zugzwang coupled with the motif of forcing the king into the corner. Unfortunately, computers are poor at both.

Having looked at the position using all the top engines (Stockfish is my favorite for such positions), they are not even close to seeing the win.

Cheers,

Steve

[lech]

The rules of chess are the only one. The method of finding the best move is based on these rules. Evoluation gives bonuses and penalties based on the 50 + 1% chance (it is a strategy). The search should verifiy it (it is a tactic). Hence there are no strange and difficult positions.
Once the Egyptian priests had authority because they had knowledge.

[JuLieN]

Last iteration before I interrupted the search.

Code: Select all

FEN: q1b4n/1p1pKp2/1p1p1Pp1/1P1P2Pp/2P4P/8/7R/1N1k4 w - - 0 1 

Predateur Bureau:
  2/2	00:00	         118	0	-7,76	Rxd6 Da4
  2/3	00:00	         218	0	-7,56	Cc3+ Rc1 Rxd6
  3/6	00:00	         910	0	-7,59	Cc3+ Rc1 Ca2+ Rd1 Rxd6
  4/8	00:00	       1 933	128 000	-7,58	Cc3+ Rc1 Ce2+ Rc2 Rxd6
  5/10	00:00	       6 073	404 000	-7,58	Cc3+ Rc1 Ca2+ Rd1 Cc3+
  6/16	00:00	      23 807	767 000	-7,63	Cc3+ Re1 Ce4 Da3 Th1+ Re2 Th2+ Re1 Cxd6
  7/16	00:00	      47 637	610 000	-7,60	Cc3+ Re1 Ta2 Db8 Ta1+ Rd2 Ce4+ Rc2 Cxd6 Dc7
  7/12	00:00	      61 479	661 000	-6,63	Rf8 Rc1 Th1+ Rb2 Rg7 Da4 Rxh8 Dxc4
  9/24	00:00	     379 555	1 282 000	-6,52	Ta2 Db8 Cc3+ Re1 Te2+ Rf1 Rd8 Da8 Te8 Da1 Rxc8 Dxc3
 10/19	00:00	     479 852	1 230 000	-5,51	Ta2 Db8 Cc3+ Re1 Te2+ Rf1 Rd8 Da8 Te8 Rg2 Txh8 Da3 Rxc8
 11/20	00:00	     577 375	1 233 000	-5,53	Ta2 Db8 Cc3+ Re1 Te2+ Rf1 Rd8 Da8 Te8 Rg2 Txh8 Da3 Rxc8 Dxc3
 12/26	00:00	     806 142	1 323 000	-5,42	Ta2 Db8 Cc3+ Re1 Te2+ Rf1 Rd8 Da8 Te8 Rg2 Txh8 Rh3 Rc7 Da3 Rxb6
 13/26	00:01	   1 170 111	1 337 000	-5,45	Ta2 Db8 Cc3+ Re1 Te2+ Rf1 Rd8 Rg1 Te8 Rg2 Txh8 Rh3 Ca4 Da7 Cxb6 Dxb6+ Rxc8
 14/28	00:01	   1 804 847	1 392 000	-5,45	Ta2 Db8 Cc3+ Re1 Te2+ Rf1 Rd8 Rg1 Te8 Rg2 Txh8 Rh3 Ca4 Da7 Cxb6 Dxb6+ Rxc8 Rxh4
 15/32	00:02	   3 739 969	1 477 000	-5,41	Ta2 Db8 Rd8 Re1 Cc3
 16/32	00:03	   5 443 783	1 457 000	-5,41	Ta2 Db8 Rd8 Re1 Cc3 Rf1 Te2 Rg1 Te8 Rg2 Txh8 Rh3 Ca4
 17/41	00:06	   9 842 794	1 468 000	-4,39	Ta2 Db8 Cc3+ Re1 Rd8 Rf1 Ca4 Da8 Rc7 Da5 Rxc8 De1 Rxb7 Dxh4 Ta1+ Rg2 Ta2+ Rg3 Cxb6
 18/38	00:10	  14 740 920	1 440 000	-4,39	Ta2 Db8 Cc3+ Rc1 Ce2+ Rb1 Cc3+ Rc1 Ce2+ Rd1 Cc3+ Rc1 Ce2+ Rb1 Cc3+ Rc1 Ce2+ Rb1 Cc3+ Rc1 Ce2+ Rd1 Cc3+ Re1
 19/43	00:21	  29 438 865	1 378 000	-4,39	Ta2 Db8 Cc3+ Re1 Rd8 Rf1 Ca4 Da8 Cc3 Db8
 22/63	37:21	2 974 053 079	1 326 000	-1,94	Ta2 Dxa2 Cc3+ Rc1 Cxa2+ Rb2 Rxd6 Rxa2 Rc7 Rb3
 23/61	1:40:49	8 256 023 327	1 364 000	-1,68	Ta2 Dxa2 Cc3+ Rc2 Cxa2 Rb3 Cc1+ Rc3 Rd8 Rxc4 Rxc8 Rxb5 Ce2 Rc4 Cf4 Rd4 Rxb7 b5 Rb6 Re3 Ch3 Rf3 Ra6 b4 Rb7
 24/67	4:03:14	19 656 938 596	1 346 000	-1,89	Ta2 Dxa2 Cc3+ Rc1 Cxa2+ Rb2 Rd8 Rxa2 Rxc8 Rb3 Rxd7 Rxc4 Rxd6 Rd4 Rc7 Rxd5 Rxb7 Rc5 Ra6 Rd4 Rxb6 Re5

Not quite the solution yet, but the eval keeps increasing...

