Jerome Gambit: Centaurs

Recently Bill Wall sent me 16 of his Jerome Gambit games that were Human + Computer vs Computer encounters. Such teamwork is sometimes referred to as advanced chess, or cyborg chess, or centaur chess.

Over the years, I have posted games from Human vs Computer matches (including the legendary 1993 Fisher-Kirshner - Knight Stalker battles, and the rolling 2006 RevvedUp - Fritz 8 - Crafty 19.19 - Hiarcs 8 - Shredder 8 - Yace Paderborn mayhem) as well as many Computer vs Computer games, but I think this is the first centaur chess I have presented.

The results are interesting - even if it is difficult to assign the relative impact that the human had on the play. Also, the time controls, which affect the strength of computer programs, are not known.

Over all, White scored 4 - 9 - 3 (34%), which would be unimpressive for a normal opening under normal circumstances, but which seems - as with all Jerome Gambit matches - a bit "high" for a many-times-refuted opening.

A little more insight is available by breaking the games down into 4-game matches.

Crafty vs Stockfish + Wall, for example, yielded 2 wins for Black when played by the team; and, likewise, 2 wins for White when played by the team. With all due respect to Dr. Robert Hyatt's computer engine, it appears it could have been simply outplayed by its stronger computer opponent. Who played what color did not seem to matter.

On the other hand, the Komodo 5 vs Rybka + Wall match, which ended with a score of 2 - 2 - 0, was composed of 4 wins by Black. Neither engine, it appears, was able to ovecome the "handicap" of playing the Jerome Gambit.

The Hiarcs 9 vs Critter + Wall match seemed a reflection of the comparative strengths of the computer programs, as Hiarcs 9 lost 2 games as White, and could only manage a draw as Black.  

Interesting, also, was the Fritz 12 vs Houdini + Wall match. The team was 1 - 0 - 1 as White, and 1 - 0 - 1 as Black, suggesting that Houdini was the brighter computer program.

Looking at a couple of lines of play, 1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+ Kxf7 5.Nxe5+ Nxe5 6.d4 scored 2 - 5 - 1, while 6.Qh5+ scored 2 - 4 - 2, not much of a difference.

I will be sharing some of the games, taking a look at what "theoretical" enlightenment they bring. 
  

The Database Upgrade

I recently upgraded The Database with games from the online playing site FICS through August, 2018, bumping the number of games it contains to just above 59,000.

With almost 20 years of games from FICS, I have a pretty good representative sample (not just wins, but losses and draws as well) of how online club players play the kind of openings this blog covers.  

There are Jerome Gambit and Jerome-related games, including:

14,353 games with the Jerome Gambit, proper, 1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5, 4.Bxf7+ (White scores 46%);

2,614 games with the Italian Four Knights Jerome Gambit, 1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.Nc3 Bc5 5.Bxf7+ (White scores 38%);

204 games with the Semi-Italian Jerome Gambit, 1.e4 e5 2.Nf3 Nc6 3.Bc4 h6 4.0-0 Bc5 5.Bxf7+ (White scores 64%);

144 games with the Semi-Italian Four Knights Jerome Gambit, 1.e4 e5 2.Nf3 Nc6 3.Bc4 h6 4.0-0 Nf6 5.Nc3 Bc5 6.Bxf7+ (White scores 65%); and

6,064 games with the Blackburne Shilling Jerome Gambit, 1.e4 e5 2.Nf3 Nc6 3.Bc4 Nd4 4.Bxf7+ (White scores 56%).

In addition, there are some miscellaneous lines, including 1.e4 e5 2.Nf3 Nc6 3.Bc4 Na5 4.Bxf7+ (2,406 games, White scores 72%); 1.e4 e5 2.Nf3 Nc6 3.Bc4 h6 4.Bxf7+ (2,061 games, White scores 40%); and various reversed Jerome-style openings, e.g. 1.e4 e5. Nf3 Bc5 3.Nxe5 Bxf2+.

So, who plays the Jerome Gambit?

