The power of a passed pawn in the opening.

[Ajedrecista]

Hello:

Continuing with my interesting-to-me games, I played today a 5+0 game as black that left a passed pawn as soon as the 10th move:

[d]1Bbqkb1r/5ppp/2p1pn2/3p4/1p1P4/2N5/4PPPP/Q3KBNR w Kk - 0 11

[pgn][Event ""]
[Site ""]
[Date "2025.08.13"]
[White ""]
[Black ""]
[TimeControl "300+0"]
[PlyCount "20"]
[Result "*"]

1. d4 d5 2. c4 e6 3. Bf4 Nf6 4. Nc3 c6 5. c5 b6 6. b4 a5 7. a3 axb4 8. axb4 Rxa1 9. Qxa1 bxc5 10. Bxb8 cxb4 *[/pgn]

The elegant statement would be that I saw a promising passed pawn chance after 9.- Qxa1 and went for 2P + passed pawn vs. B; the honest statement is that I completely overlooked BxN and had to bite the bullet, which happened to be a nice chance after all! To make it more computer chess related, I find moderately surprising that SF partly agrees with the idea of not being that bad of going down in material that early at low human level (SF would prefer other moves, of course, and would lose with my played move). Evals are from side to move POV:

Code: Select all

FEN: 1nbqkb1r/5ppp/1pp1pn2/2Pp4/1P1P1B2/2N5/4PPPP/Q3KBNR b Kk - 0 9

Stockfish-17-windows-x86-64-sse41-popcnt:
[...]
---------------------------------------------------------------------------
 30/56	01:28	  44.926.543	510.511	-1,00	b6xc5 b4xc5 Nb8d7 Ng1f3 Nf6e4 e2e3 g7g5 Bf4e5 f7f6 Be5g3 Ne4xg3 h2xg3 Qd8c7 Bf1d3 g5g4 Nf3g1 f6f5 Ng1e2 Ke8f7 OO Bf8e7 Qa1b2 h7h5 Rf1a1 h5h4 g3xh4 Be7xh4 g2g3 Bh4g5 Ra1a4 Kf7g6 Ne2f4+ Bg5xf4 e3xf4
 30/50	01:28	  44.926.543	510.511	-0,95	Nb8d7 Ng1f3 Nf6e4 e2e3 g7g5 Bf4e5 f7f6 Be5g3 Ne4xg3 h2xg3 g5g4 Nf3g1 Qd8c7 Bf1d3 f6f5 Ng1e2 Bf8g7 OO Ke8f7 Qa1b2 Bc8b7 c5xb6 Qc7xb6 Rf1a1 Rh8a8 Nc3a4 Qb6c7 Ne2f4 Bg7f8 Na4c5 Nd7xc5 b4xc5 Ra8xa1+ Qb2xa1 Bb7c8 Qa1a8
 30/57	01:28	  44.926.543	510.511	-0,18	Nb8a6 Qa1a4 Na6xb4 Qa4xb4 b6xc5 d4xc5 d5d4 Nc3b1 Nf6d5 Qb4xd4 Qd8a5+ Bf4d2 Qa5a2 Qd4d3 Bf8xc5 Ke1d1 Nd5b4 Bd2xb4 Bc5xb4 Qd3c2 Qa2a1 e2e3 OO Bf1c4 Bc8a6 Qc2a2 Qa1xa2 Bc4xa2 Rf8d8+ Kd1c1 Ba6f1 h2h4 Bf1xg2 Rh1h2 Bg2e4 Ng1e2 Bb4d6 Ne2g3 Be4d3 Rh2g2 Bd6e7 Ba2b3 Be7xh4 Bb3c2 Bh4xg3 Rg2xg3 Bd3xc2 Kc1xc2 h7h6 Nb1c3 Kg8h7 Rg3h3 Rd8a8 Rh3h5

What I find more surprising is that 9.- ..., PxP must be replied recapturing the pawn instead of going greedy capturing the knight:

