Tuesday, August 4, 2015

An astonishing oversight

What would you play in the diagrammed position?

If you said 10...Nde5! or 10...Nce5!, take full credit. White's position is already resignable after either knight move. White loses two pawns and is forced to move his king. Mega Database 2013 finds 18 games that reached the above position. In 15 of them, Black moved one of the knights to e5. He scored 14.5 points in those 15 games.

If you said anything other than 10...N(either)-e5!, please enroll in Remedial Chess Tactics 101 forthwith. But don't feel too bad: you're in good company. Garry Kasparov, arguably the greatest player of all time, also missed it. He played the lemon 10...Nb4?, which would have allowed White to survive after 11.Bd2! when White is only a little worse (-0.47 according to Komodo 9.02). Instead, White played 11.Bb1? and was steamrollered.

How long has the shot 10...Ne5! been known, you ask? For almost 90 years. It was first seen in Norman-Vidmar, Hastings 1925/26, which Black won in short order:

This trap was given in Irving Chernev's book Winning Chess Traps, first published in 1946. The entire game appears, for example, in Chernev and Reinfeld's The Fireside Book of Chess (1949), as well as in more recent works such as the Encyclopedia of Chess Miniatures (2014) and Müller and Knaak's 222 Opening Traps After 1.d4 (2008). It is scarcely believable that Kasparov did not know of this old chestnut. Even if he somehow did not, it is mind-blowing that so great a tactical genius could miss so obvious a shot.

Wednesday, February 25, 2015

Nothing new under the sun II

After playing Rhine-Boerkoel (see previous post), I found the following game, which has strikingly similar tactics. The most salient difference is that in the final position 18...Kxf7 isn't an option because of 19.Nxd8+. Had the game continued, Black would have had to take the queen, allowing White's rook to give perpetual check on g7 and h7.

Tuesday, February 24, 2015

A spectacular game

This is one of my three favorite games that I have played. The others are K.Thompson-Rhine, 1992 and Rhine-Sprenkle, 1981. This is the only one of the three that isn't published in Chess Informant I should've submitted this one too, but I got lazy. It's a beautiful game, primarily because of my opponent's efforts - he sacrificed, or offered to sacrifice, a pawn, knight, bishop, rook, and queen - literally every kind of piece except his king. The cold, dispassionate engine (Houdini 3) tells me that had I played slightly differently (23.Qc2!), I would have had a large advantage. But as "Kinghunt" once observed on chessgames.com, "This is a King's Indian Defence - Black is always objectively lost until suddenly Black wins." Or draws, in this case.

Up through 15.c5, the game was all book. My opponent's 15...c6!? was a novelty. I was quite happy after 23.Qb3, thinking that I was a pawn up for little compensation. I could meet 23...f3 with 24.g3 Qh3 25.Bf1. Boerkoel's 23...Ne3!! came as a huge shock. The main points are that Black threatens to win with both 24....Nxg2 25.Kxg2 f3+ and 24...f3 25.g3 Qh3 26.Bf1 Nxf1, and if White plays 24.fxe3, 24...fxe3 threatens 25...Qxh2#, 25...Qxe4, and 25...f2+. The subtle 23.Qc2! would have avoided this shot; since it would have guarded my knight, I could have met 23...Ne3?? with simply 24.fxe3 fxe3 25.g3 and wins. I still thought I was winning after 26.Kxf2. After 26...Bxg3+!, the light finally dawned: I had to allow perpetual check.

Sunday, February 22, 2015

Nothing new under the sun

The same tactical themes are seen again and again. Compare the finishes of these two games. They reached identical positions except for the side on which White castled. In both, 15.Nxc6! is the killer, with the point that 15...bxc6 is met by 16.Qa6#, while 15...Qxe2 is met by the unusual checkmate 16.Nxa7#. The first game is No. 314 in Chernev's classic The 1000 Best Short Games of Chess.

Over sixty years later, I played this blitz game, without knowing of its predecessor:

The astute observer will notice that in my game, 14.Nxc6! was already possible. In fairness to Albert, who was my teammate on the Lane Tech High School chess team, this result was highly atypical of our games at the time. He was much stronger. Later we both became masters. Many years later, he co-founded the chessgames.com website.

