Mate by castling

Over four years ago, I published a game where I administered checkmate by capturing en passant. Here I deliver another very rare form of checkmate: mate by castling. This appears to be comparable in rarity to mate by en passant. The only other examples of it that I've been able to find are P. Morphy-A. Morphy, New Orleans 1850, and Kvicala-NN, Prague 1875. The famous game Ed. Lasker-Thomas, London 1912 is not an instance of this, since Edward Lasker in that game chose to play 18.Kd2# rather than 18.0-0-0#. Nor, it appears, is Prins-Day, Lugano (ol) 1968, despite the score of that game given at chessgames.com, which shows the game ending with 31...0-0-0#. According to Day, the winner of the game, who should know, the game actually ended after 28...Qe4+. He writes, "And my opponent resigned, unwilling to investigate 29.K-B6 Q-B4+ 30.K-N7 Q-N3+ 31.K-R8 and Black has a choice of absurd mates." Raymond D. Keene (ed.), Learn from the Grandmasters, p. 108. Mate by castling is thus sufficiently rare that I have been unable to find an example of it occurring in the entire 20th century!

Another trap in From's Gambit

DO NOT play 1.f4 unless you have an answer ready for From's Gambit, 1...e5!? King's Gambiteers can happily play 2.e4. The line White played in the game was fine up to a point. IM Tim Taylor in his book Bird's Opening recommends 7.d4 Ng6 8.Nxg6 hxg6 9.Qd3! Nc6 10.c3! Bf5 11.e4 Qe7 12.Bg2 0-0-0 13.Be3 Bd7 14.Bf2!, when Black doesn't have enough compensation for the gambit pawn. Another line is 5.d4 g4 6.Ne5!, heading into a slightly better ending for White after 6...Bxe5 7.dxe5 Qxd1+ 8.Kxd1. Instead of playing these reasonable lines, most of my blitz opponents have played ridiculously and gotten crushed. A case in point:

Beauty and the (Silicon) Beast

I was very pleased with the following blitz game, which ends with a queen sac and a beautiful mate administered by Harrwitz bishops. But Black could have avoided mate with 8...Kd8!, though he's still quite lost. When I analyzed the game later with Houdini 3, it pointed out an even clearer win for me on move 8! Do you see it?

NN triumphs!

Contrary to the impression you may have gotten from my prior posts, I also fall into opening traps and lose games, occasionally spectacularly. And sometimes NN is the router, not the routee. In the game below, he selects the tricky Budapest Gambit. It gives White a small advantage in the main line, but he must also reckon with trappy lines like the Fajarowicz Variation, 3...Ne4!?, beloved by Dean Arond, and gambit lines with ...d6 and ...f6. The last of these proves my undoing here.

Yet another

I may not know much about real chess, but I can still mate off random fish on the Internet.

The Sicilian busted?

About 30 years ago, FM Albert Chow told me that the Sicilian Defense was unsound, and that it was a matter of time before it was refuted. I laughed. He insisted that he was serious: while White's 1.e4 is an eminently logical move, 1...c5 was a silly move that does nothing for Black's development. Chow said that I had contributed to the ultimate refutation of the Sicilian with my win over David Sprenkle. I shook my head in disbelief. How could the venerable Sicilian, beloved by great players by Fischer and Kasparov, ever be busted?

Marc Esserman's book Mayhem in the Morra! features a newspaper headline on its cover: "Sicilian Refuted! Morphy and the Romantics Run Wild!" While I'm sure Esserman would admit that this is hyperbole intended to be funny and sell more books, he makes a much stronger case for the Smith-Morra Gambit than I could have imagined. As the following game illustrates, declining the gambit with 3...Nf6 (transposing to the Alapin Sicilian) is not a panacea. Was Albert right?

Black's resignation was premature, particularly given that this was a blitz game. The variation given at the end is Houdini 3's analysis of what might have happened.

