No chance

[Lyudmil Tsvetkov]

[[d]1r3rk1/5p1p/1qb1pPp1/3pP3/p1pP4/PpP2N1R/1Pn1B2P/3RB2K w - - 0 8
I am very much afraid this is mate in 8.

Congrats, Lyudmil!

Didn't watch the thread till now, but even if you were a little bit too optimistic with the #8-announcement at this point, you indeed showed once again, that engines are blind at certain spots, especially if it comes to Zugzwang

It is not only zugzwang, Peter, there are many other engine weak spots.

For me, the main question is why do engines evaluate so wrongly the root position?

[peter]

It is not only zugzwang, Peter, there are many other engine weak spots.

For me, the main question is why do engines evaluate so wrongly the root position?

In this example because of too much pruning mainly, not only nullmove but other cuts and reductions too, of course.

If the tactical points get pruned off, eval cannot be correct, but that's a question of search here quite typically, even if search depends on eval, that's the old thing with hen and egg.


BTW, that it's nullmove mainly here you see trying engines you can switch it off.
Here is such a SF with deactivated nullmove at the position after 25...Rb7:

284: Tsvetkov,L - SF,L
[d]5rk1/1r3p2/1qb1pPpR/3pP1N1/p1pP4/PpP5/1Pn1B2P/3RB2K w - - 0 1

Analysis by SF 171114IPx 64 POPCNTe:

...
26.Bf2 Qc7 27.Rg1 Nxd4 28.Bxd4 Qxe5 29.Bxe5 d4+ 30.Bf3 Rb5 31.Bxc6 Rxe5 32.Rh7 Rf5 33.Rg7+ Kh8 34.Nxf7+ Rxf7 35.Rxf7 Kg8 36.Be8 g5 37.cxd4 c3 38.bxc3 e5 39.Rb7 exd4 40.cxd4 Rf2 41.Rxg5+ Kf8 42.Bxa4 b2 43.Rh5
+- (5.40) Depth: 30/46 00:00:33 531MN
26.Bf2
+- (5.43 ++) Depth: 31/46 00:00:34 562MN
26.Bf2
+- (5.46 ++) Depth: 31/46 00:00:35 581MN

[Vinvin]

To tell you the truth, the win was easy, as the white position is simply gorgeous. The full 2 pawns white edge are really there, but why engines do not see it? SF's e3 capture might not have been optimal, but Nc2 instead of Nf5 was really the bigger mistake. You need that knight close to the king to protect it, as it is its only defender.

Because engines lack of mate pattern recognition.
Or mate extension against a locked up king.
In this categories, Stockfish is even worse than other engines.

[Ignacio]

FEN: 5rk1/1r3p2/1qb1pPpR/3pP1N1/p1pP4/PpP5/1Pn1B2P/3RB2K w - - 0 1

Spark-0.4-win64-mp-corei:

20/44 00:27 342.762.790 12.428.398 -1,60 Ae1-f2
20/44 00:32 400.467.902 12.447.715 -0,85 Ae1-f2
20/44 00:45 571.900.004 12.517.510 0,00 Ae1-f2 Cc2-b4 c3xb4 Ac6-b5 Ae2-h5 Ab5-e8 Th6-h7 Db6-a7 Th7-g7+ Rg8-h8 Tg7-h7+ Rh8-g8
21/44 00:49 611.797.286 12.418.245 0,00 Ae1-f2 Cc2-b4 c3xb4 Ac6-b5 Ae2-h5 c4-c3 Ah5xg6 f7xg6 Th6xg6+ Rg8-h8 Tg6-h6+ Rh8-g8 Th6-g6+
22/44 00:53 668.475.585 12.411.355 0,00 Ae1-f2 Cc2-b4 c3xb4 Ac6-b5 Ae2-h5 c4-c3 Ah5xg6 f7xg6 Th6xg6+ Rg8-h8 Tg6-h6+ Rh8-g8 Th6-g6+
23/46 01:01 765.492.373 12.437.525 +0,25 Ae1-f2
23/46 01:04 800.535.970 12.447.498 +0,75 Ae1-f2
23/46 02:23 1.770.196.799 12.374.931 +5,50 Ae1-f2
23/48 03:41 2.852.042.812 12.879.644 +9,42 Ae1-f2 Db6-a5 Td1-g1 Cc2xd4 Af2xd4 Da5xc3 b2xc3 b3-b2 Tg1-b1 Tf8-b8 Th6-h7 Ac6-e8 Ad4-c5 d5-d4 Th7-g7+ Rg8-h8 Ac5xd4 Tb7-b3 Cg5xf7+ Ae8xf7 Tg7xf7 Tb3xa3 Ae2xc4 Ta3-a1 Ac4-d3 Tb8-g8 Tf7-a7 a4-a3 Rh1-g2 Ta1xb1 Ad3xb1
24/52 03:52 2.983.617.982 12.847.408 +9,42 Ae1-f2 Db6-a5 Td1-g1 Cc2xd4 Af2xd4 Da5xc3 b2xc3 b3-b2 Tg1-b1 Tf8-b8 Th6-h7 Ac6-e8 Ad4-c5 d5-d4 Th7-g7+ Rg8-h8 Ac5xd4 Tb7-b3 Cg5xf7+ Ae8xf7 Tg7xf7 Tb3xa3 Ae2xc4 Ta3-a1 Ac4-d3 Tb8-g8 Tf7-a7 a4-a3 Rh1-g2 Ta1xb1 Ad3xb1
25/54 04:18 3.328.943.095 12.890.389 +9,67 Ae1-f2
25/54 04:22 3.372.640.741 12.851.974 +10,07 Ae1-f2 Db6-a5 Td1-g1 Cc2xd4 Af2xd4 Da5xc3 b2xc3 b3-b2 Tg1-b1 Tf8-b8 Th6-h7 Ac6-e8 Ad4-c5 d5-d4 Th7-g7+ Rg8-h8 Ac5xd4 Tb7-c7 Tg7-h7+ Rh8-g8 Ae2-f3 Tb8-b5 Th7-g7+ Rg8-h8 Cg5xe6 Tc7-c8 Ce6-g5 Tc8-c7 Tg7-h7+ Rh8-g8 h2-h4 Ae8-c6 Af3xc6 Tc7xc6 Th7-g7+ Rg8-f8 Tg7xf7+ Rf8-g8
26/56 04:41 3.582.999.334 12.719.744 +10,09 Ae1-f2 Db6-a5 Td1-g1 Cc2xd4 Af2xd4 Da5xc3 b2xc3 b3-b2 Tg1-b1 Tf8-b8 Th6-h7 Ac6-e8 Ad4-c5 d5-d4 Th7-g7+ Rg8-h8 Ac5xd4 Tb7-d7 Ae2xc4 Tb8-b7 Ac4-a6 Tb7-b8 Rh1-g2 Td7-c7 Cg5xe6 Tc7-c6 Ce6-c5 Tb8-b6 Aa6-c4 Tc6-c7
27/56 05:03 3.803.488.188 12.539.812 +10,34 Ae1-f2
27/56 05:36 4.177.228.573 12.431.635 +10,62 Ae1-f2 Db6-a5 Td1-g1 Cc2xd4 Af2xd4 Da5xc3 b2xc3 b3-b2 Tg1-b1 Tf8-b8 Th6-h7 Ac6-e8 Ad4-c5 d5-d4 Th7-g7+ Rg8-h8 Ac5xd4 Tb7-d7 Ae2xc4 Tb8-b7 Ac4-a6 Tb7-b8 Rh1-g2 Td7-c7 Cg5xe6 Tc7-c6 Ce6-c5 Tb8-b6 Aa6-c4 Tc6-c7 Cc5xa4 Ae8xa4 Ad4xb6 Tc7xc4 Tg7xf7
28/62 06:41 4.918.194.839 12.259.077 +10,71 Ae1-f2 Db6-a5 Td1-g1 Cc2xd4 Af2xd4 Da5xc3 b2xc3 b3-b2 Tg1-b1 Tf8-b8 Th6-h7 Ac6-e8 Ad4-c5 d5-d4 Th7-g7+ Rg8-h8 Ac5xd4 Tb7-d7 Ae2xc4 Tb8-b7 Ac4-a6 Tb7-b8 Rh1-g2 Td7-c7 Cg5xe6 Tc7-c6 Ce6-c5 Tb8-b6 Aa6-d3 Tc6-c8 e5-e6 f7xe6
29/62 07:41 5.501.819.373 11.926.021 +10,96 Ae1-f2
29/62 08:43 6.103.094.501 11.655.802 +11,46 Ae1-f2
29/62 19:11 12.890.544.754 11.196.385 +16,21 Ae1-f2
29/68 22:51 15.274.311.133 11.136.938 +16,99 Ae1-f2 Db6-a5 Td1-g1 Cc2xd4 Af2xd4 Da5xc3 b2xc3 b3-b2 Tg1-b1 Tf8-b8 Ae2-h5 Ac6-e8 Ad4-c5 d5-d4 Ah5-f3 Ae8-b5 Af3xb7 d4xc3 Ab7-e4 Ab5-c6 Th6-h7 Tb8-b7 Th7-g7+ Rg8-h8 Tg7xf7 Tb7xf7 Cg5xf7+ Rh8-g8 Cf7-g5 Ac6-e8 Ac5-b4 c3-c2 Ae4xc2 c4-c3 Ab4xc3 Ae8-c6+ Rh1-g1
30/68 25:29 16.923.301.322 11.067.759 +17,24 Ae1-f2
30/68 28:14 18.565.173.262 10.958.050 +17,74 Ae1-f2
30/76 1:05:57 44.873.737.882 11.338.729 +28,49 Ae1-f2
30/76 1:24:08 57.190.556.251 11.329.349 +M21 Ae1-f2 Db6-a5 Td1-g1 Cc2-e3 Af2xe3 Da5xc3 b2xc3 b3-b2 Tg1-b1 Tb7-b3 Th6-h7 Tb3xa3 Tb1xb2 Ta3-b3 Th7-g7+ Rg8-h8 Cg5xf7+ Tf8xf7 Tg7xf7 Tb3xb2 Tf7-f8+ Rh8-h7 f6-f7 Tb2-b7 Tf8-h8+ Rh7xh8 f7-f8D+ Rh8-h7 Ae3-g5 Tb7-b1+ Rh1-g2 Tb1-b7 Ag5-f6 g6-g5 Ae2-g4 Tb7-b2+ Rg2-f3 Tb2-f2+ Rf3xf2 Rh7-g6 Df8-g7+

