380. 2 May 2008: Wonderful fourmover

In the latest issue of the Belgian electronic problem magazine HORIZON (4 issues a year for € 5 - in Dutch), I found an old fourmover so wonderful that I'm sorry I did not try to solve it myself.

Mate in 4
Opificius (Oswald von Krobshofer)
Schach Express 1948
Dedicated to Emil Ramin

The key is the baffling 1.Ra1 This threatens 2.Ra5 and 3.Be3 (or Rd5) mate, so the Rook must be captured (1...Bxb4 2.Ra4 or 1...Bf7 2.Nxf7): 1...Bxa1 And now a second improbable rook move: 2.Rh4 after which it turns out Black is in Zugzwang: 2...Bh(f)7 3.N(x)f7 and 4.Be3 mate, or 2...Bb2/c3 3.Bc1/d2 and now 3...Kxe5 4.BxB mate, or 3...BxB 4.Nf3 mate.

That leaves the question of why 1.Rh4 doesn't work. The threat is then 2.Bd2, again with the double mating threat. Black refutes this with 1...Bb2! with the point that c1 is not available for the double threat. That is why 1...Ba1 does not work now; after 2.Rb1! (2.Rxa1? Bf7! 3.Nxf7 stalemate) Black is again in Zugzwang, just as he is after 2.Rh4 in the main line. But after (1.Rh4?) Bb2! this Bishop can still flee to a1, preventing the attack by Bf4. This reveals the main point of the key 1.Ra1: the Bishop is lured to a1, so it cannot go there anymore, and hide from the Bf4.

NB: Noam Elkies notes that after 1.Rh4? Be1? is not a refutation; White then mates in time with 2.Be3+ Kxe5 3.Bd4+ Kxd4 4.Nf3 mate. He also remarks that the stalemate I mention above, after 1.Rh4? Ba1? 2.Rxa1 Bf7 3.Nxf7?, is not at all forced; White plays 3.exf7 and 4.Nxe6 mate.

The fact that 1.Ra1 is a 'short' threat (2.Ra5 and 3.mate, in a fourmover) and a multiple threat (2.Ra8(7) Bxb4 3.Ra4) is a small blemish. Elkies suggests adding a white pawn on a5, when 1.Ra1 only threatens 2.Ra4, 3.b5+ and 4.mate and there are no duals.

Oswald von Krobshofer (1883 - 1960) was an artist painter, born in Prague. Horizon assumes he chose a pseudonym because right after the war, many Germans preferred to appear a little less German. Opificius does not seem to be a Latin word, but it can be found as a family name.

379. 22 April 2008: Stunning computer move in a pre-computer game

Jan Greben was a strong player in Holland around 40 years ago. I once played him in a tournament, the lowly (but not so weak) "IBM-III" in Amsterdam 1967, which became memorable to me mainly because I first met and played 15-year old Jan Timman there. Later, Greben moved to South Africa where he became, apart from a physical scientist, "Pretoria's most dominant player of the 1980's." I'm quoting yesterday's Pretoria News chess column by Mark Rubery, which Greben sent to me, because it has a game of his from that old tournament. Greben beat me there in a nice game, and he also beat Timman. He didn't have the score of that game anymore and when after many years he found it again at chessgames.com and showed it to Rubery, this chess editor made some wonderful discoveries.

Timman - Greben, Amsterdam 1967
1.d4 Nf6 2.c4 g6 3.Nc3 Bg7 4.e4 d6 5.f3 c6 6.Be3 a6 7.Qd2 b5 8.Bd3 bxc4 9.Bxc4 d5 10.Bb3 dxe4 11.fxe4 Ng4 12.Nf3 Nxe3 13.Qxe3 O-O 14.O-O Qb6 15.Na4 Qa7 16.Nc5 Nd7 17.Rac1 a5 18.Ng5 Nxc5 19.Rxc5 e6 20.e5 a4 21.Bc2 Qb6 22.Ne4 Ba6 23.Nf6+ Kh8 24.Rf3 Qxb2 25.Bd1 Rad8 26.Rc3 Qb4 27.Rf4 Bh6
see diagram
Timman had been clearly better, until with 24.Rf3 (Rf2!) he started asking too much of the position. He's lost now, which he more or less acknowledged with 28.Rh4 Bxe3+ 29.Rxe3 h5 30.Bxh5 g5 31.Rhh3 Kg7 32.Rhg3 Qb1+ and White resigned.

