Jerome Gambit: Descent Into Rubble

The latest game from Cliff Hardy is a wild affair, concluding with a mad dash to the finish line - er, time control. 

In a 1-minute, no increment, bullet game, you can win, even if your opponent has a forced checkmate - if his flag falls. Just another reason for an opponent to hate the Jerome Gambit.  

I have added some diagrams and a comment or two, in blue; otherwise, the notes are by Cliff. 

Here is a bullet Jerome Gambit game I played today, which most certainly has several errors in it!

Cliff Hardy (2400) - NN (2391)
1 0, Lichess, 2019

1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+ 

4...Kxf7 5.Nxe5+ Nxe5 6.Qh5+ g6 

7.Qxe5 Qe7 8.Qf4+

I successfully remembered that taking the rook here is inferior (I think!).

Black is playing Whistler's Defense with 6...g6 and 7...Qe7. Taking the Rook at h8 is complicated, but "theoretically" dangerous for White, although The Database shows a score of over 50% in about 50 games. Check out the earlier post, "Bop!" - Rick

8...Kg7 9.d3 Nf6 10.0-0 Rf8 

Black has effectively castled-by-hand. Given enough time, he will continue to figure things out. Unfortunately, he will not have enough time. - Rick

11.Nc3 d6?!

Stockfish prefers the hard-to-see 11...d5! -++ e.g. 12.exd5?/e5? Ng4! 13.Qg5 Bxf2+.

12.Qh4 Bd4? =+

Missing White's threat - 12...h5 13.Na4 -++ would still have been strong for Black.

13.Bh6+ Kg8 14.Bxf8 Qxf8 15.Kh1 Be6?! =

The perennial inaccuracy for Black in the Jerome Gambit - this bishop move tends to run headlong into White's pawn push f2-f4-f5. Better would have been 15...Qg7 16. f4 Bd7, with a slight advantage for Black. 

16.f4 Re8 17.Nd1? =+

17.Ne2 would have been better, though I had missed that 17...Bxb2? could then have been met by 18.Rab1.

17...Ng4 18.h3 Ne3 19.Nxe3 Bxe3 20.Rf3 Bd4 

21.c3 Bg7 22.Raf1 Qe7??

22...Qf6 23.Qf2 =+

23.Qg3??

Missing the typical Jerome Gambit tactic of utilising the f-pawn with 23.Qxe7 Rxe7 24.f5 gxf5 25.exf5 Bxa2 26.f6! ++-, forking the rook and bishop.

23...Bxa2 24.f5 Bf7??

The Jerome Gambit f-pawn strikes again! Blockading it with 24...gxf5 25.exf5 Qf6 would have left Black with a clear advantage.

25.f6 Bxf6 26.Rxf6 Be6 

27.Qf4 Kg7 28.Qf2 Bg8?

The interference tactic 28...Bf5! would have been better here, though after 29.Rxf5 gxf5 30.exf5 Rf8 31.Qxa7, White would still have had a winning advantage.

29.Rf3 Qe5 30.Qxa7 Qb5 

31.Qd4+ Kh6 32.Qf6 Re5 

33.Qh4+ Kg7 34.Qf6+ Kh6 35.Rf4 Rh5? 

36.Rh4?

Black made the mistake of allowing White to mate in 2 moves with 36.Qf8+ Kg5 37.Rg4 (which White obviously missed!), though Black's position was lost in any case.

 36...Rxh4 37.Qf8+?

The game now descends into rubble. With White having 2.4 seconds left on the clock and Black having double that (exactly 4.8 seconds left), we both started to throw down any move we could think of! Clearly, 37.Qxh4+ would have been a much better move to play (though even after 37.Qf8+?, White still has a winning position).

37...Kg5 38.Qf3 Rg4?? 39.Qg3?? Rxg3 40.Kh2 Qe5 41.Rf5+ and White cheekily won on time.

