Friday, February 10, 2017

Jerome Gambit: Stockfish 8 Checks In

In the previous post I spent a little time with The Database and came up with some scoring percentages for White for different lines of play in the Jerome Gambit.

I gave two caveats:


All statistics need to be approached cautiously. In this case it is useful to remember that the vast majority of games in The Database are drawn from over a decade of games played online at FICS, regardless of outcome; so they should be somewhat representative of Jerome Gambit play among club players who contest online... 
Of course, were I to subject each of these positions to the scrutiny of, say, Stockfish 8, the computer's evaluation would likely, in some cases, vary significant from the game outcomes in The Database!


Then I decided to give Stockfish 8 a chance, after all.

1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5

If the game continues 3...Nd4 4.Bxf7+ instead, Stockfish 8 rates the position at 30 ply depth of search as -.18, about an even game.

4.Bxf7+

Stockfish 8 rates the position as -1.85 at 30 ply depth of search, meaning that Black is not quite 2 pawns better.

If Black plays 4...h6 instead, after 5.0-0 Bc5 6.Bxf7+ Stockfish 8 rates the position as -1.77 at 30 ply search depth. After 5.Nc3 Bc5 6.Bxf7+, it rates the position at 30 ply search depth as -1.45.

4...Kxf7

As for declining the Bishop, Stockfish 8 rates 4...Kf8 at 30 ply depth of search as +1.68, giving White more than a pawn and a half advantage. It rates 4...Ke7 at +2.01, giving White a two pawn advantage.

5.Nxe5+

Stockfish 8 rates this position as -2.41 at 30 ply depth of search. 

If, instead, the game continues 5.Nc3 Nf6, the game reaches the Four Knights variation by transposition. Stockfish 8 rates the position as -2.85 at 30 ply depth of search.

5...Nxe5

Stockfish 8 rates this position as -2.32 at 30 ply depth of search.

If Black replies, instead, 5...Kf8, Stockfish 8 rates the position -.56; if Black replies 5...Ke8 Stockfish 8 rates the position +.22; if Black replies 5...Ke6, Stockfish 8 rates the position +.94; if Black replies 5...Ke7, Stockfish 8 rates the position +1.04; all at 30 ply depth of search.

6.Qh5+

Stockfish 8 rates this position as -2.27 at 30 ply of depth.

If White plays 6.d4 instead, Stockfish 8 rates the position as -2.46 at 30 ply depth.

6...g6

Stockfish 8 rates this position as -1.55 at 30 ply of depth.

It rates 6...Kf8 as -1.75 ; 6...Ke6 as -2.42; and 6...Ke7 as +2.71; all at 30 ply of depth.


Wednesday, February 8, 2017

Jerome Gambit: What Are the Odds?

I spent a little time with The Database today, and came up with the following numbers.

All statistics need to be approached cautiously. In this case it is useful to remember that the vast majority of games in The Database are drawn from over a decade of games played online at FICS, regardless of outcome; so they should be somewhat representative of Jerome Gambit play among club players who contest online.



1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5

The Database has 5,307 games where the game continues 3...Nd4 4.Bxf7+ instead. White scores 56%.

4.Bxf7+

The Database has 12,662 games reaching this position. White scores 45%.

The Database has 4,589 games where Black plays 4...h6 instead. White scores 55%. This includes 187 games which continue 5.0-0 Bc5 6.Bxf7+, where White scores 65%; and 141 games which continue 5.Nc3 Bc5 6.Bxf7+, where White scores 54%.

4...Kxf7

The Database has 12,763 games reaching this position. I am not sure why this number is larger than the earlier one! White scores 45%

The Database has 272 games where Black declines the Bishop with 4...Kf8. White scores 55%. The Database has 41 games where Black declines the game with 4...Ke7. White scores 72%.

5.Nxe5+

The Database has 5,613 games reaching this position. White scores 54%.

The Database has 2,510 games where the game continues 5.Nc3 Nf6, or the game reaches the Four Knights variation by transposition. White scores 38%

5...Nxe5

The Database has 5,348 games reaching this position. White scores 54%.