In the diagram, Rubery now discovered "the cunning resource" 28.Qd2 when Bxf4? immediately loses to the devilish 29.Rh3! White could very well still have won that way, because what can Black do against the threat 29.Rh4?
    It is unlikely that any player would have found the only move with which Black could still have won: the improbable 28...Bd3!! A true computer move, which blocks the Rc3's access to h3, and the Qd2's access to d4, and therefore threatens Bxf4 as well as Qxd4+ As both 29.Qxd3 Bxf4 and 29.Rxd3 Qxd2 are pointless, White's best try would be 29.Rh4 Qxd4+ 30.Rxd4 Bxd2 31.Rxd8 Rxd8 but the resulting endgame would be completely lost.

378. 25 March 2008: New record by Alexey Khanyan - 51 consecutive checks

The history of the consecutives checks task (legal position, promoted pieces allowed) has a new chapter.

In 1974, Werner Frangen published a position with a series of 45 checks (see entry 308 in this Diary.) In August 2007 (entry 351), Sampsa Lahtonen took up this challenge and composed a position with 46 checks, immediately breaking his own record and setting it at 47. In December 2007 his compatriot, Finnish composition GM Unto Heinonen, went 2 better and showed a position with 49 consecutive checks. (Entry 374.) Almost immediately (entry 375 of January 2008), Lahtonen struck back with 50 checks.

And now Alexey Khanyan from Russia has managed to improve even that.

51 consecutive checks (legal position, promoted pieces allowed)
Alexey Khanyan
Original, March 2008

1.f8Q+ Ree7+ 2.Qf6+ Nde6+ 3.Kc4+ Nd3+ 4.Qff4+ Nexf4+ 5.Re6+ Nxe6+ 6.Qf4+ Ndxf4+ 7.Rbd3+ Nxd3+ 8.Bf4+ Ndxf4+ 9.Rad3+ Nxd3+ 10.Ne4+ Qxe4+ 11.Qd4+ Bd5+ 12.Nxd5+ Rc7+ 13.Nxc7+ Bd5+ 14.Nxd5+ Rc7+ 15.Nxc7+ Qd5+ 16.Nxd5+ Qc7+ 17.Qc5+ Ndxc5+ 18.Rd3+ Rf4+ 19.Nxf4+ Bd5+ 20.Nxd5+ Qf4+ 21.Nxf4+ Nxd3+ 22.Bxc7+ Nxc7+ 23.Nf7+ Qxf7+ 24.Ne6+ Qf4+ 25.Bxf4+ Ne5+ 26.Bxe5+
The moves can be played over on the left.

For one thing, the position must be legal. Each side has promoted six times, which can be explained by the disappearance of one white pawn and two black ones, and the Bf1 and Bf8.

Khanyan remarks that his record starts with 7 non-capturing checks when the longest such series in the previous records was 3. On the other hand, his sequence has 4 consecutive captures (moves 21-22), when none of the predecessors had more than 3 consecutive captures. Most important however is that in Khanyan's final position, 3 pieces remain (apart from the Kings) which in his opinion suggests "that 52 checks in a row is not a completely impossible result. I am going to look for them."

Others will too, I guess.

377. 11 March 2008: The Wizard outwitted

Shortest proof game
Bader al-Hajiri
Original, 2007

This diagram again - I already gave it in entry 371, asking for the shortest proof game, and in entry 373 when I published the solutions.
    The composer, Bader al-Hajiri, had given a 32.5-move game, and this number was equalled by nine solvers with a total of eleven quite different games. I wrote: "Can it be done in 31.5 moves, or even less? Among the winners who said something about this, there is a strong majority, including Al-Hajiri himself who is a proof game expert, that this is highly doubtful."

It was quite unexpected then, when I recently received this game by Mario Kamody.

