Showing posts with label Ghandybh. Show all posts
Showing posts with label Ghandybh. Show all posts

Saturday, June 13, 2015

Zeromov Gambit!?


Another recent discovery on the internet is a Serbian language video examining the famous Jerome Gambit game Amateur - Blackburne, London, 1885.

Having shown the first moves of the game - 1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+ Kxf7 5.Nxe5+ Nxe5 6.Qh5+ the presenter takes an aside to look at 6...Ke6, before continuing with Blackburne's 6...g6. There, White executes a checkmate in 17 moves - ah, well, but, anyhow, back to Blackburne...

I had to chuckle. Knowingly or not, the chess player had just shown the game Ghandybh  - ishahir, Chess.com, 2009, presented on this blog in the post "Teach / Learn" about 5 1/2 years ago.

For that matter, there had been updates in "Jerome Gambit on YouTube!?" (the game Ghandybh  - ishahir, Chess.com, 2009) and "Jerome Gambit: Winning Strategy #1".

But what really shocked me about the Serbian video was the reference to the opening as the Jerome Gambit, or the Zeromov Gambit.

I had not heard that one before. Looks like I have more research ahead of me!

Saturday, May 31, 2014

Another Provocation


In my February 11, 2010 blog post, "Provoked", I noted about a chessfriend and Jerome Gambit Gemeinde member,

Welton Vaz de Souza is a nice guy. 

Sure, from time-to-time he crushes opponents with the Jerome Gambit (1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+), but he's a nice guy.


Suggestion: don't provoke him in a game of chess.


I was reminded of that when I ran into the following game.


Richad  - NeoNunes

blitz, FICS, 2013

1.e4 e5 2.Nf3 Nc6 3.Bc4 Nd4 




The Blackburne Shilling Gambit.


4.0-0 


This is one of the orthodox treatments of the BSG, along with 4.Nxd4 or 4.c3. Not everyone wants to play the Blackburne Shilling Jerome Gambit, 4.Bxf7+!?


4...b5 


This move, however, proves to be too much provocation. White decides on a Delayed Blackburne Shilling Jerome Gambit.


5.Bxf7+ Kxf7 6.Nxe5+


As I noted in that older post,


It turns out that 5.Bxf7+ is so strong (or 4...b5 so ill-advised) that White could simply grab back a piece here with the alternative 6.Nxd4. If Black responds 6...exd4 then White has 7.Qh5+ and a few checks later the Rook on a8 will be captured.

After 6.Nxd4 Nf6, White retreats his Knight with 7.Nf3 and is better.


6...Ke6


Instead, 6...Ke8 7.Qh5+ g6 8.Nxg6 was seen in Ghandybh - BoogieKnights, Chess.com, 2010 (1-0, 16) and mikelovell - rogerlovell, FICS, 2012 (1-0, 44).


The alternative, Black's strongest move, 6...Ke7, was seen in Rub - Wall,B, Palo Alto, 1989, (0-1, 12), mukund - jantonacci, FICS, 2013 (1-0, 10) and Chabus - AlfonsoX, FICS, 2013 (1-0, 14).


7.f4 


Stronger was 7.c3, as in mukund - jantonacci, FICS, 2013 (1-0, 10) and Chabus - AlfonsoX, FICS, 2013 (1-0, 14). 


7...Nf6 8.c3 Nc6 9.d4 


 9...Nxe5

Courting danger.


10.fxe5 c6 11.exf6 gxf6 12.Qg4+ Kd6 13.Bf4+ Ke7




14.Nd2


This move is good enough, but White misses the hilarious 14.Be5!?

when Black cannot capture the Bishop, as he will be checkmated; so the cheeky prelate can go on and capture at f6.

Black's next move hangs a piece, but White is in no hurry: he builds his attack and breaks through.


14...Bg7 15.Rae1 Rg8 16.e5 d5 17.exf6+ Kf7 18.Qh5+ Kf8 19.fxg7+ Rxg7 20.Bh6+ Kg8 21.Bxg7 Kxg7 22.Rf7+ Kg8 23.Qxh7 checkmate




(May 2014 was the highest visited month in the history of this blog, going back to 2008. Many, many thanks to readers - and please feel free to continue to visit. - Rick)

Monday, February 24, 2014

Jerome Gambit: A Battle (Part 1)




After a series of Jerome Gambit games where either my opponent or I (once, both) blundered away our Queens (I will spare Readers, but include the games in The Database), I played the following exciting adventure.