The player with the most games in The Database (White and Black) is DragonTail (1,452), although I note that his last game is from 2012. Likewise, kingmaple appears 619 times, but his last game is from 2010. yorgos has 600 games, but his last game is from 2014.

As for "active" players, Bill Wall now leads the pack with 842 games in the Database, followed by perrypawnpusher with 592 games - closely followed by drumme with 553 games and Petasluk with 521. I am sure that there are others not yet on my "radar" (e.g. ZahariSokolov with 331 games and rising quickly).

Jerome Gambit: Too Fast, Too Furious (Part 3)

[continued from the previous post]

perrypawnpusher - TobiasBrunner

"Let's Play ThChess e Italian Game" tournament

Chess.com, 2018 

27.Qb5+

An example of "patzer sees check, patzer gives check", but it turns out to be a good move.

27...c5 28.Qxb7+ Ke5 

29.b3

This is funny. I was worried about Black putting a Rook on b8, with an attack on my Queen and an x-ray attack on my b-pawn. More to the point was, of course, 29.Qf3.

29...Rg8

Samuel Johnson once said “Depend upon it, Sir, when a man knows he is to be hanged in a fortnight, it concentrates his mind wonderfully.”

I could see that my opponent planned to put his Bishop on the a8-h1 diagonal to complement the Rook's attack on my King. I certainly didn't want to lose a game after winning my opponent's Queen!

30.Qf3 Bd5 

This is what I had worried about - but, with the proper "motivation" I had figured out the solution.

31.Re1+ Kd4 

Or 31...Be4 32.Rxe4+ Kd5 33.Re8+ Kd4 34.c3 checkmate 

32.c3 checkmate

Just in time, too. The second round of the "Italian Battlegreound" tournament at Chess.com was about to start...

Jerome Gambit: Too Fast, Too Furious (Part 2)

[continued from previous post]

perrypawnpusher - TobiasBrunner

"Let's Play The Italian Game" tournament

Chess.com, 2018 

Striking back, the Knight that has been kicked around targets the Queen.

I have seen the retreat 13...Nc6 three times: perrypawnpusher - Heler, Giuoco Piano tournament, Chess.com, 2015 (1-0, 33); perrypawnpusher - apurv83, Giuoco Piano tournament, Chess.com, 2015 (1/2-1/2, 29); and perrypawnpusher - Hywel2, Italian Game tournament, Chess.com, 2015 (0-1, 44).

Stockfish 9 prefers the retreat 13...Ned7; there are no games with this move in The Database.

14.Qb3+ 

Delivering the check before Black can sequester his King at g8.

14...Kf8

Better than 14...d5 15.e5 (15.h3!? Rxe4 16.hxg4 Rxg4 17.Bf4 Bxf5 =) Ne4 (15...Nxe5!? 16.dxe5 Rxe5 17.Bf4 Re8 =) 16.h3 (16.Nc3!? Ngf6 17.exf6 Nxf6) Ngf2 (16...Nxe5 17.dxe5 Rxe5) as in perrypawnpusher-whitepandora, blitz, FICS, 2011(1-0, 64) 

15.h3 Nh6

Instead, 15...Rxe4 16.hxg4 Rxg4 was seen in Vazquez,A - Carrington,W, Mexico, 2nd match 1876 (1-0, 34)

16.Bxh6 gxh6 17.Nd2 

It is rather curious that Stockfish 9, after the game, now recommended 17...b6!? for Black, giving 18.Qe3 Qe7 19.Rae1 Bb7 20.e5 dxe5 21.dxe5 Nd7 22.Qxh6+ Qg7 23.Qxg7+ Kxg7 24.e6 Nf6 25.g4 as even.

Clearly the e-pawn is poisoned, but why is that?

17...Nxe4 18.f6 

This advance and offer of a "Jerome pawn" is the reason. After 18...Nxf6 White will double his Rooks on the f-file and swing his Queen over to the g-file. As is often the case, Black will be troubled by having his Bishop on its home square, blocking entry of his other Rook.