[Milos]

I capture, of course.

[d]1r3rk1/5p2/1qb1pPpR/3pP1N1/p1pP4/PpP5/1Pn1B2P/3RB2K b - - 0 9

Still seeing nothing?

I see nothing (or precisely I see a clear DRAW) - Nxa3 or Ne3

[Lyudmil Tsvetkov]

I capture, of course.

[d]1r3rk1/5p2/1qb1pPpR/3pP1N1/p1pP4/PpP5/1Pn1B2P/3RB2K b - - 0 9

Still seeing nothing?

I see nothing (or precisely I see a clear DRAW) - Nxa3 or Ne3

Does not change anything, look a bit further. It is the structure that makes mate inevitable.

On Ne3, Rd2 Nf5 Rh7, followed by Bg4 wins, while on Na3 - same Bf2.

[Milos]

I capture, of course.

[d]1r3rk1/5p2/1qb1pPpR/3pP1N1/p1pP4/PpP5/1Pn1B2P/3RB2K b - - 0 9

Still seeing nothing?

I see nothing (or precisely I see a clear DRAW) - Nxa3 or Ne3

Does not change anything, look a bit further. It is the structure that makes mate inevitable.

On Ne3, Rd2 Nf5 Rh7, followed by Bg4 wins, while on Na3 - same Bf2.

The other line I didn't fully analyzed, and I don't yet see any win, but first line is clear draw so lets play it.
so Nxa3, Bf2, Qa5:
[d]1r3rk1/5p2/2b1pPpR/q2pP1N1/p1pP4/npP5/1P2BB1P/3R3K w - - 2 11
So your turn?

[Lyudmil Tsvetkov]

I capture, of course.

[d]1r3rk1/5p2/1qb1pPpR/3pP1N1/p1pP4/PpP5/1Pn1B2P/3RB2K b - - 0 9

Still seeing nothing?

I see nothing (or precisely I see a clear DRAW) - Nxa3 or Ne3

Does not change anything, look a bit further. It is the structure that makes mate inevitable.

On Ne3, Rd2 Nf5 Rh7, followed by Bg4 wins, while on Na3 - same Bf2.

The other line I didn't fully analyzed, and I don't yet see any win, but first line is clear draw so lets play it.
so Nxa3, Bf2, Qa5:
[d]1r3rk1/5p2/2b1pPpR/q2pP1N1/p1pP4/npP5/1P2BB1P/3R3K w - - 2 11
So your turn?

