Jerome Gambit: One Last Bot (Part 2)

 

[continued from the previous post]

perrypawnpusher - Claire bot

SparkChess, 2023

11...Qh6 

The less exact retreat 11...Qd8 was seen in Some Loser - NN, 2005 (1-0, 37); and perrypawnpusher - Chess Titans, level 5/casual game, 2019 (1-0, 42).

That earlier game appeared in the games collection Unorthodox Chess (2005), under that author's pseudonym, Some Loser. 

There are alterative 11th moves for Black:

11...Qe7 12.Qd5 checkmate, the "Buyer's Regret" variation, perrypawnpusher - intssed, blitz, FICS, 2010;

11...Qg4 in perrypawnpusher - jgknight, blitz, FICS, 2011 (1-0, 68) 

11...Nf6 12.gxh4 Black resigned, in perrypawnpusher - mikelars, blitz, FICS 2011; 

11...Ne7 in perrypawnpusher - constipatedguru, blitz, FICS, 2017 (1-0, 20); and 

11...Qf6 in Kennedy - Chess Titans, casual game, 2022 (1-0, 48) and perrypawnpusher - IlToscano, Chess.com, 2016 (1-0, 18). (Hmm, don't think I have posted that last one, yet - Rick) 

12.Rxh2 

Played too quickly and giving up the advantage.

There were two better alternatives:

Capturing two pieces quickly with 12.Qd5+ Ke7 13.Qxc5+ Kf7 14.Rxh2 ; and

Harassing the enemy King, while eventually capturing two pieces: 12.Qe5+ Kc6 13.Qd5+ Kb6 14.Qb3+ Kc6 15.Nc3 a6 16.d4 Bxd4 17.Qd5+ Kb6 18.Qxd4+ c5 19.Nd5+ Ka7 20.Qxc5+ b6 21.Qd4 Qc6 22.Rxh2. 

12...Qe6 

Removing the Queen from danger and offering an exchange, which would be beneficial for Black.

An attempt to kick-the-queen with 12...g6 fell to 13.Qe5+ Kc6 14.Rxh6 in perrypawnpusher - JTIV, blitz, FICS, 2010 (1-0, 32), but I could simply have played 13.Rxh6, as the Black g-pawn would have  been pinned. 

13.e5+ Kc6 14.Qe4+ 

14...Kb6 

A lucky break for me. Instead, blocking the check with 14...d5 or 14...Qd5 was in order.

By the way, responding to 14...d5 with the cool en passant capture 15.exd6+ would have ended the game quickly with 15...Qxe4+ 16.Kf1 Bh3+ 17.Rxh3 Qf3+ 18.Ke1 Re8 checkmate. 

Fortunately, nobody has ever called me "cool".

15.d4 Be7 

The "Jerome pawns" are ready to be disruptive, and 15...Bxd4 16. Qxd4+ c5 17.Qd3 d5 would have been the proper response.

16.d5 Qg4 17.Nc3 c5 18.d6 Bd8 

19.Nd5+ Kb5 

There is no escape: 19...Kc7 20.Nc7+ Kb6 21.Nxa8+ Ka5 22.Bd2+ Kb5 23.a4+ Ka6 24.Qc4+ b5 25.Qxb5 checkmate.

20.a4+ Ka5 21.b4+ cxb4 22.Qxb4+ Ka6 23.Qb5 checkmate

Not again! (And Again)

Since the following Jerome Gambit game* (1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+) is identical to one that I played less than two weeks ago (see "(This time I saw it)"), I'm tempted to just copy the annotations from that game, or simply give a link to perrypawnpusher - thinan, blitz FICS, 2010.

Instead, I'll add a few diagrams, make a few references to the "Jerome Gamble" – How much of any of the refutations does my opponent actually know?? – and then let you get on your way.

perrypawnpusher  - javistas
blitz FICS, 2010

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

4...Kxf7 5.Nxe5+ Nxe5 6.Qh5+ Ke6

7.Qf5+ Kd6 8.f4 Qh4+

9.g3 Nf3+

10.Kd1 Qe7

11.Qd5 checkmate

(*Two weeks after I originally composed this post, I played the identical game perrypawnpusher - fortytwooz, blitz, FICS, 2010. When I thanked my opponent afterwards for the game, he cheerfully responded with a simple "lol". That's my third example of what might be called "The Buyer's Regret Variation.")

