I have become accustomed to seeing Jerome Gambit wins. They can be difficult to come by when the defender is a computer, but for a refuted chess opening the Jerome offers surprising chances for the enterprising attacker who chooses a human as his opponent.
This all came to mind when I played over the following game between online club players. White repeatedly misses the "best" moves - in the main game and in the games in the notes - and still wins.
How to explain it? It's the Jerome Gambit.
It's also a 3-minute game, and you can only dig so far into the analysis of any move.
farhadba - MeiND
3 0 blitz, lichess.org, 2021
1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+
4...Kxf7 5.Nxe5+ Ke6
This is an odd defense. I have written about it a number of times, most recently in "Jerome Gambit: More Concrete" and "Jerome Gambit: Over the Rainbow (Part 1)"
The earliest example I have in The Database is an internet game from 2002. This can hardly be the first time that 5...Ke6 was played, however, given that The Cheltenham Examiner's chess column of Wednesday February 21, 1906, suggested the move with the dubious comment that it "should win".
Suggesting otherwise, The Database has 48 games with 5...Ke6, with White scoring 71%.
Black's strategy seems to be one of opposition: If White wants me to take the Knight, I won't; if White wants my King to retreat, it won't.
White has met this line in a number of ways.
This move makes sense: White no longer offers the piece for capture, he exchanges it.
Still, Stockfish 14, at 30 ply, rates it as only the 5th best move available.
Slightly better is 6.Nf3, which also rescinds the offer. There are no examples in The Database.
Third best is 6.0-0, which leaves the Knight en prise, playing a move that is routine in the Jerome Gambit. There are no examples in The Database.
Second best is 6.d4, leading to an even game after 6...Nxe5 7.dxc5. There are four examples in The Database, three played by stretto. White scores 50%.
Best is the forcing 6.Qg4+, which has scored 81% in 8 games. The line is discussed in Wall, Bill - Skandervitch, Internet, 2021 (1-0, 11); Wall, Bill - Guest4105968, PlayChess.com, 2018 (1/2 - 1/2, 50); and Wall, Bill - Guest13762608, PlayChess.com, 2019 (1-0 8).
Recently, Astral1119 has won a couple of short games at lichess.org with 6.f4, but he had some cooperation after 6...Nxe5 (best), as both the terrible 7.fxe5 and the creative 7.d4 (best) are strongly met by 7...Qh4+.
White could also try 6.Qh5, since after 6...Nxe5 he will have transposed into regular Jerome gambit lines, and that might have some psychological impact on Black, if he has specially prepared 5...Ke6.
Black follows the rule of thumb - capture toward the center - and surrenders his advantage.
Instead, 6...dxc6 would open the d-file so that the dark square Black Bishop and Queen could prevent d2-d4. Even more useful would be having the diagonal of his light square Bishop opened up as well.
Also, 6...Qh4 looks quirky enough to be playable in this position, but it is well met by 7.Qe2 when the strangeness continues (according to Stockfish 14) with 7...Bxf2+ 8.Kf1 bxc6 9.Qxf2 Qxe4 10.d3 with White as better.
I know, I know, I just suggested that Black should have prevented this move on his last turn, but there is a far better move available to White now.
[to be continued]