As the nursery rhyme goes (the two characters can also be found in Lewis Carroll's chess-themed Through the Looking-Glas and What Alice found There)

Tweedledum and Tweedledee
    Agreed to have a battle...  

After playing almost 500 Jerome Gambit games, I again found myself in quite a battle. 

perrypawnpusher  - vermifugo
blitz, FICS, 2014

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


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


Black was looking to hold on to both extra pieces.

7.Qf5+ Kd6 8.f4 g6 

This is not Black's best play, even if it conjurs up visions of the venerable Blackburne Defense (6...g6). Yet it leads to largely unexplored territory, and play can become very uncomfortable for White if he is not careful. It is worth exploring more in depth.

9.Qxe5+ Kc6 10.Qxh8 

At this point, the game UNPREDICTABLE - zaboulef, blitz, FICS, 2010 ended with Black's resignation. 

The earliest game (of four) that I have with this variation in The Database is perrypawnpusher - spontex, blitz, FICS, 2009 (½-½, 59), where White bypassed capturing the Rook and went after the Black King, instead, with 10.Qd5+. As I noted in "The Opponent",
It was better to take the Rook, although play remains complicated, if in White's favor.
10...d6

The more aggressive 10...Qh4+ (also reminiscent of Blackburne's play in his 1885 game) was seen in Ghandybh - DVBLTTN, Online chess, 2009 (1-0, 17). As I noted,
The problem is that Black's Kingside will fall apart faster.
Best for Black seems to be 10...Qe7.

11.Qxh7

What to do?

I have found my Queen at h8, early in the game after grabbing a Rook, about 25 times, and I have always feared it getting trapped, and eventually captured (see the introductory note to this game, although they refer to different lines of play). My first thought has always been to get Her Majesty out of confined spaces!

One lesson I should have learned from a closer look at Blackburne's Defense is that White's Queen often escapes through the middle of the board (after a well-timed Qd8) - or after blocking the diagonal of Black's dark-squared Bishop with d2-d4.

In the current game, White's best was 11.d4.

Yet, that is not the end of the discussion, as Black has several possible replies, including the straight-forward 11...Bb6, seen in UNPREDICTABLE - farhadk, blitz, FICS, 2009 (1-0, 21).

What is more fun to look at is 11.d4! Bh3!?, when Black continues to play in Blackburne style. Can White actually then take the Bishop with 12.gxh3 ?  If he does, he can face, among other things, the energetic counter-attack 12...Qh4+ 13.Kd1 Nf6!? 14.Qxa8 Qh5+




Analysis diagram





Here, Black has sacrificed a couple of Rooks for a couple of pawns, but it is hard for White to escape a draw - while avoiding checkmate (remember, this is a blitz game). 

Analyze with me (and Houdini): 15.Ke1 (Not 15.Kd2 Nxe4+ 16.Kd3 Qf3+ 17.Be3 [17.Kc4 d5#] 17...Nf2+ 18.Kd2 [18.Kc4 Qe2+ 19.Kb3 Qb5+ 20.Kc3 Qb4#; 18.Kc3 Qxe3+ 19.Kc4 d5#] 18...Ne4+ 19.Kd3 draw) 15...Qh4+ 16.Ke2! (16.Kf1 Bxd4 17.Ke2 Qf2+ 18.Kd3 Qf3+ 19.Kxd4 Qxe4+ 20.Kc3 Nd5+ 21.Kd2 Qg2+ 22.Kd3 Qf3+ 23.Kd2 draw; 16.Kd2 Qf2+ 17.Kd1 [17.Kc3 Qxd4+ 18.Kb3 Qb4#; 17.Kd3 Qxd4+ 18.Ke2 Qf2+ 19.Kd1 Qf3+ 20.Kd2 Nxe4+ 21.Ke1 Bf2+ 22.Kf1 Bg3+ 23.Kg1 Qf2#; 16.Kd1 Qh5+ 17.Ke1 Qh4+ 18.Ke2 as in the main line17...Qf3+ 18.Ke1 [18.Kd2 Nxe4+ 19.Ke1 Qxh1+ 20.Ke2 Qg2+ 21.Kd3 Qf1+ 22.Ke3 Qe1+ 23.Kf3 Qd1+ 24.Kxe4 Qxd4+ 25.Kf3 Qf2+ 26.Kg4 Qg2+ 27.Kh4 Bf2#] 18...Qxh1+ 19.Ke2 [19.Kf2 Nxe4+ 20.Ke2 Qg2+ 21.Kd3 Qf1+ 22.Ke3 Qf2+ 23.Kd3 Qxd4+ 24.Ke2 Qf2+ 25.Kd3 Qf1+ 26.Kxe4 Qe2+ 27.Be3 Qxe3#] 19...Qxh2+ 20.Kd1 Qg1+ 21.Ke2 Qg2+ 22.Ke1 Qxe4+ 23.Kd1 Qf3+ 24.Ke1 Bxd4 25.c4 a6-+) 16...Qh5+ 17.Kd3! Qf3+ 18.Be3! Qxe4+ 19.Kd2 (Not as strong is 19.Ke2 Bxd4 20.Re1 Qxe3+ 21.Kd1 Qf3+ 22.Kc1 Qxf4+ 23.Nd2 Be3 24.Rxe3 Qxe3 25.c4 Qe1+ 26.Kc2 Qxa1 27.Qf8 Qxa2 28.Qxf6 Qa4+ 29.Kc3 Qa5+ 30.Kd3 Qh5 31.Qf3+ Qxf3+ 32.Nxf3) 19...Qg2+ (19...Qxh1? 20.Nc3!? Qxa1 21.dxc5 Qh1 22.Qxa7 Qg2+ 23.Kc1 Qh1+ 24.Nd1) 20.Kc3 Qe2 21.b4! Qxe3+ 22.Kb2 Bxd4+ 23.c3 Qe2+ 24.Kb3 Qe6+ 25.Ka3 Qe3 26.Ka4! Nd5 27.Qg8 b5+ 28.Ka3 Bxc3 29.Qa8+ Kd7 30.Qxd5 Bxa1+ 31.Qb3 and it looks like White may survive. What a mess!

All of the above suggests that White should answer 11.d4! Bh3!? with the sober 12.0-0!, when his King will be safe, and all of Black's problems - down the exchange and two pawns, unsafe King, two Bishops hanging - will remain. The computer suggests 12...Nf6 13.Qxd8 Bxd4+ 14.Kh1 Rxd8 15.gxh3 Nxe4, but what fun is there in that?

11...Bg4?! 

Development toward the enemy King, but more to the point would have been 11...Qe8! and White would have to give back a couple of pawns to free his Queen - again, while negotiating some very complex play (without Houdini's help, which I had after the game) - 12.e5 (12.d3 Nf6 13.Qh4 Nxe4!? 14.Kf1 Bf2 [14...Nf2 15.b4 g5 16.Qh7 Bg4 17.Nc3 Bxb4 18.Qg7 Nxh1 19.Rb1 a5 20.a3 Bc5 21.d4 unclear, perhaps even] 15.g3 Bc5 16.h3 g5 17.fxg5 Qf7+ 18.Bf4 Qd5 19.Rh2 Bf5 20.g4 Bg6 21.dxe4 Qxe4 22.Nd2 Qxf4+ 23.Rf2 Qe5 24.Nf3 Qxb2 25.Re1 Qxa2 26.Rfe2 b5 even, whew!) 12...dxe5 13.f5 Bxf5 14.Qh4± but White is still uncomfortable, say after 14...e4 15.b3 Rd8 (15...Qe5 16.Nc3 Nf6 17.Bb2 Ng4 18.0-0-0 Bf2 [18...Nf2 19.d4] 19.Qg5 Bd4 20.Rhf1 Nxh2 21.Rf4) 16.Bb2 Qd7! 17.Qf4 g5 18.Qe5 b6 19.b4 Bxb4 20.0-0 Bc5+ 21.Kh1. 


[to be continued]

Thursday, May 30, 2013

Jerome Gambit: Winning Strategy #1


The following game illustrates the Number One "winning strategy" for the Jerome Gambiteer, what I like to call "Lose Slowly"; to which my son Jon, no stranger to the Jerome added, "Give your opponent time to make a mistake."


perrypawnpusher - trombose

blitz, FICS, 2013

1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+ 




4...Kxf7 5.Nxe5+ Nxe5 6.Qh5+ Ke6 7.Qf5+ Kd6 8.f4 Nc6




This is one of a number of ways for Black to return a piece (see "Chances", "Follow Your Friends", "No Need to Get a Big Head" and "Charlatan" for others), that, while not "best", still leaves him with an advantage.


9.Qd5+ 


This move is an improvement over 9.e5+,  which was seen in Superpippo - MattMeister, FICS, 2002 (1-0, 60) and Ghandybh - ishahir, Chess.com, 2009 (1-0, 17), the only precedents for 8...Nc6 in The Database.


9...Ke7 10.Qxc5+ d6 11.Qe3 




After the game, Houdini 3 suggested 11.Qf2 as another way of retreating the Queen.


11...Nf6 12.O-O Rf8 13.e5 


Probably premature.


13...Ng4 14.exd6+ Kxd6 




15.Qa3+ Kd7 16.Nc3 


I tend to get lost in these nebulous positions. After the game, Houdini 3 suggested 16.h3 Nf6 17.Rf3!? with the kind of attacking ideas your can find in Bill Wall's games.


16...Qe7 


A Queen exchange will blunt White's "attack", and there is little that I can do - so I try to ignore it. Curiously, for a while, so does my opponent.


17.Qa4 Qb4 18.h3 Nf6 19.b3 Qd4+ 20.Kh1 


As will be seen, 20.Kh2 was more prudent.


20...a6 21.Bb2 Qxa4 22.Nxa4 Nb4 




23.c4 


More to the point was the tactical skirmish, 23.Ba3 a5 24.c3 Ne4 25.Rfe1 Rxf4 26.cxb4 axb4 27.Bxb4 Nf2+ 28.Kg1 Nxh3+ 29.gxh3

Rxb4, when Black would have a slight edge.

23...b6 24 d4 Nd3 25.Bc3 Nh5 26.Bd2 


Focusing on trapping Black's Knight, I overlooked the loss of the exchange.


26...Ng3+ 27.Kg1 Nxf1 28.Rxf1 c5 




Instead, 28...Bb7, which must have been on Black's mind (perhaps he was spending too much time wondering what I was up to, or how he was going to retrieve his Knight), would have steadied his winning game - a Rook up.


29.Rf3


Correct was 29.Nxb6+, as White actually needs his Bishop to help his own Knight escape, when Black's edge would be minimal.


Both my opponent and I were suffering from "nerves" at this point.


29...Nb4 30.Bxb4 


Same criticism as above:  30.Nxb6+ was a better move, and winning.


30...cxb4 31.Nxb6+ Black resigned




Black must have been discouraged, seeing himself drop the exchange, as I had, earlier. The fact is, after 31...Kc6 32.Nxa8 Kb7 (or 32...Bf5) Black will gather in White's Knight, and, with a piece for three pawns, the game would be dynamically equal.


Monday, May 16, 2011

Peaceful, Not Frenetic

I have been trying for peaceful, not frenetic, Jerome Gambit games, slowly and steadily progressing toward victory. The following game is a good model for me. White gathers "small advantages" and calmly moves toward victory. But then...

Petasluk - krausepeter
blitz, FICS, 2011

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


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


7.f4 g6

Black mixes up defensive systems -- or something. The Database shows 9 wins for White against 0 losses prior to this game. 7...Qf6 or 6...d6 would have been fine for Black.

8.Qxe5+

Here Black resigned in Petasluk - Meshu, FICS. 2011, played about a week earlier.

Superpippo has played, and gotten away with 8.f5+ in Superpippo-vacaboja, FICS, 2002 (1-0, 37) and 8.Qh3+ in Superpippo-ustad, FICS, 2002 (1-0, 34), but those moves are not to be recommended.

8...Kf7 9.Qd5+


There are nudges and then there are nudges... Actually, 9.Qxh8 was a little better, as in Ghandybh-duboak, Chess.com, 2009 (1-0, 16).


9.Qxc5 was also playable, as in grobnic - vallabhan, GameKnot.com, 2004 (1-0, 24).

9...Ke8 10.Qxc5 d6 11.Qd4 Nf6 12.0-0 Qe7


White is a comfortable two pawns up. There is no need for an immediate "bashi-bazouk attack", as calm development will do.

13.Nc3 Rf8 14.d3 c5 15.Qe3 Ng4


If anyone, Black could heed the advice not to attack prematurely.

16.Qf3 h5 17.h3 Ne5 18.Qg3 h4 19.Qe3 Nc6 20.Bd2 Be6 21.Rae1 Kd7

Both sides have developed, and Black's King has found some shelter, but the story remains the same: a comfortable advantage for White.

22.a3 Rg8 23.e5 d5

Black keeps the position closed, but now White has a protected passed pawn. For now the e-pawn is blockaded, but it remains another advantage to be realized in the future.

24.Qf2

Setting his eyes on the Black pawn at h4.

Instead, Rybka 3 recommended for White the involved pawn sacrifice on the Queenside, 24.Nb5 b6 25.c4 hoping to open attacking lines against Black's King.

It's not hard to side with Petasluk's choice. Besides, he is rated higher than his opponent, and might want to let him come up with the "brilliant" ideas.

24...Raf8 25.Nd1

Perhaps White was thinking of following this up with c2-c3 and d3-d4, erecting a center wall?

In any event, danger for him is brewing on the Kingside, and White needs to strike first, as Rybka 3 points out: 25.b4 g5 26.bxc5 gxf4 27.Rb1 Kc8 28.Nb5 Bxh3 29.Nd6+ Kb8 30.Rxb7+ Qxb7 31.Nxb7 Rxg2+ 32.Qxg2 Bxg2 33.Rb1 Bh3 34.Nd6+ Ka8 35.Nb5 a6 36.Nc7+ Ka7 37.Kh2 Bg4 38.Nxd5 Nxe5 39.Bxf4 Nf3+ 40.Kg2 Whew! That's a lot to figure out in a blitz game.

25...g5

This is what Black has been aiming for. Now White has to counter-punch and mix it up; there is no more "peaceful".

26.c4 gxf4 27.Bxf4 Bxh3 28.e6+ Kd8



29.Ne3 Bxg2 30.Nxd5

In the smoke of the battle, White goes astray. Clearly time was short for him, and attacking the enemy Queen is in general a good strategy, but here there was a discovered check to deal with. 

Rybka 3's suggestion lets the air out of the position, almost like each army consuming the other, with a draw as the result: 30.Nxg2 h3 31.Qe3 hxg2 32.Rf2 Qh7 33.Rxg2 Rxg2+ 34.Kxg2 dxc4 35.dxc4 Ke8 36.Bg3 Qc2+ 37.Qe2 Qxe2+ 38.Rxe2 Rg8 39.Kh3 Nd4 40.Re4 Rg6 41.e7 Nf3 42.Bf4 Ng5+ 43.Bxg5 Rxg5.

Both White's steady peaceful build-up and Black's savagery in counter-attacking would thus have been "rewarded".30...Bxd5+ 31.Kh2 Rg2+ 32.Qxg2 Bxg2 33.Kxg2 Rg8+ 34.Kh3 Nd4 35.Be3 Ke8

Here White forfeited on time.

Saturday, August 21, 2010

One Little Pawn

The Jerome Gambit (1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+) is one of those attack-or-die kind of openings (which too often becomes an attack-and-die affair). If the attack is countered, White may not have much left for his effort and valor.

perrypawnpusher - irhall
blitz 4 10, FICS, 2010

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 Nf6


I've seen this position 20 times, scoring 75% (14-4-2). That's not bad, but it's about 10% below my total Jerome and Jerome-ish score.

