Opening Evolution - Philidor Countergambit (part two)

Howard Staunton & John Owen - Paul Morphy & Thomas Barnes
London 1858
1.e4 e5 2.Nf3 d6 3.d4 f5!?
Morphy played the Philidor nine times, which compares with three times for the Petrov and 50 times for 2...Nc6, according to ChessBase's 2019 Mega database. Eight of his Philidor opponents played 3.d4, with Morphy replying 3....f5!? four times, scoring +3=1-0, and the modern favourite 3...exd4 four times, scoring +2=1-1. The one time he faced 3.Bc4, Morphy also played 3...f5!?, with the game ending in a draw.
4.dxe5 fxe4 5.Ng5 d5 6.e6 Nh5 7.Nc3 c6 8.Ngxe4!? dxe4 9.Qh5+ g6 10.Qe5 Rg8 11.Bxh6 Bxh6 12.Rd1

The same position as reached five times in games between Atwood and Wilson in the late 1700s (see part one of this series)

*****
*****
*****
*****
*****
12...Qg5!?
Wilson always played 12...Qe7, which is at first much preferred by Stockfish10 and Komodo10. But give the engines more time and their evaluations narrow, although still preferring Wilson's move. Saveilly Tartakower in 500 Master Games Of Chess calls 12...Qe7 "passive" and the text "more active," but does not explicitly state his preference.
13.Qc7
Tartakower comments: "For Black to have permitted this intrusion, which threatens Qxc8+ as well as a mate on the move, is proof of steady nerves or most-accurate calculation."
13...Bxe6 14.Qxb7 e3
Tartakower points out that this threatens ...exf2+ followed by ...Qe6#.
15.f3 Qe7!?
The engines prefer 15...e2 16.Bxe2 Qxg2 17.Rf1 Qxh2 18.Ne4! (if 18.Qxa8?! then 18...Qg3+ 19.Rf2 Be3 20.Ne4 Bxf2+ 21.Nxf2 Qc7, when material is level but the engines give Black an edge) Qh4+ 19.Rf2 Be3 20.Qxa8 Bxf2+ 21.Nxf2 Qb4+ 22.c3 Qb6, which is very similar to the line in brackets, but this time the engines slightly prefer White.
16.Qxa8 Kf7
White has emerged from the immediate complications with rook and pawn for bishop. If the white queen can escape from the corner, White must be much better.
17...Bf4?
The engines suggest 17...Kg7, while preferring White.
18.Be2?
Komodo10 likes Johann Löwenthal's suggestion of 18.g3 for quite some time, but eventually agrees with Stockfish10 and an anonymous ChessBase annotator that best is 18.Rd4, the point being to follow with Ra4.
18...Kg7 19.0-0
No longer so strong is Rd4 as Black has 19...Rc8, when 20.Ra4?? runs into 20...Na6 21.Qxc8 Bxc8. Note that if White had played 18.Rd4, the reply 18...Rc8 would not solve Black's problems as White has the pinning 19.Bc4, removing the threat of ...Na6.
19...Qc7?!
Threatening to win the white queen and also threatening ...Bxh2+. But White finds a strong reply, so it seems that better was 19...Rc8!?, when the white queen can be saved by 20.Rd6 Bxd6 21.Nxd6 Qxd6 22.Qxa7+. The engines continue 22...Bf7 23.Qxe3, when they much prefer Black, although the position is unclear as White has three pawns for a knight.
20.Nc5! Bxh2+ 21.Kh1 Bc8
White is the exchange up but the position remains complicated as his queen is still endangered and Black has mating threats.
22.Rd4 Bg3 23.Re4?
This loses. The engines give 23.Ne4! when the game's ...Kh8? loses to 24.Nxg3 Qxg3 25.Qxa7. However, Black can save the dark-square bishop, eg 23...Be5, when the engines reckon Black has full compensation for the exchange.
23...Kh8
The fate of the white queen will be irrelevant if Black delivers mate by switching his queen to the h file via g7.
24.Rd1 Qg7 25.Rh4!?
Desperation, but it contains a point Morphy & Barnes miss.
25...Bxh4?
Best was 25...Qe7 (other moves also win), eg 26.Rdd4 Qxc5.
26.Qxb8 Ba6 27.Qh2?
White had to find 27.Qf4, which, like the text, attacks the dark-square bishop, but has the crucial extra factor of covering the h6 checking square. After 27...Bxe2 28.Rd7 Be7 (the only way to save the dark-square bishop as well as Black's queen) 29.Ne6 Qf6 30.Rxe7! Qxe7 31.Qe5+ Rg7, White has a perpetual by 32.Qb8+ etc.
27...Bxe2 28.Rd7 Qh6 29.Ne4
Nothing gets White's piece back.
29...Bc4 30.Nf6 e2 31.Re7 Qc1+ 32.Qg1 Qxg1+ 33.Kxg1 e1=Q+ 34.Rxe1 Bxe1 0-1

