Jerome Gambit: Stockfish vs Stockfish

The final game from Boris shows that Stockfish 11 cannot give itself "Jerome Gambit odds". Go figure.

I don't have suggestions on how to improve White's chances, except "Don't play the Jerome Gambit against computers".

Stockfish 11 - Stockfish 11

computer vs computer, 2020

1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+ 

4...Kxf7 5.Nxe5+ Nxe5 6.Qh5+ Ke6 

7.f4 d6 

The "silicon" or "annoying" defense, so called because it is the choice of different computer programs, and can be quite annoying to face. White can have a hard time getting something started against the centralized King.

8.fxe5 dxe5 9.Qh3+ Ke7 10.Qg3 Kf7 

11.Rf1+ Nf6 12.Qxe5 

White has 2 pawns for the sacrificed piece, and in a club level human vs human game, might have some chances. 

12...Bd4 13.Qh5+ Kg8 14.Qe2 Be6 

15.h3 c5 16.c3 Be5 17.d3 Qd6 18.Be3 Rd8 19.Nd2 Bg3+ 20.Kd1 Qxd3 21.Qxd3 Rxd3 

22.Rf3 Be5 23.Kc2 Rd8 24.a4 h6 25.a5 Kh7 26.g4 Nd7 27.Rg1 g5 28.Rf2 Kg6 29.Rfg2 Rhf8 

Black's Rooks occupy the 2 open files. White's Rooks over-protect a pawn.

30.b3 Bd6 31.Rf1 Rxf1 32.Nxf1 Rf8 33.Nd2 Bc7 34.Rg1 Bxa5 35.Ra1 Bb6 36.b4 cxb4 37.Bxb6 axb6 38.cxb4 

White's passed pawn is not sufficient compensation.

Black makes use of his extra piece.

38...Rc8+ 39.Kd3 Ne5+ 40.Kd4 Nc6+ 41.Kc3 b5 42.e5 Bd5 43.e6 Bxe6 44.Re1 Bd5 45.Nb3 Nd8+ 46.Nc5 Ne6 47.Rd1 Nf4 48.Kb2 Re8 49.Nd7 Re3 50.Nc5 Rxh3 White resigned

I suppose I'm going to have to ask Stockfish 11 to spend a bit more time trying to solve the "silicon" defense.

By the way, The Database has 214 games with the "silicon" defense, and White scores 55%. Again, go figure.

Jerome Gambit: Odds Are It's A Win

In the following game, Boris gives his opponent "Jerome Gambit odds" and finishes the game quickly against an insufficient defense.

Boris - NN

2020

1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+ 

4...Kxf7 5.Nxe5+ Nxe5 6.Qh5+ Ke6 

7.Qf5+ Kd6 8.f4 g6

A mistake. Not well thought through. When in doubt, attack the Queen.

9.Qxe5+ Kc6 10.d4 

Yes, yes, Boris knows that the Rook at h8 is en prise. Black's 8th move suggested that it was time to focus on checkmate. 

10...Bb4+ 

Black stumbles upon his best chance, when he makes his next move. He had some crazy chances right now with 10...Qh4+ 11.g3 Qh5. What follows (according to Stockfish 11) is funny: 12.f5 (threatening mate; 15.Qxh5 gxh5 16.dxc5 works, too) Bd6 13.d5+ Kb6 when 14.Qxh8?! would allow Black to hunt the White King after 14...Bxg3+ 15.Kd2 Bf4+ 16.Kd3 Qf3+ 17.Kc4 Ka6 18.Bxf4 b5+ 19.Kd4 Qxf4 20.Rd1 c5+. White would still objectively be better, but why seek out that pain? Instead, 14.Qc3 Ka6 15.Qd3+ followed by developing the Queenside and pressing the attack against the enemy King.

11.c3 Qh4+ 12.g3 Bxc3+ 13.Nxc3 Qh5 14.f5 b5 15.Qd5+ Kb6 16.Qxb5 checkmate

Jerome Gambit: Already Heading For Trouble

The following game is another good example how defenders of the Jerome Gambit (1.e4 e5 2.Bc4 Nc6 3.Nf3 Bc5 4.Bxf7+) can quickly get into trouble by playing a move that looks only indifferent, but which contains significant risk.

Boris - NN

2020

1.e4 e5 2.Bc4 Nc6 3.Nf3 Bc5 4.Bxf7+ 

4...Kxf7 5.Nxe5+ Nxe5 6.Qh5+ Ke6 

7.f4 Qf6 

We have seen this before in Boris' games. Check out "Jerome Gambit: Center Action". The move was suggested by Eric Schiller in his Unorthodox Chess Openings (1998).

8.Rf1 

The Database has 89 games with this position. White scores 65%.

An interesting alternative is 8.Nc3, which shows up 11 times in The Database, scoring 27%. See "Play The Jerome Gambit Quad".

8...g6 9.Qh3+ Kd6

Already heading for trouble. (That's the Jerome Gambit for you.) It wasn't immediately apparent, but he should have returned the piece with 9...Ng4.

