Test Position: Krogius-Niemala (Bad Lovisa) 1934

[Marc MP]

[d]r2q1rk1/pb2bp1R/1p4p1/3pP1P1/2p2P2/3BP3/PP1N2P1/R2QK3 b Q -

Black must be careful here to avoid a severe punishment. Can your favorite engine take up the challenge?

[PauloSoare]

[d]r2q1rk1/pb2bp1R/1p4p1/3pP1P1/2p2P2/3BP3/PP1N2P1/R2QK3 b Q -

Black must be careful here to avoid a severe punishment. Can your favorite engine take up the challenge?

Interesting position, Marc.
Loop 13.6 1 core

Athlon 2 cores 3800+ HT=1Gb

New game
r2q1rk1/pb2bp1R/1p4p1/3pP1P1/2p2P2/3BP3/PP1N2P1/R2QK3 b Q - 0 1

Analysis by Loop 13.6 1 core:

1...cxd3 2.Rh1 Rc8 3.Qb3 d4 4.e4 Rc2
-+ (-3.39) Depth: 6 00:00:00 49kN
1...cxd3 2.Rh3 Qd7 3.Kf2 d4 4.e4 Qb5 5.Rc1
-+ (-3.40) Depth: 7 00:00:00 144kN
1...cxd3 2.Rh3 Qd7 3.Kf2 d4 4.e4 Qb5 5.Qb1 Rac8
-+ (-3.28) Depth: 8 00:00:00 234kN
1...cxd3 2.Rh3 Bc8 3.Rh4 Qd7 4.Kf2 d4 5.e4 Qb5 6.Qb3
-+ (-3.31) Depth: 9 00:00:00 635kN
1...cxd3 2.Rh3 Bc8 3.Rh2 Qd7 4.Kf2 d4 5.e4 Qb5 6.Nb3 Be6
-+ (-3.31) Depth: 10 00:00:01 1269kN
1...cxd3 2.Rh3 Bc8 3.Rh2 Qd7 4.Kf2 d4 5.e4 Qg4 6.Qf3 Qxf3+ 7.Nxf3 Bb7 8.Rc1
-+ (-3.22) Depth: 11 00:00:02 2585kN
1...cxd3 2.Kf2 Kxh7 3.Qh1+ Kg8 4.Qh6 Bxg5 5.fxg5 Qd7 6.Rh1 f5 7.gxf6 Kf7 8.Qh7+ Ke8 9.Qxg6+
³ (-0.61) Depth: 12 00:00:05 6471kN
1...Qe8 2.Rh2 cxd3 3.Kf2 Kg7 4.f5 Bxg5 5.f6+ Bxf6 6.exf6+ Kxf6 7.Qf3+ Ke7 8.Qf4 Rc8 9.Rh7
µ (-1.39) Depth: 12 00:00:22 25515kN
1...Qe8 2.Rh2 cxd3 3.Kf2 Kg7 4.f5 Bxg5 5.f6+ Bxf6 6.exf6+ Kxf6 7.Qf3+ Ke7 8.Rc1 Rc8 9.Rxc8
-+ (-1.61) Depth: 13 00:00:26 29812kN
1...Qe8 2.Bxg6 fxg6 3.Rh6 d4 4.Nxc4 Bb4+ 5.Nd2 Kg7 6.Qb3 Bxd2+ 7.Kxd2 Bxg2 8.Rc1 Rd8 9.Rc2
-+ (-1.53) Depth: 14 00:00:49 55711kN
1...Qe8 2.Kf2 Kxh7 3.Qh1+ Kg8 4.Qh2 f5 5.gxf6 cxd3 6.Rh1 Kf7 7.Qh7+ Ke6 8.Qh3+ Kf7 9.Qh7+
= (0.00) Depth: 15 00:01:25 96929kN
1...Qe8 2.Kf2 Kxh7 3.Qh1+ Kg8 4.Qh2 f5 5.gxf6 Rxf6 6.Rh1 Rxf4+ 7.exf4 cxd3 8.f5 Bh4+ 9.Qxh4
± (0.84) Depth: 16 00:02:59 207mN
1...Bxg5 2.Bxg6 fxg6 3.Rxb7 Bh4+ 4.Kf1 Qc8 5.Nf3 Qxb7 6.Nxh4 Kg7 7.Qg4 Qf7 8.e6 Qf6 9.Rd1
= (-0.05) Depth: 16 00:04:14 298mN
1...Bxg5 2.Bxg6 fxg6 3.Rxb7 Bh4+ 4.Kf1 Qc8 5.Nf3 Qxb7 6.Nxh4 g5 7.Qg4 Qg7 8.Qe6+ Qf7 9.Qg4
= (0.00) Depth: 17 00:05:40 401mN