The Database has 202 games where Black replies, instead, 5...Kf8. White scores 52%. The Database has 103 games where Black replies 5...Ke8; White scores 59%. The Database has 24 games where Black plays 5...Ke6; White scores 65%. The Database has 12 games where Black replies 5...Ke7; White scores 67%.

6.Qh5+

The Database has 3,804 games with this position. White scores 55%.

The Database has 1,528 games where White plays 6.d4 instead. White scores 52%.

6...g6

The Database has 471 games with this position. White scores 71%.

The Database has 570 games where Black plays 6...Kf8 instead. White scores 50%.

The Database has 960 games where Black plays 6...Ke6. White scores 54%.

The Database has 1 game where Black plays 6...Ke7. White scores 100%


Of course, were I to subject each of these positions to the scrutiny of, say, Stockfish 8, the computer's evaluation would likely, in some cases, vary significant from the game outcomes in The Database!

Monday, February 6, 2017

Jerome Gambit: Black Must Castle-by-Hand

I received a new Jerome Gambit game from Vlasta Fejfar. His opponent quickly solves the problem of the light square Bishop, but, instead, his Knights trip all over each other. White's attack against the uncastled King wins.

Vlastous - koden
internet, 2017

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 

The Bishop never finds a comfortable place to stay. At e6 it encourages White to play f2-f4-f5.

10.O-O Qd7 11.f4 Bf7 12.d4 Bc4



Black moves the piece again, and may feel comfortable in that, in attacking White's Rook, he gains a tempo. The truth is that White is "forced" to make a move that he is happy to make, anyhow. In the meantime, Black's King remains a target in the center of the board.

13.Re1 N8e7 14.f5 Nf8 15.d5 



More danger for the Bishop.

White now sacrifices a pawn to open the center and fuel his attack.

15...c6 16.Na3 cxd5 17.exd5 Bxd5 18.Qd3 Bc6 19.Bg5 



Black resigned

Black will lose his Queen for a Rook and Knight. His King will still be stuck in the center, and White's attack will proceed along the helpfully opened center files.

Saturday, February 4, 2017

Jerome Gambit: "Cliff Hardy" Spectacular

Another "Cliff Hardy" spectacular! A very interesting Jerome Gambit played with a time control of one minute for each player. (That turns out to be 1 or 2 seconds a move.)

Again, I present notes from Kevin, with some information from his email. I have added diagrams. 


I couldn't resist sending you this game in the "Charlick" variation, played in the famous Jerome Gambit game, Charlick-Mann, correspondence, 1881. I am from Adelaide in Australia. Our official chronological State Championship winner board ( http://www.oocities.org/timessquare/5427/ ) starts with "1868-93 H. Charlick 1894 J. Hilton; 1895 W. J. McArthur..." and ends with "2016 Alan Goldsmith" but there is an interesting wordy preface to the "1868-93 H. Charlick" bit:
"Historical Prelude - Henry Charlick (1845-1916) was acknowledged chess champion of S.A. [South Australia] for 25 years, by common consent and by his pre-eminent ability. He confirmed this rank by winning the Australian title in Adelaide in 1887 and by taking second place the next year in Melbourne. State Championship tournaments began in 1894, after the foundation of the S.A.C.A [South Australian Chess Association] in 1892, but Charlick never took part in these contests for the title he had relinquished."

Kevin Sheldrick - NN
1 0, FICS, 2017 

1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+ 


4...Kxf7 5.Nxe5+ Nxe5 6.Qh5+ Ng6 7.Qd5+ Kf8


Charlick-Mann, correspondence, 1881, went 7...Ke8, ...1-0, 72.

8.Qxc5+ d6 9.Qe3 Nf6 10.0-0 Bd7?!



Black is now slightly better but Stockfish likes the manoeuvre 10...Kf7!, intending to artificially castle, with a clear advantage for black.

11.f4

Stockfish slightly prefers 11. d4.

11...h6?!

Doesn't do much - 11...Qe8 or 11...c5 are better.

12.d4 Qe7?!