I wish I could take credit for the move, but I had already seen it in Wall - Vijay, Chess.com, 2010 (1-0, 22). Of course, I had already played it, then, in perrypawnpusher - Bunnywarrior, Giuoco Piano thematic tournament, Chess.com, 2016 (1-0, 20)

18...Be6

This moves seems to solve Black's problems at first glance, avoiding the tempting 18...Nxd2? which led to a crush after 19.Qg3!? in Wall, - Vijay, Chess.com, 2010. It also improves on my game against Bunnywarrior, which makes me wonder - was TobiasBrunner aware of that game?

19.Qd3 Nxd2 

The Knight grabs material, abandoning his King.

After the game, Stockfish 9 said something like "Bwahaha! White now has a checkmate in 12." I didn't have any idea at the time, but I knew that White was going to have a strong attack.

What could Black have done, instead? He could have (temporarily) protected the h-pawn with 19...Ng5, although 20.h4 Bf7 21.hxg5 Bg6 22.Qf3 looks good for White. He could also have tried 19...d5, although 20.Nxe4 dxe4 21.Qe3!? targeting the advanced h-pawn, looks dangerous.

Looking at White's f-pawn, I remembered the villains' line from various Scooby Doo cartoons, "And I would have gotten away with it too, if it weren't for you meddling kids"

20.Qxh7 Qd7 21.Qh8+ Bg8 22.Qxh6+ Kf7 

Now, 23.d5! would effectively block the Black King's avenue of escape, and he could only delay checkmate by returning lots of material.

I missed the move because, two pieces down, I was mostly thinking about - material.

23.Qg7+ Ke6 24.d5+

This certainly accomplishes the task of moving the King away from his defense of the Queen, but 24.Rae1+ would have provided more of an attack, and would have been stronger. 

24...Kxd5 25.Qxd7 

25...Nxf1 26.Rxf1 Be6 

Black has a Rook and a Bishop for his Queen (and a couple of pawns), and decides that he can fight on.

[to be continued]

Jerome Gambit: Too Fast, Too Furious (Part 1)

I was playing my last game in the first round of the "Let's Play The Italian Game" tournament at Chess.com when, suddenly, two opponents made their moves against me in new games. Apparently, the computer scheduler had decided that I would progress to the second round of play, even if I lost the first round game (I didn't; I won it) - so, there I was, with two second round games.

Somewhat disconcerting, each of my opponents was playing quickly - in fact, one of my games ended after a day of play, rather fast given that the time control was one move per day.

Still, I was willing to move with speed, too, as we were contesting a Jerome Gambit - in a line that I had visited several times before. (Alas, I lost the other 3 games, so no more "next" round for more Jeromes.)

perrypawnpusher - TobiasBrunner
"Let's Play The Italian Game" tournament
Chess.com, 2018 

 1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+ 

4...Kxf7 5.Nxe5+ Nxe5 6.Qh5+ Ng6 

7.Qd5+ 

Years ago, I labeled this finesse "a nudge". It is as old as Jerome - Brownson, Iowa, US, 1875 (1-0, 28). In a blitz game, it can get Black wondering what White is up to - using up precious time.

7...Ke8 8.Qxc5 d6 9.Qe3 

After the game, I put Stockfish 9 in "blunder check" mode to look at the game. Initially, it liked 9.Qc3 Qg5 10.Qg3 Qxg3 11.hxg3, with Black's advantage less than a pawn. That seems a bit silly to me, White moving his Queen 5 times, only to let it be exchanged.

According to The Database, I have reached the diagrammed position 83 times in my games. I must have learned something about the line.

9...Nf6 10.O-O Kf7 

Black sensibly prepares to mobilize his Rook, and possibly castle-by-hand.

A major alternative is 10...Ng4.The proper response is 11.Qg3!,and this is reminiscent of the series of posts I did a while back on "Jerome Gambit Secrets". A strong and experienced Jerome Gambiteer tried 11.Qe2, instead, and perished quickly: mrjoker - weiran, ICC, 2008 (0-1, 14). Another played 11.Qc3 and needed all of his Houdini-like powers to escape and turn the tables: Wall,B - Guest1660564, PlayChess.com, 2013 (1-0, 32). 

I have played 11.Qg3 with mixed results: perrypawnpusher - lorecai, blitz, FICS, 2011 (1-0, 14), perrypawnpusher - jgknight, blitz, FICS, 2011 (1/2-1/2, 49), perrypawnpusher - FiNLiP, blitz, FICS, 2011 (0-1, 59). For a couple of more modern examples, from stronger players, see Sheldrick,K - Bhat,V, Australian Open, 2017 (1-0, 20) and Morin,L - Weston,P, Quebec Open, 2018 (1-0, 54)

11.f4 Re8 12.f5 

I have been pretty successful with this thematic move (10 - 4 - 1), chasing the Knight, so I was surprised to see, after the game, that Stockfish 9 recommended 12.Qb3+!? instead. The Database has only one game with that move, a win by Black in 25 moves.

12...Ne5 13.d4 

Here, again, the computer prefers 13.Qb3+, followed by 14.d3, which seems to lock the Queen on the Queenside. (There are no game examples in The Database.) As we will see, Stockfish's preference is based on disliking Black's next move.

13...Neg4 

[to be continued]

Jerome Gambit: The Madness Continues

So, the latest game I received from Cliff Hardy, who has been specializing in tortured Jerome Gambit games, is a 1-minute (no increment) bullet game on the website lichess.org - against an International Master.

Of course.

In a game where each player had about one second per move to think, things quickly became exciting. 

Notes by Cliff. [A few by me - Rick] Diagrams by me.

Cliff Hardy - IM Pavel Sevostianov

1 0 bullet, Lichess, 2018 

1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+ 

4....Kxf7 5.Nxe5+ 

In another bullet game recently on Lichess, I forgot this was the move to play, got confused and played 5.Ng5+?? here instead but got justifiably destroyed in short order. I'm glad I didn't make that mistake in this rare chance to play the Jerome against an IM, where I was at least hoping to put up a half-decent game and not waste the opportunity.

5...Nxe5 6.Qh5+ Ng6 

[Interesting - is the IM unfamiliar with the various Jerome Gambit refutations, or does he figure that this will be good enough? It turns out that his "second best" moves are still pretty good. - Rick ]

7.Qd5+ Kf8 8.Qxc5+ Qe7 9.Qe3 Nf6 10.Nc3 c6 

10...d5! -+ would have been the best move here, according to Stockfish, as Black should eventually be able to get the pawn back on d5 (if White were to take it), though it may then take Black a few moves of development first before he could get into a position to win it back.

11.f4? 

Which pawn to push? 11.d4 =+ would have been much better, it turns out.

11...d6? =+ 

Pushing the d-pawn further again with 11...d5! -++ would have been more aggressive.

12.O-O Kf7

13.d3?! -+ 

13.f5 Ne5 14.d4 =+ would have been a better and feistier way to play it for White. 

13...Re8 14.f5 Ne5 15.h3 Kg8 16.Bd2 d5 17.Rae1 dxe4?

Surprisingly, quite an error - 17...Qd6 -+ would have been a better move. 

18.dxe4??

18.Nxe4! +- would have been stronger, when White's threat of Nxf6+ and then attacking the pinned knight on e5 with d4 or Bc3 would have been surprisingly difficult to diffuse e.g. 18...Bd7 19.Nxf6+ Qxf6 (19...gxf6 20. d4 ++-) 20.Bc3 +-.

18...Nc4 19.Qd3 Nxd2 -++ 

19...Nxb2 -++ would also have been good.

20.Qxd2 Bd7 

21.Qf4 Qc5+ 22.Kh1 Qe5 23.Qf3 Re7 -++ 

OK, I'm getting crushed like a bug - but I've still got my Jerome pawns... [And then, there's that clock thingy - Rick]

24.g4 Rae8?? 25.Re2?? 

All right, not the best move - 25.g5! += and the black knight would have been trapped. 

25...h6 26.h4 Nh7 27.Qg2 b6 28.g5 hxg5 29.hxg5 Bc8?

Black appears to be struggling for ideas, despite his material advantage - 29...g6! -++ (intending 30.fxg6 Nxg5) was a brilliant defence to White's pawn onslaught that the computer found. 

30.f6 Rc7 31.f7+

OK, now I'm back in the game, though Stockfish found another surprising resource here in 31.Rf5!?, since 31...Bxf5?? (31...Qd4 32.f7+ = would have been similar to the game) would walk into trouble on the e-file after 32.exf5 +-. 

31...Rxf7 32.Rxf7 Kxf7 33.g6+ Kg8 34. gxh7+ Kh8! 

Black can probably take the h-pawn any time he wants and, by playing this way, keeps more options open by not allowing White to play the queen swap with 34...Kxh7 35.Qh2+.

35.Qh2 Bg4? 

...Though he allows the queen swap now anyway... - keeping the queens on with 35...Qg5 -+ would have been more attacking.

36.Rg2?? 

The last straw for White - 36.Qxe5 Rxe5 37.Re3 would have been about equal.

[White still can win if he stays alive - regardless of the material disadvantage - long enough for Black's flag to fall. He almost makes it. - Rick]

36...Bf3 37.Kg1 Bxg2 38.Qxg2 Re6 39.Kf2 Rf6+ 40.Ke2 Rh6 41.a3 Rh2 42.Kf3 Rxg2 43.Kxg2 Qg5+ 44.Kf3 Qf6+ 45.Ke2 Qg6 46.Ke3 Qh5 47.Ne2 Kxh7 48.Nf4 Qh6 49.Kf3 Qh4 50.Ke3 Qg5 

Not that it matters, but better would have been 50...Qg3+! 

51.c3 Qf6 52.Kf3 Qe5 53.Ke3 g5 54.Nh3 Qh2 

Again, better would have been 54...Qg3+! 

55.Nf2 Qg3+ 56.Ke2 g4 and Black won on time. 

(White had 1.6 seconds left on his clock when Black lost on time).

Jerome Gambit: Odd Turns (Part 2)

[continued from previous post]

perrypawnpusher - Sarantes

"Let's Play The Italian Game" tournament
Chess.com, 2018

25.d4 Kf7 26.d5 Bd7 27.e5 dxe5 

Often pawn exchanges such as these are played automatically, but there is an argument to be made for 27...Bf5, instead. Either way, a white pawn is going to advance to e6.

28.fxe5 Ke7 29.e6 Bb5 

Not as bad as it looks, but it doesn't look good.

It takes a little patience, but Black had 29...Bc8 30.Re5 (to protect the d-pawn and forestall the entrance of Black's h-rook) b5 31.c4 bxc4 32.bxc4 Kd6 33.Rg5 Re7 34.a4 b6 35.a5 bxa5 36.Ra1 Ba6 37.Rg4 Ra8 38.Rxa5 Bb7 39.Rxa8 Bxa8 40.Rxh4 c6 and the Bishop will be liberated.

analysis diagram



30.c4 Ba6 31.a4 

Locking up the Bishop while I can. It can still break out with 31...c6, but my opponent's next move is either a mouse slip (shouldn't happen when the time control is a day per move, but, even then, it does) or a slip in assessment - taking a pawn e.p. is not forced.

31...c5 

Now the game definitely swings over into White's favor, as the Bishop is out of play. Capturing the c-pawn would have been a blunder.

My plan is to move the Rooks to the f-file, threaten to enter at f7, and therefore exchange them off. White's King will then easily capture the Kingside pawns as Black dare not leave the control of the "Jerome pawns".

32.Rf1 Ref8 33.Ref3 Rxf3 34. Rxf3 Rf8 35.Rxf8 Kxf8 

36.Kf2 Kf7 37.Kf3 g5 Black resigned

The Bishop is trapped. Black's Kingside pawns will fall. One or more White pawn will Queen.