1.g3 f5 2.Bg2 Nc6 3.Bxc6 g5 4.d3 a5 5.Qd2 Kf7 6.Qxg5 h6 7.Qxe7+ Kg6 8.Bxd7 Kh5 9.Kd2 Kg4 10.Nf3 Kh3 11.Ng1+ Kg2 12.Bxf5 Kf1 13.Ke3 Ke1 14.Kf3 Nf6 15.Qxc7 Nd5 16.Kg2 Nc3 17.Bxh6 Qg5 18.Na3+ Kd2 19.Re1 Nd1 20.Kf1 b5 21.Nxb5 Kc1 22.Qxa5 Kb1 23.Rxd1+ Qc1 24.Bxc1 Bh6 25.Qe1 Bd2 26.Rxd2 Bb7 27.Qd1 Bg2+ 28.Ke1 Rh3 29.Bxh3 Bf1 30.Bxf1 Ra3 31.Nxa3+ Ka1 32.Nb1 and Al-Hajiri's position is reached.

So it could be done in one move less. Bravo! Or, as the Kuwait Wizard himself commented: "Wow!! I'm shocked!!"

376. 16 February 2008: An old lesson from dad

In his weekly chess column of today's Gooi en Eemlander, Johan Hut gives this position from a Dutch team match of a few years ago.

Black to play
A. Lont - T. van der Heijden
tt Netherlands, 2001

Black played 1...Kf6 and resigned without waiting for 2.Kf3, which will pick up the Ph4.

Some time later, Addy Lont's father Jarig found a wonderful draw for Black: 1...h5! 2.g5 h3 3.Kf3 h4 4.Kf2 Kf5 5.Kf1 Kxf4! 6.g6 Kf3 7.Kg1 Kg3 8.g7 h2+ 8.Kh1 Kh3 and White can only choose between two stalemating and two useless promotions.
It should be noted that 1...h3 also draws - after 2.Kf3 h5 3.g5 h4 it boils down to a transposition of moves.

It's a pity it didn't really happen, because this stalemating idea has never occurred in practice. But there are over 50 studies in which it was used.

White to play and draw
T. Gorgiev
Shakhmaty v SSSR, 1950

1.h6 Kf8 2.h5 Kf7 3.Ke3! and not 3.Kf4 f5 4.Kg3 Kg8 etc. - White must wait for the black King to be at the back rank 3...Kf8 4.Kf4 f5 and we have the same position as after 5.Kf1 above, colors reversed. 5.Kxf5! g3 6.Kf6 Kg8 7.Kg6 g2 8.h7+ Kh8 9.Kh6 g1Q stalemate.

White to play and draw
B. Badai
Shakhmaty v SSSR, 1965

A struggle around Black's last pawn. 1.Kf2 Nh3+ 2.Kf3! and not 2.Kg3 Nf4 3.Kg4 Ne6 4.Kf5 Kf7 and White is in Zugzwang and must allow 5.Ke5 g4 6.Ke4 Kg8 etc. 2...Nf4 3.Ke4! Kf7 (3...Ne6 4.Ke5 transposes) 4.Ke5! Again, not 4.Kf5 Ne6 and White is in Zugzwang. 4...Ne6 (4...Kg8 5.Kf5 is the same) 5.Kf5 And now Black is in Zugzwang. 5...Kg8 6.Kxe6 g4 7.Kf5 (7.Kf6?? Kh7) 7...g3 8.Kg6 g2 9.h7+ Kh8 10.Kh6 g1Q stalemate.

PS 17 February: Olli Heimo did what I should have done: check this with the tablebases. He found a cook: 3...Ne6 4.Ke5 Nd8! followed by Nf7 and Black wins the struggle to keep his last pawn.

White to play and win
F. Bondarenko and A. Kuznetsov
1st Prize, Tidskrift för Schack

Often, the idea is used to motivate an underpromotion on g8. Here, there are two underpromotions. The threat is 1...h2 and 1.Kg1 Kg3 is not good.
1.Nb7! And not 1.Nc6 h2! 2.Kxg2 Bxc6+ and Bxd7 and Black wins. 1...g1Q+ Black's best try. 2.Kxg1 Kg3 3.d8N h2+ 4.Kh1 Kh4 Best try again; after 4...Kh3 etc., Black ends up on the wrong foot, and White promotes to Queen with check 5.g6 Kh3 6.g7 h4 and now not 7.g8Q Bxb7+ 8.Nxb7 stalemate, and also not 7.g8N with a theoretical draw, but 7.g8B and White wins.

The idea seems to have originated with Karl Jänisch (1813-1872). He too, used it to show underpromotions.