Looks good but 12...Bb5 is better - a few mistakes around here but I guess this is "one minute each" chess 😃.

13.Nc3?!


13.e5! is a little better.


13...Kg8 14.f5 

Now I am equal.

14...Nf8 15.e5 de 16.de Ne8??




16...Rae8! or 16...Ng4 are better, though I am not losing a pawn after 16...Ng4 17.Qg3!, since 17...Nxe5?? loses to 18.f6.

17.Qe4??

17.Nd5! was crushing, with the threat of pushing the f-pawn.

17...Qc5??

17...Bc6 was necessary to keep black in the game, now the Jerome pawns are too strong.

18.Be3 Qc6 19.Qf4

OK, but 19.Nd5! was even stronger.

19...a6 20.e6 Bc8?!

Sacrificing with 20...Nxe6 was better but still losing for black.

21.Bd4 Nd6 22.Qg3?!

22.f6! was killing.

22...Ne8 23.f6 Nxe6 24.f7+ Kh7 25.Qd3+ g6 26.fxe8=Q Rxe8?


26...Qxe8 is better, though 27.Nd5 is still winning for white.

27.Rf7+ Kg8

27...Ng7 would still lead to a quick mate e.g. 28.Rxg7+ Kh8 29.Rxg6+ Kh7 30.Rg7+ Kh8 31.Qh7 mate

28.Qxg6 Ng7 29.Qxg7 mate



Thursday, February 2, 2017

Blackburne Shilling Jerome Gambit: Rare Miniature

One reason players enjoy the Jerome Gambit - and its relatives, like the Blackburne Shilling Jerome Gambit - is that it offers the possibility of ending the game quickly and successfully.

The following battle is very short - and I was surprised to find that it is the only example in over 5,300 BSJG games in The Database.

As always, White moves his Queen often; but Black balances this out by moving his King even more. 

Wall, Bill - NN
lichess.org, 2016

1.e4 e5 2.Nf3 Nc6 3.Bc4 Nd4

The Blackburne Shilling Gambit.

4.Bxf7+

The Blackburne Shilling Jerome Gambit. 

4...Ke7

The Blackburne Shilling Jerome Gambit Declined.

In almost 600 games in The Database, Black scores 31%.

5.Bc4 Ke8 6.Nxd4 exd4 7.Qh5+ Ke7 8.Qe5 checkmate


And that's that.

Tuesday, January 31, 2017

When White Equalizes in the Jerome Gambit, He Has the Advantage

Knowing the Jerome Gambit well gives the attacker several advantages, including the ability to "make something out of nothing" because of his understanding of typical play. Too, there is a danger for Black that, having reached an "even" position, he might lessen his attention - and invite disaster.

Wall, Bill - Ryusak
lichess.org, 2016

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



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


8.O-O Nf6 9.Nc3

9.f4 was seen as early as Jerome,A - Shinkman,W, Iowa,  (1/2-1/2, 42).

The more restrained 9.f3 was seen in  Wall,B - Guest903719, PlayChess.com 2013 (1-0, 47).

More recently Bill has tried 9.Bf4 Be6 10.Nc3 Re8 11.Rad1 Nc6 12.Qd2 Bg4 13.f3 Bh5 14.Nd5 Nxd5 15.Qxd5+ Kf8 16.Qxh5 Kg8 17.Qd5+ Kh8 18.c4 Nb4 19.Qb5 a5 20.a3 Nc2 21.Rf2 Qf6 22.Bg5 Nd4 23.Qxe8+ Rxe8 24.Bxf6 Nxf3+ 25.Rxf3 gxf6 26.Rxf6 Rxe4 27.c5 Re2 28.Rf7 Rxb2 29.cxd6 Kg8 30.Rxc7 Kf8 31.d7 Black resigned, Wall,B - Guest3687203, PlayChess.com, 2015.