Charm

This is my third Jerome Gambit (1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+) game with jgknight. The first two were relatively long draws, one last year in 58 moves, and one this year in 49 moves. As they say, "third time is a charm," and I was successful this time. There is even a little bit of "charm" in the game, although if my opponent wanted to preface with the word "lucky" I would not object.


perrypawnpusher - jgknight
blitz, FICS, 2011

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

4...Kxf7 5.Nxe5+ Nxe5 6.Qh5+ Ke6

Black's King bravely protects the Knight, giving up any thoughts of "castling-by-hand", and challenges White to justify his two piece sacrifice.

7.Qf5+ Kd6 8.f4 Qh4+

Black is headed for what I've referred to as the "Nibs Defense". See "Repairing A Variation" parts 1, 2, 3 and 4 for some background.

9.g3 Nf3+ 10.Kf1

When I posted my game with this line against JTIV last year, I wrote 

This is my current choice, although it is not clear why. The updated New Year's Database [now The Database] has 4 games (other than the current one) with this move, with White scoring 50%. I've scored 33% in 3 games (until now).

The Database has 15 games with the alternative, 10.Kd1, with White scoring 73%. That's in part to me scoring 100% in 6 of those games; but 3 of those wins are identical 11-movers with the "Buyer's Regret Variation".

Updating: The Database now has 9 games with 10.Kf1, with White scoring 78% (me, 75%). It still has 15 games with 10.Kd1, with White scoring 73% (me 100%).

I am still not sure which 10th move I will play next time, especially since I concluded at one point that 10.Kf1 "is not an improvement" over 10.Kd1.

10...Nxh2+

Tempting, as it is another forward-going, attacking move, but it probably only keeps the game in balance.

11.Kg2 Qg4

I smiled when I saw this move. I've recommended it a couple of times on this blog, when either Black "holds onto his advantage" or the game is "about even". Today I would go with the latter assessment.

The move is not mine, however, but the recommendation of Some Loser, the author of Unorthodox Chess (2005), who has occasionally played the Jerome Gambit. (Really. Read my book review.)

12.Qd5+ Ke7 13.Qxc5+ Ke8

Black's moves are interesting: he cannot protect his c-pawn with 13...Kd8, as 14.Qf8 checkmate would follow.

Yet, his best move would have been 13...d6, encouraging White to take the c-pawn, as after 14.Qxc7+ Bd7 15.Rxh2 (what else?) White's King would be defenseless against Black's Queen – and other pieces: 15...Qe2+ 16.Kg1 Qe1+ 17.Kg2 Qxe4+ 18.Kf1 Qf3+ 19.Ke1 Rc8!?. Not only is White's Queen in trouble, Black threatens the deadly 20...Kf8!

Of course, after 13...d6 14.Qf2! White would secure his King and the game would be in balance after 14...Nf6 15.Rxh2 Nxe4 16.Qe3 Qe6.

14.Rxh2

First 14.Qe3 was a bit more prudent.

14...d6 15.Qe3 Nf6

For the most part, the excitement has died down. Both Kings are relatively safe, and White is ahead by a pawn (a reasonable outcome, having faced down a refutation).

16.Nc3 c6

After the game Fritz 10 censured this move. Can you see why?

17.b3

I was looking for development, and decided to put my Bishop on the a1-h8 diagonal.

I totally overlooked 17.f5!? which threatens to trap Black's Queen. Black has a defense, of course, in 17...g5, but it costs a couple of pawns, say, after 18.Rh6 Rf8 19.d3 Rf7 20.Qxg5 and, further, Black's d-pawn is at risk, as well.

17...h5 18.Rh4

I almost always see these things a move too late.

18...Qe6 19.d3 Ng4 20.Qe2 Rh6 21.Bb2 d5

White has been more effective in rushing his troops to back up their footsoldiers.

22.exd5

Personal failing: I am always willing to "let the air out of the position" if I am even the tiniest bit of material ahead.

22...Qxe2+ 23.Nxe2 Ne3+ 24.Kf3 Nxd5

Messier, but stronger, was 24...Nxc2.

So, what is going on here?

White is better developed (including his King), is a pawn ahead, and hopes to pick up one or two more pawns.

It is essential that Black get his pieces out, cause as much disruption as possible, and keep his eyes on the possibility of transitioning to the welcoming arms of the drawish (even when down pawns) Bishops-of-opposite-colors endgame. jgknight fights, but the battle is not his today.

25.Bxg7 Bg4+ 26.Kf2 Re6 27.Be5 Bxe2 28.Kxe2 Nc3+ 29.Kf3 Nd5 30.Rxh5 Kd7

31.Rh7+ Re7 32.Rah1 Rxh7 33.Rxh7+ Ke6 34.Rxb7 Rg8 35.Rxa7 Ne7 36.g4 Ng6 37.d4 Nh4+ 38.Ke4 Rxg4 39.Rg7 Rxg7 40.Bxg7

40...Ng6 41.a4 Kf7 42.Be5 Ke7 43.a5 Kd7 44.f5 Ne7 45.c4 Nc8 46.f6 Ke6 47.d5+ cxd5+ 48.cxd5+ Kf7 49.Kf5 Na7 50.b4 Nb5 51.a6 Na7 52.Bd4 Nb5

53.a7 Nxd4+ 54.Ke4 Nb5 55.a8Q Nd6+ 56.Kf3 Kxf6 57.Qc6 Ke5 58.b5 Nf5 59.Qe6+ Kd4 60.Qxf5 Kc5 61.b6 Kxb6 62.Qe6+ Kc7 63.Qc6+ Kd8 64.Ke4 Ke7 65.Kf5 Kd8 66.Qb7 Ke8 67.Ke6 Kf8 68.Qf7 checkmate

Notice that my fingers never leave my hand...

I just finished playing a Jerome Gambit game where my opponent responded to my every move quickly and decisively. He was zooming down one of the refutation lines that had given me worries in the past, and he was just about at the point where he would make me disappear – when something unexpected happened.

perrypawnpusher - NN
blitz, FICS, 2011

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

4...Kxf7 5.Nxe5+ Nxe5 6.Qh5+ Ke6

7.Qf5+ Kd6 8.f4 Qh4+

Wow! I half-expected 8...Nf3+ as I saw in "Vortex". That's merely one of many ways to smash the Jerome.

The text move is something that I most recently wrestled with last year in "Nice Kitty..." – how close will my opponent come to playing out the full refutation in this line?

9.g3 Nf3+

Closer...

10.Kf1

I've also played 10.Kd1 and entered what I call "the buyer's regret variation".

10...Qh3+

My first thought was: Hey, I've never seen this move before...

Then, Black resigned and I suddenly realized why.

(He was such a good sport about the oversight, I saw no reason to use his handle, but opted for "NN".)

Wonderland

Sometimes a defender, having wandered into the Jerome Gambit (1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+), feels out of place, as things are not quite what was expected, as if it's a trip to Wonderland...

perrypawnpusher - anelante
blitz, FICS, 2012

1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+ 

4...Kxf7 5.Nxe5+ Nxe5 6.Qh5+ Ke6 7.Qf5+ Ke7 

The earliest example I have of this move (which might as well be called the "buyer's regret" variation) in The Database (with the addition of my game, White has scored 89% in 23 games) is SirOops - mentalGIANT, standard, FICS, 2001 (1-0, 25); although it certainly must have been played over-the-board earlier.

8.Qxe5+ Kf7 9.Qxc5

Surprisingly, I missed playing my usual "nudge" 9.Qd5+, driving the King to the back line to interfere with the Rook. 

9...Nf6 10.Nc3 d6 

White is up two pawns, and it is hard to see Black's compensation.

