In a recent book review at Chessville that I wrote about GM Nigel Davies' 10 Great Ways to Get Better at Chess (Everyman Chess, 2010), I mentioned
For an alternate “great way” – or as an addition – Chapter Two suggests Study the Endgame. Again, Davies gives game examples from his students, but this should not be necessary to persuade readers of the truth of his suggestion: most club players unconsciously imitate the play of early chess-playing computers, memorizing opening moves, concentrating on tactics at the expense of long-term planning, and playing the endings like duffers…
Davies would have been justified in giving a Chapter Three (“Study the Endgame Some More”) and a Chapter Four (“Keep Studying the Endgame”) just to reinforce the topic. It’s just really that important.
I wish I had taken more of my own advice before playing the following game.
Like two zombies dancing a waltz, my opponent and I push the pieces around the board and pretend that we are playing chess...
24.axb5
If someone had whispered "Pssst. This is a middlegame" I might have looked for and found the tactical 24.Bf4 Rb7 25.Nc7 Bf7 26.Rd2 Nf6 27.Nxa6 bxa4 28.Nxc5 Rb5 29.Nxa4 Nxe4 30.Re2
For all of my shuffling, the e-pawn will now be lost.
32.Be3+ Kh5
My opponent has grown complacent, too, or he would have played the wiser 32...Kg7
33.c3
Readers who enjoy tactical endgames surely saw the better 33.g4+ which merely threatens checkmate after 33...Kh434.Kg2, costing Black his Knight.
Playing on "auto-pilot" was hurting both of us.
33...Nxe434.Bd4
The checkmate was still on, in a slightly longer version, with 34.g4+.
34...Ng3+ 35.Kf2 Nf5
This slip costs a piece.
36.g4+ Kh4 37.gxf5 gxf5 38.Kg2
With a piece for a pawn, White is now winning – but how exactly does he win?
If he can get his King to d4 and then attack the Black pawn at c4 a second time (with the Knight) he can win it. For that to happen, it might be best to trade White's weak h-pawn for Black's f-pawn. Then, with the help of the extra piece, White's c-pawn can be "faster" than Black's h-pawn.
38...Bd5+ 39.Kh2
Protecting the pawn for now. After the game Rybka suggested 39.Nf3+, since 39...Kh5 40.Kg3 Be4 41.h4 Bd5 42.Kf4 Be6 43.Bf2 Kg6 44.Ne5+ Kh5 45.Nc6 Bd7 46.Ne7 would allow White to win the f-pawn. If Black exchanges off the Knight instead with 39...Bxf3+ 40.Kxf3 then after 40...Kxh3 41.Kf4 White's King will wander over and win the Black c-pawn and Queen his own; the Black h-pawn will not be an issue.
39...Be6 40.Nf3+ Kh5 41.Kg3 f4+
A mistaken offer to exchange pawns that I should have accepted.
42.Kh2 Kg6 43.Kg2 Kf5 44.Ne5 Bd5+ 45.Nf3Bg8
Black is content to sit and wait.
It is not clear to me that there is a win any more for White, as Black's advanced King and h-pawn make it dangerous for White's King to leave that area to help win the c-pawn.
My "winning try" falls to the inevitability of the Bishops-of-opposite-colors endgame.
So much of chess is trusting your own plans and being skeptical of your opponent's. Yet, misperceptions arise, and our play is often guided by "illusions" of what we expect or what we fear, instead of what actually is happening (or will happen) on the board. perrypawnpusher - yuriko blitz, FICS, 2011
This is an improvement over our earlier game, which continued 13...Nb4, perrypawnpusher - yuriko, blitz, FICS, 2010 (1-0,31). Black's pieces are developed, his King is safe, and he has the advantage.
Simply an oversight. Necessary was to protect the pawn with 16.c4. Fortunately, my opponent either overlooked the capture, or took my error as deep thought.
16...Rf817.Bxf6 Rxf6 18.Qe4 Rf7 19.f5
Deciding that if there is any play for White (there isn't) it has to be on the Kingside.
19...Qf820.Qe6 Kh8 21.g4h5 22.h3 hxg4 23.hxg4 Re8
At first glance it looks as if White's Queen is taking on the combined forces of Black's Queen, two Rooks and Knight.
In actuality, Black's dallying pawn moves on the Kingside (instead of re-positioning his Knight to f6, via g8) have given White an even game, and Black's last move has given White a clear advantage.