The win is there, the order of moves might be different, but the win is there. I must concentrate now on the other game, but later might come back to this.

Please note that I do not have to consider anything, as I know white's advantage in terms of eval after Qf6 is some 5-6 full pawns, meaning mate is close by. White just keeps the structure. Wherever the rabious black knight is sacrificed, a3,b4,e7,d6, white does not capture and simply increases threats on the king. If the knight is sacrificed on e4, after passing through d6, white captures it, and immediately after that closes the game again with Be3.

White combines the following threats in different lines:
- battery of rooks along the h file
- mating with Bh6-g7 and Rh8, with a single rook on the h file needed
- if the f8 rook goes away, f7 pawn falls and white f6 pawn decides the game; that is why SF outwardly strange moves like Rb7, to protect f7, before moving the f8 rook
- in a number of lines, checking with Rg7 first, so the king goes to h8, and then Rh3 is immediately lethal
- if h5 is played, then of course Bh5 wins easily

[Milos]

[d]1r3rk1/5p2/2b1pPpR/q2pP1N1/p1pP4/npP5/1P2BB1P/3R3K w - - 2 11
So your turn?

The win is there, the order of moves might be different, but the win is there. I must concentrate now on the other game, but later might come back to this.

Please note that I do not have to consider anything, as I know white's advantage in terms of eval after Qf6 is some 5-6 full pawns, meaning mate is close by. White just keeps the structure. Wherever the rabious black knight is sacrificed, a3,b4,e7,d6, white does not capture and simply increases threats on the king. If the knight is sacrificed on e4, after passing through d6, white captures it, and immediately after that closes the game again with Be3.

White combines the following threats in different lines:
- battery of rooks along the h file
- mating with Bh6-g7 and Rh8, with a single rook on the h file needed
- if the f8 rook goes away, f7 pawn falls and white f6 pawn decides the game; that is why SF outwardly strange moves like Rb7, to protect f7, before moving the f8 rook
- in a number of lines, checking with Rg7 first, so the king goes to h8, and then Rh3 is immediately lethal
- if h5 is played, then of course Bh5 wins easily

What you obviously fail to observe is that no matter what you do, before white manages to mate its c3 pawn falls and white will either lose rook and bishop in exchange for black queen and knight, or open b file for black tying its rook indefinitely. So no mate picture since white doesn't have material to perform it (there was never one, but you simply don't evaluate threats correctly).

But anyway, take your time, and then come back when you can, and try demonstrating us that mate of yours, I'll congratulate you even if you manage to mate in 100 moves (not 8 as what you were originally claiming), just play the moves instead of talking .

[Terry McCracken]

[d]1r3rk1/5p2/2b1pPpR/q2pP1N1/p1pP4/npP5/1P2BB1P/3R3K w - - 2 11
So your turn?

The win is there, the order of moves might be different, but the win is there. I must concentrate now on the other game, but later might come back to this.

Please note that I do not have to consider anything, as I know white's advantage in terms of eval after Qf6 is some 5-6 full pawns, meaning mate is close by. White just keeps the structure. Wherever the rabious black knight is sacrificed, a3,b4,e7,d6, white does not capture and simply increases threats on the king. If the knight is sacrificed on e4, after passing through d6, white captures it, and immediately after that closes the game again with Be3.

White combines the following threats in different lines:
- battery of rooks along the h file
- mating with Bh6-g7 and Rh8, with a single rook on the h file needed
- if the f8 rook goes away, f7 pawn falls and white f6 pawn decides the game; that is why SF outwardly strange moves like Rb7, to protect f7, before moving the f8 rook
- in a number of lines, checking with Rg7 first, so the king goes to h8, and then Rh3 is immediately lethal
- if h5 is played, then of course Bh5 wins easily

What you obviously fail to observe is that no matter what you do, before white manages to mate its c3 pawn falls and white will either lose rook and bishop in exchange for black queen and knight, or open b file for black tying its rook indefinitely. So no mate picture since white doesn't have material to perform it (there was never one, but you simply don't evaluate threats correctly).

But anyway, take your time, and then come back when you can, and try demonstrating us that mate of yours, I'll congratulate you even if you manage to mate in 100 moves (not 8 as what you were originally claiming), just play the moves instead of talking .

Black hasn't the time, the h-file will silence Black.