9...b6

One of a number of defenses that Bill has faced:

9...Re8 10.Bg5 (10.f4 Nc6 11.Qd3 Kg8 12.Bd2 Kh8 13.f5 Ne5 14.Qd4 c5 15.Qf2 Rf8 16.Rae1 Nfg4 17.Qg3 Qb6 18.b3 c4+ 19.Kh1 cxb3 20.Nd5 Qd8 21.axb3 b6 22.c4 Bb7 23.Nf4 Nf6 24.Ne6 Qe7 25.Nxf8 Rxf8 26.Bg5 Qf7 27.Bxf6 Qxf6 28.Re2 g6 29.Kg1 a5 30.Qf2 a4 31.bxa4 Nxc4 32.fxg6 Qxf2+ 33.Rfxf2 Rxf2 34.Kxf2 hxg6 35.Rc2 Ba6 36.Kg3 Kg7 37.Kf4 Kf6 38.h4 Ne5 39.Rd2 Nd3+ 40.Ke3 Ke5 41.Rxd3 Black resigned, Wall,B - Ferreira,J, lichess.org 2016) 10...h6 (10...Kg8 11.f4 (11.Nd5 Be6 12.Bxf6 gxf6 13.Rad1 c6 14.Nf4 Qb6 15.Qc3 Qc7 16.Rd4 c5 17.Rd2 Rad8 18.Rfd1 Qe7 19.f3 b6 20.b3 Ng6 21.Qe3 f5 22.Nxg6 hxg6 23.exf5 gxf5 24.Qh6 Qg7 25.Qxg7+ Kxg7 26.Rxd6 Rxd6 27.Rxd6 Kf6 28.c4 Ke5 29.Rd2 a6 30.Kf2 b5 31.cxb5 axb5 32.Re2+ Kf6 33.Kg3 c4 34.bxc4 bxc4 35.h4 c3 36.Rc2 Rc8 37.Kf4 Bxa2 38.Rc1 c2 39.Ke3 Bb3 40.Kd4 f4 41.Kd3 Rd8+ White resigned,Wall,B - Guest4809124, PlayChess.com 2013) 11...Nc6 12.Qa4 (12.Qd3 Nb4 13.Qc4+ Be6 14.Qxb4 b6 15.Rae1 c5 16.Qb5 d5 17.e5 d4 18.exf6 a6 19.Qc6 Bd7 20.Qd5+ Kh8 21.fxg7+ Kxg7 22.Bxd8 Black resigned, Wall,B - Foman, Chess.com 2010) 12...Bd7 13.Rae1 h6 14.Bh4 Ne5 15.Qb3+ Nf7 16.e5 dxe5 17.fxe5 Be6 18.Qa4 Bd7 19.Qd4 g5 20.exf6 gxh4 21.Rd1 Nd6 22.Qd3 Bf5 23.Rxf5 Nxf5 24.Qxf5 Qxd1+ 25.Nxd1 Re1+ 26.Kf2 Rxd1 27.Qg6+ Kf8 28.Qg7+ Ke8 29.Qe7 checkmate, Wall,B - U80, PlayChess.com, 2016) 11.Bh4 c5 (11...Be6 12.f4 c5 13.Qa4 Ng6 14.Bxf6 gxf615.f5 b5 16.Qxb5 Bd7 17.fxg6+ Kxg6 18.Qd3 Kg7 19.Qxd6 Rc8 20.Rad1 Rc7 21.Qg3+ Kh7 22.Nd5 Rc6 23.Nxf6+ Rxf6 24.Rxf6 Qxf6 25.Rxd7+ Black resigned, Wall,B - Guest1561957, PlayChess.com 2014) 12.Qd2 Be6 13.Rad1 Nc4 14.Qc1 Qb6 15.b3 Ne5 16.Bxf6 gxf6 17.f4 Ng4 18.f5 c4+ 19.Kh1 Nf2+ 20.Rxf2 Qxf2 21.fxe6+ Rxe6 22.Rf1 Qd4 23.Qxh6 Qxc3 24.Qh7+ Ke8 25.Qg8+ Kd7 26.Qxa8 cxb3 27.Qxb7+ Qc7 28.Qxc7+ Kxc7 29.cxb3 Rxe4 30.Kg1 Re6 31.h4 d5 32.h5 Kd6 33.g4 Ke7 34.Kf2 Kf7 35.Rc1 Re7 36.Kf3 d4 37.Rc6 Rd7 38.Ke2 d3+ 39.Kd2 Rd4 40.Rc7+ Kg8 41.Rxa7 Rxg4 42.Kxd3 Rh4 43.a4 Rxh5 44.a5 Rb5 45.Kc4 Rb8 46.a6 f5 47.Rb7 Ra8 48.a7 Black resigned, Wall,B - Guest5111265, PlayChess.com 2014;