Tuesday, February 17, 2015

A typical sacrifice

In the game below, White was already in big trouble when he attempted to exploit the pin on my knight with 8.Ne4?? 8...Nxe4! was a typical sham sacrifice. After 9.Bxd8 Bb4+, Black will regain the sacrificed material with interest. In the final position, the threat of 15...Nd2 double check and mate was enough to induce my opponent to resign.

Tuesday, February 3, 2015

Still another Englund Gambit trap

I have previously written several posts about traps in the Englund Gambit (1.d4 e5? or, to be charitable, 1.d4 e5?!). There are two "official" refutations of the main line of the gambit: after 2.dxe5 Nc6 3.Nf3 Qe7, one begins with 4.Bf4 (or 4.Bg5) Qb4+ 5.Bd2 Qxc3 6.Nc3!, the other with 4.Qd5. As I've previously noted, indiscriminately combining the two can lead to disaster. Below is another instance of this. After 5.Bg5?? (5.Nc3!) Qb4+!, White is losing after either 6.Bd2 Qxb2 7.Bc3 Bb4, as in the game, or 6.Qd2 Qxb2 7.Qc3 Bb4. The final position is very reminiscent of that arising after the classic Englund trap, 4.Bf4 Qb4+ 5.Bd2 Qxb2 6.Bc3?? Bb4! 7.Qd2 Bxc3 8.Qxc3 Qc1#.

Thursday, January 22, 2015

Three Times a Lady

In the B Group at the Tata Steel (Wijk aan Zee) tournament, the world's current youngest grandmaster, 14-year-old Samuel Sevian, performed a feat that has surely occurred very rarely in chess history. He sacrificed three queens on three consecutive moves! He thereby achieved an easily winning rook versus pawns ending, which his opponent promptly resigned. Very cool.

Saturday, January 3, 2015

A trap in Larsen's Opening

The following game illustrates what happens if White tries to grab a pawn in Larsen's Opening with 6.fxe5 fxe5 7.Bxc6+? bxc6 8.Bxe5?? Black wins by force. Instead, White should play 6.Qh5+ g6 7.Qh4 exf4 8.exf4 Bd7 9.Nf3 Nb4 10.Bxd7+ Qxd7 11.Na3 O-O-O or 6.fxe5 fxe5 7.Nf3 Bg4 8.h3 Bxf3 9.Qxf3 Nf6 10.Nc3 a6, in either case with approximate equality. Note that in the game continuation, 14...Bf5!, staying in the middlegame, was even better than allowing White to go into a lost endgame.

Note also that 10...Qxh1!, as played, was much better the tempting 10...Bg4?! I think that I once had the White side of this many years ago against James Fagan in a blitz tournament at the Elbo Room. White survives after 11.Nc3!, and now (a) 11...Qg6 12.Bxh8! Bxd1 13.Nxd1!, when White is only a little worse, or (b) 11...Qxh1 12.Qxg4 Qxg1+ 13.Ke2 Qxa1 14.Nxd5 Qh1 15.Qe6+ Be7 [15...Kd8 16.Nf6 (16.Bf6+ also draws) 16...Qg2+ 17.Kd3 Qf1+ 18.Ke4 Qh1+ 19.Kd3 with a draw by perpetual check] 16.Nxc7+ Kd8 17.Nxa8.

ADDENDUM: White can also grab the pawn immediately on move 7 without first playing Bxc6+. It turns out that this is a better try, although still weak. After 7.Bxe5 Qh4+ 8.g3 Qe4 9.Bxg7 Qxh1 10.Qh5+ Ke7 11.Qh4+ Kf7 12.Bxh8 Qxg1+ 13.Bf1 Bf5 14.Nc3 Nge7 15.Ne2 Qh1 16.Bc3 Ne5 17.O-O-O Qe4 18.Qxe4 dxe4 Black is up a piece for two pawns. The alternative 11.Kf1 is clearly winning for Black after 11...Be6 12.Be2 Be5 13.Bxh8 [or 13.Bxe5 Nxe5 14.Qxe5 Rf8+ 15.Bf3 Rxf3+ 16.Ke2 Qxg1 17.Qxc7+ (17.Nc3 Qg2+ 18.Kd3 Nf6) 17... Kf6] 13...Rf8+ 14.Bf3 Bxh8.