Sunday, February 20, 2011

Serious About Silliness

I have been watching the early movies of the Marx Brothers, and I have to say that they are pretty serious about their silliness. In a similar way (except for my glaring lack of talent) I can be rather serious about the silliness that is the Jerome Gambit (and its relatives). 

Fortunately, not everyone is.

HauntedKnight  - sarahdaniel
blitz, FICS, 2011

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


The good old Jerome Gambit. I've researched its history, analyzed its lines, annotated games and gathered a large database.

As Chico might say, "Why a duck?"

4...Kxf7 5.Nxe5+ Nxe5 6.d4 Bd6


This unusual defense seems to be borrowed from the "fork trick".

The Database has 41 games with this line. White scores 61%.

7.dxe5 Bxe5 8.Qh5+


This is the strongest line for White. It is interesting to note, however, that in the 24 games with 8.Qh5+ in The Database, White has scored 54%.

Yes, that's right: the best move scored worse than the lesser alternatives.

8...Ke6 9.Qf5+


Again, this is the strongest continuation of the attack for White; yet it appeared in less than 1/4 of the games in the 6...Bd6 line: 10. In those, White scored only 50%, a result less succesful than that of the worse alternative moves combined.

Happily for HauntedKnight his opponent now responded with

9...Resigns

Other games from The Database continued. We will look at them, too.

9...Kd6

10.Na3

This is White's strongest move, planning a Knight fork to the enemy King and Bishop, with check, winning a piece. It is not easily met, but after 10...Bf6 11.e5+ Ke7 12.exf6+ Nxf6 13.Bg5 d5 14.Bxf6+ gxf6 15.Qf3 followed by 0-0 or 0-0-0 White will have equalized in material while having the better pawn structure and safer King.




analysis diagram








Instead of the text, 10.f4 is very Jerome-ish. With 10...Bf6 Black can return the piece (i.e. 11.e5+) for an even game. The contest wzeller - bourgas, FICS, 2005 instead entered the world of "optical illusions" with 10...Qf6 11.fxe5+ Qxe5+ 12.Bf4 and White won (1-0, 62).

10.Nc3 is not dangerous enough, although White won against 10...a6 in jfhumphrey - jrhicksdotnet, FICS, 2008 (1-0, 14). The better defense 10...Qf6 brought Black victory in jfhumphrey-biryuk, FICS, 2010 (0-1,65) and duanezz - pawelek, FICS, 2005 (0-1, 38).

Likewise, 10.0-0 also falls to 10...Qf6 but not as often as expected. White's plan is to continue his attack on the King with 11.Rd1+, and after 11...Ke7 12.Bg5 he was successful ("optical illusions" again!) twice: EvanJunior - BeNumberOne, FICS, 2008 (1-0, 22) and AGRPlayer - trebbor, FIC, 2010 (1-0, 19). However, Black's King should escape the Rook check to fresh air with 11...Kc6 instead and then White cannot conjure up either an attack or a recovery of material, as seen in carlkrull - bodensee, FICS, 2009 (0-1, 28).

10.b3 contains a great little trap – if Black now takes the Rook with 10...Bxa1 he is checkmated after 11.Ba3+ c5 12.Qxc5+ Ke6 13.Qf5# – but the proper response, 10...Qf6 White could not escape in ljar - Castled, FICS, 2006 (0-1, 39)


10...Qf6

Not the strongest response, as indicated in the note to 10.Na3, but certainly one that White can expect to see.

11.Nc4+ Ke7

Falling in with his opponent's plan too quickly. He should have at least tried 11...Kc5, since after 12.Nxe5 Qxf5 13.exf5 White's advantage is only a doubled pawn. After the better 12.Qxe5+, however, Black needs to cooperate with 12...Qxe5 13.Nxe5, as the obstreperous 12...Kxc4 – which may have been Black's intent – falls to chekmate after 13.b3+ Kb4 14.c3#.

12.Qxe5+ Qxe5 13.Nxe5 d6


White is a safe pawn up in a quiet position.

However, in the Jerome Gambit "safe" and "quiet" are relative terms, and Black eventually won in noatun - Papaflesas, FICS, 2008 (0-1, 48)...

(By the way, all of the games mentioned here are available in The Database – just email me and ask.)




graphic by Jeff Bucchino, the WizardofDraws

Saturday, February 19, 2011

It's hard to know...

It's hard to know what progress, if any, the following game shows for my recent play. Several times when things were getting interesting, the bottom dropped out of my opponent's play...