[lech]

Last iteration before I interrupted the search.

Code: Select all

FEN: q1b4n/1p1pKp2/1p1p1Pp1/1P1P2Pp/2P4P/8/7R/1N1k4 w - - 0 1 

Predateur Bureau:
  2/2	00:00	         118	0	-7,76	Rxd6 Da4
  2/3	00:00	         218	0	-7,56	Cc3+ Rc1 Rxd6
  3/6	00:00	         910	0	-7,59	Cc3+ Rc1 Ca2+ Rd1 Rxd6
  4/8	00:00	       1 933	128 000	-7,58	Cc3+ Rc1 Ce2+ Rc2 Rxd6
  5/10	00:00	       6 073	404 000	-7,58	Cc3+ Rc1 Ca2+ Rd1 Cc3+
  6/16	00:00	      23 807	767 000	-7,63	Cc3+ Re1 Ce4 Da3 Th1+ Re2 Th2+ Re1 Cxd6
  7/16	00:00	      47 637	610 000	-7,60	Cc3+ Re1 Ta2 Db8 Ta1+ Rd2 Ce4+ Rc2 Cxd6 Dc7
  7/12	00:00	      61 479	661 000	-6,63	Rf8 Rc1 Th1+ Rb2 Rg7 Da4 Rxh8 Dxc4
  9/24	00:00	     379 555	1 282 000	-6,52	Ta2 Db8 Cc3+ Re1 Te2+ Rf1 Rd8 Da8 Te8 Da1 Rxc8 Dxc3
 10/19	00:00	     479 852	1 230 000	-5,51	Ta2 Db8 Cc3+ Re1 Te2+ Rf1 Rd8 Da8 Te8 Rg2 Txh8 Da3 Rxc8
 11/20	00:00	     577 375	1 233 000	-5,53	Ta2 Db8 Cc3+ Re1 Te2+ Rf1 Rd8 Da8 Te8 Rg2 Txh8 Da3 Rxc8 Dxc3
 12/26	00:00	     806 142	1 323 000	-5,42	Ta2 Db8 Cc3+ Re1 Te2+ Rf1 Rd8 Da8 Te8 Rg2 Txh8 Rh3 Rc7 Da3 Rxb6
 13/26	00:01	   1 170 111	1 337 000	-5,45	Ta2 Db8 Cc3+ Re1 Te2+ Rf1 Rd8 Rg1 Te8 Rg2 Txh8 Rh3 Ca4 Da7 Cxb6 Dxb6+ Rxc8
 14/28	00:01	   1 804 847	1 392 000	-5,45	Ta2 Db8 Cc3+ Re1 Te2+ Rf1 Rd8 Rg1 Te8 Rg2 Txh8 Rh3 Ca4 Da7 Cxb6 Dxb6+ Rxc8 Rxh4
 15/32	00:02	   3 739 969	1 477 000	-5,41	Ta2 Db8 Rd8 Re1 Cc3
 16/32	00:03	   5 443 783	1 457 000	-5,41	Ta2 Db8 Rd8 Re1 Cc3 Rf1 Te2 Rg1 Te8 Rg2 Txh8 Rh3 Ca4
 17/41	00:06	   9 842 794	1 468 000	-4,39	Ta2 Db8 Cc3+ Re1 Rd8 Rf1 Ca4 Da8 Rc7 Da5 Rxc8 De1 Rxb7 Dxh4 Ta1+ Rg2 Ta2+ Rg3 Cxb6
 18/38	00:10	  14 740 920	1 440 000	-4,39	Ta2 Db8 Cc3+ Rc1 Ce2+ Rb1 Cc3+ Rc1 Ce2+ Rd1 Cc3+ Rc1 Ce2+ Rb1 Cc3+ Rc1 Ce2+ Rb1 Cc3+ Rc1 Ce2+ Rd1 Cc3+ Re1
 19/43	00:21	  29 438 865	1 378 000	-4,39	Ta2 Db8 Cc3+ Re1 Rd8 Rf1 Ca4 Da8 Cc3 Db8
 22/63	37:21	2 974 053 079	1 326 000	-1,94	Ta2 Dxa2 Cc3+ Rc1 Cxa2+ Rb2 Rxd6 Rxa2 Rc7 Rb3
 23/61	1:40:49	8 256 023 327	1 364 000	-1,68	Ta2 Dxa2 Cc3+ Rc2 Cxa2 Rb3 Cc1+ Rc3 Rd8 Rxc4 Rxc8 Rxb5 Ce2 Rc4 Cf4 Rd4 Rxb7 b5 Rb6 Re3 Ch3 Rf3 Ra6 b4 Rb7
 24/67	4:03:14	19 656 938 596	1 346 000	-1,89	Ta2 Dxa2 Cc3+ Rc1 Cxa2+ Rb2 Rd8 Rxa2 Rxc8 Rb3 Rxd7 Rxc4 Rxd6 Rd4 Rc7 Rxd5 Rxb7 Rc5 Ra6 Rd4 Rxb6 Re5

Not quite the solution yet, but the eval keeps increasing...

Because the last score was under zero we don't know why the move Qb8 (after 1. Ra2) was rejected. It is very likely that Qb8 was marked as draw, enough.

[Jouni]

In case you are not reading CSS-Forum there is solution by who else Houdini 2.0c with + score after 26m!