Code: Select all

FEN: 1nbqkb1r/5ppp/2p1pn2/2pp4/1P1P1B2/2N5/4PPPP/Q3KBNR w Kk - 0 10

Stockfish-17-windows-x86-64-sse41-popcnt:
[...]
---------------------------------------------------------------------------
 31/46	01:14	  34.731.027	467.329	-1,84	Bf4xb8 c5xd4 Nc3a2 Qd8b6 Bb8f4 e6e5 Bf4d2 Bf8d6 e2e3 d4xe3 f2xe3 OO Bf1e2 Bc8e6 Qa1b2 c6c5 Ng1f3 d5d4 OO Rf8a8 Na2c1 c5xb4 Nf3g5 Be6d5 e3xd4 h7h6 Ng5h3 b4b3 Nc1xb3 Bd5xb3 Rf1b1
 31/45	01:14	  34.731.027	467.329	+0,10	d4xc5 Nb8a6 Qa1a4 d5d4 Nc3a2 Qd8d5 Ng1f3 Qd5e4 Bf4c1 Bf8e7 e2e3 d4xe3 f2xe3 Na6xc5 Qa4a8 OO b4xc5 Be7xc5 Bf1e2 Nf6g4 Qa8a5 Bc5xe3 Na2c3 Qe4c2 Bc1xe3 Ng4xe3 Rh1g1 e6e5 Qa5xe5 Ne3g4 Qe5d4 Qc2c1+ Nc3d1 Rf8e8 Ke1f1 Ng4f6 g2g4 Nf6d5 Qd4d2 Qc1xd2 Nf3xd2 Nd5f4
 31/54	01:14	  34.731.027	467.329	+1,07	b4xc5 Nb8d7 Ng1f3 Nf6e4 e2e3 g7g5 Bf4e5 Nd7xe5 Nf3xe5 Ne4xc3 Qa1xc3 Qd8c7 Bf1e2 Bf8e7 OO OO Rf1a1 f7f6 Ne5d3 e6e5 Qc3a3 Bc8f5 Nd3b4 Qc7b7 Be2a6 Qb7c7 Ba6f1 e5xd4 e3xd4 Qc7b8

Then, capturing towards the centre is the best move according to SF, being clearly up. I deliberately avoided that to get the passed pawn, which was also defended with my DSB:

Code: Select all

FEN: 1Bbqkb1r/5ppp/2p1pn2/2pp4/1P1P4/2N5/4PPPP/Q3KBNR b Kk - 0 10

Stockfish-17-windows-x86-64-sse41-popcnt:
[...]
---------------------------------------------------------------------------
 32/48	01:30	  45.333.773	500.886	-0,02	c5c4 Qa1a4 Qd8b6 Bb8a7 Qb6b7 Ba7c5 Bf8xc5 d4xc5 OO e2e3 e6e5 Ng1f3 Bc8g4 Qa4a3 Rf8a8 Qa3b2 d5d4 Bf1xc4 d4xc3 Qb2xc3 Nf6e4 Qc3b2 Ra8d8 Bc4e2 Bg4xf3 g2xf3 Ne4d2 Rh1g1 g7g6 Rg1g5 f7f6 Rg5g3 Kg8g7 h2h4 Qb7d7 e3e4 Qd7b7 h4h5 Rd8d4 Qb2c3 Qb7d7
 32/65	01:30	  45.333.773	500.886	+0,32	c5xb4 Nc3b1 b4b3 Bb8f4 e6e5 d4xe5 Nf6e4 Bf4d2 Bf8b4 e2e3 Qd8b6 Ng1f3 b3b2 Qa1a2 Bc8f5 Bf1e2 OO OO Bb4xd2 Nf3xd2 Ne4xd2 Nb1xd2 Rf8b8 Qa2a3 Qb6b4 Qa3xb4 Rb8xb4 f2f3 Bf5c2 Nd2b1 Rb4a4 Be2d1 Bc2xb1 Bd1xa4 Bb1d3 Rf1d1 c6c5 e3e4 d5d4 Ba4b3 b2b1Q Rd1xb1 Bd3xb1 Bb3c4 Bb1c2 Kg1f2 Kg8f8 Kf2e2
 32/56	01:30	  45.333.773	500.886	+1,99	c5xd4 Nc3a2 Bc8a6 Bb8e5 Ba6c4 e2e3 Bc4xa2 Qa1xa2 Bf8xb4+ Ke1d1 OO Be5xd4 Nf6e4 Ng1h3 c6c5 Bd4a1 c5c4 f2f3 Qd8b6 Kd1e2 Ne4d6 g2g4 Bb4c5 Ba1e5 Nd6b5 Ke2f2 Bc5xe3+ Kf2g2 Be3d4 Be5xd4 Qb6xd4 Qa2f2

