2 questionable test positions

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

Moderator: Ras

Vinvin
Posts: 5305
Joined: Thu Mar 09, 2006 9:40 am
Full name: Vincent Lejeune

2 questionable test positions

Post by Vinvin »

Is Bc8 really better than Nc6 or e6 ?
After 1 hours (6cores@4Ghz) Nc6 still selected but the score is decreasing.

Code: Select all

    40/61+	52:28	 28.655.209k	9.101k	+0,17	1. ... Cc6  2.Dxb7
    40/61-	53:58	 29.480.955k	9.103k	+0,01	1. ... Cc6
    40/61 	58:16	 31.840.944k	9.106k	 0,00	1. ... Cc6  2.Dxb7  Fd7  3.Db3  Tb8  4.Dc2  Cb4  5.Db1  Dc8  6.Fd3  Cxd3+  7.Dxd3  Ff5  8.Dd1  e6  9.Cf3  Fb4  10.Da4+  Cd7  11.0-0  0-0  12.Fd2  Fd6  13.b3  Db7  14.Tfc1  h6  15.Da5  Fb4  16.Da4  Fd6
[d]rn1qkb1r/pp2pppp/5n2/3p1b2/3P4/1QN1P3/PP3PPP/R1B1KBNR b KQkq - bm Bc8

Second position : fortress from Serper vs. Nakamura.
Nxe4 is the way to draw by locking our own king in the corner. But Ng4 seems drawish too.

Code: Select all

    103/101	 46:51	 24.949.029k	8.873k	 0,00	1.Cg4  Rd6  2.Ch6  Re6  3.Cg4  Rd6
    104/101-	52:43	 28.606.068k	9.041k	-0,06	1.Cg4  Fh8
    104/101-	52:47	 28.661.760k	9.049k	-0,12	1.Cg4  Fh8
    104/101-	53:19	 29.128.344k	9.105k	-0,35	1.Cg4  Fh8
    104/101+	54:28	 29.780.403k	9.110k	-0,24	1.Cg4
    104/101-	55:06	 30.399.688k	9.195k	-0,40	1.Cg4  Fg7
    104/101-	55:41	 30.868.049k	9.238k	-0,87	1.Cg4  Fg7
[d]8/8/8/3k2p1/4p3/2b1P1pP/4KNP1/8 w - - bm Nxe4
Top
Paloma
Posts: 1214
Joined: Thu Dec 25, 2008 9:07 pm
Full name: Herbert L

Re: 2 questionable test positions

Post by Paloma »

Hi Vincent,

yes it seems that Ng4 draws

82. Ng4 Bg7 83. Kd2 Kc4 (83... Kc5 84. Kd1 Kc4 85. Kc2 {A (transposition)
}) 84. Kc2 {A} Kb4 85. Kd2! Kb3 86. Kc1 Kc3 87. Kd1! Kd3 88. Ke1! Bc3+ 89. Kf1! Kd2 90. Nh6!

859: Serper - Nakamura, San Diego 2004
8/8/8/6p1/4p1N1/2b1P1pP/3k2P1/5K2 w - - 0 1

Analysis by Stockfish 261114 64:

1. = (0.00): 90.Nh6 Bf6 91.Nf5 Be5 92.Nh6 Bf4 93.Nf5 Be5
2. -+ (-123.44): 90.h4 gxh4 91.Nh6 Kxe3 92.Nf5+ Kd3 93.Nxh4 e3 94.Nf3 e2+ 95.Kg1 Bd4+ 96.Kh1 Bc5
Top
Paloma
Posts: 1214
Joined: Thu Dec 25, 2008 9:07 pm
Full name: Herbert L

Re: 2 questionable test positions

Post by Paloma »

To N°1 :
Image
Top
Vinvin
Posts: 5305
Joined: Thu Mar 09, 2006 9:40 am
Full name: Vincent Lejeune

Re: 2 questionable test positions

Post by Vinvin »

Interesting stat but do you have the Elo perf ?
18 games is not a lot ...
Top
Ferdy
Posts: 4851
Joined: Sun Aug 10, 2008 3:15 pm
Location: Philippines

Re: 2 questionable test positions

Post by Ferdy »

Vinvin wrote:Is Bc8 really better than Nc6 or e6 ?
After 1 hours (6cores@4Ghz) Nc6 still selected but the score is decreasing.

