Jerome Gambit: Crazyhouse

Dear Readers,

The latest email from Cliff Hardy is a doozie.

Can you imagine playing the Jerome Gambit in a "crazyhouse" game of chess?

Mr. Hardy did.

The following game comes with a warning - "crazyhouse" chess, is kind of "crazy"!

Hi Rick

I'm sorry, so sorry, please accept my apology.

In saying that, I don't mean to offend Brenda Lee, but it's just that I should not be sending you this game under any circumstances (especially as, for one thing, it is heading on a tangent away from what your blog is about) and I apologise profusely for any anxiety or vexation caused by it.

But I couldn't help it and, besides, you would have figured out sooner or later (if you haven't already) that I belong in the crazy house.

Anyway, speaking of crazyhouse, this is a game I played in a Lichess crazyhouse tournament (crazyhouse is the chess variant where, once you capture a piece, you are able to put that piece back on the board on any empty square at any particular time, though placing it counts as a move).

Lichess actually annotates crazyhouse games using Stockfish.

Lol, I understand if you want to reply to me with something like, "Thanks for the game but please never, ever, ever, ever, ever send me a crazyhouse game again!!" 😉

Cliff Hardy (1884) - NN (1686), Lichess Crazyhouse Tournament, 3 + 2, 2018

1. e4 e5

2. Nf3 Nc6

3. Bc4 Bc5

This is a common crazyhouse opening variation.

4. Bxf7+ Kxf7                [1xP          1xB]

My note on the right there in square brackets means that since White has captured one pawn (i.e 1xP), he now has that pawn "in hand" and can place it on the board when he wishes. The "1xB" in bold is saying that, likewise, as Black has captured one bishop he can place that bishop on the board when he wishes.

5. Nxe5+                       [2xP          1xB]

Stockfish doesn't like the Jerome Gambit at all and actually likes it a lot less in crazyhouse than it does it normal chess! It rates the position as around -9.7, but what's a queen advantage between friends?

5.  ...   Nxe5                  [2xP         1xB; 1xN]

6. Qh5+ N@g6?           [2xP         1xB]

      

Stockfish much prefers 6...Kf8, similarly to how it does when White plays the Jerome Gambit in normal chess, but nevertheless still rates this crazyhouse position as -6.6, in Black's favour.

["N@g6" means that Black places the knight that he had "in hand" on the board for his move. Naturally, that means he now only has a bishop "in hand" left (i.e. 1xB)].

7. @f5       Qf6               [1xP         1xB]

[@f5 means to place one of white's pawns that he had "in hand" on the board at f5].

8. d4?                            [1xP         1xB]   

White sacrifices a pawn to speed up development but attacking the queen with @g5 was better, with Black still holding a -4.3 pawn advantage, though with my low standard of crazyhouse play, a piece here or there is not that significant. Cashing in too early with 8. fxg6+?? is even worse than the move I played in the game, since White would then needlessly take the pressure off Black and may allow Black a big counter-attack e.g. 8...hxg6 9. Qxh8?? Qxf2+ 10. Kd1 B@e2 mate.

8.  ....       Bxd4              [1xP         1xP; 1xB]

9. @g5    Qb6                [               1xP; 1xB]

10. fxg6+  hxg6!            [1xN         2xP; 1xB]

During the game, I thought this was a mistake as it left the rook on h8 unguarded, but Stockfish likes the rook sacrifice for Black as he now has a strong attack on the vulnerable white king.

   

11. Qxh8   Bxf2+        [1xN; 1xR     3xP; 1xB]

12. Kd1?

Black's advantage blows out from -11.1 to -25.9 after this move. I had been afraid of playing 12. Kf1 because I thought that black could then have gotten a big attack with 12...@e2+? 13. Kxe2 B@g4? but after blocking the check with 14. @f3 (I think I forgot that

White would have had a pawn available to put on f3 as it is the pawn White took with 13. Kxe2), White would then have had a +13.7 advantage, with his and knight and rook "in hand" and safe king .

12.  ...     Qd4+?        [1xN; 1xR     3xP; 1xB]

The position is fairly even after this move but 12...B@g4+ (-25.9) was much stronger since if White then blocks the check with 13. N@f3, Black can take the knight on f3 with check when he wishes and place it on the board with check, to keep the attack going (the initiative, by way of continuous checks, is often of supreme importance in crazyhouse, in order to obtain a forced checkmate).