With the analysis of the position of the diagramme being:

Code: Select all

FEN: 1Bbqkb1r/5ppp/2p1pn2/3p4/1p1P4/2N5/4PPPP/Q3KBNR w Kk - 0 11

Stockfish-17-windows-x86-64-sse41-popcnt:
[...]
---------------------------------------------------------------------------
 35/52	03:43	 109.304.704	490.619	-3,56	Bb8g3 b4xc3 e2e3 Qd8b6 Bf1d3 Qb6b2 Qa1xb2 c3xb2 Bd3b1 Nf6e4 Ke1d1 Bc8a6 Ng1f3 Bf8a3 Kd1c2 Ke8e7 Kc2b3 Ba3d6 Kb3xb2 Rh8b8+ Kb2a1 Ne4xg3 h2xg3 Ba6c4 Bb1c2 Bd6a3 Rh1xh7 Ba3b2+ Ka1b1 Bb2xd4+ Kb1c1 Bd4c3 g3g4 c6c5 g4g5 Ke7d6 g5g6 f7xg6 Nf3g5 Bc3b2+ Kc1d2
 35/58	03:43	 109.304.704	490.619	-1,33	Nc3d1 b4b3 f2f3 Bf8b4+ Ke1f2 Qd8b6 Bb8f4 c6c5 d4xc5 Bb4xc5+ e2e3 d5d4 Bf1c4 e6e5 Bf4xe5 d4xe3+ Kf2f1 OO Ng1e2 Rf8d8 Nd1b2 Bc8e6 Bc4d3 Bc5e7 Qa1c1 Rd8c8 Qc1b1 Nf6d5 g2g4 h7h5 h2h3 Nd5b4 Be5d4 Be7g5
 35/62	03:43	 109.304.704	490.619	-0,29	Nc3b1 b4b3 Bb8f4 e6e5 d4xe5 Nf6e4 Bf4d2 Qd8b6 e2e3 Bf8b4 Ng1f3 b3b2 Qa1a2 Bc8f5 Bf1e2 Ne4xd2 Nf3xd2 OO OO Bb4xd2 Nb1xd2 Rf8b8 Nd2b1 Qb6c5 h2h4 h7h6 Qa2a4 Qc5c2 Qa4a7 Rb8f8 Be2f3 Bf5d3 Rf1d1 c6c5 Bf3xd5

A summary of the evals (from white POV) of the positions of interest is:

Code: Select all

========================================================================
  AFTER                BEST              2nd BEST            3rd BEST
========================================================================
 9.- Qxa1              +0.18              +0.95               +1.00
                  (9.- ..., Na6)     (9.- ..., Nbd7)     (9.- ..., bxc5)
------------------------------------------------------------------------
 9.- ..., bxc5         +1.07              +0.10               -1.84
                    (10.- bxc5)        (10.- dxc5)          (10.- Bxb8)
------------------------------------------------------------------------
10.- Bxb8              -1.99              -0.32               +0.02
                  (10.- ..., cxd4)   (10.- ..., cxb4)    (10.- ..., c4)
------------------------------------------------------------------------
10.- ..., cxb4         -0.29              -1.33               -3.56
                    (11.- Nb1)         (11.- Nd1)          (11.- Bg3)

