Jerome Gambit: The Science of the Draw?! (Part 1)

I just completed my second Jerome Gambit game in the ongoing "Italian Battleground" tournament at Chess.com. It put me very much in mind of the questions raised in the recent post "Jerome Gambit: The Eternal Question of Draws".

It felt, at first, like I was dropping 1/2 a point, but by the end of the game it appeared like I had won 1/2 a point. In any event, it is appropriate to speak well of my opponent, Krisstianes_017, whose scientific approach to defending against the opening was successful - if you call achieving a draw against a "refuted" opening a success. (Another "eternal question".)

(Although my "escape" into a drawn endgame a pawn down was very much "scientific" as well, as it involved knowledge and experience with Bishops-of-opposite-colors.)  

perrypawnpusher - Krisstianes_017

"Italian Battleground", Chess.com, 2018

1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+ 

4...Kxf7 5.Nxe5+ Nxe5 6.Qh5+ Ng6 

The same line chosen by Abhishek29 in my other Jerome Gambit (so far) in the same tournament. For a short while I was worried that I would be playing the same game, twice.

7.Qd5+

Again, the "nudge".

7...Ke8 8.Qxc5 d6 9.Qe3 Nf6 10.O-O Be6

Black is developing his minor pieces, but this move is probably not best, as it summons the move f2-f4 (threatening a pawn fork with f4-f5) from White.

After the game, in preparing this blog post, I asked Stockfish 9 to evaluate the move 10...Be6. At a depth of 30 ply, it saw the resulting position as equal. That is worth thinking about. Should Black be happy? Should White be happy?

It is funny to note, however, that after White plays 11.f4, Stockfish 9 - again, at 30 ply analysis depth - recommends returning the Bishop with 11...Bc8!? and evaluates the result as a slight edge to Black!

I have just recently begun using version 9 of Stockfish to help me understand my completed games, and this is not the only quirky output it has provide me.  

In any event, the computer prefers 10...Nh5!? (preventing f2-f4?) instead of the text.

11.f4 Ne7 

Interesting and good.

I faced 11...Bc4 in perrypawnpusher - avgur, blitz, FICS, 2009 (1-0, 43) as well as 11...Ng4 in perrypawnpusher - Verlen, blitz, FICS, 2010 (1-0, 23) and 11...Kd7 in perrypawnpusher - dirceu, blitz, FICS, 2010 (1-0, 16).

12.d4

Stockfish 9's recommendation is 12.f5 Bc8 13.c4 Nc6 14.d3, apparently seeing ...d7-d5 as worth preventing.

12...Kd7 

Courageously planning to castle-by-hand on the queenside. I can see why my opponent would not want to strike at the center until his King was safe.

Still, the computer's suggestion is 12...d5, and I faced that move in perrypawnpusher - nmuffjgp, blitz, FICS, 2010 (1-0, 22).

13.d5 Bf7 14.c4 Kc8 

15.e5 

My plan was to open the center and use my "Jerome pawns" to cause havoc amongst Black's minor pieces.

Amazingly, after the game Stockfish 9 recommended the line 15.Bd2 Kb8 16.Nc3 h5 17.Rae1 a5. I was reminded of my ancient Fidelity Chess Challenger 7, that, when it assessed its King to be safe (wherever it was) and the position balanced, couldn't "think" of what to do, and so would advance its Rook pawns...  

15...dxe5 

This seems natural to me, but the computer doesn't like it, and prefers 15...Ng4 with an even game.

16.fxe5 Nfxd5 

Again, in the post mortem Stockfish 9 had a fit with this idea. I thought it was rather "scientific" to give back the sacrificed material and leave me with an isolated pawn.

17.cxd5 

Sigh. After the game, the compter had a fit with this move, too. I am glad I wasn't working with the old "talking" Fritz computer program that had a whole CD full of insults and sarcasm!

[to be continued] 

One Step Behind

My opponent's play was bold enough to give me trouble, and even if he did not solve all of the mysteries of the Jerome Gambit, he played a game which left me feeling constantly as if I were one step behind. 

 perrypawnpusher - magza
blitz, FICS, 2011

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

4...Kxf7 5.Nxe5+ Nxe5 6.Qh5+ Ng6

7.Qd5+ Ke8 8.Qxc5 d6 9.Qe3 Be6

This move provokes f2-f4 by White, but it is only the beginning of Black's sharp play.

10.0-0

Or 10.f4 as in mrjoker - Melbourne, blitz, ICC, 2008 (1-0, 38), mrjoker - tomnoah, ICC, 2009 (0-1, 42), perrypawnpusher - GabrielChime, blitz, FICS, 2010 (1-0, 33); perrypawnpusher - udofink, blitz, FICS, 2010 (1-0, 29), and perrypawnpusher - Kingsmeal, blitz, FICS, 2010 (0-1, 25).

10...Kd7

My opponent practically screams "Come and get me!"

Alternatives include 10...Qf6, as in perrypawnpusher - OverwiseMan, blitz, FICS, 2010 (1-0, 24); 10...Nf6, as in perrypawnpusher - dirceu, blitz, FICS, 2010 (1-0, 16) and perrypawnpusher - nmuffjgp, blitz, FICS, 2010 (1-0, 22); and 10...h6 as in MrJoker - pds1, ICC, 2011 (1-0, 28).

(All of the games referred to can be found in The Database.)


11.f4 N6e7 12.f5 Bf7 13.d4 Bc4 14.Rd1 Nf6

Black's pieces keep close watch on White's "Jerome pawns."

"Just a couple more moves," I told myself, "and I'll have an even game."

15.b3

Either 15.e5 or 15.Na3 might have been a tiny bit better.

15...Ba6 16.Nc3

A curious move.

I know that I did not play 16.e5 because I was worried about 16...Nxf5, but that was just poor "analysis", as 17.Qf3 wins back the sacrificed piece with advantage to White. Black does better to answer 16.e5 with 16...Ng4, and after 17.Qg5 he holds his own (or better) in the tactics on the Kingside.

Consistent was 16.c4 followed by 17.Nc3.

16...Re8

This move was my "opportunity", the one that usually comes knocking in the Jerome Gambit. Black's best was 16...Qf8 working against the line given concerning 15.e5. In fact, White's next move should be 17.e5.

17.Qg5 Rg8 18.e5 Nfd5

19.Bb2

Developing the Bishop, protecting the Knight, uniting the Rooks: and it all still feels too slow.

After the game, Houdini suggested 19.Nxd5 Nxd5 20.Qg4 (coming soon: c2-c4) with an edge for White.

19...h6 20.Qd2 Nxc3 21.Bxc3 Nxf5

Bleh. A score-and-one moves, and White has little to say about his position, while Black's pieces are beginning to glow again.

22.d5 Qe7

Possible, and better, was 22...dxe5

23.e6+ Kc8 24.Bb2 b6 25.c4 Kb7

Black has castled-by-hand on the Queenside, and his light-squared Bishop looks a little bit odd. White has his dark-squared Bishop on a strong diagonal, and a protected, passed pawn at e6.

Houdini gives Black only a slight edge, but the question is: Where will White get his play? Black's pieces are well-placed to defend against the one plan that suggests itself, infiltrating along the f-file to f7: 26.Rf1 Raf8 27.Rf2 g6 28.Raf1 h5.

26.b4

Opting to stir things up on the Queenside.

26...Bxc4 27.Rdc1

Making the same kind of mistake referred to in the notes to White's 16th move, concerning 16.e5. Here the right move was 27.Qf4, although after 27...Bxd5 28.Rxd5 g6 Black has returned his extra piece for some pawns, and is a pawn ahead.

White's e-pawn would be weak, not strong, and that would give Black the advantage in the long run.

White's game now just flows from bad to worse.

27...b5 28.a4 a6 29.a5 g6 30.Rxc4 bxc4 31.Rc1 Qg5 32.Qxg5 hxg5 33.Rxc4 Ne3

White's sacrifice of the exchange only means that he is now a Rook behind.

34.Re4 Nxd5 35.Rg4 Rae8 36.Rxg5 Nxb4 37.Rg4 Nd3 38.Bc3 Rxe6 39.Rd4 Nc5 40.Rb4+ Kc6 White resigned

Crushed

After I was crushed in this game, I looked it over: both my opening and my opponent had given me chances for play. The Jerome Gambit did not let me down – I let it down.

 perrypawnpusher - Kingsmeal
blitz, FICS, 2010

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

4...Kxf7 5.Nxe5+ Nxe5 6.Qh5+ Ng6

7.Qd5+ Ke8 8.Qxc5 d6 9.Qe3 Be6

Black's last move encourages White's next.

10.f4

Or 10.0-0 Nf6 11.f4 Kd7 12.f5 Bxf5 13.exf5 Re8 14.Qg5 Ne7 15.Qxg7 c6 16.Qxf6 Black resigned, perrypawnpusher - dirceu, blitz, FICS, 2010.

10...Bf7

The problem with White's 10th move is that Black can play, instead, 10...Qh4+ 11.g3 Qh3, causing problems on the light squares for White. Better, as in my play against dirceau, was 10.0-0.

11.0-0 Qh4

12.Nc3

Here my play could have been guided by a little logic. Does Black's Queen pose a threat? No. What would make it dangerous? If a Knight got to g4. Can I stop ...Ng8-f6-g4? No, but you can take the sting out of it with 12.d4 Nf6 13.Qd3 Ng4 14.h3. What else? Don't chase the other Black Knight to g4. How to I do that? Start with 12.d4, so that after 13.f5 the Knight can't find its way to g4.

12...N8e7

13.f5 Ne5 14.d4

Driving the Knight to where I don't want it to go. Thus: 14.h3 is better. My play was too stereotyped.

14...Ng4 15.Qg3

Completing the farce, but Rybka's recommendation after the game – 15.Qh3 Qxh3 16.gxh3 Nf6 17.e5 – leaves Black better, as well.

15...Qxg3 16.hxg3 Kd7

The Knight on g4 is a serious pain in my position.

17.Bf4 Bc4 18.Rfe1 Raf8 19.b3 Ba6 20.Ne2

It is difficult getting my pieces to cooperate. Black, on the other hand, is ready to tear things apart.

20...g6 21.fxg6 Nxg6

Even stronger was 21...hxg6, adding the other Rook to the attack.

22.c4 b5 23.c5 b4 24.cxd6 cxd6

25.Bd2

It is hard to believe Rybka's post mortem suggestion, although perhaps White's best is to eliminate almost all of the pawns: 25.a3 Bxe2 26.Rxe2 Nxf4 27.gxf4 Rxf4 28.axb4 Ra8 29.Ra6 d5 30.exd5 Rxd4 31.g3 Rd3 32.Kg2 Ne3+ 33.Kh3 Nxd5 34.Rf2 Ke7 35.Rh6 Nxb4 36.Rxh7+ Kd6 37.Rff7 Rxb3 38.Rxa7

analysis diagram







 25...Rf2 White resigned

Pawn-ful Lessons

A pawn advance seems to sew up the game for White, but it turns out upon further inspection that not as many were needed, only the proper amount of support from the pieces...

perrypawnpusher  - nmuffjgp
blitz FICS, 2010

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

4...Kxf7 5.Nxe5+ Nxe5 6.Qh5+ Ng6

This is one of the recognized refutations of the Jerome Gambit, but one reason I am pleased to face it is that White gets to eliminate Black's dark-squared Bishop which would otherwise hold up f2-f4.

7.Qd5+ Ke8 8.Qxc5 d6 9.Qe3 Be6

10.0-0 Nf6 11.f4

An interesting alternative: 11.d3 Qd7 12.b3 Kf7 13.Bb2 Rhf8 14.Nd2 Kg8 15.f4 Ng4 16.Qg3 Bf7 17.h3 Nf6 18.f5 Ne5 19.Bxe5 dxe5 20.Qxe5 Rfe8 21.Qg3 Rad8 22.Nf3 Qc6 23.Rf2 Qc5 24.Raf1 a5 25.a4 b5 26.axb5 Qxb5 27.Kh1 c5 28.Ne5 a4 29.bxa4 Qxa4 30.Nxf7 Kxf7 31.Qc7+ Rd7 32.Qxc5 Kg8 33.e5 Nd5 34.f6 gxf6 35.exf6 Qa8 36.f7+ Kh8 37.f8Q+ Black resigned, fehim - BoardStupid, FICS, 2009.