Yet another trap in the King's Gambit

I have previously shown (here and here) that in the King's Gambit Accepted after 3.Nf3, 3...f6? 4.Bc4 g5? 5.Nxg5! is crushing. My opponent in the following game found a slightly different ...g5 and ...f6 "defense," with equally grisly results. Incidentally, Houdini 3 points out 5...h5! (stopping Qh5+) 6.Nh3 d5! as preferable to accepting the knight, although White still gets a huge advantage.

The play in the game continuation is reminiscent of the main line of Damiano's Defense, which goes 1.e4 e5 2.Nf3 f6? 3.Nxe5! fxe5 4.Qh5+ Ke7 5.Qxe5+ Kf7 6.Bc4+ d5 (best) 7.Bxd5+ Kg6 8.h4! In both lines White's queen, king bishop, and h-pawn kick Black's poor king all over the board.

My play was slightly inaccurate. White mates two moves faster with 11.h5+! Kh6 12.Qxf4+ Qg5 (12...Kg7 13.h6+! and mate next) 13.Qxf8+ Qg7 and now White can choose between the classical 14.d3# (Allgaier-NN, 1807) and the modern preference 14.d4# (Green-Kempfe, Minnesota 1982). By the same token, 15.h6+! would have forced mate a move faster than my 15.Qf7+.

One would be tempted to say, "Never play ...f6 when defending against the King's Gambit!", but that would be too dogmatic. As Bobby Fischer observed in his famous article A Bust to the King's Gambit, after 1.e4 e5 2.f4 exf4 3.Nf3 d6 (Fischer's Defense to the King's Gambit) 4.d4 g5 5.h4 g4 6.Ng5?!, 6...f6! 7.Nh3 gxh3 8.Qh5+ Kd7 wins a piece without too much trouble. But be very leery of ...f6 - it weakens Black's king-side, takes away the king knight's best square, and does nothing for Black's development.

The Art of Positional Play

Not really. I was in a bad mood because I had failed in my previous game, a Smith-Morra, in my attempt to "be like Marc." A positional player would have played 5...Nf6 (instead of my 5...e4!?) and 10...Bc5 11.Bd3 Qe7 (instead of my 10...g5?!). In either case, says Houdini, I would have had the advantage, whereas my "cowabunga" play could have enabled White to equalize. But hey, it worked!

Bishop and knight versus rook and pawn

Every beginner learns that a bishop or knight is worth three pawns, while a rook is worth five. It follows that a bishop and knight are worth six pawns, as is a rook and pawn, so exchanging bishop and knight for rook and pawn is an even exchange. In practice, however, it is rarely so simple.

Often a player has the opportunity to make such an exchange in the opening by giving up a bishop and knight for rook and pawn on the f7 or f2 square. This is rarely a good idea: the opponent's bishop and knight can get very active before the "extra" rook has a chance to get into play.

These themes are illustrated by the following game, where my minor pieces and queen whipped up a winning attack while my opponent's rook sat, unmoved, on a1. In the final position my opponent flagged while futilely searching for a defense against my dual threats of 24...Qxh2# and 24...Nxf2+ 25.Kg1 Qxg2#.

Knight forks tutorial

Since we're on the subject of forks, here's an amusing game. White's play was execrable, but the game features two knight forks, one in a variation, the other in the game continuation. White played a sort of demented Kopec System against my Sicilian. Houdini 3 doesn't like my 5...Ne5, winning the bishop pair but exchanging off my opponent's bad bishop, preferring 5..Nf6 with a small advantage. White should have responded to my 6...Bh6!? (see diagram) with 7.Qe2, when the position is about equal. After his 7.Qg3??, I already had a winning advantage! Do you see how?

After 7...Bf4! the bishop was immune since 8.Qxf4? Nxd3+ would fork White's king and queen. So 8.Qh3 was forced, when 8...d5! discovered an attack on White's queen. Now the only way to save White's bishop would have been the abject 9.g4 Bxg4 10.Qf1. He missed this, playing 9.Qh4??, when 9...Nxd3+ won a piece and left White's queen in mortal danger. He fell into a knight fork with 10.Kd1? g5! 11.Qh5 Bg4+! 12.Qxg4 Nxf2+, and resigned.