(Soares, Residencia 22.05.2007)

[rightrook]

[d]r2q1rk1/pb2bp1p/1p4p1/3pP1P1/2p2P2/3BP3/PP1N2P1/R2QK2R w Q - 0 1

If we go back one move...we have another good test move for white......Rxh7...!



regards

Robert

[Marc MP]

[d]r2q1rk1/pb2bp1R/1p4p1/3pP1P1/2p2P2/3BP3/PP1N2P1/R2QK3 b Q -

Black must be careful here to avoid a severe punishment. Can your favorite engine take up the challenge?

Interesting position, Marc.
Loop 13.6 1 core

Athlon 2 cores 3800+ HT=1Gb

New game
r2q1rk1/pb2bp1R/1p4p1/3pP1P1/2p2P2/3BP3/PP1N2P1/R2QK3 b Q - 0 1

Analysis by Loop 13.6 1 core:

1...cxd3 2.Rh1 Rc8 3.Qb3 d4 4.e4 Rc2
-+ (-3.39) Depth: 6 00:00:00 49kN
1...cxd3 2.Rh3 Qd7 3.Kf2 d4 4.e4 Qb5 5.Rc1
-+ (-3.40) Depth: 7 00:00:00 144kN
1...cxd3 2.Rh3 Qd7 3.Kf2 d4 4.e4 Qb5 5.Qb1 Rac8
-+ (-3.28) Depth: 8 00:00:00 234kN
1...cxd3 2.Rh3 Bc8 3.Rh4 Qd7 4.Kf2 d4 5.e4 Qb5 6.Qb3
-+ (-3.31) Depth: 9 00:00:00 635kN
1...cxd3 2.Rh3 Bc8 3.Rh2 Qd7 4.Kf2 d4 5.e4 Qb5 6.Nb3 Be6
-+ (-3.31) Depth: 10 00:00:01 1269kN
1...cxd3 2.Rh3 Bc8 3.Rh2 Qd7 4.Kf2 d4 5.e4 Qg4 6.Qf3 Qxf3+ 7.Nxf3 Bb7 8.Rc1
-+ (-3.22) Depth: 11 00:00:02 2585kN
1...cxd3 2.Kf2 Kxh7 3.Qh1+ Kg8 4.Qh6 Bxg5 5.fxg5 Qd7 6.Rh1 f5 7.gxf6 Kf7 8.Qh7+ Ke8 9.Qxg6+
³ (-0.61) Depth: 12 00:00:05 6471kN
1...Qe8 2.Rh2 cxd3 3.Kf2 Kg7 4.f5 Bxg5 5.f6+ Bxf6 6.exf6+ Kxf6 7.Qf3+ Ke7 8.Qf4 Rc8 9.Rh7
µ (-1.39) Depth: 12 00:00:22 25515kN
1...Qe8 2.Rh2 cxd3 3.Kf2 Kg7 4.f5 Bxg5 5.f6+ Bxf6 6.exf6+ Kxf6 7.Qf3+ Ke7 8.Rc1 Rc8 9.Rxc8
-+ (-1.61) Depth: 13 00:00:26 29812kN
1...Qe8 2.Bxg6 fxg6 3.Rh6 d4 4.Nxc4 Bb4+ 5.Nd2 Kg7 6.Qb3 Bxd2+ 7.Kxd2 Bxg2 8.Rc1 Rd8 9.Rc2
-+ (-1.53) Depth: 14 00:00:49 55711kN
1...Qe8 2.Kf2 Kxh7 3.Qh1+ Kg8 4.Qh2 f5 5.gxf6 cxd3 6.Rh1 Kf7 7.Qh7+ Ke6 8.Qh3+ Kf7 9.Qh7+
= (0.00) Depth: 15 00:01:25 96929kN
1...Qe8 2.Kf2 Kxh7 3.Qh1+ Kg8 4.Qh2 f5 5.gxf6 Rxf6 6.Rh1 Rxf4+ 7.exf4 cxd3 8.f5 Bh4+ 9.Qxh4
± (0.84) Depth: 16 00:02:59 207mN
1...Bxg5 2.Bxg6 fxg6 3.Rxb7 Bh4+ 4.Kf1 Qc8 5.Nf3 Qxb7 6.Nxh4 Kg7 7.Qg4 Qf7 8.e6 Qf6 9.Rd1
= (-0.05) Depth: 16 00:04:14 298mN
1...Bxg5 2.Bxg6 fxg6 3.Rxb7 Bh4+ 4.Kf1 Qc8 5.Nf3 Qxb7 6.Nxh4 g5 7.Qg4 Qg7 8.Qe6+ Qf7 9.Qg4
= (0.00) Depth: 17 00:05:40 401mN

(Soares, Residencia 22.05.2007)

Hi Paulo,

Yes I thought the position was interesting because usually engines initially think that cxd3 wins for black, then switch to Qe8 thinking it is also much better for black, then see the danger in the line beginning with Qe8 Kf2! and finally switch to the correct Bxg5! So there are two temping moves to avoid.

Have a good day,

[Cubeman]

Tested some PDA programs on the first position.
Toga ll 1.2.1a was the only one to find Bxg5 and took 30 minutes.
Hiarcs 9.6 has been on it for about 2 hours and thinks cxd3 is the best with a 1.90 advantge to black.
PocketFritz2 has also been on it for 2 hours and likes cxd3 with a 3.10 advantage to black.

[PauloSoare]

Hi Paulo,

Yes I thought the position was interesting because usually engines initially think that cxd3 wins for black, then switch to Qe8 thinking it is also much better for black, then see the danger in the line beginning with Qe8 Kf2! and finally switch to the correct Bxg5! So there are two temping moves to avoid.

Have a good day,

Some programs analyses d4, that loses.

Paulo Soares

[PauloSoare]

I don´t know if it´s a good test move, because I am not sure that Rxh7
it´s the best move for whites.

Paulo Soares

[Cubeman]

Hi Paulo,

Yes I thought the position was interesting because usually engines initially think that cxd3 wins for black, then switch to Qe8 thinking it is also much better for black, then see the danger in the line beginning with Qe8 Kf2! and finally switch to the correct Bxg5! So there are two temping moves to avoid.

Have a good day,

Some programs analyses d4, that loses.

Paulo Soares

Just not sure that d4 loses, as White can force a perpetual after Qg4 Kxh7 Qh5+ Kg8 Bxg6 fxg6 Qxg6+ etc.There is no time for the Kf2 manoever or Ke2.I think alot of engines like Black in the Bxg5 variation but to me it is unclear as Black is the exchange ahead but 2 pawns down.
Please give a winning line for White after d4?!

[Cubeman]

[d]r2q1rk1/pb2bp1R/1p4p1/3pP1P1/2p2P2/3BP3/PP1N2P1/R2QK3 b Q -

Black must be careful here to avoid a severe punishment. Can your favorite engine take up the challenge?

Interesting position, Marc.
Loop 13.6 1 core

Athlon 2 cores 3800+ HT=1Gb

New game
r2q1rk1/pb2bp1R/1p4p1/3pP1P1/2p2P2/3BP3/PP1N2P1/R2QK3 b Q - 0 1

Analysis by Loop 13.6 1 core:

1...cxd3 2.Rh1 Rc8 3.Qb3 d4 4.e4 Rc2
-+ (-3.39) Depth: 6 00:00:00 49kN
1...cxd3 2.Rh3 Qd7 3.Kf2 d4 4.e4 Qb5 5.Rc1
-+ (-3.40) Depth: 7 00:00:00 144kN
1...cxd3 2.Rh3 Qd7 3.Kf2 d4 4.e4 Qb5 5.Qb1 Rac8
-+ (-3.28) Depth: 8 00:00:00 234kN
1...cxd3 2.Rh3 Bc8 3.Rh4 Qd7 4.Kf2 d4 5.e4 Qb5 6.Qb3
-+ (-3.31) Depth: 9 00:00:00 635kN
1...cxd3 2.Rh3 Bc8 3.Rh2 Qd7 4.Kf2 d4 5.e4 Qb5 6.Nb3 Be6
-+ (-3.31) Depth: 10 00:00:01 1269kN
1...cxd3 2.Rh3 Bc8 3.Rh2 Qd7 4.Kf2 d4 5.e4 Qg4 6.Qf3 Qxf3+ 7.Nxf3 Bb7 8.Rc1
-+ (-3.22) Depth: 11 00:00:02 2585kN
1...cxd3 2.Kf2 Kxh7 3.Qh1+ Kg8 4.Qh6 Bxg5 5.fxg5 Qd7 6.Rh1 f5 7.gxf6 Kf7 8.Qh7+ Ke8 9.Qxg6+
³ (-0.61) Depth: 12 00:00:05 6471kN
1...Qe8 2.Rh2 cxd3 3.Kf2 Kg7 4.f5 Bxg5 5.f6+ Bxf6 6.exf6+ Kxf6 7.Qf3+ Ke7 8.Qf4 Rc8 9.Rh7
µ (-1.39) Depth: 12 00:00:22 25515kN
1...Qe8 2.Rh2 cxd3 3.Kf2 Kg7 4.f5 Bxg5 5.f6+ Bxf6 6.exf6+ Kxf6 7.Qf3+ Ke7 8.Rc1 Rc8 9.Rxc8
-+ (-1.61) Depth: 13 00:00:26 29812kN
1...Qe8 2.Bxg6 fxg6 3.Rh6 d4 4.Nxc4 Bb4+ 5.Nd2 Kg7 6.Qb3 Bxd2+ 7.Kxd2 Bxg2 8.Rc1 Rd8 9.Rc2
-+ (-1.53) Depth: 14 00:00:49 55711kN
1...Qe8 2.Kf2 Kxh7 3.Qh1+ Kg8 4.Qh2 f5 5.gxf6 cxd3 6.Rh1 Kf7 7.Qh7+ Ke6 8.Qh3+ Kf7 9.Qh7+
= (0.00) Depth: 15 00:01:25 96929kN
1...Qe8 2.Kf2 Kxh7 3.Qh1+ Kg8 4.Qh2 f5 5.gxf6 Rxf6 6.Rh1 Rxf4+ 7.exf4 cxd3 8.f5 Bh4+ 9.Qxh4
± (0.84) Depth: 16 00:02:59 207mN
1...Bxg5 2.Bxg6 fxg6 3.Rxb7 Bh4+ 4.Kf1 Qc8 5.Nf3 Qxb7 6.Nxh4 Kg7 7.Qg4 Qf7 8.e6 Qf6 9.Rd1
= (-0.05) Depth: 16 00:04:14 298mN
1...Bxg5 2.Bxg6 fxg6 3.Rxb7 Bh4+ 4.Kf1 Qc8 5.Nf3 Qxb7 6.Nxh4 g5 7.Qg4 Qg7 8.Qe6+ Qf7 9.Qg4
= (0.00) Depth: 17 00:05:40 401mN

(Soares, Residencia 22.05.2007)

Hi Paulo,

Yes I thought the position was interesting because usually engines initially think that cxd3 wins for black, then switch to Qe8 thinking it is also much better for black, then see the danger in the line beginning with Qe8 Kf2! and finally switch to the correct Bxg5! So there are two temping moves to avoid.

Have a good day,

The analysis at the end of the above variation after Qc8 Nf3 Qxb7 Nxh4 g5? which is weak Qg4 Qg7 Qe6+?(better is Nf5!) is not good for black either.But instead of g5? better is Qc6.I think black can hold this but I would want to play White.

[Marc MP]

Hi Paulo,

Yes I thought the position was interesting because usually engines initially think that cxd3 wins for black, then switch to Qe8 thinking it is also much better for black, then see the danger in the line beginning with Qe8 Kf2! and finally switch to the correct Bxg5! So there are two temping moves to avoid.

Have a good day,

Some programs analyses d4, that loses.

Paulo Soares

Just not sure that d4 loses, as White can force a perpetual after Qg4 Kxh7 Qh5+ Kg8 Bxg6 fxg6 Qxg6+ etc.There is no time for the Kf2 manoever or Ke2.I think alot of engines like Black in the Bxg5 variation but to me it is unclear as Black is the exchange ahead but 2 pawns down.
Please give a winning line for White after d4?!