Code: Select all

    40/61+	52:28	 28.655.209k	9.101k	+0,17	1. ... Cc6  2.Dxb7
    40/61-	53:58	 29.480.955k	9.103k	+0,01	1. ... Cc6
    40/61 	58:16	 31.840.944k	9.106k	 0,00	1. ... Cc6  2.Dxb7  Fd7  3.Db3  Tb8  4.Dc2  Cb4  5.Db1  Dc8  6.Fd3  Cxd3+  7.Dxd3  Ff5  8.Dd1  e6  9.Cf3  Fb4  10.Da4+  Cd7  11.0-0  0-0  12.Fd2  Fd6  13.b3  Db7  14.Tfc1  h6  15.Da5  Fb4  16.Da4  Fd6
[d]rn1qkb1r/pp2pppp/5n2/3p1b2/3P4/1QN1P3/PP3PPP/R1B1KBNR b KQkq - bm Bc8
The score is close between Bc8 and Nc6. Nc6 gives away the pawn in b7 in return for active pieces and inactive white's c1 bishop. Can be a sort of a gambit.
After a couple of thousand positions (IDEAPos) so far using sf 301114. DBGames and Correspondence games references are only up to 2012. IDEA scores have been minimaxed.
Image
Top
Ferdy
Posts: 4851
Joined: Sun Aug 10, 2008 3:15 pm
Location: Philippines

Re: 2 questionable test positions

Post by Ferdy »

Bc8 is now considered as best.

Image
Top
Vinvin
Posts: 5305
Joined: Thu Mar 09, 2006 9:40 am
Full name: Vincent Lejeune

Re: 2 questionable test positions

Post by Vinvin »

Vinvin wrote:Is Bc8 really better than Nc6 or e6 ?
After 1 hours (6cores@4Ghz) Nc6 still selected but the score is decreasing.

Code: Select all

    40/61+	52:28	 28.655.209k	9.101k	+0,17	1. ... Cc6  2.Dxb7
    40/61-	53:58	 29.480.955k	9.103k	+0,01	1. ... Cc6
    40/61 	58:16	 31.840.944k	9.106k	 0,00	1. ... Cc6  2.Dxb7  Fd7  3.Db3  Tb8  4.Dc2  Cb4  5.Db1  Dc8  6.Fd3  Cxd3+  7.Dxd3  Ff5  8.Dd1  e6  9.Cf3  Fb4  10.Da4+  Cd7  11.0-0  0-0  12.Fd2  Fd6  13.b3  Db7  14.Tfc1  h6  15.Da5  Fb4  16.Da4  Fd6
[d]rn1qkb1r/pp2pppp/5n2/3p1b2/3P4/1QN1P3/PP3PPP/R1B1KBNR b KQkq - bm Bc8
...
After 17 hours (depth 46/76) with 3 best lines, SF still prefers Nc6 (+0.13) over Bc8 (+0.25) and e6 (+0.23).