Jerome Gambit: More GMs? (Part 1)

I received an interesting email from "Cliff Hardy" the other day. In it, he notes

In the post about my game vs the Cuban GM, "A GM Faces the Jerome Gambit (Part 2)", you mentioned the virtually mistake-free game:  ChessCoachUA - PArnaudov, 3 0 blitz, lichess.org, 2017: 1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+ Kxf7 5.Nxe5+ Nxe5 6.Qh5+ Kf8 7.Qxe5 d6 8.Qg3 Nf6 9.d3 Qd7 10.Nc3 Qg4 11.Qxg4 Bxg4 12.h3 Be6 13.Be3 Bxe3 14.fxe3 Ke7 15.O-O Rhf8 16.Nb5 c6 17.Nd4 Nd7 18.a4 Rxf1+ 19.Rxf1 Rf8 20.Rxf8 Nxf8 21.Kf2 g6 22.Nf3 Nd7 23.b3 a5 24.Ke2 Nc5 25.e5 dxe5 26.Nxe5 Kd6 27.d4 Nd7 28.Nd3 Bf5 29.g4 Bxd3+ 30.Kxd3 b5 31.axb5 cxb5 32.e4 a4 33.bxa4 bxa4 34.h4 a3 35.Kc3 Nb6 36.Kb3 Nc4 37.c3 Nd2+ 38.Kxa3 Nxe4 39.c4 Black lost on time; in which black maintains a winning advantage the whole game (from move 4), according to the Stockfish feature on Lichess (but loses on time in a winning position). I mentioned in a comment to the game that, in a slight anticlimax to my game being a rare example of a GM facing the Jerome Gambit, the black player PArnaudov is also a GM! https://lichess.org/@/PArnaudov

This certainly got me thinking! I responded quickly

The game ChessCoachUA - PArnaudov, 3 0 blitz, lichess.org, 2017, is a good example of why some skeptical people see the Jerome Gambit as only playable in very quick games. Black can figure it all out, but it might take too much time. Certainly angelcamina has found this out with his 1 0 games.

I was interested in learning that PArnaudov is a GM... 

For fun, I went to The Database and did a search for games with 1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+ and "ratings > 2500". I found 25 games, with 6 games where White was >2500. 

Wow! I was going to check out the games where the 2500+ player played the Jerome Gambit, first. The relevant games in The Database were all played online, not over-the-board.

One game was by BlackDemon, at FICS; it's a computer account; still, it might be fun to see if the Jerome is/was part of its book, the way "Brause" was set up to play the Halloween Gambit... 

Two of the games were by Topper76, at redhotpawn.com, in a Kentucky Opening tournament; I will have to track him down. Oddly enough, he won one Jerome Gambit game, on time, in 7 moves, and lost the other to a player rated 2421. 

Three of the games were by ChessCoachUA, at lichess.org - in fact, he was rated 2394, 2453, 2519, 2530, 2514 and 2459 on the site, for 6 Jeromes. Alas, ChessCoachUA's lichess.org account was listed as "closed", when I tried to check up on him.

Concerning the FICS computer account BlackDemon, I went to the online FICS games database, and found only 2 Jerome Gambits that it played with White, both in TheDatabase, one where the computer was "only" rated 2347 on the site. As a silicon beast, it couldn't be a grandmaster "in real life", anyhow.

I could not find a Topper76, at redhotpawn.com or elsewhere. Of course, those Jerome Gambit games were from 8 years ago, so the player may have moved on.

The ChessCoachUA story is a bit more complicated, and I had slipped up in my response to Cliff, above. The Database has 2 games by ChessCoachUA, as well as 2 games by ChessCoach_UA, and 4 games by ChessCoach1985. All of the players have lichess.org accounts listed as "closed", so it was not possible to do the kind of research that Cliff Hardy did with PArnaudov (above), to see if we were dealing with a GM in real life.

I tried, though.

I could find nothing on ChessCoach1985.

On the other hand, ChessCoach_UA, could well be Vasily Burishin, of Kiev, Ukraine. I could not find an FIDE rating.

I did find a ChessCoachUA at Chess.com - Alexey Pugach, of Dnepr, Ukraine, whose FIDE Chess Profile lists his rating as 1985. Of course I messaged him.