More Smith-Morra fun

The Deadly Isolani

Positions where White has an isolated pawn on d4 (IQP) arise in many openings, including the Queen's Gambit Accepted; Nimzo-Indian; Caro-Kann Defense, Panov-Botvinnik Variation; and Alapin Sicilian. In the game below, the IQP occurred in a Smith-Morra Gambit Declined, which transposed into an Alapin Sicilian. If Black wants to decline the Smith-Morra Gambit, 3...d5 is not the most effective way to do so; Black ends up in an Alapin Sicilian where he has exchanged on d4 prematurely. I recommend either accepting the gambit or playing 3...Nf6.

A lot of players are leery of taking on an IQP, fearing that it will be weak in an ending. However, as Tarrasch said, "Before the endgame, the gods have placed the middlegame." He also said, "He who fears an isolated queen's pawn should give up chess." The IQP supports a knight outpost on e5 and often results in a king-side attack for White.

Another advantage of the IQP is that a sudden liquidation of the center with d5 is sometimes dangerous for Black, as seen in the game below. Although Black's opening moves were very plausible, 12.d5! already left him tactically busted in view of his loose pieces on c6 and b4, and the unfortunate vis-à-vis of Black's queen and White's rook on the d-file. Houdini 3 gives Black's best line as 12...Na5 13.dxe6 (unleashing a discovered attack on the queen) Qe7 14.exf7+ Kh8. After 12...exd5 13.Nxd5 Nxd5 14.Bxd5, White threatened both 15.Bxc6 and 15.Bxf7+, winning Black's queen. Black tried to defend everything with 14...Qc7, but I unleashed a whirlwind attack on his king with 15.Qe4! Bb7 16.Ng5! g6 17.Qh4!, forcing the ugly 17...h5. Houdini says that I should have continued with 18.Qe4! (threatening Qxg6+) Kh8 (even worse is 18...Kg7 19.Bxf7! Rxf7? 20.Ne6+, winning the queen) 19.Bf4! (not 19.Bxc6? Qxc6! 20.Qxb4?? Qxg2#) Qe7 20.Bxc6, winning a piece. My 18.Bf4?! was less accurate, when Black could have played on with 18...Qe7. After 18...Bd6??, Black resigned before I could play 19.Bxd6 Qxd6 20.Bxf7+! Rxf7 21.Rxd6.

Black to play and win

The discerning reader will ask why I didn't play 5...h6! Alas, I had premoved 5...Bxc5. My opponent obligingly allowed me to trap his knight with ...h6 two moves later. The rest of the game was a rout.

More Morra Mayhem

Marc Esserman has demonstrated both in his book Mayhem in the Morra! and in tournament practice that piece sacrifices on d5 can be very strong in the Smith-Morra Gambit (1.e4 c5 2.d4 cxd4 3.c3). In both Esserman-Van Wely, 2011 and Esserman-Sarkar, 2008 he successfully sacrificed a knight on d5, annihilating a GM and an IM in 26 and 16 moves, respectively. In the following game I tried to "be like Marc," but maybe my 10.Nd5!? was premature. The silicon monster thinks that Black is better if he declines the knight with 10...Na5! My opponent made the mistake of accepting it instead and was swiftly dispatched.

A novelty!