11.Qe3 

Probably not best, given that it leaves White's Queen and King on the same file that Black's Rook can quickly go to. Nothing bad happens, as a result, but this reinforces the point that White should have "nudged" when he had the chance.

11...Rf8

Black prudently castles-by-hand. After the game Rybka suggested the wild 11...d5 12.Qf4 g5!? 13.Qg3 (13.Qxg5 Rg8 14.Qf4 Rg4 15.Qe5 dxe4 16.0-0) 13...dxe4  to reduce White's advantage.

12.0-0 Kg8 13.d4 b6 14.f4 Bb7 15.e5 Re8 

16.Qd3 Ng4 17.Qg3 

Or 17.h3 as Rybka later suggested.

17...Bc8 18.f5 dxe5 19.Qxg4 exd4 

20.Bh6 

Missing the better 20.Ne4

20...Qf6 21.Bg5 Qc6 22.Nd1 Black forfeited on time

Jerome Gambit: One Last Bot (Part 1)

One more Jerome Gambit (1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+) game against a bot, this time played online at SparkChess.com.

The line features a go-for-the-throat response by Black.

perrypawnpusher - Claire bot

Spark Chess, 2023

1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+ 

4...Kxf7 5.Nxe5+ Nxe5 6.Qh5+ Ke6

7.Qf5+ Kd6 8.f4 Qh4+ 

Black counter-attacks.

As I mentioned in the post "Jerome Gambit Tournament: Chapter II", what we have here is

An interesting line that I refer to as "His Nib's Defense," which made its first appearance in the American Chess Magazine of June 1899 in a comical article lampooning a new pastime, chess by phone. 

9.g3 Nf3+ 

This move surprised me, especially played by a bot. There is no way that it could be in Claire bot's opening book, is there?

Stockfish 15.1 likes the Knight check, assessing that (at 31 ply) Black is almost 3 pawns better.

Practical results, however, show a different story.

There are 206 games with this position in The Database, with White scoring 57%.

For that matter, The Database has 25 of my games with this position, and I have scored 88%.

Verdict: It's complicated, and familiarity helps.

10.Kf1 

Stockfish 16, at 46 ply, prefers 10.Kd1 over 10.Kf1 by .17 of a pawn: not much. 

Consulting The Database again, there are 128 games with 10.Kd1, with White scoring 63%.

This compares with 43 games with 10.Kf1, with White scoring 71%.

So, despite the computer's assessment of winning for Black, in practical online club level play, White does well with either 10.Kd1 or 10.Kf1; a bit better with 10.Kf1.

I have played 7 games with 10.Kd1 and won them all, including 3 games with the "Buyer's Regret variation", 10...Qe7 11.Qd5# ; while playing 18 games with 10.Kf1, scoring 83%. 

Again: It's complicated, and familiarity helps.

10...Nxh2+

Another surprise.

I had expected 10...Qf6, as in perrypawnpusher - james042665, Chess.com, 2008 (0-1, 18); perrypawnpusher - HarlemKnight, 6 13 blitz, FICS, 2014 (0-1, 24); and perrypawnpusher - vasbur, Chess.com, 2018 (1-0, 34). 

I wouldn't have been totally put off to see 10...Qh6 as in perrypawnpusher - Temmo, Jerome Gambit 3 thematic tournament, ChessWorld.net, 2008 (0-1, 43). 

11.Kg2 

Despite all the attention that it has attracted, White's King is the safer one.

In the meantime, Black stands to lose his advanced Knight or his developed Bishop, and his King is sitting uneasy.  

Still, at 35 ply, Stockfish 15.1 still sees Black as about 9/10 of a pawn better.

[to be continued]

Nice Kitty...