Of course, it was nothing like our first game.


perrypawnpusher - VGxdys
blitz, FICS, 2011

1.e4 e5 2.Nf3 Nc6 3.Bc4 h6

The Semi-Italian Opening.

4.O-O Bc5 5.Bxf7+


The Semi-Italian Jerome Gambit.

5...Kxf7 6.Nxe5+ Nxe5 7.Qh5+ Ng6


8.Qd5+ Ke8 9.Qxc5 d6 10.Qe3 N8e7


11.f4

I have had similar positions coming out of the regular Jerome Gambit line (i.e. no ...h6, not yet 0-0) and it is clear that White benefits here from the Semi-Italian move order.

11...Qd7

A bit puzzling: I really don't know what this move was about. Black can't afford to throw tempi about. Instead, perrypawnpusher - apinheiro, blitz, FICS, 2011, continued 11...Rf8 12.f5 Ne5 13.d4 Nc4 14.Qg3 Rg8 (1-0, 25)

12.f5 Ne5 13.d4 Nc4



14.Qe2

Grabbing a tempo by attacking the Knight, while staying on the e-file with the enemy King (and off the d-file with the enemy Queen). After the game Rybka expressed a preference for 14.Qg3.

14...b5 15.b3

Another kick at the Knight, while preparing to put my Bishop on the a1-h8 diagonal (a plan that I subsequently dropped).

An alternate idea was to undermine Black's Knight with 15.a4. It looks like after the consistent liquidation 15...d5 16.axb5 Nd6 17.Nc3 dxe4 18.Nxe4 Nxe4 19.Qxe4 Qd5 20.Qxd5 Nxd5 21.c4 Nf6 
White's 3 pawns balance out Black's Knight.




analysis diagram









15... Ba6

A strange oversight. Retreating the Knight to b6 was the move.

16.bxc4 bxc4 17.Nc3 Rf8


The nice thing for Black about the Jerome Gambit is that he can often drop a piece, and still have a tolerable game.

Here, White is better (up a pawn, more center control, safer King) but the game is far from over. My plan was to break through with e4-e5, but only after I had made sure that I could do so safely and with proper development.

18.Ba3

Linking my Rooks and targetting e7 and f8.

18...Kd8 19. g4 Nc6 20. Rad1


20...g5 21.e5 Nxd4

Again, incomprehensible.

22.Rxd4 h5

A spirited notion: how about some counterplay on the h-file?

23.h3

This was an I-don't-want-to-think-about-it-right-now move. Too bad I missed 23.Bxd6.

23...hxg4


24...Qh7

Planning to back the Queen up with ...Rh8 if he gets a chance.

He doesn't.

25.exd6 Kc8



If, instead, 25...c5, one of the few moves that does not lead directly to checkmate, I had 26.Qe7+, ending all counterplay and winning the exchange as well with 26...Qxe7 27.dxe7+ Kxe7 28.Bxc5+

Now it all tumbles to an end.

26.d7+ Kb8 27.d8=Q+ Rxd8 28.Rxd8+ Kb7 29.Qe4+ c6 30.Rxa8 Kxa8 31.Qxc6+ Bb7 32.Qe8+ Black resigned




Friday, February 18, 2011

A Tale of Two Pawns

I really like this game, but I have to admit that my attack was made possible by my opponent keeping his d-pawn un-moved for too long. After the game, Houdini and Rybka also criticized me for leaving my f-pawn still – true, but when I pushed it, things really started to move!


perrypawnpusher - Frele
blitz, FICS, 2011

1.e4 e5 2.Nf3 Nc6 3.Bc4 h6


The Semi-Italian Opening.

4.0-0 Nf6 5.Nc3 Bc5


The Semi-Itlian Four Knights Game.

6.Bxf7+

The Semi-Italian Four Knights Jerome Gambit.

6...Kxf7 7.Nxe5+ Nxe5 8.d4 Re8


One of a number of "okay" defenses in this line (solid strategic idea, keeps Black's advantage) that still are not "best". Black develops his Rook and prepares to castle-by-hand.

9.dxc5 Kg8

Black immediately went after the forward White c-pawn with 9...b6 in  perrypawnpusher - catri, blitz, FICS, 2010 (½-½, 42) and with 9...d6 in perrypawnpusher - louarn, blitz, FICS, 2010 (1-0, 49).