10.0-0 Ne5

Transferring the Knight closer to my Kingside, where it will annoy my Queen. It was better to castle-by-hand with 10...Kf7, ...Re8 and ...Kf8.

11.d4

Probably better than 11.f4 as in perrypawnpusher - mikelars, blitz, FICS, 2010 (1-0, 26), and perrypawnpusher - Kotimatka, blitz, FICS, 2010 (1-0, 19).

11...Neg4 12.Qf3

White had an even game and went on to win with the slightly better 12.Qb3 b6 13.h3 Nh6 14.e5 dxe5 15.dxe5 Nfg8 16.Qf3 Be6 17.Rd1 Qc8 18.Nc3 Ne7 19.Nb5 Rf8 20.Qxa8 Kf7 21.Qxc8 Rxc8 22.Bxh6 gxh6 23.Nxa7 Rg8 24.Kh2 Nf5 25.g4 Ne7 26.Nb5 h5 27.Nxc7 hxg4 28.Rd6 Bf5 29.Rf6+ Kg7 30.hxg4 Bg6 31.Rd1 Bxc2 32.Rd7 Black resigned, Ghandybh - Noel1942, Chess.com, 2009.

12...Bd7


13.h3 Nh6 14.Bxh6 gxh6


The Knight's adventures are at an end.

15.Nc3 Qe7

White has an even game, but no more than that. It was important for Black to get something going before the first player can take advantage of Black's open Kingside. Both 15...Rg8 and 15...Bc6 come to mind. 

16.e5 dxe5 17.dxe5 Bc6


Black protects his Queenside, gives his Bishop better development, and returns the gambit piece.

18.Qxf6 Qxf6 19.exf6 Rg8


It seems that for all his attacking efforts, White has achieved only one little pawn on f6, something that Black can blockade with his King and then capture. He does have a winning endgame advantage on the Kingside, but he faces a coordinated defense from Black's Bishop and Rook.

20.Rfe1+ Kd8


Suddenly, the future looked brighter for me.

After 20...Kf7 21.Ne4 Rae8 22.Ng3 Kxf6 Black can fight on. 

21.g3

It turns out that this defensive move is not necessary: White can play instead 21.Rad1+ Kc8 22.Nd5

21...Rg6

Picking on the poor defenseless little pawn. After the game Rybka suggested, instead, 21...Rf8, although White still has the familiar 22.Rad1+ Kc8 23.Nd5 with advantage.

22.Rad1+ Kc8 23.f7


23...Rf6

Capitulating, but 23...b6  can be met by either 24.Rd8+, 24.Re8+ or 24.f8/Q+.

24.Re8+ Bxe8 25.fxe8Q checkmate




Saturday, August 7, 2010

Knights

    Jerome Gambit vs those Knights.

perrypawnpusher - Riversider
blitz, FICS, 2010

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 Nf6


A normal setup for the Jerome Gambit. Often the Black Knights like to kick the White Queen.

10.0-0 Ng4 11.Qg3

Better than the dangerous 11.Qe2 Qh4 12.h3 Nf4 13.Qf3 Ne5 14.Qc3 Ne2+ White resigned, mrjoker - weiran, ICC, 2008.

11...h5



Or 11...Qh4 12.Qxh4 Nxh4 as in tedhort - popasile, FICS, 2009 (1-0, 30) and Ghandy - julianaajedrez, FICS, 2009 (0-1,26)

12.h3 h4 13.Qb3

I didn't like taking the Queen away from the Kingside – but, those Knights!

13...N4e5


Black missed a dynamic attacking line that Rybka pointed out after the game: 13...Nf4!? 14.Qf3 Rf8 15.hxg4 Qg5 16.Kh2 Bxg4 17.Qe3 Kd7 18.d3 Bd1 19.Rg1 Bxc3 and Black has an edge in the resulting messy position.




analysis diagram







14.f4 Nc6 15.Nc3 Qf6 16.d3 Qd4+ 17.Kh1 Nf8


Surprisingly, this re-deployment is fatal. The Knight needed to go to e7.

18.Nd5 Kd8


Black played his move, and then Resigned, seeing that his Queen is trapped after 19.Be3.