24.Qe4
Giving the Queen the chance to move to the h-file with Qe4-h1, an idea that only gets partial credit. 24.g5 was the correct move, with the plan of advancing to g6 if possible, setting up a mating net that will close when White doubles rooks on the h-file.
Or at least that's what Rybka suggested after the game.
24...Kg8 25.Qe6 Kh8 26.Qe4
Toying with the idea of a draw by repetition. This was underestimating my chances.
My opponent decided not to repeat, but keeps the position balanced.
26...g6 27.Qd4+ Kg8 28.f6 Nc8
29.g5
Logical, but 29.Re6 gave White an edge.
Time was getting short for both of us.
29...Rxe1 30.Rxe1 Nb6 31.Qe4
Aiming at the pawn at g6, a major weakness in Black's King's shelter.
31...Qc8
Having been sold on the illusion of my attack for many moves, my opponent now conjures an illusion of counter-attack for himself. While objectively this move is wrong, because it allows checkmate, it turns out to be a game-saver for yuriko.
Chess is strange.
There was a likely drawn Rook + pawns vs Rook + pawns endgame after 31...Nd7 32.Qxg6+ Kh8 33.Re3 Nxf6 34.gxf6 Qg8 35.Rh3+ Rh7 36.Qxg8+ Kxg8.
32.Qxg6+ Kf8 33.Qh6+ Kg8
Now all that I had to see was 34.g6, the triumph of the "Jerome pawns" (with f6-f7+, winning a Rook, and then checkmate, following) and the game was mine.
Instead, I worried that 34.g6 could be answered by 34...Qg4+ (that was why Black had played 33...Qc8) with either counterplay, or a draw, anyway – so with little thought (time was ticking) I forced the draw myself.
This game is the latest in a series of my attempts to understand a move that Rybka has recommended in various post mortems. Although I still quite don't understand it, I played it – and my opponent responded as if he didn't understand it, either – and suddenly the game was over.
perrypawnpusher - LeChinois blitz, FICS, 2011
1.e4 e5 2.Nf3 Nc6 3.Bc4 h6
The Semi-Italian Opening.
4.0-0 Nf6 5.Nc3 Bc5 6.Bxf7+
The Semi-Italian Four Knights Jerome Gambit.
6...Kxf7 7.Nxe5+ Nxe5 8.d4
8...Bxd4 9.Qxd4 Nc6
More frequently seen is 9...d6 10.f4 Nc6, but I do not see a problem with the text.
After the game I fed the moves to Rybka in blundercheck mode, and here (and for subsequent moves) it recommended instead 11.Nd5.
I know that Nd5 is the kind of move that you play here (at least according to Rybka, which has made that recommendation several times in past games) but I did not like what would happen with Black's Queen at e7: 11...Qxe4 12.Qxe4 Nxe4 13.Nxc7 Rb8. I was hoping for more than exchanging my e-pawn for my opponent's c-pawn.
Okay, with my e-pawn protected with this move, now I'm ready for Nd5.
13...a5
Of course Rybka preferred 13...Rae8– followed by14.Nd5 (and 15.Bc3).
14.Nd5
Tah dah!
White also had 14.e5 dxe5 15.fxe5 Nxe5 16.Qd5+ Kg6 17.Qxe5 Qxe5 18.Rxe5 which equalized, but which took much of the dynamism out of the game.
14...Be6
Well, that settles everything...
Rybka pointed out the essential 14...Qd8 and then suggested that White could/should draw with 15.Nxf6 Qxf6 16.Bc3 Qe7 17.e5 Be8 18.Re3 Kg8 19.Qd5+ Kh7 20.Qe3+ Kg8 21.Qd5+ etc.
Not too long ago I received an email from Stipe Drinovac who had noticed (from a post at Chessgames.com) that I was interested in Jerome Gambit games. He sent me a game that he had played last year at a club event. My eyes popped.
Drinovac - NN club game, 2010
1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+
4...Kxf7 5.Nxe5+ Nxe5 6.Qh5+ Ke6
7.f4 Qf6 8.Rf1
I found only a couple dozen games in The Database with this move (White scored 54%), but the players were impressive: Jerome Gambit Gemeinde members mrjoker, Petasluk and Gary_Seven, as well as paladins Fritz 5.32, Fritz 8, Junior 7 and Zappa 1.1.
8...Nh69. Nc3 d6
10.Nd5 Qd8
And, just like that, White is winning, according to Rybka, which preferred the offer to exchange Queens with 10...Qf7.
11.d4 Bxd4 12. fxe5 dxe5
13.Bg5 Qg8 14.c3Bb615.Nxb6axb6
16.Qh3+ Kd6 17.Qd3+ Kc6 18.Rf8
If Black had not realized until this point that he was doomed, this probably delivered the message.