So - the search to find a human GM (FIDE rating, not online site rating) has not (yet) been successful.

[to be continued]

A GM Faces the Jerome Gambit (Part 2)

[continued from previous post]

Cliff Hardy (2285) - GM Yasser Quesada Perez (2275), 

Lichess, 1 0, 2019

notes by Cliff Hardy and Rick

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

Finally! A GM has walked into my trap!

4.Bxf7+ 

4...Kxf7 5.Nxe5+ Nxe5 6.Qh5+ 

Interesting. Some players - including Alonzo Wheeler Jerome, at first - prefer 6.d4, instead. The line is less forcing than the text, and gives Black more chances to go astray. Would the GM have found the sharpest reply, 6...Qh4!? That is a tale for another day, another game.-Rick

6...Kf8 

So, is this the revealed wisdom of the grandmaster? Is Black's best defense to the Jerome Gambit the Jerome Defense, seen in two Daniel Jaeger - Alonzo Wheeler Jerome correspondence games,  from 1880, both wins by Black?

Not necessarily.

There are at least 3 ways to address this issue.

First, we can ask another "grandmaster", Stockfish 10, what its choice is. It prefers 6...Ke6 over 6...Kf8 by about 3/4 of a pawn, although, of course it sees both moves as decisive for Black.

Secondly, we can check with The Database, which will give feedback on the results of different defensive moves, drawn from online club level player games. There are 529 games with 6...g6, which can lead to either the Blackburne or Whistler's Defenses, and against which White scores a surprising 72% - no doubt because of the complexity of the play. There are 2,071 games with 6...Ng6, and 1,101 games with 6...Ke6; against each White scores 53%. Finally, there are 663 games with 6...Kf8, against which White scores 49%.

So, if the grandmaster is "thinking like a club player", he will choose 6...Kf8.

I am not sure that GM Quesada Perez has ever given a thought to the Jerome Gambit, let alone prepared a special "best" defense against it. I suspect that when confronted by the Jerome in a bullet game, he simply grabbed a move that looked okay, that didn't advance his King further into danger, and that didn't possibly weaken his pawn structure.

To be fair, if the GM had 8 seconds left on his clock at the end of the game, then he used up 52 seconds for 54 moves - an average of slightly less than a second per move. That he spent twice that average - a whole 2 seconds - on 6...Kf8, might be significant only in humorous commentary. (I recall a master suggesting that in over-the-board blitz games, it was best to choose moves on the side of the board closest to the clock.)- Rick

7.Qxe5 d6 8.Qg3 Nf6 9.Nc3 Qd7

Perez's 9...Qd7 caught me on the hop. With more time, I probably would have played 10.h3 here to stop black going for a queen swap with 10...Qg4. Keeping queens on the board, I think gives White better attacking prospects, though I'd be interested in what the database has to say about the success rates of White's different options on this tenth move!

9...Qd7 is a novelty, according to The Database, although there are a couple of related games that this game will transpose into:

ChessCoachUA - PArnaudov, 3 0 blitz, lichess.org, 2017: 1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+ Kxf7 5.Nxe5+ Nxe5 6.Qh5+ Kf8 7.Qxe5 d6 8.Qg3 Nf6 9.d3 Qd7 10.Nc3 Qg4 11.Qxg4 Bxg4 12.h3 Be6 13.Be3 Bxe3 14.fxe3 Ke7 15.O-O Rhf8 16.Nb5 c6 17.Nd4 Nd7 18.a4 Rxf1+ 19.Rxf1 Rf8 20.Rxf8 Nxf8 21.Kf2 g6 22.Nf3 Nd7 23.b3 a5 24.Ke2 Nc5 25.e5 dxe5 26.Nxe5 Kd6 27.d4 Nd7 28.Nd3 Bf5 29.g4 Bxd3+ 30.Kxd3 b5 31.axb5 cxb5 32.e4 a4 33.bxa4 bxa4 34.h4 a3 35.Kc3 Nb6 36.Kb3 Nc4 37.c3 Nd2+ 38.Kxa3 Nxe4 39.c4 Black lost on time; and