Morphy v The Petrov (part six)

PAUL Morphy's final game against the Petrov was part of a five-board simul against strong opposition, including Johann Löwenthal and Thomas Barnes.
Morphy - Barnes
St James's Chess Club, London 1859
Petrov Defence Boden–Kieseritzky Gambit
1.e4 e5 2.Nf3 Nf6 3.Bc4!?
This game was played just six days after their previous encounter, in which Morphy chose the Classical Attack (3.Nxe5).
3...Nxe4 4.Nf3 Nxc3
The main move, and surely the real test of the Boden–Kieseritzky.
5.dxc3 f6 6.0-0!?
More popular in ChessBase's 2019 Mega database is 6.Nh4, but David Hooper in A Complete Defence To 1 P-K4: A Study Of Petroff's Defence (Pergamon Press, 1967) says 6.Nh4 "commits rather too soon to one line of attack."
6...Qe7
In a previous game between the same players in 1858, the same position was reached from a Bishop's Opening move-order, and in that game Barnes played 6...Nc6, so the text was presumably a prepared improvement.
7.Nh4
Most popular, but "not effective," according to Hooper. He preferred 7.Re1!?, which is the choice of Stockfish10 and Komodo10.
7...d6?
Hooper gave 7...g6, which is much preferred by the engines.
8.Qh5+ Kd8
The displacement of the black king, combined with White's lead in development, make for fair compensation for the pawn-minus.
9.f4 Be6
The point of 7...d6?, but the engines prefer 9...g6 10.Nxg6 Qe8 11.Qh4 Qxg6 12.fxe5 Bf5 13.exf6 Nd7, with a small edge (Komodo10), or 9...a5 10.f5 Bd7 11.Be3 c6 12.Ng6 hxg6 13.Qxh8 gxf5, again with a small edge (Stockfish10). However, in both cases there is a lot of play left in the position.
10.Bxe6 Qxe6 11.fxe5 dxe5 12.Ng6+
12.Be3!? was played in a 1998 Spanish game.
12...Bc5+ 13.Kh1 Re8 14.Qxh7
Hooper reckons the position is equal, as does Komodo10, but Stockfish10 gives White the upper hand.
14...Qg8 15.Qh5 Nd7 16.b4 Bd6 17.Bd2?!
Lowenthal: "Intending, probably, to advance c4, but it would have been much stronger play to post this bishop at e3." The engines agree.
17...Qf7 18.Qg4 Qe6 19.Qe4 Nb6!?
Lowenthal: "A very subtle, clever move, the true motive of which White entirely overlooked."
20.Qxb7!?
Löwenthal: "Curiously enough, this move, as Black foresaw, cost White his knight." But the text is the engines' top choice, so who is trapping whom?
20...Qg4 21.a4?
This is the real culprit. Komodo10 gives 21.c4!? Qxc4 22.Be3, with a slight edge for Black (although the more-volatile Stockfish10 has Black close to winning). Stockfish10 reckons best was 21.Rfe1!? Qxg6 22.c4 Ke7 23.c5 Kf8 24.cxd6 cxd6, with slightly the better position for Black.
21...Rc8!
White's knight is still trapped, and his compensation is inadequate.
22.Rad1 Qxg6 23.Be3 Nc4 24.Qc6 Qf7
Löwenthal points out that 24...Nxe3?? loses to 25.Rxd6.
25.Bxa7 e4?
Black is much better after 25...Rh8 or 25...Qe6. The text is faulty because it allows White's d1 rook strongly into the attack.
26.Rd4?
But this is not the way to menace the black knight as Black can ignore White's threat. Correct was 26.Rd5, with the engines giving best play as 26...Re5 27.Rd4 Rh5 28.Qxc4 Rxh2+ 29.Kg1, when White seems to be better despite the precarious-looking situation of his king. In this line, if Black tries 27...Ne3!?, White has 28.Rxd6+! cxd6 29.Qxd6+, with Be3 to come.