------------

I did not save the game and tried to rebuild it from scratch until I succeeded! You can laugh at some blunders by both sides, but nobody said we were GMs:

[pgn][Event ""]
[Site ""]
[Date "2025.08.13"]
[White ""]
[Black ""]
[TimeControl "300+0"]
[PlyCount "64"]
[Result "0-1"]

1. d4 d5 2. c4 e6 3. Bf4 Nf6 4. Nc3 c6 5. c5 b6 6. b4 a5 7. a3 axb4 8. axb4 Rxa1 9. Qxa1 bxc5 10. Bxb8 cxb4 11. Nd1 Qb6 12. Qa7 Qxa7 13. Bxa7 b3 14. Bc5 Bxc5 15. dxc5 Ne4 16. e3 Nxc5 17. Nf3 O-O 18. Nd4 e5 19. Nxc6 Bd7 20. Nxe5 Rb8 21. Nb2 Na4 22. Ned3 Nxb2 23. Nxb2 Ra8 24. Be2 Ra2 25. Nd1 b2 26. Nc3 Ra1+ 27. Bd1 Ba4 28. O-O Bxd1 29. Rxd1 Rxd1+ 30. Nxd1 b1=Q 31. g3 Qxd1+ 32. Kg2 d4 0-1[/pgn]

It is notorious to see that white is completely underdeveloped in the kingside and made the first move on that side at the 17th move (17.- ..., e3)! This definitively helped me. Analysis before 17.- ..., e3:

Code: Select all

FEN: 2b1k2r/5ppp/2p1p3/2Pp4/4n3/1p6/4PPPP/3NKBNR w Kk - 1 16

Stockfish-17-windows-x86-64-sse41-popcnt:
[...]
---------------------------------------------------------------------------
 37/49	05:48	 146.009.967	419.532	-0,65	e2e3 e6e5 Ng1f3 f7f6 Bf1d3 Ne4xc5 Bd3b1 Ke8e7 Nd1b2 Bc8e6 h2h4 Ke7d6 Ke1d2 Nc5a6 h4h5 Na6b4 Kd2e2 Nb4a2 Nf3d2 Na2c3+ Ke2d3 Nc3b5 Nd2xb3 c6c5 Nb3d2 c5c4+ Nb2xc4+ d5xc4+ Nd2xc4+ Kd6e7 Nc4d2 Rh8c8 f2f3 f6f5 Rh1h4 Rc8c3+ Kd3e2 Rc3c1 Rh4b4 Nb5c3+ Ke2d3 Ke7f6 Bb1c2 Nc3d5
 37/63	05:48	 146.009.967	419.532	-0,58	h2h4 e6e5 Ng1f3 f7f6 Nf3d2 Ne4xc5 g2g4 Bc8e6 Rh1h3 d5d4 Rh3g3 Ke8d7 e2e3 Rh8a8 e3xd4 e5xd4 Bf1c4 Be6xc4 Nd2xc4 Ra8a1 Rg3f3 Kd7e6 Ke1d2 g7g6 h4h5 g6xh5 g4xh5 f6f5 Nc4b2 Ke6f6 Kd2e1 Kf6g5 Rf3g3+ Kg5xh5 Rg3g8 Ra1c1 Rg8d8 Nc5e6 Rd8b8 c6c5 Nb2d3 Rc1a1
 37/63	05:48	 146.009.967	419.532	-0,57	Ng1f3 Ke8e7 e2e3 Ne4xc5 Ke1d2 f7f6 Kd2c3 e6e5 Nf3d2 Bc8e6 Nd1b2 Rh8a8 Bf1d3 Ra8a2 Bd3xh7 f6f5 Rh1b1 Ke7f6 Kc3b4 Nc5e4 Nd2xe4+ d5xe4 g2g4 g7g6 g4xf5 g6xf5 Nb2a4 Ra2xf2 h2h4 Rf2f3 Rb1e1 Rf3h3 Na4c5 Rh3xh4 Nc5xe6 Rh4xh7 Re1g1

I immediately captured the c5 pawn and got the 3P vs. N imbalance, which lasted few. Then, the blunderfest started with me doing the second-to-last blunder. The last blunder was castle kingside (in the 28th move!), which is usually a good move, but not here.