10.d4 Bxd4 

This line demands more creativity. Instead, 10...Nd3+ 11.cxd3 Qxd4 takes advantage of White's King being placed in the center. The position would be unclear

11.fxe5+ Bxe5 12.Rxf6+ Nxf6 

White has a Queen for a Rook and a Bishop.

13.Bg5 Bxb2 14.c3

Trapping the Bishop. Also playable was 14.Nc3, as 14...Bxa1 15.e5+ Kxe5 (15...Kc6 was better, but still losing) would lead to mate after 16.Qe3+.

14...Nxe4 

Careless. Boris doesn't usually leave pieces lying around.

15.Qd3+ Kc6 16.Qxe4+ d5 17.Qa4+ b5 18.Qa5 Bxa1 19.Bf4 

19...Re8+ 20.Kd2 Re7 21.Na3 Rb8 22.Nc2 Rb7 23.Nxa1 Rf7 

24.Be3 Bf5 25.Bxa7 Kd6 26.Nc2 Bxc2 27.Kxc2 Rxa7 28.Qxa7 Kc6

Unless the clock is a factor, the game is largely over.

29.a4 bxa4 30.Qxa4+ Kb6 31.c4 Rf2+ 32.Kc3 dxc4 33.Qxc4 Rxg2 

It is true that a Queen vs Rook endgame, with no pawns, can be very difficult to win; perhaps this is why Black is hanging on.

34.Qd4+ Ka6 35.Qa4+ Kb6 36.Qd4+ Ka6 37.Qf4 Rg1 38.Qh6 Black resigned

Jerome Gambit: Very Complicated and Dangerous

In the following Jerome Gambit (1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+) game, play enters the very complicated and dangerous (for both players) historical "Nibs" variation. Only when Black overlooks a few things can White take control of the game, and take the full point. The notes only touch on a few of the difficulties.

For the  origin story of the line, see "Repairing A Variation (Part 2)"

Boris - NN

2020 

1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+ 

4...Kxf7 5.Nxe5+ Nxe5 6.Qh5+ Ke6 

7.Qf5+ Kd6 8.f4 Qh4+ 

Sometimes Black plays this move to weaken the White Kingside, before retreating his Queen to f6.

9.g3 

9...Nf3+ 

Black's counter attack is very strong. I have had my own troubles with the line (see "Jerome Gambit: Against the Titans"), and my record of 16 - 3 is a bit misleading.

Stockfish 11 shows the tiniest preference (about 1/4 pawn) at 30 ply for this move over 9...Qf6. 

10.Kd1 

The first big decision. His Majesty enters A Dangerous Place.

The Database has 26 games with 10...Kd1; White scores 65%. The Database has 19 games with 10...Kf1; White scores 74%. (Don't bother with 10...Ke2. White has scored 1 - 9 - 1 - 14% - with the move, according to The Database. See "Jerome Gambit: Amusement".)

Both Komodo 10 and Stockfish 11 clearly favor 10.Kf1. How does that fit with the practical results reflected in The Database? As always, the player more comfortable and more familiar with Jerome Gambit positions tends to do better, despite "objective" measurements.

10...Ne7 

Another developing move. For a long time I was convinced that this was the strongest continuation, although lately the computer engines have been pushing 10...Qh6 as even stronger. (The Database shows two games, with White scoring 50%).

11.e5+

White goes after the King. The computer engines suggest that he sigh deeply and give up on the idea of rescuing his Queen, settling for 11.Qxc5+ Kxc5 12.gxh4 with Black clearly for choice. 

11...Kc6 

The only move.

12.Qe4+ d5 

13.exd6+ 

Boris noted

Here my opponent was shocked, spent much time and lost his confidence - he was waiting for Qxf3 to execute a totally crushing move Bg4!!!, but he completely forgot the possibility of en-passant. 

Black has only one move that keeps the advantage, and he doesn't find it.

13...Kxd6 

Through the fog he had to see 13...Nd5 14.gxh4 Bg4, with the advantage, for example 15.c4 Ng5+ 16.Kc2 Nxe4 17.cxd5+ Kxd6 and White's extra pawns are too weak, while Black's development (and extra piece) likely give him a winning advantage.

14.gxh4 Bg4 

The position still looks very scary for White, but he has a new response that keeps Black's advantage small.

15.f5 Nxh2+ 16.Ke1 Nf3+ 

Rescuing the Knight, with check, looks reasonable, but this is where the complications overwhelm Black. He needed to find 16...Nxf5 17.Rxh2 Rae8 18.d3 Rxe4+ 19.dxe4 Nd4 when his lead in development counts more than White's material advantage of the exchange.

17.Kf1 Bxf5 18.Qxf3 Rhf8 19.Ke1 Rae8 

White's Queen is facing two Bishops, a Knight and two Rooks. Boris gives the proper perspective, though

Funny position - he seems to have two horrific discoveries, however there are no good attacks with the knight and lightsquared bishop, so I just smiled and made a room for my bishop, also taking e4 away from his bishop. 