9...c6 10.f4 Ng6 (10...c5 11.Qf2 Neg4 12.Qd2 Rf8 13.h3 Nh6 14.e5 Nh5 15.g4 Ng3 16.Rf3 Qh4 17.Kg2 Bxg4 18.hxg4 Nxg4 19.Qd5+ Ke7 20.Qxb7+ Ke6 21.Qd5+ Kf5 22.Qd3+ Ke6 23.Qxd6+ Kf7 24.Qd5+ Kg6 25.Rxg3 Qh2+ 26.Kf3 Qf2+ 27.Kxg4 Rxf4+ 28.Bxf4 h5+ 29.Kh4+ Kh7 30.Qe4+ g6 31.Qxg6+ Kh8 32.Qg7 checkmate, Wall,B - Guest708676, PlayChess.com, 2016) 11.Be3 Re8 12.Rad1 Bg4 13.Qc4+ Kf8 14.Rxd6 Nxe4 15.Rxd8 Raxd8 16.Nxe4 b5 17.Bc5+ Ne7 18.Bxe7+ Rxe7 19.Qxc6 Be2 20.Re1 Bc4 21.b3 Bd5 22.Qc5 Rdd7 23.Ng5 Rc7 24.Qd6 Rcd7 25.Qb8+ Black resigned, Wall,B - Uscanac, lichess.org, 2016;

9...c5 10.Qd2 Nc4 11.Qd3 Be6 12.f4 Bg4 13.Qxc4+ Be6 14.Qb5 Black resigned, Wall,B - TryToBeg, lichess.org, 2016; and

9...Rf8 10.f4 Nc6 11.Qd3 Kg8 12.Bd2 Be6 13.Rad1 a6 (13...Bf7 14.Nd5 Nxd5 15.exd5 Bg6 16.f5 Ne5 17.Qh3 Be8 18.Rf4 Qf6 19.Qg3 Bb5 20.a4 Bd7 21.Rdf1 Rac8 22.b3 Rce8 23.c4 b6 24.Bc3 Rf7 25.h3 c6 26.dxc6 Bxc6 27.Rd1 Qe7 28.Bb4 Rf6 29.Kh2 Bb7 30.Qf2 g6 31.Qg3 Kh8 32.Rfd4 Rxf5 33.Rxd6 Rg5 34.R6d2 Rxg3 35.Bxe7 Rxb3 36.Bf6+ Kg8 37.Rd8 Kf7 38.Rxe8 Kxf6 39.Rf1+ Kg5 40.Rxe5+ Kh6 41.Rf7 Ba6 42.g4 Rb2+ 43.Kg3 Rb3+ 44.Kf4 g5+ 45.Rxg5 Bxc4 46.Rf6 checkmate, Wall,B - Anonymous, lichess.org, 2016) 14.Nd5 Bg4 15.Rde1 Bh5 16.Bc3 Nxd5 17.exd5 Ne7 18.f5 Bf7 19.Qd4 Black resigned, Wall,B - Anonymous, lichess.org, 2016.

10.f4 Ned7 11.e5 dxe5 12.fxe5 Nxe5 13.Qxe5 Re8


Black has returned the sacrificed piece for what he believes is an equal game.

He has forgotten the old saying: When White equalizes in the Jerome Gambit, he has the advantage. How else to explain the quick end to the game?

14.Qg3 Qd4+ 15.Kh1 Ba6 16.Bh6 


A nice sacrifice that moves the advantage in White's direction.

16...gxh6 17.Rad1 Qg4 

Black had to try 17...Bxf1 18.Rxd4 Bxg2+, although White would still have an edge after 19.Kg1!?

