Best move?

[Werewolf]

In the book "Modern Chess Analysis", Robin Smith cites this game below.


[Site "London"]
[Date "1862.06.??"]
[White "Dubois, Serafino"]
[Black "Steinitz, William"]
[Result "0-1"]
[Annotator "Tartakower/du Mont"]
[EventDate "1862.06.??"]

He claims no computer has ever found the best move 8...h5!!


1. e4 e5 2. Nf3 Nc6 3.Bc4 Bc5 4. O-O Nf6 5. d3 d6 6. Bg5 h6 7. Bh4 g5 8. Bg3

[d]r1bqk2r/ppp2p2/2np1n1p/2b1p1p1/2B1P3/3P1NB1/PPP2PPP/RN1Q1RK1 b kq - 0 8

However, his analysis only includes accepting the sacrifice with 8...h5 9.Ng5 which indeed is bad for white. But white can decline, as Dubois did, with 9.h4. No winning line is given though Steinitz won the game thus "refuting" 9.h4 as well. I'm just not sure everyone is right, white seems OK to me with good play after 9.h4.

Thoughts?

[MikeB]

In the book "Modern Chess Analysis", Robin Smith cites this game below.


[Site "London"]
[Date "1862.06.??"]
[White "Dubois, Serafino"]
[Black "Steinitz, William"]
[Result "0-1"]
[Annotator "Tartakower/du Mont"]
[EventDate "1862.06.??"]

He claims no computer has ever found the best move 8...h5!!


1. e4 e5 2. Nf3 Nc6 3.Bc4 Bc5 4. O-O Nf6 5. d3 d6 6. Bg5 h6 7. Bh4 g5 8. Bg3

[d]r1bqk2r/ppp2p2/2np1n1p/2b1p1p1/2B1P3/3P1NB1/PPP2PPP/RN1Q1RK1 b kq - 0 8

However, his analysis only includes accepting the sacrifice with 8...h5 9.Ng5 which indeed is bad for white. But white can decline, as Dubois did, with 9.h4. No winning line is given though Steinitz won the game thus "refuting" 9.h4 as well. I'm just not sure everyone is right, white seems OK to me with good play after 9.h4.

Thoughts?

I agree with his claim that no computer has ever found it, and disagree with his claim that it is the best move. 9.h4 is an excellent move after 8. ..h5

[Dann Corbit]

I think h4 is the best move, and h5 is the best response. The result of which (with perfect play) is equality for the position.

However, I think that h4 is especially good against a human because they will be sorely tempted by Nxg5. That's what I would have done.

And if they do not do the Nxg5 mistake, b4 is waiting in the wings as a more sophisticated (wrong) temptation.

And if they do find h4, they have only achieved equality, the same as they would have gotten if black had originally chose one of the plodding move choices, but those choices do not introduce the steaming peril of h5.

[Eelco de Groot]

If we can believe Kaissa (still very buggy IV, but maybe just a bit less because of synergy of bugs with the new Stockfish code) analysis, it has shown by now that 9. h4 is in fact the only playable move for White. So you only have to refute one move and Robin Smith would still be right. Or even if your computer can't refute it right now, such a situation where only one move does not lose according to computer, everything else falls down if you look longer, you can still think there is a probability the computer does not see it very well and that one move also loses. It is only 8 moves in the opening so unless h4 very good, one line line of Giuoco Piano can be deemed unplayable. I have not checked what the theory says.

[d]r1bqk2r/ppp2p2/2np1n2/2b1p1pp/2B1P3/3P1NB1/PPP2PPP/RN1Q1RK1 w kq -

Engine: Kaissa IV (Totally untested, based on yesterday's Stockfish master code, no tablebases, 1 thread, i7 6700, 512 MB)
by T. Romstad, M. Costalba, J. Kiiski, G. Linscott

50 499:06 +1.07 9.h4 Nh7 10.c3 Bb6 11.a4 a5 12.Nbd2 g4
13.Nh2 Qf6 14.Kh1 Qg6 15.f3 gxf3
16.Rxf3 Be6 17.Bxe6 fxe6 18.Nc4 Ba7
19.Qb3 O-O-O 20.Raf1 Rhf8 21.Qb5 Rxf3
22.Rxf3 (50.736.684.239) 1694

50 499:06 -0.97 9.c3 h4 10.d4 exd4 11.cxd4 hxg3
12.dxc5 gxh2+ 13.Kh1 g4 14.Ng5 Ne5
15.Qb3 O-O 16.f4 Nxc4 17.Qxc4 d5
18.exd5 Qxd5 19.Qxd5 Nxd5 20.Kxh2 Ne3
21.Rf2 Kg7 22.Re2 (50.736.684.239) 1694

50 499:06 -0.98 9.b4 Bb6 10.b5 Na5 11.Bxf7+ Kxf7
12.Nxg5+ Kg7 13.Bh4 a6 14.Nc3 Bd4
15.Qd2 axb5 16.Nd5 Bxa1 17.Rxa1 Nxd5
18.Ne6+ Bxe6 19.Bxd8 Rhxd8 20.Qg5+ Kf7
21.exd5 Bxd5 22.Qxh5+ (50.736.684.239) 1694

50 499:06 -1.44 9.Nxg5 h4 10.Nxf7 Qe7 11.Nxh8 hxg3
12.Kh1 Bxf2 13.Rxf2 gxf2 14.Qf3 Qg7
15.Qxf2 Qxh8 16.Nc3 Ng4 17.Qf7+ Kd8
18.h3 Ne3 19.Nd5 Nxd5 20.Bxd5 Ne7
21.Rf1 Bd7 22.Be6 (50.736.684.239) 1694

50 499:06 -2.69 9.Bxf7+ Kxf7 10.Nxg5+ Kg7 11.h4 Bg4
12.Nf3 Nd4 13.Nxd4 Bxd4 14.Qc1 Rf8
15.Na3 Be6 16.Nb1 Ng4 17.Nd2 Bb6
18.Nf3 Qf6 19.a4 a5 20.Qg5+ Qxg5
21.hxg5 Rf7 22.Rad1 (50.736.684.239) 1694

If anyone finds a refutation of 9. h4 maybe we can dedicate it to Robin

[Eelco de Groot]

The situation at depth 59 was about the same for the 5 best moves. 9. h4 got a little better, 9. c3 dropped out and in its stead came 9 Bb5.

r1bqk2r/ppp2p2/2np1n2/2b1p1pp/2B1P3/3P1NB1/PPP2PPP/RN1Q1RK1 w kq -

Engine: Kaissa IV (512 MB)
by T. Romstad, M. Costalba, J. Kiiski, G. Linscott

59 2485:00+1.22 9.h4 Bg4 10.Nbd2 Nd4 11.c3 Nxf3+
12.Nxf3 Nd7 13.d4 Bb6 14.Qd3 Bxf3
15.Qxf3 Qf6 16.Qxf6 Nxf6 17.Rfe1 gxh4
18.Bxh4 Nd7 19.Rad1 c6 20.f4 exf4
21.Rf1 Bd8 22.Bxd8 (256.659.659.331) 1721

59 2485:00-0.57 9.b4 Bb6 10.h4 Bg4 11.c3 a5 12.b5 Ne7
13.Nbd2 Nh7 14.Qb3 O-O 15.Nh2 a4
16.Qc2 Bd7 17.hxg5 Ng6 18.Ndf3 Nxg5
19.Qd2 Nxf3+ 20.Nxf3 Kg7 21.Bh2 h4
22.Kh1 (256.659.659.331) 1721

59 2485:00-1.47 9.Nxg5 h4 10.Nxf7 Qe7 11.Nxh8 hxg3
12.Kh1 Bxf2 13.Rxf2 gxf2 14.Qf3 Qg7
15.Qxf2 Qxh8 16.Nc3 Ng4 17.Qf7+ Kd8
18.h3 Ne3 19.Nd5 Nxd5 20.Bxd5 Ne7
21.Rf1 Bd7 22.Be6 (256.659.659.331) 1721

59 2485:00-2.60 9.Bb5 h4 10.Bxe5 dxe5 11.Nxe5 Qd6
12.Nxc6 bxc6 13.e5 Qd7 14.Bc4 Nh5
15.Nd2 Nf4 16.Qf3 Rb8 17.Rab1 Qg4
18.e6 Bxe6 19.Bxe6 Qxf3 20.Bxf7+ Kxf7
21.Nxf3 Kf6 22.g3 (256.659.659.331) 1721

59 2485:00-2.67 9.Bxf7+ Kxf7 10.Nxg5+ Kg7 11.h4 Bg4
12.Nf3 Nd4 13.Nxd4 Bxd4 14.Qc1 Rg8
15.Nd2 a6 16.c3 Ba7 17.d4 exd4
18.cxd4 Bxd4 19.Qc4 c5 20.Nb3 d5
21.exd5 Qxd5 22.Qxd5 (256.659.659.331) 1721