18..Qe6 19.Rxh8 b5 20.a4 bxa4 21.Re8
Another deflecting sacrifice, offering the Rook a second time.
21.Qxe8
Drinking the poison. Declining with something like 21...Qb3 left Black down the exchange, with his King still in danger. Rybka likes 22.Bd8, putting pressure on c7 and planning to sacrifice White's Bishop there, if nessary, as a response.
Many readers are familiar with the game young Paul Morphy played against his uncle, Alonzo Morphy, where he gave his elder the odds of a Rook, and still checkmated him– with the last move being Castles:
Paul Morphy - Alonzo Morphy New Orleans, 1850 Remove White's Queen Rook
If it's true that "bad things come in threes" (see "Struggling" and "Tactical Sloppiness") that would be a good explanation for the following win that I recently scored. It's the kind of game you might as well have if your opponent is going to have difficulty with the clock...
A rarely-played move (according to The Database) in a standard position. Rybka and Fritz8 almost always suggest Nc3-d5 now for White, but I was lured by the opportunity to start my Kingside pawns moving "for free". Instead, I chased the Black Bishop back to a better square.
14.h3 Be615.g4
Like I said in the note to move 13, better was 15.Nd5. The reason has little to do with the weakening of White's pawn structure, and more to to with dropping the c-pawn.
15...Nb4 16.Qg3 Nxc2 17.Re2
This should cost the exchange after 17...Bc4.
17...Nd418.Rg2 Qd7 19.f5
I figured at this point that whatever chances I had to save the game (slim & none, actually) would lie in a Kingside attack.
19...Bc4 20.Re1 Nc2
Much better was 20...Qc6.
21.Rc1 Nd422.g5hxg5 23.Bxg5 Rf8
Another lucky break for me: Black believes that my attack is worth worrying about. He should have simply smashed my center with 23...Nxe4.
24.Bxf6 Rxf6 25.Rd1
Missing a key tactical point that would have almost equalized the game: after 25.b3 Black's Bishop has nowhere safe to go (even 25...Ba6 allows the strong attacking 26.Nd5 with threats to c7) so after 25...d5 26.Nxd5 Bxd5 27.Rxc7 Nf3+ 28.Kh1 Rc6 29.Rxd7 Rc1+ 30.Rg1 Rxg1+ 31.Qxg1 Nxg1 32.exd5 Ne2 Black's advantage is much smaller than in the game.
25...c5 26.e5
There is nothing but bluff in this.
26...Rxf5 27.exd6 Rf6 28.Ne4
Simply hoping to keep Black busy enough until his flag falls.
28...Rh6
Unfortunately this reduces his chances to a draw, while 28...Nf3+ would have killed.
29.Nxc5 Qxd6
Dropping the whole point, which seems unfair.
After the game Rybka suggested instead 29...Ne2+ 30.Rxe2 Qf5 with a possible continuation being 31.Re5 Qc2 32.Rde1 Rf8 33.Ne4 Qxb2 34.Re7 Bf7 35.d7 Rg6 36.Ng5 Qd4+ 37.R1e3 Qd1+ 38.Re1 Qd4+ drawing. Not easy to see with a few seconds left.
Here we have a second straight loss for me, the kind better blamed on my tactical sloppiness than on the Jerome Gambit itself, which certainly gave me chances for success.
perrypawnpusher - wbrandl blitz, FICS, 2011
1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+
4...Kxf7 5.Nxe5+ Nxe5 6.Qh5+ Ng6
7.Qd5+ Kf6 8.Qxc5
What difference, if any, does it make that Black's King is on f6?
Before I played 8.Qxc5, I toyed with playing 8.d4, with the threat of 9.B (or Q) g5+ winning Black's Queen. I had obviously forgotten that I had played the move successfully several years ago in perrypawnpusher - whitepandora, blitz, FICS, 2008, (1-0, 41). For the same reason, 8.f4 was also playable, and also better than the text.
After the game Rybka concluded that White had equalized.
24.Nxd5 Bg4 25.Nf4
The Knight needed to go to f6 with the possibility of continuing on to g8. The position would be complicated, but White at worst can expect to draw by repetition.
My move was made with little analysis other than thinking "this looks like a good square" and it allowed Black to bring his Rook into play, winning quickly.
25...Rf8
26.Qxb7 Rxf427.Qxa7 Rxf1+ 28.Kxf1 Qf4+ White resigned