13. Bd2     Qxe4??     [1xN; 1xR     4xP; 1xB]

13...@d3 is much better and is a typical attacking move in crazyhouse; placing a pawn menacingly on any empty square near the opponent's king.

  

14. N@d8+                    [1xR     4xP; 1xB]

Very strong but the simple fork of king and queen with 14. R@f4!+ (+14.2) was even better, but I had missed that it was a possibility.

14.   ...      Ke7  

15. Nc3

I didn't notice 15. Qxg7+, which is a bit stronger and more forcing.

15. ...       @e2+         [1xR     3xP; 1xB]

Black tries in vain to wrest the initiative from White but it is hard work as the white king is actually the much safer of the two kings.

16. Nxe2     @d3       [1xP; 1xR     2xP; 1xB]

17. Qxg7+   Kd6        [2xP; 1xR     2xP; 1xB]

17...B@f7 was an alternative, though after 18. @f6+, White would still have had a huge initiative.

  

18. Qf8+    Kd5??          [2xP; 1xR     2xP; 1xB]

18...B@e7! would have slowed White's initiative down and made things very unclear.

19. Nc3+

Stockfish finds a mate in 4 here with 19. R@b5+ c5 20. Nc3+ Kc4 21. b3+ Kd4 22. Qxc5 mate.

19. ...     Kd4

Black is forced to bring his king even further down the board into an extremely precarious position.

20. Qxf2+               [2xP; 1xB; 1xR     2xP; 1xB]

Stockfish finds another mate in 4 here with 20. Nb5+ Kd5 21. Nxc7+ Kd4 22. c3+ Kc4 23. b3 mate.

20. ...      @e3        [2xP; 1xB; 1xR     1xP; 1xB]

21. Bxe3+ Qxe3     [3xP; 1xB; 1xR     2xP; 1xB]

Or 21...Kc4 22. R@c5+ Kb4 23. a3 mate.

22. Qxe3+ Kxe3     [3xP; 1xB; 1xR; 1xQ     2xP; 1xB; 1xQ]

23. Nd5+

There were several mates in 2 here e.g. 23. R@e4+ Kf2 24. Q@f1 mate.

23.  ...         Kd4

24. B@e3+   Kxd5          [3xP; 1xR; 1xQ     2xP; 1xB; 1xQ]

24...Ke4 25. Q@d4+ Kf5 26. Qf4 mate would not have been much better.

25. Q@d4 mate

  

Sorry again!

Bye,

Me ☺

Jerome Gambit: Play Chess

It may seem hard to imagine, but the Jerome Gambit has its share of "quiet" positions, and they can put an extra burden on the defender, in that there are no "forced" lines to tumble into, shining light on where to go, even when tumbling down a mine shaft. Instead, Black has to figure out what to do on his own, as it were, and sometimes "playing chess" can be difficult.

Wall, Bill - Guest273475
PlayChess.com, 2017

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

4...Kxf7 5.Nxe5+ Nxe5 6.Qh5+ Kf8

The Jerome Variation of the Jerome Gambit, played by Alonzo Wheeler Jerome against David Jaeger in correspondence, 1880.

7.Qxe5 d6 8.Qg3 Nf6

Also 8...h5 appeared in Wall,B - Thieveyen, Chess.com, 2010 (1-0, 61), while 8...d5 was played in billwall - bfcace, Chess.com, 2012 (1-0, 25). 

9.Nc3 Be6

Instead, 9...Nh5 led to an exchange of Queens after 10.Qf3+ Qf6 11.d3 Qxf3 12.gxf3 in Wall,B - Ahmadi,S, Chess.com, 2010 (0-1, 59). A rare loss by Bill in the Jerome Gambit, so it is worth checking out.

10.O-O Kf7 11.d3 Rf8 12.Na4

It is not always smash! and crash! in the Jerome Gambit. Sometimes you just grab the "minor exchange" of Bishop for Knight.

On the other hand, there is always 12.Be3 Nh5 13.Qf3+ Kg6 14.Qe2 Nf4 as in Wall,B - Guest874250, PlayChess.com, 2014, (0-1, 32).

12...Bb6 13.Nxb6 axb6 14.a4 Kg8 

Black has castled-by-hand and still has his piece for a couple of pawns. What is White to do? Advance the "Jerome pawns", of course.

15.f4 Qd7 16.f5 Bf7 17.b3 Rfe8 18.Bb2 Qe7 19.Rae1 Kh8 

White has stubbed his toe in this variation before, so he prefers now to put his pawns and pieces on the right squares and see what happens. (Stockfish 8 agrees with him, rating the position, at 28 ply, as "0.00" - even.) Sometimes these "small" moves put pressure on Black to play chess, not just react to threats. 

20.c4 Ra5 21.Bc3 Raa8 22.Re3 Qf8

It is interesting that the computer considers this move an error, and recommends, instead, opening the Queenside with 22...b5 23.axb5 Ra3. It then sees a way, through pawn and Queen exchanges, to an edge for Black, despite leaving White with three pawns (two of them passed) for a piece: 24.Rb1 Ra2 25.Rbe1 Nh5 26.Qg4 Rc2 27.Bd4 Nf6 28.Qd1 Ra2 29.e5 dxe5 30.Bxe5 Qd7 31.Qf3 c6 32.bxc6 Qxc6 33.Qg3 Nh5 34.Qf3 Qxf3 35.Rxf3 Nf6 36.h3 b5 37.cxb5 Bxb3 38.Rfe3. This is a subtle position, one the better player will navigate more easily.

23.Qh4

This is the move the computer recommends, and I find it highly ironic. Not too long ago, humans used to adopt "anti-computer" strategies (with White) that included locking the position up with pawns (say, with the Stonewall opening) followed by a slow advance of Kingside pawns toward the enemy monarch. It is as if Stockfish has learned this lesson and suggests that White move his Queen and unblock his g-pawn.

It is interesting that Bill pursues a different line of play - until he is helped by what one famous computer (HAL) would call "human error". 

23...Nd7 24.Rh3 Qg8 25.Rg3 Ne5 26.d4 Nd7 27.e5 

Black was pretty sure that he had prevented this move, as the advance costs White a pawn. However, it also leads to the opening of the deadly a1-h8 diagonal for White's Bishop. (Looking at the current position, it is easy to overlook this fact, as the diagonal looks clogged with pawns, and it will soon inherit a Knight - but the piece will be unable to stay there.) Add that to the pressure along the g-file, and that spells disaster.

27...dxe5 28.dxe5 Nxe5 29.Qf4 

Aha! The Knight on e5 is attacked twice and defended once...

29...Ng6

Hoping to block the g-file and protect the vulnerable g7 square. Best was probably 29...Nc6, when 30.Bxg7+ Qxg7 31.Rxg7 Kxg7 would lead to a very interesting position where Black would have a Rook and two minor pieces of his Queen. Would that be enough?

Probably not. After 32.f6+! White could show that there are two more features to the position - White's advanced pawn, and the unsafe position of Black's King. Therefore 32...Kh8 (forced) 33.Qg3 Rg8 34.Qxc7 follows, and what is Black to do?

The line 34...Bg6 35.Qxb7 Ne5 36.Qxb6 reminds me of the arcane expression "nibbled to death by ducks"; Black will not be able to coordinate his pieces, protect his King, and deal with White's 4 passers.

Instead, Black could try 34...Nd8, although White could continue to apply pressure to the contorted positions with 35.Rd1, e.g. 35...Rf8 36.Qxb6 Kg8 37.Rd7 Bg6 38.Qc7 Nf7 39.Qxb7 (or 39.h4). Again, White's passed pawns would be decisive.

However, the text move, returning a piece, falls tactically.

30.fxg6 Bxg6 31.Rxg6 

You saw this move, right?

31...hxg6 32.Rf3 Black resigned

Black can defend against the threat down the h-file with 32...Ra5, but that simply costs a Rook after 33.Bxa5, as he does not have time for 33...bxa5 because of 34.Rh3+, losing his Queen. Instead, he can take advantage of the diverted Bishop by giving his King some breathing space with 33...Qe6, but after 34.Bc3 (of course) he would be a piece down and vulnerable after 34...Qe7 35.Qf7 Qxf7 36.Rxf7 or 34...Kg8 35.Qxc7.

Jerome Gambit: Choices

I suspect that one of the reasons that Bill Wall opts for a main line in the Jerome Gambit - 6.d4 - that gives Black a lot of choices is specifically because it gives Black a lot of choices. Choice can mean "taking up clock time to calculate". It also can mean "making the wrong choice". Why force Black to make the right choice?

Wall, Bill - Guest_5402
chesstempo.com, 2018

1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+ 

4...Kxf7 5.Nxe5+ Nxe5 6.d4 Bxd4 7.Qxd4 d6 

8.Qd5+ 

Black has too many choices. To block the check, or not? To move the King, or not? To protect the b-pawn or not? 

"Objectively", Black is fine after any choice, but it hasn't worked out that way in "real life". The Database has 13 games with this position. White scores 69%.

8...Ke8 

The Database has 5 games with this position. White scores 80%.

The alternative was to let go of the b-pawn: 8...Be6 9.Qxb7 Nf6. 

9.Nc3 Nf6 10.Qb3 b6 

11.O-O Bb7 12.f4 Nc6

13.e5 dxe5 14.fxe5 Qd4+ 15.Kh1 Nxe5 

Black's idea, but not a good one.

16.Qe6+ Kd8 

Not 16...Kf8 17.Rxf6+ gxf6 18.Bh6 checkmate 

17.Rd1 Qxd1+ 18.Nxd1 Re8 19.Qh3 

Black has a piece and a Rook for his Queen.

19...Neg4 20.Qg3 Nh5  

This slip ends the game.

21.Bg5+ Kc8 22.Qxg4+ Kb8 23.Nc3 Black resigned

White's material advantage is too great.

Jerome Gambit: Quick Lesson

The latest Jerome Gambit from Cliff Hardy is a 1 0 lightning wonder. Pay attention to his notes, below - look how much information can be packed in a game that averaged less than 2 seconds a move!

Amazing.

Cliff is pretty hard on himself and his opponent, given the time control, but it only shows how challenging the Jerome Gambit can be.

Hardy, Cliff - NN
1 0 bullet, Italian Game, lichess.org
(notes by Cliff)

1.e4 e5 2.Nf3 Nc6 3.Bc4

Forced starting position.

3...Bc5 4.Bxf7+ 

4...Kxf7 5.Nxe5+ Nxe5 6.Qh5+ Ng6?! 

6...Kf8 would have been better. 

7.Qd5+ Kf8 8.Qxc5+ d6 9.Qe3 Nf6 

10.O-O Ke7? =/+ 

10...Kf7 would have been a better way to artificially castle. 

11.f4 Rf8?  +/=

11...Re8 =/+ would have been a more consistent move to enable artificial castling after a later ...Kf8. 

12.d4?! =/+ 

Seizing space by pushing the kingside Jerome pawns immediately with 12.f5 Ne5 13.d4 +/= would have been better. 

12...Kf7 13.Nc3? -+

Again, seizing space by 13.f5 = would have been better. 

13...Kg8 14.f5 Ne7 15.Qg3? -+

White thereby curtailed his own ability to push the Jerome pawns since if White were to continue after this move to push the e-pawn to e5, Black may respond with the tricky intermezzo ...Nxf5!, attacking the queen on g3; 15.Qf3 -+ would have been better.

15...Kh8? =/+

Best would have been 15...d5! 16.e5 Nxf5 17.Rxf5 Bxf5 since then 18.exf6 Qxf6 -+ would have left Black a comfortable exchange ahead. 

16.Bg5 c6 17.e5?

Over zealous. 17.Rae1 =/+ would have been better.

17...dxe5 18.dxe5 Nfd5??  

18...Nxf5 19.Rxf5 Bxf5 20.exf6 Qd4+! -+ or 20...gxf6 -+ would have been much better. 

19.f6?? = 

As Black's knight on d5 defends the square f6, the best continuation here would have been to eliminate it with 19.Nxd5 cxd5 20.f6 +-
when Black couldn't have adequately saved his knight on e7 with ...Ng6 due to 21.fxg7+ Kxg7 22.Bxd8.

19...gxf6 20.Bxf6+? -+

After 20.exf6 = or 20.Nxd5 fxg5 21.Rxf8+ Qxf8 22.Nc7 $5 Rb8 23.e6 = White would have had compensation for the material deficit, due to his cramping advanced Jerome pawn. 

20...Nxf6 21.exf6 Ng6 22.Rad1 Qb6+ 23.Kh1 Be6 24.Rd6? 

24.b3 or 24.h4 would have been better tries. 

24...Bf7?! -+

24...Bc4 -+ would have been more attacking. 

25.Ne4 Rad8? =/+

The greedy 25...Qxb2 26.Ng5 Qe5! -+ would have been much better. 

26.Ng5?? -+ 

26.Qh3! Rxd6 27.Nxd6 would have put Black under a bit of pressure, with the threat of Qh3-h6-g7 mate, though ...Qc7! would then still have left Black with a slight advantage. 

26...Rxd6 27.Qxd6 Qxb2?? = 

27...Kg8! or 27...Qe3! in either case with a winning position for Black, were better.