Fun fact: Huntsman evaluates the last position (Q+4P vs. 4P) with not losing evals as late as depth 15/16:

Code: Select all

FEN: 6k1/5ppp/8/8/3p4/4P1P1/5PKP/3q4 w - - 0 33

Huntsman:
Found 145 tablebases
classical evaluation enabled
[...]
 15/16	00:00	     593.219	2.951.338	-0,85	e3xd4 g7g5 f2f4 Qd1e2+ Kg2g1 g5g4 d4d5 Kg8g7 d5d6 Kg7f6 d6d7 Qe2e1+ Kg1g2 Kf6e7 f4f5 Ke7xd7
 16/23	00:00	   1.912.780	3.421.788	-62,56	e3xd4 g7g5 h2h3 h7h5 f2f4 Qd1e2+ Kg2g1 g5xf4 g3xf4 Qe2d1+ Kg1h2 Qd1xd4 f4f5 Qd4e4 f5f6 Qe4f3 Kh2g1
 17/21	00:00	   1.972.958	3.413.422	-62,71	e3xd4 g7g5 h2h3 h7h5 d4d5 Qd1xd5+ Kg2h2 Qd5e4 Kh2g1 Qe4e1+ Kg1g2 g5g4 h3xg4 h5xg4 f2f4 Qe1e2+ Kg2g1 Qe2e3+ Kg1g2 Qe3f3+ Kg2h2
 18/26	00:00	   2.415.898	3.486.144	-63,79	e3xd4 g7g5 d4d5 Qd1xd5+ Kg2h3 h7h5 f2f4 g5xf4 Kh3h4 Qd5f3 Kh4g5 f4xg3 h2h3 Qf3g4+ Kg5f6 Qg4xh3 Kf6e5 Qh3g4
 19/24	00:01	   2.845.975	3.517.892	-152,45	e3xd4 g7g5 h2h3 h7h5 Kg2h2 Qd1f1 f2f3 Qf1f2+ Kh2h1 Qf2xf3+ Kh1g1 Qf3xg3+ Kg1f1 Qg3xh3+ Kf1f2 Qh3h4+ Kf2e3 Qh4f4+ Ke3d3 Qf4xd4+ Kd3c2 Qd4d1+ Kc2c3 Qd1c2+ Kc3d4
 20/27	00:01	   3.124.476	3.491.034	-152,52	e3xd4 g7g5 f2f3 Qd1e2+ Kg2g1 Qe2xf3 g3g4 Qf3xg4+ Kg1f1 Qg4d1+ Kf1g2 Qd1xd4 Kg2f3 Qd4f4+ Kf3e2 Qf4xh2+ Ke2d3 Qh2f4 Kd3c3 Qf4d4+ Kc3b3 Qd4c3+ Kb3a2 Qc3b3+ Ka2a1 Qb3b1+ Ka1xb1
[...]
 34/35	00:13	  55.032.689	4.073.779	-152,61	e3xd4 g7g5 h2h4 g5xh4 g3xh4 Qd1xd4 h4h5 Qd4xf2+ Kg2xf2
 35/46	00:16	  66.373.432	4.102.443	-M15	e3xd4 g7g5 h2h3 h7h5 g3g4 h5xg4 h3xg4 Qd1xg4+ Kg2f1 Qg4xd4 Kf1g2 Qd4g4+ Kg2f1 Qg4f3 Kf1e1 g5g4 Ke1d2 Qf3xf2+ Kd2d3 g4g3 Kd3c3 g3g2 Kc3c4 g2g1Q Kc4b3 Qf2d4 Kb3a2 Qg1a1+ Ka2b3 Qa1a4+
 36/31	00:20	  80.444.001	3.985.335	-M14	e3xd4 g7g5 h2h3 h7h5 g3g4 h5xg4 h3xg4 Qd1xg4+ Kg2f1 Qg4xd4 Kf1g2 Qd4d3 f2f3 Qd3e3 Kg2f1 Qe3xf3+ Kf1e1 g5g4 Ke1d2 g4g3 Kd2c2 g3g2 Kc2b2 g2g1Q Kb2c2 Qg1h2+ Kc2b1 Qf3f1+
[...]
 40/29	00:45	 176.288.288	3.881.804	-M14	e3xd4 g7g5 h2h3 h7h5 f2f3 Qd1e2+ Kg2g1 Qe2xf3 d4d5 Qf3xg3+ Kg1f1 Qg3xh3+ Kf1e2 Qh3g2+ Ke2e3 Qg2xd5 Ke3e2 h5h4 Ke2e3 h4h3 Ke3e2 h3h2 Ke2e3 h2h1Q Ke3e2 Qh1d1+ Ke2e3 Qd1e1+
 41/43	01:16	 293.501.343	3.849.803	-M12	e3xd4 g7g5 h2h3 h7h5 f2f3 Qd1xd4 Kg2f1 Qd4e3 Kf1g2 Qe3e2+ Kg2g1 Qe2xf3 h3h4 g5xh4 g3xh4 Qf3e2 Kg1h1 f7f5 Kh1g1 f5f4 Kg1h1 f4f3 Kh1g1 Qe2g2+
[...]
 45/25	01:52	 430.565.836	3.832.361	-M12	e3xd4 g7g5 h2h3 Qd1xd4 f2f3 h7h5 Kg2f1 Qd4e3 Kf1g2 Qe3e2+ Kg2g1 Qe2xf3 h3h4 g5xh4 g3xh4 f7f5 Kg1h2 Qf3g4 Kh2h1 f5f4 Kh1h2 f4f3 Kh2h1 Qg4g2+

Regards from Spain.

Ajedrecista.

[Dann Corbit]

It reminds me of WAC.210, and of Alex Szabo's refutation of WAC.210 as well (both based on powerful passed pawns).

[Eelco de Groot]

In case of a computer variation with 10...cxd4 I think it is fair to say, White would be totally lost after accepting the Bishop gambit. Is a line with Rebel Extreme just a PV but partially checcked with Jim's compile of Crystal, not everything checked though.

[pgn]
[Event "?"]
[Site "?"]
[Date "2025.08.21"]
[Round "?"]
[White "?"]
[Black "?"]
[Result "*"]

1. d4 d5 2. c4 e6 3. Bf4 Nf6 4. Nc3 c6 5. c5 b6 6. b4 a5
7. a3 axb4 8. axb4 Rxa1 9. Qxa1 bxc5 10. Bxb8 cxb4
(10... cxd4 {Rebel Extreme 1.1 -2.59/35 24:01m} 11. Na2 Qb6 12. Bf4 e5
13. Bd2 Bd6 14. Nf3 O-O 15. g3 Ba6 16. e3 c5 17. bxc5 Bxc5
18. Nc1 dxe3 19. fxe3 Bb7 20. Be2 Bxe3 21. Qxe5 Bxd2+
22. Nxd2 Rc8 23. Ncb3 Re8 24. Qd4 Qe6) *
[/pgn]

The endposition of the Rebel Extreme PV plus two more moves. White never castling in main line:

[d]4r1k1/1b3ppp/4qn2/3p4/3Q4/1N4P1/3NB2P/4K2R w K -

Engine: Crystal 9 JA dev-20250802-a37b38bd (1024 MB)
gemaakt door the Stockfish developers (see AUTHORS f

