On engine eval scores

[MikeGL]

I followed this game yesterday being relayed live.
Everybody on the site, where I was watching this game, was surprised as to why black handled by GM Granda Zuniga resigned in a position with 1 pawn down against Carlsen.

It turns out that the score of SF8 keeps on rising a bit if I play the game with engine assistance and the strong GM from Peru seems to be correct, considering of course that GM Granda-Zuniga is a strong positional and endgame player.

My question is this, why can't engines give a convincing winning score for the final position of the below winning game?
I have seen eval scores of +2, +4, +8 even +28.00 in a real game GM position analysis posted on this forum but which only ends up in draw!! While this winning position of Carlsen below is given just around +1.80 but was in fact really winning considering the B is of the same color.

[pgn]
[Event "Chess-com Isle of Man Masters"]
[Site "Isle of Man"]
[Date "?"]
[Round "?"]
[White "Carlsen"]
[Black "Granda Zuniga"]
[Result "1-0"]
[TimeControl "?"]
[Termination "unterminated"]
[PlyCount "55"]
[WhiteType "human"]
[BlackType "human"]

1. e4 e5 2. Nf3 Nf6 3. Nxe5 d6 4. Nf3 Nxe4 5. d4 d5 6. Bd3 Be7 7. O-O O-O 8. c4 Nf6 9. Nc3 Nc6 10. cxd5 Nb4 11. Re1 Nbxd5 12. h3 c6 13. a3 Re8 14.Qc2 h6 15. Nxd5 Nxd5 16. Ne5 Bd6 17. Bh7+ Kf8 18. Bd2 Be6 19. Re2 Nf6 20.Rae1 Qc7 21. Ng6+ fxg6 22. Rxe6 Nxh7 23. Rxe8+ Rxe8 24. Rxe8+ Kxe8 25.Qxg6+ Kd8 26. Qxh7 Qe7 27. g3 Kc7 28. Qg6 1-0
[/pgn]


Final position after white played 28.Qg6
[d]8/ppk1q1p1/2pb2Qp/8/3P4/P5PP/1P1B1P2/6K1 b - - 2 28
Are there engines which can give a convincing winning score for this final position between Carlsen-Granda Zuniga?

[Vinvin]

Eval by Stockfish seems realistic :
3 best moves for black :

Code: Select all

35/70	03:56	 2.111.693k	8.920k	+1,91	28. ... Rd7 29.d5 cxd5 30.Fc3 De6 31.Dxg7+ Rc6 32.h4 h5 33.b4 a6 34.Fd2 
35/51	03:56	 2.111.693k	8.920k	+1,90	28. ... a6 29.h4 Dd7 30.Fc3 Ff8 31.d5 cxd5 32.Fd4 Rc8 33.b4 Rc7 34.Rg2 Rb8 35.h5 
36/49	03:56	 2.111.693k	8.920k	+1,75	28. ... Dd7 29.Fc3 Ff8 30.d5 cxd5 31.Fd4 b6 32.h4 Rc8 33.b4 Rd8 34.Rh2 Db7 35.a4

Several possibilities :
1) Granda Zuniga underestimate his chance to draw this game
2) Granda Zuniga doesn't want to waste his energy for a draw against a very strong play
3) Granda Zuniga blundered in his analyse and he saw a forced quick lost

[Houdini]

While the position is probably lost, it definitely feels like a premature resignation.
I guess that a top engine would probably hold this ending with Black against Magnus.

[Uri Blass]

While the position is probably lost, it definitely feels like a premature resignation.
I guess that a top engine would probably hold this ending with Black against Magnus.

I wonder if you consider the following as premature resignation for kasparov with white against karpov

[D]4qbk1/5ppp/6b1/2nB4/1p6/1N5P/5PP1/B1Q3K1 b - - 0 1


I read in an old chess book that karpov wrote his secret move Nd3 and kasparov resigned because analysis showed that after trading queens at e1 white will have to sacrifice a piece to prevent promotion.

I do not understand it for some reasons:
1)How can kasparov know that karpov wrote Nd3.

Nd3 is not stockfish's move

2)How can black trade queens at e1 after Nd3 Qe3(Qe3 is the computer move after Nd3 and it seems that Nd3 is a clear mistake based on stockfish's evaluation because the score drop from 2.08 to 1.41)

[D]4qbk1/5ppp/6b1/2nB4/1p6/1N5P/5PP1/B1Q3K1 b - - 0 1


Stockfish_17092216_x64_modern:

34/52- 01:18 309,058k 3,931k +1.83 1. ... Be4 2.Bxf7+
34/52- 01:20 316,570k 3,923k +1.75 1. ... Be4 2.Bxf7+
34/73+ 01:21 320,844k 3,921k +1.83 1. ... Be4
34/73+ 01:30 353,682k 3,918k +1.99 1. ... Be4
34/73 01:53 442,049k 3,888k +1.99 1. ... Be4 2.Bxf7+ Qxf7 3.Nxc5 Bxg2 4.Kxg2 Qd5+ 5.Kg1 Bxc5 6.Qb2 Bf8 7.Qe2 Qg5+ 8.Kh2 Bd6+ 9.Kh1 Qd5+ 10.Kg1 h5 11.Qe3 Bc5 12.Qf4 b3 13.Qg3 Qd1+ 14.Kg2 Qxa1 15.Qxb3+ Kh8 16.Qd5 Qd4 17.Qxh5+ Kg8 18.Qe8+ Bf8 19.Qe6+ Kh8 20.Qf5 Qd6 21.Qg4 Qf6 22.Qh5+ Kg8 23.Qf3 Qxf3+ 24.Kxf3 Kh7 25.Kg4 Kg6 26.f3 Kf6 27.Kf4 Bd6+ 28.Ke4 Bc5 29.Kd5
35/47- 02:09 502,233k 3,888k +1.91 1. ... Be4 2.Bxe4
35/47+ 02:10 505,002k 3,882k +1.99 1. ... Be4
35/48 02:20 541,654k 3,863k +1.87 1. ... Be4 2.Bxe4 Nxe4 3.Bd4 Qe6 4.Qc2 Nc3 5.Be3 Qc4 6.Nc1 g6 7.g3 Bc5 8.Bh6 f6 9.Qd2 Kf7 10.Be3 Be7 11.h4 Nd5 12.Bd4 Bd6 13.Qd1 Be5 14.Bxe5 fxe5 15.h5 Kf6 16.hxg6 hxg6 17.Nb3 Ke7 18.Na5 Qb5 19.Nb3 Nc3
36/49+ 02:25 560,682k 3,856k +1.95 1. ... Be4
36/49 02:35 600,864k 3,853k +2.00 1. ... Be4 2.Bxe4 Nxe4 3.Bd4 Qe6 4.Qc2 Nc3 5.Nc1 Qc4 6.Be3 g6 7.g3 Bc5 8.Bh6 f6 9.Qd2 Kf7 10.Be3 Be7 11.h4 h5 12.Bd4 Ne4 13.Qe3 Nd6 14.Qd2 Kg7 15.Qd1 Ne4 16.Bb2 Bc5 17.Nd3 Bb6 18.Qf3 b3 19.Kg2 Qd5 20.Kg1 Bd4 21.Qf4 Bxb2
37/51- 02:45 637,852k 3,846k +1.92 1. ... Be4 2.Bxe4
37/51 02:46 641,680k 3,847k +1.94 1. ... Be4 2.Bxe4 Nxe4 3.Bd4 Qe6 4.Qe3 Qd5 5.h4 Nc3 6.Nc1 Nb5 7.Be5 Qd1+ 8.Kh2 Na3 9.Bf4 Nc4 10.Qb3 Qd4 11.Qf3 Ne5 12.Bxe5 Qxe5+ 13.g3 Qc3 14.Qd1 g6 15.Nb3 Bc5 16.Kg2 Bb6 17.Qd5 Kg7 18.Nd2 Qf6 19.Ne4 Qd4 20.Qxd4+ Bxd4 21.Nd2 Kf6 22.Kf3 Ke6 23.g4
38/51- 03:06 710,300k 3,806k +1.87 1. ... Be4 2.Bc4
38/51- 03:13 733,798k 3,796k +1.79 1. ... Be4 2.Bc4
38/55+ 03:31 794,678k 3,765k +1.87 1. ... Be4
38/58+ 03:49 861,908k 3,762k +2.02 1. ... Be4
38/70 05:16 1,192,596k 3,762k +2.11 1. ... Be4 2.Bc4 Nxb3 3.Bxb3 Qa8 4.f3 Bd5 5.Bd1 b3 6.Bb2 Be6 7.Qc7 Qd5 8.Be2 f6 9.Qb8 Qc5+ 10.Kh1 Bf7 11.Qa8 g6 12.Qd8 Qd6 13.Qxd6 Bxd6 14.Bxf6 Bf8 15.Bd3 Bg7 16.Bxg7 Kxg7 17.f4 b2 18.Kg1 g5 19.f5 Ba2 20.g4 Kf6 21.Kf2 Ke5 22.Ke3 h6 23.Be4 b1Q 24.Bxb1 Bxb1 25.Kf2 Be4 26.Kg3 Kf6
39/60- 05:53 1,331,980k 3,772k +2.04 1. ... Be4 2.Bxe4
39/60+ 06:04 1,376,726k 3,773k +2.11 1. ... Be4
39/60 06:35 1,492,971k 3,771k +2.08 1. ... Be4 2.Bxe4 Nxe4 3.Bd4 Qe6 4.Qc2 Nc3 5.Bxc3 Qc4 6.Qd1 bxc3 7.Na1 g6 8.Nc2 Bc5 9.Qd8+ Kg7 10.Qd1 Qe4 11.Kf1 Qf4 12.Qe2 Qc1+ 13.Ne1 Bd4 14.Qc4 Qd2 15.Qe2 Bf6 16.Qe4 Qd1 17.Qc2 Qd5 18.Ke2 Qh5+ 19.Kf1 Qg5 20.Qe2 Qd5 21.Nf3 Qc5 22.Ne1 Bd8 23.Nc2 Bb6
40/57- 07:01 1,596,451k 3,788k +2.00 1. ... Be4 2.Qe3
40/65- 07:10 1,628,082k 3,785k +1.93 1. ... Be4 2.Qe3
40/65 07:41 1,745,719k 3,785k +1.93 1. ... Be4 2.Qe3 Qd7 3.Bc4 Qd1+ 4.Kh2 Nxb3 5.Bxb3 Bc2 6.Bxg7 Bxb3 7.Bxf8 Kxf8 8.Qc5+ Ke8 9.Qxb4 Qd5 10.Qb8+ Ke7 11.Qc7+ Kf6 12.Qf4+ Qf5 13.Qh6+ Ke7 14.Qe3+ Be6 15.f3 Qd5 16.g4 f6 17.h4 Kf7 18.h5 Kg7 19.Kg3 Qd6+ 20.Kg2 Bf7 21.Kh3 Bd5 22.f4 Kf7 23.Kh4 Qe6 24.Qa7+ Qe7
41/55+ 08:09 1,857,456k 3,797k +2.00 1. ... Be4
41/55+ 08:22 1,909,507k 3,803k +2.08 1. ... Be4
41/66 08:51 2,029,643k 3,820k +2.08 1. ... Be4 2.Qe3 Qd7 3.Bc4 Qd1+ 4.Kh2 Nxb3 5.Bxb3 Bc2 6.Bxg7 Bxb3 7.Bxf8 Kxf8 8.Qc5+ Ke8 9.Qxb4 Qd5 10.Qf4 Kd7 11.f3 Bc2 12.Kg3 Ke6 13.Qg4+ Bf5 14.Qb4 Kf6 15.Qf8 Qe5+ 16.Kf2 Qd4+ 17.Kg3 Qe3 18.Qd6+ Kg7 19.Qf4 Qxf4+ 20.Kxf4 Kf6 21.g4 Bd7 22.h4 Ba4 23.Ke4 Bd1 24.Kf4 Bc2 25.Ke3 Ke5 26.h5 Bb3 27.f4+ Kd5 28.h6


[D]4qbk1/5ppp/6b1/3B4/1p6/1N1n3P/5PP1/B1Q3K1 w - - 1 2

Stockfish_17092216_x64_modern:

39/47- 02:06 446,400k 3,530k -1.41 2.Qe3 Qb5
39/47+ 02:09 458,698k 3,530k -1.32 2.Qe3
39/47 02:12 467,054k 3,529k -1.30 2.Qe3 Qc8 3.Bd4 h6 4.Nd2 Qc1+ 5.Kh2 Qc7+ 6.Kg1 Qc2 7.g3 Nc5 8.Kg2 Bf5 9.Bxc5 Qxc5 10.Qb3 Qe7 11.g4 Be6 12.Bxe6 fxe6 13.Kf1 Qd6 14.Ke2 Qe5+ 15.Kd1 g5 16.Kc2 Be7 17.Qd3 Kg7 18.Qe4 Qd5 19.Qxd5 exd5 20.Nf3 Kf6 21.Nd4 Bc5 22.Nc6
40/58 02:35 547,168k 3,527k -1.30 2.Qe3 Qb5 3.Qg5 Qb6 4.Qe3 Qc7 5.g3 Qc8 6.h4 Qc2 7.Bd4 Qd1+ 8.Kh2 Ne1 9.Nd2 Nc2 10.Qf4 Nxd4 11.Qxd4 Qc2 12.Bc4 Bf5 13.g4 Be6 14.Bxe6 fxe6 15.Qd7 Qa2 16.Kg3 Qd5 17.Qxd5 exd5 18.h5 Bd6+ 19.f4 Kf7 20.Nb3 Bc7 21.Kf3 Bb6 22.Ke2 Ke6 23.Kd3 Kf6 24.Ke2 Kf7 25.Nd2 Bc5 26.Nb3 Bd6 27.Kf3 Bc7 28.Ke3
41/58 03:00 621,402k 3,440k -1.30 2.Qe3 Qb5 3.Qg5 Qb6 4.Qe3 Qc7 5.g3 Qc8 6.h4 Qc2 7.Bd4 Qd1+ 8.Kh2 Ne1 9.Nd2 Nc2 10.Qf4 Nxd4 11.Qxd4 Qc2 12.Bc4 Bf5 13.g4 Be6 14.Bxe6 fxe6 15.Qd7 Qa2 16.Kg3 Qd5 17.Qxd5 exd5 18.h5 Bd6+ 19.f4 Kf7 20.Nb3 Bc7 21.Kf3 Bb6 22.Ke2 Ke6 23.Kd3 Kf6 24.Nd2 Bc5 25.Nb3 Bd6 26.Ke3 g6 27.hxg6
42/54- 03:03 633,140k 3,441k -1.38 2.Qe3 Qb5
42/60- 04:10 880,424k 3,511k -1.45 2.Qe3 Qb5
42/60+ 05:02 1,072,700k 3,541k -1.38 2.Qe3
42/62 06:00 1,288,441k 3,570k -1.41 2.Qe3 Qb5 3.Qg5 Qb6 4.Bd4 Qc7 5.Qe3 Nf4 6.Bf3 Ne6 7.Bd5 Qd7 8.Bc4 Nxd4 9.Qxd4 Qxd4 10.Nxd4 Bc5 11.Nb3 Bb6 12.Kf1 Kf8 13.Nc1 f6 14.Ke2 Bf7 15.Bxf7 Kxf7 16.Kd3 Bxf2 17.Kc4 Be1 18.Nd3 Bd2 19.Nxb4 Bxb4 20.Kxb4 Ke6 21.Kc4 Ke5 22.h4 Kf4 23.Kd5 Kg3 24.Ke6 Kxg2 25.h5 Kg3

3)Suppose black trade queens at e1 after Nd3 Qd1 Qe1+ Qxe1 Nxe1
What is the reason to think that white is losing.

I can see only a score of +1.1 for black.

39/60 08:44 2,025,334k 3,863k -1.17 4.g3 Bf5 5.h4 Nc2 6.Be5 Bd3 7.Kg2 Be2 8.Bc7 g6 9.Bb6 Bg7 10.Nc1 Ne1+ 11.Kg1 Bb5 12.f3 Bc3 13.Kf2 Nc2 14.f4 Kg7 15.Bc5 Kf6 16.Kf3 Bd7 17.Ne2 Bd2 18.Nd4 Be3 19.Nb3 Bxc5 20.Nxc5 Bg4+ 21.Ke4 Bf5+ 22.Kf3 Ke7 23.Bc4 Bg4+ 24.Kxg4 Ne3+ 25.Kf3 Nxc4
40/52+ 09:14 2,142,660k 3,861k -1.09 4.g3
40/52 09:21 2,169,687k 3,861k -1.10 4.g3 Bf5 5.h4 Nc2 6.Be5 Bd3 7.Kg2 Be2 8.Bf4 Ne1+ 9.Kg1 g6 10.Nd4 Bd1 11.Kf1 Nd3 12.Bg5 Bg7 13.Be3 Kf8 14.f3 Bf6 15.Bc4 Ne5 16.Bd5 Ke7 17.Ke1 Kd6 18.Bxf7 Bxf3 19.Bg8 Be4 20.Ke2 Ng4 21.Bxh7 Be5 22.Nb3 Nxe3 23.Kxe3 Bd5 24.Nd2 Bxg3 25.Bxg6 Bxh4 26.Be4

[Lyudmil Tsvetkov]

I followed this game yesterday being relayed live.
Everybody on the site, where I was watching this game, was surprised as to why black handled by GM Granda Zuniga resigned in a position with 1 pawn down against Carlsen.

