Jerome Gambit: This "Refuted" Thing (Part 2, What to Do?)

 

[continued from previous post]

So - what do you do, if your favorite opening is not appreciated by grandmasters or chess computers?

Of course, the easiest thing is to do nothing. Win. Lose. Draw. Set the pieces up again. Play some more. No need to get excited. It's only a game. A few minutes after you leave the board, it's all forgotten.

Similarly, if you are having success with the Jerome Gambit against humans, but are getting crushed by computers - don't play against computers. Some people like the challenge (see "Beating Beth Harmon With the Worst Opening Ever" and "Jerome Gambit: How He Did It") but others find it exhausting and unrewarding. 

As we saw in the previous post, it is possible to take advantage of  factors that tilt things more in the Jerome Gambit player's favor. Two of these are time and knowledge.

For example, as we saw in a couple of earlier posts (see "Jerome Gambit: GM vs GM!?" and "Jerome Gambit: Oops, He Did It Again?!") Grandmaster Hikaru Nakamura benefitted from limiting his opponent's thinking time - Nakamura is one of the top blitz players in the world - and benefitted from the fact that his Grandmaster opponent was not familiar with the Jerome Gambit to score two wins in a couple of 3-minute games.

On the other hand, when GM Nakamura later faced Grandmaster Peter Michalik (see "Jerome Gambit: GM vs GM, Upset" and "Jerome Gambit: Gm vs GM, Upset Again"), GM Michalik seemed to know enough about the Jerome Gambit that the result was a couple of losses for White.

(I am not sure where time and knowledge for Grandmasters fit into Chandler's "blunder table".)

Of course, one thing that the Jerome Gambit player can do is examine his own games. Which moves seemed to advance his cause, which moves seemed to set him back. While we are always competing against our opponent, we also are also playing a game against ourselves. Increasing our own knowledge decreases the defender's, relatively.

The next step is to play over more Jerome Gambit games, like the ones on this blog, simply to absorb the strategy and play. Familiarity is a form of knowledge, and the more the gambiteer knows abut the opening, the better his chances.

On top of that, it is sometimes useful to examine the notes to the games, and to analyze the play of the game. 

It is also possible to create (see "You, too, can add to Jerome Gambit theory!") your own solutions to some of the Jerome Gambit's problems. A good example of how this is done is the series of posts " JG: The New in Its Opening Theory, in Its Psychology  (Parts 1, 2, 3, 4 & 5)" by Yury Bukayev, as he examines a new idea for White after 1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+ Kxf7 5.Nxe5+ Nxe5 6.Qh5+ Ke6.

  

Jerome Gambit: This "Refuted" Thing (Part 1, A Basic Truth)

I recently received an email from a Reader of this blog who has been playing the Jerome Gambit (1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+) successfully, and who has been enjoying the opening quite a bit.

There was this one line, however, that gave him trouble when he played it against a computer engine (set at a reasonble level of strength) - did I have any ideas for how he could succeed against it?

Other Readers have written to me with similar concerns.

Some Readers have written expressing puzzlement, that while the Jerome Gambit had led to a series of smashing, fun wins for them - a computer chess engine evaluated the opening as just plain no good, so how could that be? 

Those questions sent me into a deep think...

I have some answers. Please read along.

Let me start with a basic truth: The Jerome Gambit is a refuted opening. But - what does that mean, exactly?

What that means is: If your opponent is familiar with the Jerome Gambit and has enough time to think things through, he will probably have the advantage and may very well defeat you.

Of course, it's not that simple.

It is worth going back 13 years to a blog post titled "But – Is this stuff  playable?? (Part I)" where I reframed the discussion

Maybe a more useful question would be --
Under what conditions might the Jerome Gambit be playable?

I then shared some of Grandmaster Nigel Davies' wisdom from his book Gambiteer I (2007)

Having examined literally thousands of club players’ games over the years, I have noticed several things:
1) The player with the more active pieces tends to win.
2) A pawn or even several pawns is rarely a decisive advantage.
3) Nobody knows much theory.
4) When faced with aggressive play, the usual reaction is to cower.

 It is important to quickly pair GM Davies's thoughts with Geoff Chandler's "Blunder Table" (substitute "gambit" for "blunder")

Here is a one-move blunder table showing how severe the blunder needs to be in a game between two players of the same grade.

All players should be able to spot their opponent leaving a mate in one on.

A 1200 player should win if an opponent blunders a Queen or a Rook. But not necessarily if they pick up a Bishop or Knight.

1500 players often convert piece-up games into a win, but this is not the case if a pawn or two up.

An 1800 player usually wins if they are two pawns up.

In a game between two 2000+ players a blundered pawn is usually enough to win.

One conclusion to be drawn here is that an opening that is "refuted" at the Grandmaster level is not necessarily "refuted" at the beginner or club player level.

Another conclusion is that if you play the Jerome Gambit in a blitz or bullet game, your opponent may not have enough time to figure things out - much to his disadvantage.

Finally, if your opponent is unfamiliar with the Jerome Gambit - and many players are not familiar - he is more likely to make errors, much to your advantage. (As GM Nakamura cautioned "Once everybody knows about it, it won't be as good... as it has been")

So - how to respond to the questions mentioned at the start of this post? The answers range from the very simple, to the rather complex.

[to be continued]

BSJG Improvement

A day after our first game, HanPeng and I met up again at FICS. He was comfortable playing the Blackburne Shilling Gambit. I was comfortable playing the Blackburne Shilling Jerome Gambit.

My goal was to avoid hanging my Queen or a Rook this time.

perrypawnpusher - HanPeng

10 0 blitz, FICS, 2021

1.e4 e5 2.Nf3 Nc6 3.Bc4 Nd4 

The Blackburne Shilling Gambit.

4.Bxf7+ Kxf7 5. Nxe5+ Ke7 

My opponent varies first. He is not interested in 5...Ke8 6.Qh5+ g6 7.Nxg6 again.

I decided to take my own advice, offered a few posts ago

Against 5...Ke7 or 5...Ke8, 6.c3 is also a good idea...

6.c3 

Not 6.Qh5, as in perrypawnpusher - vlas, blitz, FICS, 2009 (0-1, 26) because of 6...Qe8.

6...Ne6

Keeping the Knight closer to the Kingside, but  6...Nc6 was probably a bit better.

Instead, 6...d6 7.cxd4 dxe5 8.dxe5 Bg4 was seen in perrypawnpusher - michon, blitz, FICS, 2013 (1-0, 18).

7.d4 d6 

An improvement over 7...Nf6 seen in perrypawnpusher-Macgregr, blitz, FICS, 2010 (1-0, 56)

8.Nd3 h6 

The center strike 8...d5 looks like the best move. 

A different idea was the Kingside fianchetto 8...g6 9.f4 Ke8 10.O-O Bg7 in perrypawnpusher - ZekeTheWolf, blitz, FICS, 2014 (1-0, 22). 

9.O-O g5

This move fails in its intention to stop White's f-pawn. Black comes up with another idea - to return the sacrificed material - as a fall back plan.

10.f4 Nf6 11.f5 Ng7 12.e5 dxe5 13.dxe5 Bxf5 14.exf6+ Kxf6 

White's Knight at d3 is under pressure, and the undeveloped state of White's Queenside is a problem - despite the fact that I can now win a piece.

15.g4 Qd5 16.Qf3 

Leading to nothing more than an even position. A little creatvivity would have produced 16.Be3 with the idea if 16...Rd8 17.Bd4+.

16...Qxf3 17.Rxf3 

17...Kg6 

In turn, Back wants to untangle his Kingside.

After the game Stockfish 13 gave a lesson in the value of development. The line is long, but well worth playing over: 17...Rd8 18.Nf2 Bc5 19.Nd2 Kg6 20.gxf5+ Nxf5 21.Kf1 Ne3+ 22.Ke2 Rhe8 23.Nfe4 Nd5 24.Kf1 Bb6 25.a4 a5 26.b4 axb4 27.cxb4 Ne3+ 28.Ke2 Nc2 29.Rf6+ Kh7 30.Rb1 Be3 31.Rb2 Bxd2 32.Rxc2 Rxe4+ 33.Kf3 Re3+ 34.Kf2 Bxc1 35.Rxc7+ Kg8 36.Rg6+ Kf8 37.Rxc1 Rh3 and the game is even.

18.gxf5+ Nxf5 19.Nd2 Bd6 20.Ne4 Rad8 21.Nxd6 Rxd6 22.Ne5+ Kf6 23.Ng4+ Kg6 24.Ne3 Nxe3 25.Bxe3 

Untangled.

25...Rd3 26.Raf1 Rhd8 27.Bd4 Rd2 28.Rf6+ Kh5 29.R1f2 Rd1+ 30.Kg2 b6 31.Re2 c5 32.Bf2 

32...Ra1 33.Ree6 Rxa2 

Hastening the end.

34.Rxh6+ Kg4 35.h3+ Kf5 36.Rhf6 checkmate

BSJG Dumpster Fire

It has been almost 2 years since I played a Jerome Gambit game at blitz speed. That is kind of sad, as a quick peek at early blog posts (say, perhaps the first 2,500 or so) would show a lot of quick games by perrypawnpusher.

I recently started to try some blitz at FICS, and the following game - an embarassing dumpster fire - shows that I need much more practice to return to successful quick play.

perrypawnpusher - HanPeng

10 0 blitz, FICS, 2021

1.e4 e5 2.Nf3 Nc6 3.Bc4 Nd4 

The Blackburne Shilling Gambit.

4.Bxf7+ 

The Blackburne Shilling Jerome Gambit, played instantly. I have played the line 44 times previously.

4...Kxf7 5.Nxe5+ Ke8 6.Qh5+ 

6... g6 7.Nxg6 Nf6 

Stockfish 13 prefers 7...hxg6 8.Qxh8 Nxc2+ 9.Kd1 Nxa1 10.Qxg8 Qg5 and gives Black the advantage in a messy position.

8.Qe5+ Ne6 9.Nxh8 Bg7 

10.d4 Bxh8 11.d5 d6 12.Qd4??

When Willy Hendriks wrote Move First, Think Later (2012), this is not the kind of move he had in mind.

Well, at least The Database says it was not a TN. I was ready to resign when...

12...Nc5?

We both staggered on.

13.Bg5 b6 14.0–0 Ncxe4 15.Re1 Kf7 16.Rxe4

Missing the fact that the Knight is not really pinned.

16...Nxe4 17.Bxd8 Bxd4 

From here I played 22 more moves, hoping that I could hustle my opponent into losing on time, but it was all for naught and I had to resign a move before being checkmated.

Well, I will try to not do that again.

Jerome Gambit: Inspiring

One thing I like about the Jerome Gambit (1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+) is that it often inspires (for example, see "Jerome Gambit-Inspired Play" Parts 1, 2, 3, 4, 5, 6, 7 & 8 ) bold, attacking play, even if it arises from slightly different openings.

The following game is fun to play over and examine. It reminds me of the quote by Mikhail Tal "You must take your opponent into a deep dark forest where 2+2=5, and the path leading out is only wide enough for one."

tokyoghoul - NN

Chess.com, 2020

1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 

The Italian Game.

4.a3 

Fascinating. Is White preparing to play a "reversed" Vienna game? The move brings to mind 1.e4 e5 2.a3, Mengarini's Opening, and Hugh E. Myers' Reversed king pawns: Mengarini's opening (1977) which looked at 1.e4 e5 2.Nc3 Nf6 3.a3.

Tokyoghoul might not have been aware that he was following in the path of Windmueller - Schallopp, blindfold simultaneous exhibition, Hamburg, 1868 (0-1, 50) - at least for 4 moves. 

4...b6 

Possibly taking White's last move as a sign of slow and peaceful intent. White 4.a3 can have its benefits for the first player, as we shall see, the benefit of 4...b6 for the defender is a bit less clear, showing up mostly in lines not played. Certainly there was little wrong with 4...Nf6.

5.Bxf7+ 

Suddenly: Jerome!

5... Kxf7 6.Nxe5+ Nxe5 7.Qh5+ g6 8.Qxe5 Qe7 

Wow. The game now resembles Whistler's Defense to the Jerome Gambit (1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+ Kxf7 5.Nxe5+ Ne5 6.Qh5+ g6 7.Qxe5 Qe7), with the addition of a2-a3 for White (which might prove helpful) and ...b7-b6 for Black (which creates weaknesses).

9.Qd5+ 

Following the major rule when facing Whistler's - do not take the Rook, for 9.Qxh8 Qxe4+ creates great danger for White's King.

9...Kg7 

It is too easy (and likely incorrect) to psychologize and decide that Black's move unconsciously blocks a possible enemy Queen attack on the h8 Rook, along the a1-h8 diagonal.

Even more, the correct move, 9...Qe6 would have lost the other Rook, although Stockfish 13 likes Black after 10.Qxa8 Ne7 11.Nc3 Nc6 12.b4 Bd4 13.b5 Ba6 14.Qxh8 Bxh8 15.bxa6 Bxc3 16.dxc3 Qxe4+ 17.Be3 Qc4 18.O-O-O d6 as Black's Queen would be better at grabbing extra material than White's Rooks would be.

10.b4 

This works.

10...Ba6

This does not. 

Black could have solved some of his problems with 10...c6 11.Bb2+ Nf6. Instead, he sets another trap for White.  

11.Bb2+ 

White did a wonderful job of avoiding difficulties, this time 11.Qxa8 Nf6 12.Qxh8+ Kxh8 13.d3 Bd4 and Black's better development and two Bishops would win the day.

11...Kh6 12.Qxa8 

Yes.

This is playble now, because of White's Bishop on b2, which was possible because of 10.b4 and 4.a3. Nice.

12...Nf6 13.Qxh8 Nxe4 

13...Qxe4+ was no longer a great danger. Now White drains some of the excitment (danger) from the position.

14.Qg7+ Qxg7 15.Bxg7+ Kxg7 16.bxc5 Nxc5 

White can wrap up the game 

17.f3 Bb7 18.Kf2 h5 19.Re1 Bd5 20.Re7+ Kh6 21.Nc3 Bc6 22.Rae1

I do not understand Black and White's next couple of moves, but it does not matter: for all practical purposes, the game is already over.

22...Nd3+ 23.Ke2 Nxe1 24.Kxe1 d5 25.g4 hxg4 26.fxg4 g5 

I suspect that the clock had become a factor at this point.

27.h3

See the above note. 

27...Bb7 28.Ne2 

Black resigned

A Rook is a Rook is a Rook.

Nice game.

Jerome Gambit: First Steps (Part 8)

 

Ah, the Jerome Gambit: 1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+ Kxf7 5.Nxe5+ Nxe5

The two main continuations now are 6.Qh5+ and 6.d4. Which move you chose is mostly a question of the kind of game you like.

At a depth of 41 ply, Stockfish 13 prefers 6.Qh5+ by less than 1/5 of a pawn. Not much of a difference.

The Database has 5,036 games with 6.Qh5+, with White scoring 58%. It has 3,099 games with 6.d4, with White scoring 56%. Again, not much of a difference.

Some direction: 

6.Qh5+ Ng6  and 6...Kf8 allow White to immediately recover one of his two sacrificed pieces. He will want to castle and play f2-f4 or d2-d4, or both, counting on his "Jerome pawns".

6.Qh5+ g6 also allows White to capture a piece, and can lead to two complex defenses, after 7.Qxe5 Qe7 or 7...d6. The better prepared player is more likely to win. Unprepared defenders are likely to drop at least a Rook, however.

6.Qh5+ Ke6 is the sign of a player who is familiar with the Jerome Gambit, and who has prepared a particular defense - or it is the sign of a player who is unfamiliar with the Jerome and who hopes against hope that he can hang on to all of the captured material and survive.

6.d4 Bxd4 Qxd4 is less forcing than the 6.Qh5+ lines, and this can actually be difficult for Black - having to figure things out on his own. In the meantime, White develops, decides which pawns to advance, and even which side to castle on.

Other ways for Black to deal with 6.d4 run into the same problem, that of the defender having to put a whole defense together without much help from White.

Which lines do I like or loathe? I prefer 6.Qh5+, because I am uneasy with the positions after 6.d4 Qh4!? - the one time I tried the line, perrypawnpusher - 4xel, Chess.com, 2017 (1/2-1/2, 29), I had to weasel my way out of a difficult situation. Bill Wall, on the other hand, plays 6.d4 regularly, and rarely sees 6...Qh4, anyhow. I am not afraid of 6.Qh5+ g6, but I always follow my rule: Do not take the Rook at h8. I am happy to see 6.Qh5+ Ke6, but after the possible move 7.f4, often the answer 7...d6 creates difficulties.

A balanced diet of playing and studying the Jerome Gambit should provide plenty of enjoyment in your chessic activitied.

Jerome Gambit: First Steps (Part 7)

 

You might think that by now it would be time to discuss the Jerome Gambit proper, but there is one more distinction to make.

After 1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+ Kxf7

Should White capture the e-pawn?

That was never a question for Alonzo Wheeler Jerome or his followers. Of course White should play 5.Nxe5. That is what the Jerome Gambit is all about.

The Database, however, has thousands of "not-5.Nxe5" games.

Some of this can be because White is not totally familar with the Jerome Gambit.

Some of this can be attributed to a sudden outbreak of caution - White is willing to risk 1 piece, but not 2.

Some of this can be credited to earlier computer chess engines that analyzed the position and urged White to stop being so generous. (Stockfish 13 has no such qualms.)

I could have differentiated the 5.Nxe5 / not-5.Nxe5 games as "full" or "decapitated", but I settled upon "classical" and "modern".

My preference and recommendation is the classical Jerome Gambit, 1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+ Kxf7 5.Nxe5+

By the way, please resist the siren song of 5.Ng5+?. It introduces the "trap" 5...Qxg5 6.d4, where White attacks both the Queen and the Bishop. The problem for White is that the reply 6...Qxg2 is simply crushing - at that point the computers asses Black as having over a 2 Rook advantage, and The Database shows White scoring less than 10%.