28.h4?? -+ 

Stockfish found the interesting 28.Qxf8+! Nxf8 29.Nxf7+ Kg8 30.Nh6+ with a draw by perpetual check, but I decided to play for the win (mind you, I didn't find this line myself during the game anyway so couldn't have played even it if I'd wanted to?) 

28...Kg8! 29.h5 Nh4?? +- 

Black had 2 good moves here, the smart fork 29...Qb5! or the safe 29...Qe5! - in either case he'd have then had a winning position. 

30.Nxf7 Rxf7 31.Qd8+ Rf8 32.f7+ Kg7 33.Qg5+?? =/+

Surprisingly a blunder, as Black could now have restored material parity and gotten his king safe; 33.h6+ Kg6 34.Qxf8 or 33.Qxh4 both would have left white with an easily won position. 

33...Kh8?? 

The final blunder; instead, after 33...Ng6 34.hxg6 hxg6 =/+ White would have been fortunate enough if he went on to hang in for a draw.

34.Qxh4 Qg7 35.h6 Qg6? 

Though 35...Qe5 36.Re1 Qd6 37.Qg5 Qd4 38.Qe7 +- still would have won for White, or 35...Qc3 36.Qe7 +- 

36.Qd4+ Qg7 37.Qxg7 checkmate

Jerome Gambit: Bad Moves

Some commentators have suggested that Emanuel Lasker would purposely play "bad" moves in order to upset his opponents, making it easier for him to win a game. As someone who plays the Jerome Gambit, I guess I understand that argument.

It is also possible that Lasker won because he understood, better than his opponents, the positions that he played, both "bad" and "good".

The following game has several times when I can not understand what was going on - only that Bill understood the positions better than his opponent, and that was enough to fashion a win. 

Wall, Bill - Guest4919498
PlayChess.com, 2018

1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+ 

4...Kxf7 5.Nxe5+ Nxe5 6.d4 Bxd4 7.Qxd4 d6 8.Qd2 

Okay, I don't get it. I would suggest "mouse slip", but Bill has played the move before. Given that there's about a 500 rating point difference between him and me, I have to give him the benefit of the doubt and continue to try to figure it out.

8...Nf6  

Bill has also faced 8...Be6 9.O-O Nf6, in Wall,B - Guest477, ICC, 2017, (1-0, 25). 

9.Nc3 c6 10.f4 Ng6 11.O-O 

Bill has suggested that here, and the next few moves, he could have played Qd3. This re-positioning suggests that 8.Qd2 might have been a search for novelty, as he had seen the position after 7...d6 in over 70 games.

11...Re8 12.Re1 b5

When you have the advantage, you are obliged to attack, said Wilhelm Steinitz.

So, once again, the Jerome Gambiteer tests the attacking prowess of the defender.

13.a3 Qb6+ 14.Kh1 Ng4 15.h3 Nf2+ 

16.Kh2 d5 17.exd5 

17...Rxe1

Bill points out that 17...Bf5 was stronger.

18.Qxe1 Bd7 

Stockfish 8 gives a mind-numbing, attacking, alternative: 18...Bxh3!? 19.Kg3!? Nd3 (the piece was lost, any way) 20.cxd3 Bf5 21.Qe3 Qd8 22.Kh2!? cxd5 23.Nxb5 Bd7 24.f5 Kg8!? 25.Kg1!? Nh4 26.Nd4 Qf6 27.Bd2 Re8 28.Qf2 Nxf5 29.Bc3 Qg6 30.Nf3 Bb5 31.a4 Ba6 32.Ne5 Qb6 33.d4 Qf6 34.b4 with an even position. Um, sure.

19.Be3 Ng4+ 

Black has to give up his advanced Knight, but at least gets to dent White's pawn structure.

20.hxg4 Qc7 21.Kg3 

The King has to help out on defense.

21...Qd8 

Eyeing h4.

22.Qh1 Qe7 23.Re1 Qf6 

Now Black's game crashes. He had better chances of survival after 23...Kg8 24.dxc6 Bxc6 25.f5 Qe5+ 26.Kh3 Nf4+ 27.Bxf4 Qxf4 but White would still have the advantage.

24.Ne4 Qd8 25.Ng5+ Kf6 26.Bd4+ Black resigned

Black faces checkmate, the quickest of which is 26...Ne5 27.Rxe5 Bxg4 28.Re6+ Kf5 29.Qxh7+ g6 30.Qxg6 checkmate. Ouch!