ZahariSokolov - puthoor, 5 0 blitz, FICS, 2017: 1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+ Kxf7 5.Nxe5+ Nxe5 6.Qh5+ Kf8 7.Qxe5 d6 8.Qd5 Nf6 9.Qd3 Qd7 10.O-O Qg4 11.Nc3 Be6 12.h3 Qg6 13.Qf3 Ke7 14.Nb5 Bb6 15.d3 a6 16.Na3 Rhf8 17.Qe2 Rae8 18.Be3 Bxe3 19.Qxe3 Kd8 20.Qa7 Bxh3 21.g3 Bc8 22.Qd4 Bd7 23.Rfe1 c6 24.Nc4 Ng4 25.Nxd6 Re6 26.Nxb7+ Kc8 27.Nc5 Rd6 28.Qb4 Qh6 29.Nxd7 Qh2+ White resigned - Rick

10.0-0 Qg4 11.Qxg4

Escaping the queen swap with 11.Qd3?! looked a little silly, with the queen blocking the d-pawn. (Yes, see the ZahariSokolov - puthoor game, above - Rick)

11...Bxg4 12.h3 Be6 13.d3 d5 14.e5 Nd7 15.Kh1

15...g6

For a second, I thought I'd botched the game, as I missed that 15...Nxe5?? was a possibility. However, if that had been played, 16.Re1 Bd6? 17.Bf4! would then have been a strong line for White, winning the piece back.

16.Bh6+ Ke7 17.f4 Bd4 

I am not sure what is going on here. It looks like Black wants to attack White's pawn at e5, to make the advance of the "Jerome pawns" more difficult. Komodo 10 gives two interesting lines, both with the second player sacrificing the exchange and both leading to equality: 18.g4 Raf8!? 19.Bxf8 Rxf8 20.Ne2 Bb6 21.d4 c5 22.c3 and 18.g4 Be3 19.Nd1 Bb6 20.f5 gxf5 21.gxf5 Raf8!? 22.Bxf8 Rxf8 23.Nc3 Rxf5 24.Rxf5 Bxf5 25.Nxd5+ Ke6 26.Nxb6 Nxb6. - Rick

18.Bg5+? -++

The bishop was actually better placed on h6, where it would be hard for Black to remove it. On g5, it allows Black the chance of a bit of kingside counterplay by moving the h-pawn. 18.Rae1 would have been a better move, though Black would still have had a clear advantage. Unfortunately, after this move, I didn't really get back into the game. My 'trap' is starting to look not quite so good 😉.

18...Kf7 19.Rab1 h5 20.Ne2 Bb6 21.c3 

Hoping to get in d3-d4, but his opponent does not allow it.- Rick

21...Bf5 22.Rbd1 Nc5 23.Nc1 Ne6 24.Bh4 Be3 25.Ne2 c5 26.Bf2 Bxf2 27.Rxf2 d4 

White's extra pawns are blockaded. Slowly, Black makes his extra piece tell. There is a lot of serious strategy in this 1-minute game.- Rick

28.c4 h4 29.Kg1 a6 30.Kf1 b5 31.Ng1 Rab8 32.b3 bxc4 33.dxc4 a5 34.Nf3 a4 

35.Ng5+ Nxg5 36.fxg5 Ke6 37.Rf3 axb3 38.axb3 Rb7 39.Re1 Rhb8 40.Rf2 Rxb3 41.Ra2 Rb2 42.Ra6+ R8b6 43.Ra7 R2b3 44.Rh7 Rb1 45.Rxh4 Rxe1+ 46.Kxe1 

White still has two pawns for a piece, but Black has the dangerous passer.

46...Rb1+ 47.Kf2 Rb2+ 48.Kg3 Rb3+ 49.Kh2 Rb2 50.Rh7 Kxe5 51.Ra7 Kf4 52.Ra6 Ke3 53.Rd6 Rc2 54.Rf6 Rxc4 and White lost on time.

Wow.

Thanks, Cliff, for fighting the good fight - and then sharing it with us. - Rick