10.f4 Nc6


This retreat seems okay to me, but after the game Houdini preferred 10...Nf7. The advance 10...Nc4 proved unfortunate after 11.e5 Nh7 12.Qd5+ Kh8 13.Qxc4 in Wall - Santiago, FICS, 2010 (1-0, 28).


11.e5 Nh7 12.Qd5+

I admit that I was mostly looking for the cheapo 12...Re6? 13.f5 with this move; but the check does little harm.

12...Kh8

White has locked down Black's d-pawn, and that is why most of Houdini's post mortem suggestions for the second player included playing ...d7-d6, anyway, even if it required temporarily sacrificing a pawn. Black cannot afford to play the game without his Queen Bishop and Queen Rook.

13.Bd2 Ne7

While Houdini holds that this still gives Black an edge, it prefers 13...d6 14.cxd6 cxd6 15.Qxd6 Qxd6 16.exd6 Bf5 17.Nd5 Rad8 18.Rfe1 Nf6 19.Nxf6 Rxe1+ 20.Bxe1 gxf6 21.Rd1 Bxc2 when Black's extra piece outweighs White's two pawns.






analysis diagram







14.Qd3 Rf8


The recommendation was still 14...d6.


15.f5 Nc6



The loss of time with this Knight gives White tactical chances.

Still, against both 15...b6 and 15...d6 White now has 16.f6!? with an even game, maybe a bit more.

16.Rae1

Houdini suggested, instead, 16.f6 – a move that I was preparing, something that I didn't want (like in earlier games) to rush into.

Its analysis shows White taking advantage of a lead in development and the available tactical nuances – to achieve a winning endgame: 16...gxf6 (here, the computer dismissed the pawn grab that attacks my Queen, 16...Nxe5, with 17.fxg7+ Kxg7 18.Qg3+ Ng6 19.Nd5 vacating c3 for the dark-squared Bishop Ng5 20.Bc3+ Kg8 21.Qd3 Nh4 22.Kh1 Ne6 23.Bf6 Qe8 24.Ne7+ Wow) 17.exf6 d6 18.Nd5 Be6 19.Bxh6 (the real reason behind 16.f5-f6) Rf7 20.Rad1 Qd7 21.Bg7+ Rxg7 22.fxg7+ Qxg7 23.cxd6 cxd6 24.Nf4 Rg8 25.Rd2 Bg4 26.h3 Ne5 27.Qc3 Bc8 28.Rxd6 Ng5 29.Ng6+ Nxg6 30.Rxg6 Qxc3 31.Rxg8+ Kxg8 32.bxc3.





analysis diagram








Um... sure... right...

Okay, I guess that proves I'm not a computer... Back to the game.

16...Ng5


My opponent hurries to reposition his Knight at f7, for defensive purposes.

You can imagine that Houdini, as a computer, was doubly horrified: Black doesn't play the necessary 16...d6 (leading to only a slight edge for White) and his move allows for the killer 17.f6.

(Silicon Dude! This is a blitz game between two club players! Relax!)

17.Nd5

Of course, everyone saw that 17.f6 again forced a winning endgame, right? After 17...gxf6 18.exf6 Rxf6 19.Rxf6 Qxf6 20.Re8+ Kg7 21.Nd5 Qd4+ 22.Qxd4+ Nxd4 23.Nxc7 Rb8 24.Bc3 Kf7 25.Bxd4 Ne6 26.Be5 Nxc7 27.Rh8 Ra8 28.Bxc7 Kg7 and there you have it...

17...Nf7 18.Bf4



Protecting the e-pawn so that I can play f5-f6, but this was unnecessary, as after 18.f6 Nfxe5 19.fxg7+ Kxg7 20.Qg3+ White's attack would be strong.

18...Re8 19.f6

If it could sigh, I sure that Houdini did when I finally played this move.

19...gxf6


After the proper defense 19...Rg8 White can apply pressure with 20.Qg6 although after 20...Qf8 it would take another tactical mele, starting with 21.Nxc7 to demonstrate my advantage (eventually finding my way to an ending where my Rook and several pawns outweights Black's two extra Knights).

Sometimes it feels that if it were not for mistakes, chess would be too hard.


20.Nxf6


The correct recapture. My 17th move not only kept Black's Queen at home with the threat of Nxc7, it also strongly supported the (eventual) advance f5-f6. Now White's pieces will swarm the Kingside.

20...Kg7 21.Qh7+ Kf8 22.Bxh6+ Ke7 23.Nd5+ Ke6 24.Nf4+


A little nervousness, missing the mate in one, but it did not get away, and I saw it the second time around. 


24...Ke7 25.Ng6+ Ke6 26.Qxf7 checkmate

Thursday, February 17, 2011

BSG Talk: Hmmmm....

I was doing a Google search on "4.Bxf7+" to see what was out there on the internet these days, and I ran across a discussion of the Blackburne Shilling Gambit (1.e4 e5 2.Nf3 Nc6 4.Bc4 Nd4) at Wikipedia titled "Talk:Italian Game, Blackburne Shilling Gambit", first visited by this blog in "Talk" a couple of years ago.

The writers referenced Wikipedia's article on the BSG, and one opined
Another suggestion not mentioned in the article is that after 1.P-K4,P-K4 2.N-KB3,N-QB3 3.B-B4,N-Q5?...I should have thought the most enterprising way of punishing black's loss of tempo is with 4.BxPcheck!! (eg. ....KXB .NxPcheck etc. with ample compensation for the piece and a probably winning attack.
This was met at first with skepticism

Disagree. After 5. ... Ke7 white has no way to continue the attack or even getting more piece in to contine. Moving the Queen out 6. Qg4 d6 7. Qg5+ Nf6 =+ or 6. Qh5 can be met with Nf6 =+. Black is better and white's 4. Bxf7 is an unsound sacrifice.
However, after the correction
Mi scusi, per favore. But what about the reply 5. Nxe5 to 4. Bxf7+?! Ke7!?
The skeptic relented
Much to my surprise I agree. After 4. Bxf7+ Ke7 5. Nxe5 White is better, White could of even retreated the bishop by playing 5. Bc4, so 4...Ke7 is not enough.
There is much to comment upon, here...

In "Talk" I had written

I like the thoughts contained in the first paragraph, especially "ample compensation for the piece," although "and a probably winning attack" may be over-stating it a bit.
I still agree – if "ample compensation" means "an even game", which is what White gets against best play by Black.

But what about that line 1.e4 e5 2.Nf3 Nc6 3.Bc4 Nd4 4.Bxf7+ Ke7 ? Is there more to it than a "Jedi Mind Trick"?

I checked The Database and ran an Opening Report. There are 299 games with that line, and "Black scores badly (31%)." No surprise, there.

What of 5.Nxe5 ? Well, in 125 games "White scores above average (60%)" That sounds okay, right?

But, wait a minute.

With 5.Nxd4 (56 games) White "scores well (72%)".

For that matter, with 5.Bxg8 (60 games) White "scores excellently (78%)" and also with 5.Bc4 (20 games) White "scores excellently (80%)". Likewise with 5.c3 (5 games) White "scores excellently (80%)".

Further, with 5.Bb3 (5 games) White "scores excellently (100%)".

It is important not to get too carried away by the Opening Report, which also says that with 5.0-0 and 5.d3 (5 games each) which allow 5...Kxf7, White still "scores excellently (100%)" — both moves are clearly inferior choices (they lose a piece). For that matter, in The Database the strong 5.Bd5 (6 games) only "scores below average (50%)" for White.

So, play 5.Nxe5 for variety, but retreat (or exchange) the Bishop for the best play.

Wednesday, February 16, 2011

Things are looking up

With a game like the following, I get the sense that it is possible to play the Jerome Gambit effectively (at least every once-in-a-while) and even have a bit of fun in the process.
perrypawnpusher - Jaqueperpetuo
blitz, FICS, 2011

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


4...Kxf7 5.Nxe5+ Nxe5 6.Qh5+ Ng6


7.Qd5+ Kf8 8.Qxc5+ N8e7


Not a frequent response (usually seen is 8...d6), but understandable: Black develops a piece even as he blocks White's check.

9.0-0 d6 10.Qe3 b6


This reminds me of my game with jdvatty.