White to play and win
C. Jänisch
Schachzeitung, 1850

1.Qxg5+ Rxg5 2.fxg5 h2 3.g6 h4 4.g7 Kh3 5.g8B! and White wins. But not 5.g8N Kg3 6.Nf6 Kf3 7.Nd5 Ke4 8.Nc3+ Kd3 9.Nb5 Kc2 10.Nxa3+ Kb2 with a draw.

Adding wPb3 and bPb4 changes the story; now 5.g8B does not win (it even loses) but 5.g8N does.

375. 28 January 2007: Sampsa Lahtonen reclaims his record - 50 consecutive checks

Hardly had I published Heinonen's record of 49 consecutive checks (entry below), when Sampsa Lahtonen, who previously held this record with 47 (entry 351), sent me this position.

50 consecutive checks (legal position, promoted pieces allowed)
Sampsa Lahtonen
Original, January 2008

1.Qa5+ Nac5+ 2.Nxc5+ Rb3+ 3.Nxb3+ Nc5+ 4.Nxc5+ Rb3+ 5.Nxb3+ Bb5+ 6.Nxb5+ Rd6+ 7.e4+ Nxe4+ 8.Nd2+ Qb3+ 9.Qxb3+ Ke5+ 10.Qxd6+ Nxd6+ 11.R4f5+ Nxf5+ 12.Bfd6+ Nxd6+ 13.Rf5+ Nxf5+ 14.Qd5+ Bxd5+ 15.Nf3+ Bxf3+ 16.Bd6+ Nxd6+ 17.Re4+ Bhxe4+ 18.Rxe4+ Nxe4+ 19.Nd7+ Qdxd7+ 20.Nd4+ Qb5+ 21.Qxb5+ Nc5+ 22.Qxc5+ Bd5+ 23.Nf3+ Qxf3+ 24.Qe3+ Qe4+ 25.Qxe4+ Bxe4+

50 consecutive checks - a wonderful new record. "You will notice that the final blizzard slightly resembles that of Heinonen's 49," Lahtonen writes, adding: "There are many ideas out to try, and I would be surprised if this stands as a record for long."

It should also be noted that Lahtonen uses the same number of pieces (27) as Heinonen and that this is the first construction in this domain with only three pieces in the final position - the two Kings, and the piece that gave the last check.

Lahtonen's own proof game for the legality of his position can be played over on the left.

374. 12 January 2008: New checks record

Not so long ago, in entry 351 on 12 August, I gave a dazzling new record by Sampsa Lahtonen, of 47 consecutive checks (legal position, promoted pieces allowed.) Now this record has been broken by his countryman, Finnish composition Grandmaster Unto Heinonen, with not one, but two checks more, which brings it to forty-nine.

49 consecutive checks (legal position, promoted pieces allowed)
Unto Heinonen
Suomen Tehtäväniekat, December 2007

1.Ndc4+ Nd6+ 2.Ndf4+ Nxf4+ 3.Ng2+ Be2+ 4.Rxe2+ Ne3+ 5.Ncxe3+ Nd5+ 6.Qe4+ Rxe4+ 7.Kg5+ Rg4+ 8.Nexg4+ Q3e3+ 9.Qbxe3+ Qgxe3+ 10.Ngf4+ Nxf4+ 11.Qd5+ Nxd5+ 12.Nf4+ Nxf4+ 13.Nd5+ Nxg6+ 14.Ngxe3+ Rg4+ 15.Nxg4+ Qe3+ 16.Nf4+ Nxf4+ 17.Qg6+ Nxg6+ 18.Nxe3+ Qf5+ 19.Nxf5+ Be3+ 20.Rxe3+ Qxe3+ 21.Qxe3+ Ne4+ 22.Qxe4+ Ne5+ 23.Ng7+ Qxg7+ 24.Qg6+ Qf6+ 25.Qxf6+

However, Heinonen sees it as a blemish that on move 10 and 12, White can switch his Knight moves to f4, all the other checks remaining the same. I don't see this as a blemish myself - this is about a length record. But Heinonen also wants to keep the following position where no such switches are possible, be it at the cost of one check.