20.d3 

If you are not chuckling at this move, you have not been paying attention. Boris has ice water in his veins.

20...h6 

Stockfish 11 worries that Black has allowed the game to slip away. White's Queen-for-Knight material advantage is now the most important factor.

21.Nc3 a6 22.Bf4+ Kd7 23. Kd2 Kc8 

Black's King is safe, but so is White's.

24.Rae1 Bb4 25.a3 Bxc3+ 26.bxc3 Rf6 27.Rhg1 Bg6 

28.Re2

Boris effectively described the situation

Black has no useful move. 

28...Kb8 29.Re3

Boris, again

At first I wanted to double rooks and win his knight, but now I realised that a rook-queen battery is better, because I can give up my bishop for his bishop with an invasion on the 7-th rank. 

29...Kc8 30.Qe2 Rxf4 31.Rxe7 Black resigned

Whew! That was quite a game. Great win, Boris!

Jerome Gambit: Under Pressure

The following game is a good example of the pressure that White can inflict upon the enemy King in a Jerome-related line. 

Boris - NN

2020

1.e4 e5 2.Nf3 Nc6 3.Bc4 Na5 

In this side line, Black immediately moves to win "the minor exchange" after 4.Bb3 Nxb3.

White can now grab a pawn with 4.Nxe5 Nxc4 5.Nxc4, but he prefers the "Jerome treatment".

4.Bxf7+ Kxf7 5.Nxe5+ Ke6 

A bit stronger might be 5...Ke7. See "Revision".

6.Qh5 

This is very Jerome-like play: first the Bishop is sacrificed, then the Queen goes to h5. The result is a very dynamic position. 

It is worth looking at the alternatives, however, as, surprisingly the fussy Komodo 10 rates other moves higher. Slightly.

First, it is important to mention that The Database has 87 games with 6.Qh5, with White scoring 68%, so it's a good move from a practical point of view, even if the computer evaluates it as -/=.

What does Komodo 10 prefer?

It's favorite (at 25 ply) is 6.d4, which it evaluates as +/=. The Database has 476 games with that move, with White scoring 73%. Fair enough.

The next in preference is 6.Nc3, evaluated as =. The Database has only 1 game, with White scoring 100%. Practically a novelty, or worthy of "Jerome Gambit Secrets #14"

Then comes 6.Qg4+, also evaluated as =. The Database has 55 games, with White scoring 64%.

Following is 6.Nf3, rated as =. The Database has 11 games, with White scoring 82%

Next is 6.0-0, evaluated as =. The Database has 2 games, with White scoring 100%. Again, practically worthy of "Jerome Gambit Secrets #15".

Rated as -/=, with 17 games in The Database, is 6.f4, although it scores 71%.

One way of deciding is to ask yourself, Do you want to sacrifice the Knight at e5? If Yes, you should look at 6.Nc3, 6.Qg4+ and 6.0-0. If No, you should look at 6.d4, 6.Nf3, 6.f4 or 6.Qh5.

It is very important to note that Komodo 10 (like The Database) is only an assistant in making a move choice. The difference in evaluation between its favorite, and the text move, is only about 1 pawn.

6...b5

This is a bit of a surprise, and justifies White's move. Strongest was probably 6...Nf6.

7.Qf5+ Ke7 8.Qf7+ Kd6 

9.Qd5+ Ke7 10.d4 

Capturing the Rook was possible, but White is looking to capture the King. 

10...c6 11.Bg5+ Nf6 12.Qf7+ Kd6 13.Bd2 Nb7 

14.a4 

Looking to add his Rook to the attack. Also possible was the material grab, 14.Ng6 Kc7 15.Nxh8 as 14...hxg6 would be answered by 15.Bf4#.

14...Qe8 

Looking to evict or exchange the Queen. Safer was to withdraw his King with 14...Kc7.

15.axb5 Nd8 

More of the same. His position is a mess. For example, if 15...cxb5, 16.Qxe8 Nxe8 17.Nf7+ will lead to the win of the Rook.

16.Bb4+ Kc7 17.Ba5+ Kb8 18.Qb3

Still focused on the King. He could be temporarily bought off by exchanging Queens on e8, then capturing the Knight on d8, but he wants more.

18...Nb7 19.bxc6 dxc6 20.Nd2 Bd6 

21.Ndc4 Bxe5 22.Nxe5 Nxe4 23.O-O Bd7 

Black still has a  piece for a couple of pawns, but the pressure on his King is intense.

24.Rfe1 Ned6 25.Bb4 Kc8 26.Ra6 h5 

If this is the start of a counter-attack, it is a bit optimistic.

27.Nxc6 Qf7 28.Qg3 Rh6 29.d5 h4 

30.Qc3 Bxc6 31.Rxc6+ Kb8 32.Bxd6+ Nxd6 33.Qb4+ Nb7 

34.Re7 Qxd5 35.Qxb7 checkmate