48 101:49 -3.98 25.Pc1 La6 26.Kf2 Lxe2 27.Te1 Df5+
28.Df4 Dxf4+ 29.gxf4 Pg4+ 30.Kg2 Lb5
31.Txe8+ Lxe8 32.Pa2 Kf8 33.Kg3 Ld7
34.Pc3 Pf6 35.Kf2 Ke7 36.Ke3 Kd6
37.Kd4 Lc6 38.Pf1 (3.486.954.563) 570

48 101:49 -4.11 25.O-O Dxe2 26.Tf2 Dh5 27.Dd3 Lc8
28.Pd4 h6 29.Te2 Txe2 30.Dxe2 Lg4
31.De1 Lh3 32.De2 Dg5 33.Kf2 Pg4+
34.Kg1 Dd8 35.P2f3 Pf6 36.Pe1 Db6
37.De3 Db1 38.Dc3 (3.486.954.563) 570

48 101:49 -5.11 25.Df2 Pg4 26.Df3 d4 27.Pc5 Lxf3
28.Pxe6 Lxh1 29.Pxd4 Pxh2 30.Kf2 Tb8
31.g4 Tb4 32.Ke3 Lg2 33.g5 g6 34.Kd3 Ld5
35.Ke3 Kg7 36.Kd3 Ta4 37.Pb5 Pg4
38.Pc3 (3.486.954.563) 570

[Eelco de Groot]

Just a single other Crystal side-line where White plays 14. g3 Ne4 and 16.Bg2. Because of all the computer time invested in the endposition to get a win score. It is not a Bishop sacrifice but a Knight sacrifice! Sorry for confusion of course.

[pgn]
[Event "?"]
[Site "?"]
[Date "2025.08.25"]
[Round "?"]
[White "?"]
[Black "?"]
[Result "*"]

1. d4 d5 2. c4 e6 3. Bf4 Nf6 4. Nc3 c6 5. c5 b6 6. b4 a5
7. a3 axb4 8. axb4 Rxa1 9. Qxa1 bxc5 10. Bxb8 cxb4
(10... cxd4 {-2.59/35 24:01m} 11. Na2 Qb6 12. Bf4 e5
13. Bd2 Bd6 14. Nf3 (14. g3 {-2.88/53 353:44m} Ne4 15. Nf3
O-O 16. Bg2 Ba6 17. Qd1 d3 18. e3 Bc4 19. Nc3 Bxb4 20. Nb1
Nxd2 21. Nfxd2 e4 22. O-O Ba6 23. Qc1 c5 24. Nc3 Rd8
25. Rd1 Bb7 26. Ndb1 f5 27. h4 {White seems to be winning.
See diagram below}) 14... O-O 15. g3 Ba6 16. e3
c5 17. bxc5 Bxc5 18. Nc1 dxe3 19. fxe3 Bb7 20. Be2 Bxe3
21. Qxe5 Bxd2+ 22. Nxd2 Rc8 23. Ncb3 Re8 24. Qd4 Qe6) *
[/pgn]

The endposition in this line analyzed 30 hours with Crystal best of three moves:

[d]3r2k1/1b4pp/1q6/2pp1p2/1b2p2P/2NpP1P1/5PB1/1NQR2K1 b - -

Engine: Crystal 9 JA dev-20250802-a37b38bd (1024 MB)
gemaakt door the Stockfish developers (see AUTHORS f

56 1825:00 -3.45 27...Dd6 28.Lh3 g6 29.Lg2 Lc6 30.Kh2 Tb8
31.Tf1 La5 32.Pd2 Lc7 33.Pd1 c4
34.h5 gxh5 35.f4 Ta8 36.Db2 Ld8
37.Lh3 Ld7 38.Pc3 Db6 39.Dc1 Le6
40.Pdb1 Tb8 (58.484.346.902) 534