In the line with d4, white can castle long instead of playing Kf2 or Ke2 as SlowChess suggests:

SlowChess Blitz WV2.1 (190 MB hash) PIV 3.2 Prescott
FEN: r2q1rk1/pb2bp1R/1p4p1/4P1P1/2pp1P2/3BP3/PP1N2P1/R2QK3 w Q -

[15] 2.60 18:15.6 Qg4 Kxh7 Qh5+ Kg8 Bxg6 fxg6 Qxg6+ Kh8 O-O-O Bxg5 Rh1+ Bh4 Qh6+ Kg8 Rxh4 Qxh4 Qxh4 dxe3 Qg3+ Kf7 Qxe3 Bxg2 Qe2 Bd5 Nxc4
[14] 2.60 08:37.7 Qg4 Kxh7 Qh5+ Kg8 Bxg6 fxg6 Qxg6+ Kh8 O-O-O Bxg5 Rh1+ Bh4 Qh5+ Kg7 Rxh4 Qxh4 Qxh4 dxe3 Qg3+ Kf7 Qxe3 Rad8 f5
[13] 2.52 03:59.1 Qg4 Kxh7 Qh5+ Kg8 Bxg6 fxg6 Qxg6+ Kh8 O-O-O Bxg5 Rh1+ Bh4 Qh5+ Kg7 Rxh4 Qxh4 Qxh4 c3 Qg4+ Kh8 Qh5+ Kg7 Qg5+ Kf7 Qf6+ Kg8 Qg6+ Kh8 Qh6+ Kg8 Nc4
[12] 1.56 01:05.1 Qg4 Kxh7 Qh5+ Kg8 Bxg6 fxg6 Qxg6+ Kh8 O-O-O Bxg5 Rh1+ Bh4 Qh5+ Kg7 Rxh4 Rg8 e6 Qf6 Qh7+ Kf8 Qxb7
[11] 1.20 00:30.2 Qg4 Kxh7 Qh5+ Kg7 Qh6+ Kg8 Bxg6 fxg6 Qxg6+ Kh8 O-O-O Bxg5 Rh1+ Bh4 Qh5+ Kg7 Rxh4 Rg8 Qh7+ Kf8 Qxb7 Rg7
[10] 0.41 00:12.1 Qg4 Kxh7 Qh5+ Kg7 Qh6+ Kg8 Bxg6 fxg6 Qxg6+ Kh8 O-O-O Bxg5 Rh1+ Bh4 Qh6+ Kg8 Qg6+ Kh8 <draw>
[9] 0.41 00:08.8 Qg4 Kxh7 Qh5+ Kg7 Qh6+ <tt>
[8] -0.76 00:03.0 Qg4 dxe3 Qh4 exd2+ Kd1 Bf6 gxf6 Bf3+ gxf3 Qxd3
[8] -0.80 00:02.7 Qg4 {++}
[8] -0.80 00:01.7 Nxc4 b5 Rh6 bxc4 Bxc4 Qb6 Qxd4 Bxg2
[7] -1.52 00:00.5 Nxc4 b5 Nd6 Bxd6 Bxg6 Bb4+ Ke2 dxe3 Kxe3 Bxg2
[6] -1.52 00:00.3 Nxc4 b5 Nd6 Bxd6 Bxg6 Bb4+ Kf2 dxe3+ Kxe3 Bxg2
[6] -2.84 00:00.1 Nxc4 {++}
[6] -2.84 00:00.0 Bxg6 fxg6 Rh4 c3 bxc3 dxc3
[5] -2.68 00:00.0 Bxg6 fxg6 Rh4 Qd5 Qe2
[4] -2.08 00:00.0 Bxg6 fxg6 Rh6 Kg7 Nxc4 dxe3
[3] -2.40 00:00.0 Bxg6 fxg6 Rh6
[2] -1.76 00:00.0 Bxg6 fxg6

Also, It looks like white is still (much?) better in the line with Bxg5 but this is clearly Black's best chance. I had Glaurung-2-minus-epsilon-over-two analyze overnight the position after Bxg5 Bxg6 fxg6 Rxb7, and it evaluates it as +1.03 for white at depth=22. Unfortunately I closed the GUI before I could copy the analysis, but Glaurung was intending to play Bh4+ at this point.