Shockingly, I encountered an Internet opponent tonight who doesn't read Chicago Chess Blog! As such, he had failed to read my post from last night demonstrating that 3...f6 and 4...g5 is not an effective response to the King's Gambit. But I wasn't content with 7.Qf7+ and mate in two more moves (7...Kd6 8.Qd5+ Ke7 9.Qe5#) as in last night's partie, and uncorked the novelty (at least for me) 7.Qxg5+! I learned this trick from Wall-Greenwalt, Dayton 1983. If Black responds with 7...Kd6, he gets mated the same way - just without his g-pawn - by 8.Qd5+ Ke7 9.Qe5#. Or he can play 7...Ke8, and get mated by 8.Qh5+ Ke7 9.Qe5#, as Greenwalt did. But NN played a novelty himself with 7...Nf6, allowing me to mate a move faster with 8.Qe5#!

Silly Internet Game No. 1048

This King's Gambit concludes with a sacrificial flurry. Had Black played 24...fxg6, 25.d6+ would have mated quickly; ditto with 24...Bg7 25.Rxg7+ Kxg7 26.Nxf7+. Incidentally, in the opening 5.e5 Nd5 6.Bc4 was better than my 5.d5, when Black would have been OK after 5...Ne7. On move 16, Houdini 3 points out the powerful shot 16.Nxf7! with a winning advantage.

A trap in the King's Gambit

In the King's Knight's Gambit (1.e4 e5 2.f4 exf4 3.Nf3), 3...h6 intending 4...g5 is a respectable line. 3...f6 with the same idea doesn't work so well. This game duplicates Trap No. 133 in Irving Chernev's book Winning Chess Traps, except that in that game Black played 3...g5 and 4...f6?

The Halloween Gambit

Here's a game with that scariest of openings, the Halloween Gambit (1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6 4.Nxe5?!). It may be dubious if Black responds perfectly. But if not, White gets frightening attacking chances:

A trap in the London System

In the game below, 4...e5!, forking White's knight and bishop, wins a piece since Black's pawn, seemingly hanging, is indirectly guarded by ...Qa5+. My inspiration for this trap was NN-Pandolfini, a Kan Sicilian with the same motif: 1.e4 c5 2.Nf3 e6 3.d4 cxd4 4.Nxd4 a6 5.Bf4?? e5! The concept was also seen in Zaitsik-Zichulidze, Tiflis 1976 , which is the same as my game, but with colors reversed and White having the extra move g3.

A trap in the Vienna Game

If Black plays symmetrically against the Vienna Game (1.e4 e5 2.Nc3) with 2...Nc6 3.Bc4 Bc5, White can set a nice trap with the primitive 4.Qg4! threatening 5.Qxg7. This is surprisingly awkward for Black to meet: 4...Bf8 is obviously unattractive, 4...g6 is weakening, and no one wants to move their king with 4...Kf8 (ugly, but perhaps best). The most plausible move, and indeed Black's most common move, is 4...Qf6?, meeting queen move with queen move and counterattacking the pawn on f2. However, White gets a large advantage with the shocking 5.Nd5!, threatening both 6.Nxf6+ and 6.Nxc7+. Since the queen can't defend g7 and c7 at the same time, and 5...Bxf2+? would be met by 6.Kf1!, winning material, 5...Qxf2+ is forced.

After White's also forced reply 6.Kd1, White's king is surprisingly safe, while Black's is in trouble. White again threatens both 7.Qxg7 and 7.Nxc7+. 6...g6!?, hoping for counterplay with 7.Nxc7+ Kd8 8.Nxa8 d5! threatening White's queen and bishop simultaneously, is met by 7.Nh3! Qd4 8.d3 (threatening 9.c3, trapping the queen) Bb6 (8...d6 9.Qf3) 9.Qf3 (threatening 10.Nxb6 axb6 11.Qxf7+ Kd8 12.Qf8#) d6 10.c3 Qc5 11.b4, winning material, as in Emms-Hawkworth, where future grandmaster Emms upset his higher-rated opponent:

Instead, Black usually tries to meet both of White's threats with 7...Kf8. Then 8.Nh3 forces 8...Qd4 (8...d6? 9.Nxf2 Bxg4+ 10.Nxg4 wins a piece), when 9.d3 threatens to trap the queen with 10.c3. The following games well illustrate White's possibilities: