Jerome Gambit: Silicon Clash (Part 1)

A while ago, Sebastian Pereira sent me the following Jerome Gambit (1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+) game between Stockfish 14 and Stockfsh 14. He went to the chess.com website and chose the computer at engine level 25 (rated 3200). In a separate tab, he set up the Stockfish 14 engine to play against it.

Who won the game? Well, Stockfish 14, of course.

Along the way, it had some interesting things to show about our favorite opening. 

Stockfish 14 - Stockfish 14

Chess.com, 2021

1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+ 

4...Kxf7 5.Nxe5+ Nxe5 6.Qh5+ Ke6 

I find it interesting that Stockfish 14 prefers this move to 6...g6. I know that in its evaluations it is skeptical of Blackburne's defense, 7.Qxe5 d6, but it rates Whistler's defense, 7.Qxe5 Qe7 highly. Perhaps the computer is still in "book", and the move choice reflects the choices in place there. 

7.Qf5+ Kd6 8.f4 Qh4+ 9.g3 Qf6 

Again, Stockfish avoids a wild move - this time, 9...Nf3+!? leading to a Queen sacrifice - and plays a move that is solid and reliable for Black.

10.Qxe5+ Qxe5 11.fxe5+ Kxe5 12.b4

The Database has 4 previous examples of this move, all intenet games, all wins for White. One of them is a game that I played, and one was played by GM Hikaru Nakamura. I would happily give the name of the earliest player to use 12.b4, but it is given as NN in a 2004 game. At about the time of the current game, detonacasano won a rapid game at lichess.org.

It is also very relevant to read up on the Jerome-Mills variations, as analyzed by Yury V. Bukayev.

12...Bd4 

Of course, taking the pawn with 12...Bxb4 leads to instant trouble because of 13.Bb2+, as in detonacasano - yokucht, 15 10 rapid, lichess.org, 2021: 13.Bb2+ Ke6 14.Bxg7 Nf6 15.Bxh8 Nxe4 16.a3 Bc5 17.d3 Nf2 18.0–0? (18...Rf1) 18...Nxd3+ 19.Kg2 b6 20.cxd3 Bb7+ 21.Kh3 Rxh8 22.Nc3 Ke7 23.g4 Ba6 24.Rad1 Bd4 25.Ne4 Rg8 26.Ng3 Bc5 27.a4 Rg6 28.Kh4 Rh6+ 29.Nh5 Re6?! 30.g5 Re2 31.h3 Bf2+ 32.Kg4 Ra2 33.Nf4 Kd6 34.Ra1 Rd2 35.h4 Bd4 36.Rae1 Ra2 37.h5 Rxa4 38.h6 Ra3 39.Re4 Bc3?! 40.g6 hxg6 41.d4 Bb7 42.h7 Bxe4 43.h8Q Bf5+ 44.Kg5 White won on time

Simply backing up the Bishop with 12...Bb6 was reasonable, and was seen in NN - Gain,K, internet, 2004, (1-0, 55). 

Then there is the interesting 12...Bf8, protecting against the coming threat along the a1-h8 diagonal, as in perrypawnpusher - Kevin the fruitbat, Jerome Gambit thematic, ChessWorld.net, 2008 (1-0, 38).

13.c3 Bb6 14.Bb2 

About equally strong is 14.d4+ as seen in the grandmaster vs grandmaster blitz game, SanitationEngineer - GM_dmitrij, 3 0 blitz, Chess.com 2020, (1-0, 28). 

14...Kxe4 

This is the kind of risky move that a human would make. I am not sure what Stockfish 14 saw, or didn't see - and Black is still better after this move; but moves like 14...Nf6, 14...d6 or 14...c6 seem more reasonable.

15.Rf1 Nf6 16.Na3 Ke5 

Stockfish 14.1, given enough thinking time, recommends 16...Re8 17.O-O-O Kd3 which can lead to some wild play. Despite silicon calculations, I think it is a good idea to move the Black King closer to home.

[to be continued]