Now Black loses his Queen, and checkmate is forced.

18.Rxf6+ Kxf6 19.Qxg4 Black resigned



Sunday, January 29, 2017

Jerome Gambit: Old McDonnell

Kevin Sheldrick ("Cliff Hardy") sends one of his latest Jerome Gambits. He gets into trouble, but he calls upon his creativity and his "Jerome pawns" save him. The comments are mostly Kevin's. I have added a few diagrams and some notes in blue - Rick


Old McDonnell Had a Pawn Storm

A kingside pawn storm features in many Jerome Gambit games, as is clearly evident in the games on this blog. In that vein, I recently played a Jerome Gambit game that reminded me somewhat of a famous position from an old game between two classic masters, Alexander McDonnell and Louis-Charles de La Bourdonnais in London in 1834. These were two of the strongest players in the world at the time and the final position in that game I have heard being referred to as, arguably, the most famous position in chess history.

  

La Bourdonnais, as black, has just played 37...e2, and so McDonnell resigned in a lost position, despite being the player who currently has a queen on the board.

My game: 

dsjkljdkdjkf - beba buba beba542
15 0, instantchess.com, 25/1/17

1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+ Kxf7 5.Nxe5+ Kf8 6.Qh5??


6.Nxc6! would have been better. (See gfdgfd - leleos
InstantChess.com, 2016.)

6...Nxe5 7.Qxe5 d6 8.Qg3 Nf6 9.d3 h5 10.Bg5 Bd4?!


The odd-looking 10...Qd7!, blocking his bishop on c8 but breaking the pin on the knight on f6, was best. (10...Be6 was perrypawnpusher - ulisimbolon, Giuoco Piano Thematic, Chess.com, 2016 [0-1, 22])

11.c3 Be5??

11...Bb6 would still leave black with a clear advantage.

12.f4

Old McDonnell had a bishop!

12...h4 13. Bxh4??

13.Qf3! Bg4 14.Qe3 and white will win the trapped bishop on e5, with a winning position.

13...Nh5!

13...Nh5 was good for a slight advantage for black (saves his bishop on e5) but 13...Nxe4! was slightly better than that again

14.Qf2 Bf6 15.Bxf6 Qxf6 16.0-0 Bd7 17.Nd2 Ke7 18.d4 Raf8 19.f5 Kd8 20.Rae1 Qg5 21.Nf3 Qg4 22.h3 Qg3



23.Qxg3?

23.Re3 would yield an approximately equal position.

23...Nxg3 24.Rf2 Rhg8?

24...Re8, attacking the e-pawn, would lead to a winning advantage for black, whereas after 24...Rhg8?, white has a slight advantage.

25.Kh2 Nh5 26.g4 Nf4 27.Kg3??


E - I - E - I - O! I missed the fork.

27...Nd3 28.e5 Nxe1 29.Nxe1 de 30.de Bc6 31.g5 Kc8 32.Kg4 Rd8 33.h4 Rd1 34.Re2 Rgd8?

 

Black is still better but 34...Bb5! 35.Re3 Rd2, with a powerful rook on the seventh rank, would lead to a really crushing position for black.

35.e6

With a push-push here and a push-push there of my e- and f-pawns, I was hoping to pull off a swindle, though 35. h5! (i.e. everywhere a push-push!), with a four-pawn storm, would give me better chances to fluke a win from my bad position.

35...R8d2 36.Rxd2 Rxd2 37.f6 gf 38.gf Bd5 39.Kf5 Rf2+ 40.Ke5 Bxa2 41.f7 Re2+??

41...Kd8! was actually the only winning move for black here.

42.Kf4

Now white has a winning position!

42...Rf2+ 43.Nf3 Bxe6 44.f8=Q+ Kd7 45.Qg7+ Kd6 46.Qd4+ Bd5 47.Qxf2 c6 48.Qxa7 Bxf3 49.Kxf3 and black resigned.


Take care of your "Jerome pawns" and they will take care of you. Wonderful, Kevin!