Code: Select all

Stockfish_14112600_x64_modern:
...
 44/70	7:01:29	237.317.644k	9.384k	+0,25	1. ... Fc8 2.Cf3 e6 3.Fd3 Cc6 4.0-0 Fe7 5.Fd2 0-0 6.Tfc1 Fd7 7.a3 a6 8.h3 Tc8 9.Ce5 Fd6 10.Cxd7 Dxd7 11.Ca4 Tc7 12.Cb6 De7 13.Tc2 Cd7 14.Tac1 Tb8 15.Cxd7 Dxd7 16.Da4 g6 17.b4 Tbc8 18.Db3 De7 19.a4 Df6 20.b5 axb5 21.axb5 Ce7 22.Txc7 Txc7 23.Txc7 Fxc7
 44/70	7:01:29	237.317.644k	9.384k	+0,23	1. ... e6 2.Dxb7 Cbd7 3.Cf3 Fe7 4.Da6 0-0 5.Fe2 Tb8 6.0-0 Dc8 7.Cd2 Fb4 8.Cb3 Dxa6 9.Fxa6 Tb6 10.Fe2 Fxc3 11.bxc3 Ce4 12.Te1 Cxc3 13.Ff1 h6 14.Fa3 Tfb8 15.f3 Fg6 16.Cc5 Cxc5 17.Fxc5 Tb2 18.Fxa7 Ta8 19.Fc5 Taxa2 20.Txa2 Cxa2 21.Ta1 Cc3 22.h4 Ce2+ 23.Rh2 Cc3 24.Ta8+ Rh7 25.Rg3 Ce2+ 26.Rg4
 44/70	7:01:29	237.317.644k	9.384k	+0,13	1. ... Cc6 2.Dxb7 Fd7 3.Db3 e5 4.dxe5 Cxe5 5.Fb5 Fc5 6.Fxd7+ Dxd7 7.Dd1 0-0 8.Cf3 Cxf3+ 9.Dxf3 Fb4 10.De2 Tfc8 11.0-0 Fxc3 12.bxc3 Ce4 13.Fd2 Tc5 14.Dd3 Tac8 15.Tab1 Cxc3 16.Tbc1 Ce4 17.f3 Cxd2 18.Dxd2 Dc7 19.Txc5 Dxc5 20.Td1 a5 21.h3 a4 22.Dd4 h6 23.Dxc5 Txc5 24.Td4 Tc1+ 25.Rh2 Tc3 26.Rg3 Ta3 27.Td2 Txe3
---------------------------------------------------------------------------
 45/70	9:34:16	322.223.960k	9.352k	+0,44	1. ... e6 2.Dxb7 Cbd7 3.Da6 Tb8 4.Cf3 Fe7 5.Fb5 0-0 6.Fd2 e5 7.Fxd7 Dxd7 8.Cxe5 Dc8 9.Dxc8 Tfxc8 10.Fc1 Fb4 11.0-0 Fxc3 12.bxc3 Txc3 13.a4 Ce8 14.Fa3 f6 15.Cf3 a5 16.Tfc1 Tc4 17.Fc5 Txc1+ 18.Txc1 Tb2 19.Ta1 Fd7 20.h3 Cc7 21.g4 Ce6 22.Rg2 Cxc5 23.dxc5 Tc2 24.Tb1 Txc5 25.Tb8+ Tc8 26.Txc8+ Fxc8 27.Cd4 Fd7
 45/70	9:34:16	322.223.960k	9.352k	+0,24	1. ... Fc8 2.Cf3 Cc6 3.Fd2 e6 4.Fd3 Fd7 5.0-0 Fe7 6.a3 0-0 7.Tfc1 a6 8.h3 Tc8 9.Ce5 Fd6 10.Cxd7 Dxd7 11.Ca4 Tc7 12.Cb6 De7 13.Tc2 Cd7 14.a4 Cxb6 15.Dxb6 Tfc8 16.Tac1 Dd7 17.b4 Ce7 18.b5 Txc2 19.Txc2 Txc2 20.Fxc2 axb5 21.Dxb5 Cc6 22.Fd3 g6 23.g3 h5 24.Rg2 Rg7 25.Fe2 Dc7 26.h4 Rf8 27.Db3 Rg7
 45/70	9:34:16	322.223.960k	9.352k	+0,13	1. ... Cc6 2.Dxb7 Fd7 3.Db3 Tb8 4.Dd1 e5 5.dxe5 Cxe5 6.Fe2 Dc7 7.Cf3 Cxf3+ 8.Fxf3 Fd6 9.Fd2 Txb2 10.Cxd5 Cxd5 11.Fxd5 Dc5 12.Ff3 Da3 13.0-0 Fxh2+ 14.Rxh2 Dd6+ 15.Rg1 Txd2 16.Db3 0-0 17.Tad1 Fe6 18.Txd2 Dxd2 19.Td1 Da5 20.Fd5 Td8 21.Fxe6 Txd1+ 22.Dxd1 fxe6 23.a4 Dd5 24.Dxd5 exd5 25.Rf1 Rf7 26.Re2 Re6 27.Rd3
---------------------------------------------------------------------------
 46/76	17:07:29	577.283.145k	9.364k	+0,30	1. ... e6
 46/76	17:07:29	577.283.145k	9.364k	+0,27	1. ... Fc8 2.Cf3 Cc6 3.Fd2 e6 4.Fd3 Fd7 5.0-0 Fe7 6.Tfc1 0-0 7.a3 a6 8.Dd1 Tc8 9.Ce5 g6 10.Cxd7 Cxd7 11.Fe1 Cb6 12.Ce2 Fd6 13.Db3 Te8 14.g3 Dc7 15.Fb4 Fxb4 16.axb4 Ta8 17.b5 axb5 18.Fxb5 Txa1 19.Txa1 Cd7 20.Cf4 Tc8 21.Cd3 Rg7 22.Rg2 Db6 23.Tc1 Cf6 24.Cc5 Ce7 25.Ta1
 46/76	17:07:29	577.283.145k	9.364k	+0,12	1. ... Cc6 2.Dxb7 Fd7 3.Db3 Tb8 4.Dd1 e5 5.Cf3 Fd6 6.dxe5 Cxe5 7.Cd4 0-0 8.Fe2 Ce4 9.0-0 Cxc3 10.bxc3 Dh4 11.f4 Cg4 12.Fxg4 Fxg4 13.Dd3 Fd7 14.Fd2 Tb6 15.Tfb1 De7 16.Tb3 Fc5 17.Te1 Te8 18.Teb1 Tc8 19.Fe1 De4 20.Dxe4 dxe4 21.f5 f6 22.g4 Rf7 23.h3 g6 24.fxg6+ hxg6 25.Rg2 f5 26.Txb6 Fxb6 27.gxf5 gxf5 28.Fd2 Rf6
Top
Ferdy
Posts: 4851
Joined: Sun Aug 10, 2008 3:15 pm
Location: Philippines

Re: 2 questionable test positions

Post by Ferdy »

Nc6 is favored, after 100k + positions.
Image
Top

Return to “Computer Chess Club: General Topics”