56 1825:00 -3.45 27...Ta8 28.Db2 c4 29.Td2 Da5 30.Lh3 g6
31.Pd1 Lxd2 32.Dxb7 Lb4 33.h5 Ta7
34.Db8+ Lf8 35.Pbc3 Dc7 36.De8 De7
37.Db8 Td7 38.Db5 Dd8 39.Da4 Lg7
40.g4 Td6 (58.484.346.902) 534

56 1825:00 -3.40 27...Kh8 28.Db2 Df6 29.Db3 Lc6
30.Tc1 Dd6 31.Pa2 La5 32.Da3 Lb6
33.Pac3 c4 34.Dxd6 Txd6 35.Pd1 Td8
36.Pbc3 g6 37.Tb1 Lc5 38.Kf1 Ta8
39.f3 Kg7 40.g4 Kf6 (58.484.346.902) 534

[Eelco de Groot]

Prizewinning (see Mark Young's tournament in other forum) Rebel Extreme 1.1 has about the same score as Crystal I think you can compare the scores as related to win percentages. But I did not really study it.
Rebel/CSTal can't do Multi PV so it chooses just a single line as PV wich means it needs a little less time to reach comparable depths not having to do three PVs, braching reported wil be different than Crystal, depthfigures I mean but actual depths you can not really get from iteration numbers. The same position after about ten hours:

[d]3r2k1/1b4pp/1q6/2pp1p2/1b2p2P/2NpP1P1/5PB1/1NQR2K1 b - -

Engine: Rebel-EXTREME 1.1 (1024 MB)
gemaakt door Ed Schroder and Chris Whittington (C) 2
.
.
.
41/72 504:04 -3.59 27...Ta8 28.Kh2 Dd6 29.Pb5 De5
30.P5c3 De7 31.Kg1 Lc6 32.Pxd5 Lxd5
33.Txd3 De6 34.Td1 Lb3 35.Lf1 Kh8
36.Td2 h6 37.h5 Tc8 38.Te2 c4 39.Pc3 De5
40.Pa2 Le7 (19.263.168.687)

42/68 864:13 -3.59++ 27...Ta8 28.Lh3 (33.340.408.205)

42/82 1133:40-3.59 27...Ta8 28.Lh3 Df6 29.Db2 Lc8
30.Lf1 Le6 31.Dd2 f4 32.exf4 Ta1
33.Kh2 c4 34.Pxe4 Lxd2 35.Pxf6+ gxf6
36.Txd2 Txb1 37.Lxd3 cxd3 38.Txd3 Kf7
39.f3 Tb2+ 40.Kg1 Ke7 (43.484.673.369)

[Ajedrecista]

Hello:

I am happy that this position raised some interest. Thank you.

I wrote a typo at least twice in the OP, where I said that the first white move on the kingside was 17.- ..., e3, when it was 16.- e3, obviously. I wonder how many serious games have this particularity (i.e. not proof games or troll players), of being 15 full moves (or any other large number of moves) neither moving a single piece of a kingside or queenside nor being captured a single piece by the opponent. I guess that pgn-extract can do the task in a database: reading a topic on Chess StackExchange, you might search for some FEN patterns after a given number of moves, something like:

Code: Select all

FENPatternI "rnbq????/pppp????/*/*/*/*/*/*"
FENPatternI "????kbnr/????pppp/*/*/*/*/*/*"
FENPatternI "*/*/*/*/*/*/PPPP????/RNBQ????"
FENPatternI "*/*/*/*/*/*/????PPPP/????KBNR"

This will not catch things like Ng1-Nf3-Ng1, but there will be some real games. Since I do no expect the number of returned games to be high after 14 moves, 15 moves, 16 moves and so on, those games probably could be checked manually. Just thinking out loud.

Regards from Spain.

Ajedrecista.

[Wisely]

Most GrandMasters (GMs) will let you have a passed pawn, only to be blockaded by let's say, a Queen/Knight/Bishop. Then the endgame will be about pawn structures and pawn islands.

I will try my best to analyze your position using a very fast machine.