It turns out that the score of SF8 keeps on rising a bit if I play the game with engine assistance and the strong GM from Peru seems to be correct, considering of course that GM Granda-Zuniga is a strong positional and endgame player.

My question is this, why can't engines give a convincing winning score for the final position of the below winning game?
I have seen eval scores of +2, +4, +8 even +28.00 in a real game GM position analysis posted on this forum but which only ends up in draw!! While this winning position of Carlsen below is given just around +1.80 but was in fact really winning considering the B is of the same color.

[pgn]
[Event "Chess-com Isle of Man Masters"]
[Site "Isle of Man"]
[Date "?"]
[Round "?"]
[White "Carlsen"]
[Black "Granda Zuniga"]
[Result "1-0"]
[TimeControl "?"]
[Termination "unterminated"]
[PlyCount "55"]
[WhiteType "human"]
[BlackType "human"]

1. e4 e5 2. Nf3 Nf6 3. Nxe5 d6 4. Nf3 Nxe4 5. d4 d5 6. Bd3 Be7 7. O-O O-O 8. c4 Nf6 9. Nc3 Nc6 10. cxd5 Nb4 11. Re1 Nbxd5 12. h3 c6 13. a3 Re8 14.Qc2 h6 15. Nxd5 Nxd5 16. Ne5 Bd6 17. Bh7+ Kf8 18. Bd2 Be6 19. Re2 Nf6 20.Rae1 Qc7 21. Ng6+ fxg6 22. Rxe6 Nxh7 23. Rxe8+ Rxe8 24. Rxe8+ Kxe8 25.Qxg6+ Kd8 26. Qxh7 Qe7 27. g3 Kc7 28. Qg6 1-0
[/pgn]


Final position after white played 28.Qg6
[d]8/ppk1q1p1/2pb2Qp/8/3P4/P5PP/1P1B1P2/6K1 b - - 2 28
Are there engines which can give a convincing winning score for this final position between Carlsen-Granda Zuniga?

this position is not playable against Carlsen, even if black is SF.

maybe Zuniga the Grande was under time pressure, too.

[Lyudmil Tsvetkov]

While the position is probably lost, it definitely feels like a premature resignation.
I guess that a top engine would probably hold this ending with Black against Magnus.

oh, no.

[Lyudmil Tsvetkov]

While the position is probably lost, it definitely feels like a premature resignation.
I guess that a top engine would probably hold this ending with Black against Magnus.

I wonder if you consider the following as premature resignation for kasparov with white against karpov

[D]4qbk1/5ppp/6b1/2nB4/1p6/1N5P/5PP1/B1Q3K1 b - - 0 1


I read in an old chess book that karpov wrote his secret move Nd3 and kasparov resigned because analysis showed that after trading queens at e1 white will have to sacrifice a piece to prevent promotion.

I do not understand it for some reasons:
1)How can kasparov know that karpov wrote Nd3.

Nd3 is not stockfish's move

2)How can black trade queens at e1 after Nd3 Qe3(Qe3 is the computer move after Nd3 and it seems that Nd3 is a clear mistake based on stockfish's evaluation because the score drop from 2.08 to 1.41)

[D]4qbk1/5ppp/6b1/2nB4/1p6/1N5P/5PP1/B1Q3K1 b - - 0 1


Stockfish_17092216_x64_modern:

34/52- 01:18 309,058k 3,931k +1.83 1. ... Be4 2.Bxf7+
34/52- 01:20 316,570k 3,923k +1.75 1. ... Be4 2.Bxf7+
34/73+ 01:21 320,844k 3,921k +1.83 1. ... Be4
34/73+ 01:30 353,682k 3,918k +1.99 1. ... Be4
34/73 01:53 442,049k 3,888k +1.99 1. ... Be4 2.Bxf7+ Qxf7 3.Nxc5 Bxg2 4.Kxg2 Qd5+ 5.Kg1 Bxc5 6.Qb2 Bf8 7.Qe2 Qg5+ 8.Kh2 Bd6+ 9.Kh1 Qd5+ 10.Kg1 h5 11.Qe3 Bc5 12.Qf4 b3 13.Qg3 Qd1+ 14.Kg2 Qxa1 15.Qxb3+ Kh8 16.Qd5 Qd4 17.Qxh5+ Kg8 18.Qe8+ Bf8 19.Qe6+ Kh8 20.Qf5 Qd6 21.Qg4 Qf6 22.Qh5+ Kg8 23.Qf3 Qxf3+ 24.Kxf3 Kh7 25.Kg4 Kg6 26.f3 Kf6 27.Kf4 Bd6+ 28.Ke4 Bc5 29.Kd5
35/47- 02:09 502,233k 3,888k +1.91 1. ... Be4 2.Bxe4
35/47+ 02:10 505,002k 3,882k +1.99 1. ... Be4
35/48 02:20 541,654k 3,863k +1.87 1. ... Be4 2.Bxe4 Nxe4 3.Bd4 Qe6 4.Qc2 Nc3 5.Be3 Qc4 6.Nc1 g6 7.g3 Bc5 8.Bh6 f6 9.Qd2 Kf7 10.Be3 Be7 11.h4 Nd5 12.Bd4 Bd6 13.Qd1 Be5 14.Bxe5 fxe5 15.h5 Kf6 16.hxg6 hxg6 17.Nb3 Ke7 18.Na5 Qb5 19.Nb3 Nc3
36/49+ 02:25 560,682k 3,856k +1.95 1. ... Be4
36/49 02:35 600,864k 3,853k +2.00 1. ... Be4 2.Bxe4 Nxe4 3.Bd4 Qe6 4.Qc2 Nc3 5.Nc1 Qc4 6.Be3 g6 7.g3 Bc5 8.Bh6 f6 9.Qd2 Kf7 10.Be3 Be7 11.h4 h5 12.Bd4 Ne4 13.Qe3 Nd6 14.Qd2 Kg7 15.Qd1 Ne4 16.Bb2 Bc5 17.Nd3 Bb6 18.Qf3 b3 19.Kg2 Qd5 20.Kg1 Bd4 21.Qf4 Bxb2
37/51- 02:45 637,852k 3,846k +1.92 1. ... Be4 2.Bxe4
37/51 02:46 641,680k 3,847k +1.94 1. ... Be4 2.Bxe4 Nxe4 3.Bd4 Qe6 4.Qe3 Qd5 5.h4 Nc3 6.Nc1 Nb5 7.Be5 Qd1+ 8.Kh2 Na3 9.Bf4 Nc4 10.Qb3 Qd4 11.Qf3 Ne5 12.Bxe5 Qxe5+ 13.g3 Qc3 14.Qd1 g6 15.Nb3 Bc5 16.Kg2 Bb6 17.Qd5 Kg7 18.Nd2 Qf6 19.Ne4 Qd4 20.Qxd4+ Bxd4 21.Nd2 Kf6 22.Kf3 Ke6 23.g4
38/51- 03:06 710,300k 3,806k +1.87 1. ... Be4 2.Bc4
38/51- 03:13 733,798k 3,796k +1.79 1. ... Be4 2.Bc4
38/55+ 03:31 794,678k 3,765k +1.87 1. ... Be4
38/58+ 03:49 861,908k 3,762k +2.02 1. ... Be4
38/70 05:16 1,192,596k 3,762k +2.11 1. ... Be4 2.Bc4 Nxb3 3.Bxb3 Qa8 4.f3 Bd5 5.Bd1 b3 6.Bb2 Be6 7.Qc7 Qd5 8.Be2 f6 9.Qb8 Qc5+ 10.Kh1 Bf7 11.Qa8 g6 12.Qd8 Qd6 13.Qxd6 Bxd6 14.Bxf6 Bf8 15.Bd3 Bg7 16.Bxg7 Kxg7 17.f4 b2 18.Kg1 g5 19.f5 Ba2 20.g4 Kf6 21.Kf2 Ke5 22.Ke3 h6 23.Be4 b1Q 24.Bxb1 Bxb1 25.Kf2 Be4 26.Kg3 Kf6
39/60- 05:53 1,331,980k 3,772k +2.04 1. ... Be4 2.Bxe4
39/60+ 06:04 1,376,726k 3,773k +2.11 1. ... Be4
39/60 06:35 1,492,971k 3,771k +2.08 1. ... Be4 2.Bxe4 Nxe4 3.Bd4 Qe6 4.Qc2 Nc3 5.Bxc3 Qc4 6.Qd1 bxc3 7.Na1 g6 8.Nc2 Bc5 9.Qd8+ Kg7 10.Qd1 Qe4 11.Kf1 Qf4 12.Qe2 Qc1+ 13.Ne1 Bd4 14.Qc4 Qd2 15.Qe2 Bf6 16.Qe4 Qd1 17.Qc2 Qd5 18.Ke2 Qh5+ 19.Kf1 Qg5 20.Qe2 Qd5 21.Nf3 Qc5 22.Ne1 Bd8 23.Nc2 Bb6
40/57- 07:01 1,596,451k 3,788k +2.00 1. ... Be4 2.Qe3
40/65- 07:10 1,628,082k 3,785k +1.93 1. ... Be4 2.Qe3
40/65 07:41 1,745,719k 3,785k +1.93 1. ... Be4 2.Qe3 Qd7 3.Bc4 Qd1+ 4.Kh2 Nxb3 5.Bxb3 Bc2 6.Bxg7 Bxb3 7.Bxf8 Kxf8 8.Qc5+ Ke8 9.Qxb4 Qd5 10.Qb8+ Ke7 11.Qc7+ Kf6 12.Qf4+ Qf5 13.Qh6+ Ke7 14.Qe3+ Be6 15.f3 Qd5 16.g4 f6 17.h4 Kf7 18.h5 Kg7 19.Kg3 Qd6+ 20.Kg2 Bf7 21.Kh3 Bd5 22.f4 Kf7 23.Kh4 Qe6 24.Qa7+ Qe7
41/55+ 08:09 1,857,456k 3,797k +2.00 1. ... Be4
41/55+ 08:22 1,909,507k 3,803k +2.08 1. ... Be4
41/66 08:51 2,029,643k 3,820k +2.08 1. ... Be4 2.Qe3 Qd7 3.Bc4 Qd1+ 4.Kh2 Nxb3 5.Bxb3 Bc2 6.Bxg7 Bxb3 7.Bxf8 Kxf8 8.Qc5+ Ke8 9.Qxb4 Qd5 10.Qf4 Kd7 11.f3 Bc2 12.Kg3 Ke6 13.Qg4+ Bf5 14.Qb4 Kf6 15.Qf8 Qe5+ 16.Kf2 Qd4+ 17.Kg3 Qe3 18.Qd6+ Kg7 19.Qf4 Qxf4+ 20.Kxf4 Kf6 21.g4 Bd7 22.h4 Ba4 23.Ke4 Bd1 24.Kf4 Bc2 25.Ke3 Ke5 26.h5 Bb3 27.f4+ Kd5 28.h6


[D]4qbk1/5ppp/6b1/3B4/1p6/1N1n3P/5PP1/B1Q3K1 w - - 1 2

Stockfish_17092216_x64_modern:

39/47- 02:06 446,400k 3,530k -1.41 2.Qe3 Qb5
39/47+ 02:09 458,698k 3,530k -1.32 2.Qe3
39/47 02:12 467,054k 3,529k -1.30 2.Qe3 Qc8 3.Bd4 h6 4.Nd2 Qc1+ 5.Kh2 Qc7+ 6.Kg1 Qc2 7.g3 Nc5 8.Kg2 Bf5 9.Bxc5 Qxc5 10.Qb3 Qe7 11.g4 Be6 12.Bxe6 fxe6 13.Kf1 Qd6 14.Ke2 Qe5+ 15.Kd1 g5 16.Kc2 Be7 17.Qd3 Kg7 18.Qe4 Qd5 19.Qxd5 exd5 20.Nf3 Kf6 21.Nd4 Bc5 22.Nc6
40/58 02:35 547,168k 3,527k -1.30 2.Qe3 Qb5 3.Qg5 Qb6 4.Qe3 Qc7 5.g3 Qc8 6.h4 Qc2 7.Bd4 Qd1+ 8.Kh2 Ne1 9.Nd2 Nc2 10.Qf4 Nxd4 11.Qxd4 Qc2 12.Bc4 Bf5 13.g4 Be6 14.Bxe6 fxe6 15.Qd7 Qa2 16.Kg3 Qd5 17.Qxd5 exd5 18.h5 Bd6+ 19.f4 Kf7 20.Nb3 Bc7 21.Kf3 Bb6 22.Ke2 Ke6 23.Kd3 Kf6 24.Ke2 Kf7 25.Nd2 Bc5 26.Nb3 Bd6 27.Kf3 Bc7 28.Ke3
41/58 03:00 621,402k 3,440k -1.30 2.Qe3 Qb5 3.Qg5 Qb6 4.Qe3 Qc7 5.g3 Qc8 6.h4 Qc2 7.Bd4 Qd1+ 8.Kh2 Ne1 9.Nd2 Nc2 10.Qf4 Nxd4 11.Qxd4 Qc2 12.Bc4 Bf5 13.g4 Be6 14.Bxe6 fxe6 15.Qd7 Qa2 16.Kg3 Qd5 17.Qxd5 exd5 18.h5 Bd6+ 19.f4 Kf7 20.Nb3 Bc7 21.Kf3 Bb6 22.Ke2 Ke6 23.Kd3 Kf6 24.Nd2 Bc5 25.Nb3 Bd6 26.Ke3 g6 27.hxg6
42/54- 03:03 633,140k 3,441k -1.38 2.Qe3 Qb5
42/60- 04:10 880,424k 3,511k -1.45 2.Qe3 Qb5
42/60+ 05:02 1,072,700k 3,541k -1.38 2.Qe3
42/62 06:00 1,288,441k 3,570k -1.41 2.Qe3 Qb5 3.Qg5 Qb6 4.Bd4 Qc7 5.Qe3 Nf4 6.Bf3 Ne6 7.Bd5 Qd7 8.Bc4 Nxd4 9.Qxd4 Qxd4 10.Nxd4 Bc5 11.Nb3 Bb6 12.Kf1 Kf8 13.Nc1 f6 14.Ke2 Bf7 15.Bxf7 Kxf7 16.Kd3 Bxf2 17.Kc4 Be1 18.Nd3 Bd2 19.Nxb4 Bxb4 20.Kxb4 Ke6 21.Kc4 Ke5 22.h4 Kf4 23.Kd5 Kg3 24.Ke6 Kxg2 25.h5 Kg3

3)Suppose black trade queens at e1 after Nd3 Qd1 Qe1+ Qxe1 Nxe1
What is the reason to think that white is losing.

I can see only a score of +1.1 for black.

39/60 08:44 2,025,334k 3,863k -1.17 4.g3 Bf5 5.h4 Nc2 6.Be5 Bd3 7.Kg2 Be2 8.Bc7 g6 9.Bb6 Bg7 10.Nc1 Ne1+ 11.Kg1 Bb5 12.f3 Bc3 13.Kf2 Nc2 14.f4 Kg7 15.Bc5 Kf6 16.Kf3 Bd7 17.Ne2 Bd2 18.Nd4 Be3 19.Nb3 Bxc5 20.Nxc5 Bg4+ 21.Ke4 Bf5+ 22.Kf3 Ke7 23.Bc4 Bg4+ 24.Kxg4 Ne3+ 25.Kf3 Nxc4
40/52+ 09:14 2,142,660k 3,861k -1.09 4.g3
40/52 09:21 2,169,687k 3,861k -1.10 4.g3 Bf5 5.h4 Nc2 6.Be5 Bd3 7.Kg2 Be2 8.Bf4 Ne1+ 9.Kg1 g6 10.Nd4 Bd1 11.Kf1 Nd3 12.Bg5 Bg7 13.Be3 Kf8 14.f3 Bf6 15.Bc4 Ne5 16.Bd5 Ke7 17.Ke1 Kd6 18.Bxf7 Bxf3 19.Bg8 Be4 20.Ke2 Ng4 21.Bxh7 Be5 22.Nb3 Nxe3 23.Kxe3 Bd5 24.Nd2 Bxg3 25.Bxg6 Bxh4 26.Be4

this one is interesting.

why not Nb3 Bb3 Bf5, followed by Be6 and trading of light-square bishops?