Jerome Gambit Gemeinde member Louis Morin (mrjoker, guest) has the most experience with the 8...N8e7 defense:

10...Ke8 11.d4 Rf8 12.f4 d5 13.f5 Nh8 14.exd5 Rxf5 15.Rxf5 Bxf5 16.Nc3 Qd7 17.Bd2 Bxc2 18.Qe5 Qf5 19.Qxg7 Kd7 20.Re1 Nf7 21.Rf1 Qg6 22.Rxf7 Qxg7 23.Rxg7 Kd6 24.Nb5+ Kd7 25.Bf4 Rc8 26.Bxc7 a6 27.d6 axb5 28.Rxe7+ Kc6 29.d5+ Kxd5 30.d7 Rxc7 31.d8Q+ Black resigned, guest645 - guest1600, Internet Chess Club 2001;

10...Bd7 11.f4 Bc6 12.d4 Ke8 13.f5 Nf8 14.c4 d5 15.cxd5 Bb5 16.Rf2 c6 17.Nc3 Ba6 18.Bd2 cxd5 19.e5 Nd7 20.f6 Ng6 21.fxg7 Rg8 22.Nxd5 Rxg7 23.e6 Nb6 24.Nf6+ Ke7 25.Bb4+ Black resigned, guest4097 - guest4686, Internet Chess Club, 2004;

10...Be6 11.f4 Bf7 12.d4 Kg8 13.f5 Nf8 14.b3 Nc6 15.c4 Qf6 16.Bb2 Qe7 17.Nd2 Re8 18.Rae1 Nb4 19.Qg3 Nd7 20.e5 dxe5 21.dxe5 Kf8 22.e6 Nf6 23.exf7 Qxf7 24.Rxe8+ Qxe8 25.Re1 Qd7 26.Ne4 Qxf5 27.Nxf6 gxf6 28.Bxf6 Rg8 29.Re8+ Kxe8 30.Qxg8+ Kd7 31.Qf7+ Kc6 32.Qe8+ Kb6 33.Bd4+ c5 34.Qb5+ Kc7 35.Qxc5+ Qxc5 36.Bxc5 Nxa2 37.Bxa7 Kc6 38.Be3 Nb4 39.Kf2 Nc2 40.Ke2 b5 41.cxb5+ Kxb5 42.g4 Kb4 43.h4 Kxb3 44.g5 Kc4 45.h5 Nxe3 46.Kxe3 Kd5 47.g6 hxg6 48.h6 Ke6 49.h7 Kf7 50.h8Q Black resigned, mrjoker - cherryhead, Internet Chess Club, 2008;

10...h6 11.d4 Kg8 12.f4 c5 13.c3 cxd4 14.cxd4 Kh7 15.f5 Nf8 16.f6 gxf6 17.Qxh6+ Kg8 18.Qxf6 Neg6 19.Qf7 checkmate, mrjoker - hp9000, Internet Chess Club, 2009.

11.f4 Bb7 12.f5 Ne5 13.d4


13...Nc4

This gives White a useful tempo without getting anything in return. Black had better chances from harassing the Queen with 13...Ng4 although after 14.Qf4 Nf6 15.Nd2 his advantage would be minimal.

14.Qd3 Na5 15.f6


After the game, Fritz10 preferred the positional 15.d5, when White would have a solid grip on the light squares. That is all well and good, but I wanted to attack the King!

15...gxf6 16.Rxf6+ Ke8 17.Nc3 Rf8


Black is skeptical of White's plan: can't the Rook just be traded off?

His assessment is largely correct – White needs more pieces to join the attack before it can become successful – but the answer was to counter-attack to keep the first player occupied. Rybka and Fritz10 like the idea of ...Qc8 followed by ...Qg4, combined with ...Rg8.

As it goes, White can exchange Rooks and use a gained tempo to bring his other Rook into play, something Black cannot imitate.

18.Rxf8+ Kxf8 19.Bh6+ Ke8


It was not at first clear that 19...Kg8 was the better King move.

Of course, after 19...Kg8, the f8 square would be protected three times (King, Queen, Rook) instead of once, as it is after the text. That would be a defensive improvement, but White could still make progress, taking advantage of the weakened light squares, with 20.b4.  If Black's Knight retreats with 20...Nc6, then 21.Qc4+
has to be met with 21...d5, when 22.exd5 maintains the attack on the Knight at c6 while threatening the one at d7 with a possible d5-d6+. White will keep a small advantage.

20.Rf1 Ng6 21.Qf3 Qe7


Black is holding on, although his extra piece is not doing much at a5.

White's most consistent course now would be to play 22.h4, with the idea of driving off Black's Knight on g6 which is one of the protectors of the f8 square. If Black answers 22...Nc6, bringing his offside Knight back into play, then 23.h5 Nxd4 24.Qf2 allows White to win one of the Knights.

Instead, I was content to keep Black's position cramped and torture his Knights a bit more.

22.Nd5 Bxd5 23.exd5


White's initiative will last into the endgame, for example: 23...Kd8 24.Qf5 Qd7 25.Bg5+ Kc8 26.Qxd7+ (another exchange to gain a tempo) Kxd7 27.Rf7+ Kc8 28.Rxh7 when White's Kingside pawns will allow him to triumph. 

However, Black now moves to free his un-moved Rook and comes to grief.

23...Kd7 24.Qf5+ Kd8

If 24...Ke8 then 25.Bg5 wins as well, e.g. 25...Qg7 26.Re1+, etc. 

25.Bg5 Black resigned

Tuesday, February 15, 2011

The amazing Chess World

Chess players with a taste for the Jerome Gambit (1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+) might have in interest in other unorthodox openings – in which case they would be interested in the blog The amazing Chess World hosted by Davide Rozzoni.

Not only does he include various "Tips for the improving UCO players rated between 1500 and 1800", he has links to interesting software (e.g. free engines Spike 1.4 Leiden and Houdini 1.5a; as well as software to play against, Chess Hero). 

Check it out.

Monday, February 14, 2011

No Way to Make A Living


Using my "extra" pawns to chase down a draw or a win against Black's extra piece in the Jerome Gambit can sometimes be an exhausting task. "That's no way to make a living," as they say. Even if your opponent slips at the end...

perrypawnpusher  - vladchess
blitz, FICS, 2011

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


4...Kxf7 5.Nxe5+ Kf8

One way to change the flow of the game and try to keep his King out of trouble.

6.Nxc6 dxc6 7.O-O


I've had success after 7.d3 Qf6 8.O-O in two games in which I reversed my 7th and 8th moves:  perrypawnpusher - fortytwooz, blitz, FICS, 2010 (1-0, 29) and perrypawnpusher - Lark, blitz, FICS, 2011 (1-0, 12).

7...Nf6

Also seen: 7... Bd6 perrypawnpusher - Lark, blitz, FICS, 2009 (1-0, 18) and 7... Be6, perrypawnpusher - CorH, blitz, FICS, 2009 (0-1, 74).

8.d3 h6

8...Bg4 was seen in Lowe - Parker, corr, 1879, (1-0, 37)

9.Be3 Qe7

Exchanging Bishops was also okay.

10.d4 Bb6 11.f3 Kg8


Computers like Rybka tend to like solid pawn centers like the one White has and rate it accordingly, but there is a lot to do to get "something" out of the position for White.

12.Nc3

Moving another pawn with 12.c4 might have been better, at least according to Houdini 1.5a.

12... Kh7

The cute 12...Nd5 (note the pin on the e-pawn) was a bit better.

13. Qd3 g6 14.f4

Instead, Houdini prefers 14.Na4 Ba5 15.c4 b6 16.a3 c5 17.dxc5 Rd8 18.Qc2 bxc5 19.Bxc5 Bb6 20.Qf2 Qe8 21.Nxb6 axb6 22. Be3 Qe7 23.Rfd1 Be6 24.Qc2 Nd7 25.Qc3 Qg7 26.Bd4 Qf7 when Black's advantage is minimal.

Okay, if you say so... (The next book that I review for Chessville will be a strategy book!)

14...Rg8 15.Rae1 Ng4 16.h3 Nxe3 17.Qxe3 c5 18.Nd5 Qd6 19.Nxb6 axb6

20.dxc5 bxc5 21.Rd1 Qb6 22.Qb3 Be6 23.Qxb6 cxb6

If there was a thread in this game, I've probably lost it...

My only chance is to activate my Rooks.

24.a3 c4 25.Rd6 Bc8 26.Rxb6 Re8 27.e5 c3 28.bxc3 Rxa3 29.Rb3 Ra2 30.Rf2 Re7 31.Rd2 Rf7


32.g3

Simply overlooking the loss of a pawn.

32...Bxh3 33.Rb5 Re7 34.Kf2 Bf5 35.Kf3 h5 36.Rbd5 Ra3


Black is still better, but 36...Rc7 here would have been stronger.

Time was getting short for both of us. My plan was to continue to harass Black until one of us blundered.

37.Rc5 b6 38.Rc6 Rb7 39.Kf2 b5 40.e6  Re7 41.Rd7


Going for the swindle.

41...Rxd7 42.exd7 Bxd7

Simply 42... Ra8 was Game Over.

43.Rc7 Kh6 44.Rxd7 Rxc3 45.Rd2 b4 46.Kg2 draw.


This game reminds me of the saying about airplane landings: any one that you can walk away from is a good one...