11...Ne7

Alternatives: 11...Bc4 12.d3 Bb5 13.c4 Bc6 14.f5 Ne5 15.Nc3 Qd7 16.b3 Qf7 17.d4 Neg4 18.Qe2 Qe7 19.Bg5 h6 20.Bxf6 Nxf6 21.Rae1 h5 22.e5 dxe5 23.dxe5 Ng4 24.h3 Nh6 25.Qf2 Kd7 26.Qd4+ Ke8 27.f6 Qe6 28.fxg7 Rg8 29.Rf6 Qd7 30.Qxd7+ Kxd7 31.e6+ Kd6 32.e7+ Kc5 33.Rxh6 Rxg7 34.Rxh5+ Kd4 35.Nb5+ Kd3 36.g4 Rag8 37.Rhe5 Rh7 38.e8Q Bxe8 39.Rxe8 Rgg7 40.Rd8+ Kc2 41.Re2+ Kc1 42.Na3 Rxh3 43.Rc2 checkmate, perrypawnpusher - avgur, blitz, FICS 2009;

11...Bf7 12.d4 Kf8 13.b3 h5 14.f5 Ng4 15.Qg3 Ne7 16.Bg5 Nf6 17.Nd2 Qd7 18.Rae1 Re8 19.c4 c5 20.d5 Ng4 21.Nf3 b5 22.h3 Nf6 23.Nh4 bxc4 24.bxc4 Qa4 25.Bxf6 gxf6 26.Ng6+ Bxg6 27.fxg6 Kg7 28.Qxd6 Rhf8 29.Qxc5 Qxa2 30.d6 Ng8 31.Ra1 Qe2 32.Rxa7+ Kxg6 33.Qf5+ Kh6 34.Rh7 checkmate, mrjoker - PhlebasP, ICC, 2009; and

11...Kd7 12.f5 Bxf5 13.exf5 Re8 14.Qg5 Ne7 15.Qxg7 c6 16.Qxf6 Black resigned, perrypawnpusher - dirceu, blitz, FICS, 2010.

12.d4

A dynamic alternative was 12.f5, as in 12...Bf7 13.d3 c6 14.h3 Qb6 15.g4 Qxe3+ 16.Bxe3 Kd7 17.Nd2 b6 18.Kg2 g6 19.g5 Nh5 20.f6 Nc8 21.d4 b5 22.b3 Nb6 23.Rac1 a5 24.c4 bxc4 25.Nxc4 Rhb8 26.Kf3 Nxc4 27.bxc4 Rb2 28.Rf2 Rab8 29.d5 c5 30.Rcc2 Rxc2 31.Rxc2 a4 32.Rc3 Kc7 33.Ra3 Be8 34.e5 dxe5 35.Bxc5 Rb2 36.Ke4 Re2+ 37.Be3 Bd7 38.f7 Black resigned, mrjoker - Melbourne, ICC, 2008.

12...d5 13.e5

Falling in with Black's plan. After the game Rybka suggested that White give up a Rook and a pawn for a couple of pieces: 13.f5 Bd7 14.e5 Nxf5 15.Rxf5 Ng4 16.Qf4 Bxf5 17.Qxf5 when Black has a small edge. 

13...Ne4

14.Nd2 Nxd2 15.Bxd2 Nc6

This move lets White's pawns run free. Instead, 15...Qd7 covers the f5 square.

16.f5 Bf7 17.e6 Bh5

18.h3

Preparing for the advance of more pawns. After the game Rybka showed that this was unnecessary. Instead, 18.Qh3 was strong, and after 18...Be2 White could afford 19.Rae1!? as 19...Bxf1 would be well met by 20.Qh5+.

Indeed, no further pawns need advance, if White's pieces are active.

18...Kf8

Better 18...Qf6 to answer 19.g4 with 19...Qxd4 although the Bishop is still lost after 20.gxh5.

19.g4

Here 19.f6 gxf6 20.Qh6+ Kg8 21.Qxh5 is stronger. 

19...Be8 20.g5

Again, this works, but stronger was 20.f6 gxf6 21.Qh6+

20...Ne7 21.f6

Here, 21.Bb4 would have tightened the screws before the pawn advance.

Black's resistance crumbles.

21...Ng6 22.e7+ Qxe7 Black resigned

More Than Meets the Eye

One of the joys of chess, and this has to be true even for that reprobate of openings, the Jerome Gambit (1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+) and its relatives, is that often things on the board are not as they first appear to be: there is more than meets the eye. 

perrypawnpusher - udofink
blitz FICS, 2010

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

4...Kxf7 5.Nxe5+ Nxe5 6.Qh5+ Ng6

This is a solid response for Black – computers will rate the second player about 1 1/4 pawns better.

Familiarity has its benefits, however: my score with White against this position in 36 games is 93%.

7.Qd5+ Ke8 8.Qxc5 d6 9.Qe3

9...Be6

This move is not as bad as I thought it to be, and my response is not as good as I thought, either.

10.f4

Black's response should cure me of this premature move.

I was more on target in an earlier game with 10.0-0 Nf6 11.f4 Kd7 12.f5 Bxf5 13.exf5 Re8 14.Qg5 Ne7 15.Qxg7 c6 16.Qxf6 Black resigned, perrypawnpusher - dirceu, blitz FICS, 2010

10...Qh4+

Thematic, and an improvement over 10...N6e7 11.0-0 Nc6 (11...Nf6 12.f5 Bf7 13.d3 c6 14.h3 Qb6 15.g4 Qxe3+ 16.Bxe3 Kd7 17.Nd2 b6 18.Kg2 g6 19.g5 Nh5 20.f6 Nc8 21.d4 b5 22.b3 Nb6 23.Rac1 a5 24.c4 bxc4 25.Nxc4 Rhb8 26.Kf3 Nxc4 27.bxc4 Rb2 28.Rf2 Rab8 29.d5 c5 30.Rcc2 Rxc2 31.Rxc2 a4 32.Rc3 Kc7 33.Ra3 Be8 34.e5 dxe5 35.Bxc5 Rb2 36.Ke4 Re2+ 37.Be3 Bd7 38.f7 1-0, mrjoker - Melbourne, ICC 2 12 blitz, 2008,) 12.d4 Bf7 13.Nc3 Qd7 14.e5 dxe5 15.dxe5 Qe6 16.f5 Qxe5 17.Qxe5+ Nxe5 18.Re1 Rd8 19.Bf4 Kd7 20.Bxe5 Nf6 21.Bxf6 gxf6 22.Ne4 Kc8 23.Nxf6 Rd6 24.Ne4 Rdd8 25.Rad1 Rxd1 26.Rxd1 Re8 27.Nc3 Bc4 28.Kf2 Rf8 29.g4 c6 30.b3 Bf7 31.Ne4 Bd5 32.Nc5 Rf6 33.c4 Black forfeited on time, perrypawnpusher - GabrielChime, blitz FICS, 2010

11.g3 Qg4

Here Black's followup is inaccurate, however. Certainly the Queen can be more annoying at h3.

After 11...Qh3, White might as well continue developing with 12.Nc3, as the pawn fork 12.f5, instead, would win a piece after 12...Bxf5 13.exf5+  but allow after 13...Kf8 the deadly move ...Rae8.  

12.f5

Now the pawn fork works, however.

12...Bf7

A bit better, but still leading to an edge for White, was 12...Nf6, which is best answered by 13.d3.

The problem now with 12...Bxf5 for Black is that after 13.exf5+ Kd7 (threatening ...Rae8, as in the not to White's 12th move) White can safely castle out of the threatened pin of his Queen. 

With a Black Queen on h3, that would not have been possible.

13.fxg6 Bxg6 14.d3 Kd7 15.0-0 Nf6 16.Nc3 c6

White is a pawn up.

17.Qf4

An exchange of Queens would make my King safer: Black has 4 pieces on the Kingside (with another Rook waiting in the wings) to my 2 pieces.

17...Qe6 18.Bd2 Rhf8 19.Qg5 Rf7 20.Rae1 Raf8

21.e5 dxe5 22.Qxe5

22...Qh3

Black declines to exchange Queens (his best choice), seeing his chances for an attack on the enemy King (starting with pressure on f1) to be good.

23.Bf4

It is Black's King, on the Queenside, however, who is at greater risk. 

23...Ne8

Not so good, but Rybka's suggestion after the game – to exchange Queens and give up a piece – is not attractive, either: 23...Qxh2+ 24.Kxh2 Ng4+ 25.Kg2 Nxe5 26.Rxe5.





analysis diagram





24.Qd4+ Kc8 25.Qxa7

26...Nd6 26.Qa8+

Noticing that capturing Black's Knight now leads to me being checkmated, I was a bit too nervous to work out the cool Knight sacrifice at b5, but after the game Rybka did: 26.Nb5 cxb5 (26...Nxb5 27.Qb8+ Kd7 28.Qxb7+ Kd8 29.Bg5+ Rf6 30.Qxg7) 27.Qc5+ Kd8 28.Qxd6+.

26...Kd7 27.Qa3

27...Nc8

White's Bishop is making things too hot, and Black needed instead to sacrifice the exchange with 27...Rxf4 in order to stay in the game.


28.Qc5 b6

Swatting at my Queen as if she were a bothersome mosquito, Black hastens his end.

29.Qd4+ Black resigned

Checkmate is unavoidable.