Black to play and win

*****
*****
*****
*****
*****
26...Qh5
Löwenthal: "Menacing mate, and capitally played."
27.Rf4
There is no salvation in 27.g3, eg 27...Rh8 28.Rf2 e3 29.Rg2 e2 30.Re4 Qf3+ 31.Rg2 Qf1+ 32.Rg1 e1=Q 33.Rxf1 Rxh2+! 34.Kxh2 Qxg3+ 35.Kh1 Qh2#
27...e3 28.g4 e2 29.gxh5 e1=Q+ 30.Kg2 Re2+ 31.Kh3 f5?!
Several other moves win more easily, but this is good enough.
32.Qxc4?!
More challenging was 32.Rxc4. Black then has to find 32...g5 33.hxg6 Ke7, when the white pieces cannot get at the black king.
32...Qf1+! 33.Kh4
33.Rxf1?? Rxh2#
33...Rxh2+ 34.Kg5 Qg2+??
Correct was 34...Qg1+, and if, as in the game, 35.Rg4, then 35...Qe3+ leads to mate.
35.Rg4! fxg4 36.Qf7??
White wins with 36.Rxd6+ (a recurring theme in this game, which makes it all the more surprising that Morphy missed it) cxd6 37.Qg8+ Kd7 38.Qxg7+ Ke8 (other moves allow a quick mate) 39.Qg6+ Ke7 40.Qf6+ Ke8 41.Qe6+ Kf8 42.Qxd6+ Ke8 43.Qe6+ Kd8 44.Bb6+ Rc7 45.Qd6+ Ke8 46.Qxc7 - a long line, but much of it forced.
36...Qxc6
Making a sac on d6 ineffective, and thus winning.
37.b5 Qd7 38.Rxd6 Qxd6 0-1
Even at this late stage it was still possible to go wrong, ie 38...cxd6?? 39.Bb6+ wins for White.

Morphy v The Petrov (part five)

PAUL Morphy's final two games against the Petrov featured the English master Thomas Barnes.
Morphy - Barnes
St George's Chess Club, London 1859
Petrov Defence Classical Attack
1.e4 e5 2.Nf3 Nf6 3.Nxe5
In this game, part of an eight-board blindfold simul, Morphy sticks with the Classical Attack that he used to such good effect earlier in the year in Paris (see part four of this series).
3...d6 4.Nf3 Nxe4 5.d4 d5 6.Bd3 Bd6!?
Barnes is the first to vary. This aggressive continuation has become a favourite of Caruana.
7.0-0 0-0 8.c4 Nf6!?
8...c6 is overwhelmingly more popular today, and was already the mainline when David Hooper write his 1969 classic A Complete Defence to 1 P-K4: A Study Of Petroff's Defence. Hooper credited the text to Pillsbury, and called it "worth consideration."
9.Bg5 Be6
The mainline runs 9...dxc4 10.Bxc4, transposing to a game from the 1834 first match between Louis de Labourdonnais and Alexander McDonnell.
10.Qb3 dxc4

White to make his 11th move

*****
*****
*****
*****
*****
11.Bxc4
Johann Löwenthal in Morphy's Games Of Chess points out that 11.Qxb7?? loses to 11...Bd5, eg 12.Bxf6 gxf6 13.Qxd5 Bxh2+ etc.
11...Bxc4 12.Qxc4 Nc6 13.Nc3 h6 14.Bh4 g5!? 15.Bg3 Qd7 16.Ne5 Qe6 17.Qxe6 fxe6 18.Nxc6 bxc6
Bearing in mind his airy king's position, Black can be pleased at getting the queens off, but, as Löwenthal points out, "White has somewhat the better game as Black's pawns are evidently weak."
19.Rfe1 Rfe8 20.Rac1 Rab8 21.b3 Re7 22.Na4 Rb4 23.Be5 Bxe5 24.dxe5 Nd5 25.Rxc6
White wins a pawn but, as in his game against Löwenthal, Morphy's endgame technique proves insufficient.
25...Rd4 26.Rc2 Nb4 27.Rce2 Nd3
Löwenthal  states that "from this position the game ought, with correct play, to be drawn," but that is highly debatable.
28.Rd2?!
Better seems to be 28.Rd1, preventing the little combination Black plays in the game.
28...Rxa4! 29.Rxd3?
And here White should have played 29.Re3 Ra5 30.Rexd3 Rxe5 31.h3, with winning chances thanks to Black's three isolated pawns.
29...Rxa2 30.Rc1 Rf7 31.f3 ½–½