The following game is another visit with a Jerome Gambit (1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+) defense / refutation that has sharp teeth and claws – at least in theory. When entering this jungle, I always feel a mixture of excitement and dread as I wonder what will happen this time? See "Closer... But not there yet" and "Jerome Gambit Tournament: Chapter XVI".

perrypawnpusher  - JTIV
blitz, FICS, 2010

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

4...Kxf7

My opponent played the Jerome Gambit successfully 3 days before our game: 5.d4 exd4 6.e5 h6 7.c3 dxc3 8.Qb3+ Ke8 9.Nxc3 d6 10.exd6 Qxd6 11.0-0 b6 12.Re1+ Kd8 13.Qf7 Qf6 14.Re8, checkmate, JTIV - NyteFork, FICS, 2010.

5.Nxe5+ Nxe5 6.Qh5+ Ke6

7.Qf5+ Kd6 8.f4 Qh4+

Whenever I see this move, I suspect that someone has been reading this blog... (Not that that's a bad thing.)

9.g3 Nf3+

And this move almost confirms it...

Interestingly enough, the updated New Year's Database has 25 games (other than the current one) with this position, with White scoring 58%. My own experience with the line is scoring 78% in 9 games.

10.Kf1

This is my current choice, although it is not clear why. The updated New Year's Database has 4 games (other than the current one) with this move, with White scoring 50%.  I've scored 33% in 3 games (until now).

The Database has 15 games with the alternative, 10.Kd1, with White scoring 73%. That's in part to me scoring 100% in 6 of those games; but 3 of those wins are identical 11-movers with the "Buyer's Regret Variation".

10.Kf1 Nxh2+

This is an okay followup, although the "magic move" is 10...Ne7. Black has had success with two other moves, as well:

10...Qf6 11.Qd5+ Ke7 12.Qxc5+ Kd8 13.Kg2 d6 14.Qd5 Ne7 15.Qd3 Nd4 16.e5 dxe5 17.c3 Bf5 18.fxe5 Qc6+ White resigned, perrypawnpusher-james042665/Chess.com 2008 ;

10...Qh6 11.Qd5+ Ke7 12.Qxc5+ Qd6 13.Qxd6+ cxd6 14.c3 g5 15.Kf2 g4 16.h3 d5 17.hxg4 dxe4 18.d3 d5 19.g5 h6 20.dxe4 dxe4 21.Nd2 Nxd2 22.Bxd2 Bg4 23.Be3 Ke6 24.Rh4 Kf5 25.Bd4 Rh7 26.Rah1 h5 27.Ke3 Ne7 28.R4h2 Nd5+ 29.Kd2 b6 30.Re1 Re8 31.Be3 Rd8 32.Kc2 Rc7 33.a3 Rcd7 34.Bd4 b5 35.Be5 Rc8 36.Kb1 a5 37.g6 b4 38.axb4 axb4 39.g7 b3 40.Bd4 Kg6 41.Kc1 Kf5 42.Rhh1 Nb4 43.Kd2 Rxd4+ White resigned, perrypawnpusher - Temmo, JG3 thematic, ChessWorld, 2008.

11.Kg2 Qh6

Best is the tricky 11...Qg4 and Black holds onto his advantage.

Worst is 11...Qe7 12.Qd5 checkmate, perrypawnpusher - intssed, blitz, FICS, 2010.

12.Rxh2

This move allows Black the chance to slip out of his difficulties. The straight-forward 12.Qd5+ Ke7 13.Qxc5+ Qd6 14.Qxd6+ Kxd6 15.Rxh2 gives White the edge.

12...g6

Necessary was 12...Qe6. The text move turns the game over to White.

13.Qe5+

White's Queen escapes attack with check. Even stronger was 13.Rxh6, since the Black g-pawn would then be pinned.

13...Kc6 14.Rxh6 Nxh6 15.Qxh8

15...Ng4 16.d4 Be7 17.d5+ Kd6 18.Qd4 c6 19.e5+ Kc7 20.d6+

20...Bxd6 21.exd6+ Kb8 22.Nc3 b6 23.a4 Ba6 24.a5 b5 25.b4 Kb7

26.Ne4 Re8 27.Nc5+ Kc8 28.Kf3 Nh2+ 29.Kg2 Re2+ 30.Kh3 h5 31.Qh8+ Re8 32.Qxe8 checkmate