48 consecutive checks (legal position, promoted pieces allowed)
Black to play
Unto Heinonen
Suomen Tehtäväniekat, December 2007

1...Nd5+ 2.Nef4+ Ngxf4+ 3.Ng2+ Be2+ 4.Rxe2+ Ne3+ 5.Ncxe3+ Nd5+ 6.Qe4+ Rxe4+ 7.Kg5+ Rg4+ 8.Nexg4+ Q3e3+ 9.Qbxe3+ Qgxe3+ 10.Ngf4+ Nxf4+ 11.Bd5+ Nxd5+ 12.Nf4+ Nxf4+ 13.Nd5+ Nxg6+ 14.Ngxe3+ Rg4+ 15.Nxg4+ Qe3+ 16.Nf4+ Nxf4+ 17.Qg6+ Nxg6+ 18.Nxe3+ Qf5+ 19.Nxf5+ Be3+ 20.Rxe3+ Qxe3+ 21.Qxe3+ Ne4+ 22.Qxe4+ Ne5+ 23.Ng7+ Qxg7+ 24.Qg6+ Qf6+ 25.Qxf6+

All those checks can be played over on the Palview board.

Thanks to Bader al-Hajiri who offers proof games for the legality of both positions (which can also be played over on the left) and to Joose Norri for bringing these positions to my attention.

373. 9 January 2008: The Christmas puzzle solved

In item 371, I asked for proof games for this position.

Proof game?
Bader al-Hajiri
Original, 2007

A well-known grandmaster wrote: "There is simply no way that that position is legal - NO WAY" - but I would never send my readers on impossible missions and of course, when he sent me the position, Bader Al-Hajiri had already included his own proof game of 32,5 moves, or 65 ply.

1.d3 h6 2.Bxh6 f5 3.Qd2 f4 4.Qxf4 a5 5.Qxc7 Kf7 6.g3 Kg6 7.Bg2 Kh5 8.Bxb7 Kg4 9.Nf3 Kh3 10.Bxc8 e5 11.Bxg7 e4 12.Kd2 e3+ 13.Kxe3 Kg2 14.Ng1 Kf1 15.Kf3 Ke1 16.Qxa5+ Bb4 17.Nc3+ Kd2 18.Rf1 Rh3 19.Bxd7 Nh6 20.Nd1 Kc1 21.Bxh6+ Kb1 22.Bc1 Na6 23.Kg2 Rc8 24.Bxh3 Rc3 25.Nxc3+ Ka1 26.Nb1 Nc5 27.Rd1 Be1 28.Qxe1 Ne4 29.Kf1 Nd2+ 30.Rxd2 Qd5 31.Qd1 Qg2+ 32.Ke1 Qf1+ 33.Bxf1

I received twenty entries. Nine solvers tied for first with games where the position was also reached on move 33 - one of them, Mario Richter from Germany, producing three different ones. The others with 33 were Thomas Beuman, Martin van Essen, Paul Boxebeld (Netherlands), Olli Heimo, Sampsa Lahtonen (Finland), Aaron Guthrie (Australia), Renny Bosch (USA) and Olav Bakke (Norway). Congratulations to them all - I'll send a little prize. Other proof games were in 34, 35 (3), 36, 38, 39 (2), 44, 47 and 49 moves.
    It was amazing to see in how many different ways Al-Hajiri's position can be reached. In the 32.5-move games, all the white Kings take the same route back to e1, but Van Essen's King picks up two pieces in the process, while most do it without a capture. The Rd2 gets there by moving, and by capturing a Queen, a Bishop or a Knight. The white Queen makes from only 4 moves (Beuman, Boxebeld) to as many as 10 (Bosch, Richter-2), capturing between 2 and 7 pieces. The black King reaches the back rank as early as on move 10 (Van Essen) and as late as on move 20 (Richter-1.) White's last moves are BxQf1; BxBf1; BxNf1; Bf1 and Nb1. Al-Hajiri notes he is the only one with Ra1-f1; all the others use Ra1-e1.
    In all the solutions, moves can be interchanged. But then again, as Al-Hajiri says, this was not about a unique solution, as is normally required for proof games, but about the shortest possible solution.
    Can it be done in 31.5 moves, or even less? Among the winners who said something about this, there is a strong majority, including Al-Hajiri himself who is a proof game expert, that this is highly doubtful. However, seeing all the black waiting moves near the end, especially in Heimo, Bakke, Richter-1 and 2 and Boxebeld, you get the impression that it might just be done. Perhaps someone, seeing all the tricks used here, will manage.
    The 11 winning proof games can be played over on the left, and can be downloaded here as a PGN-file.

372. 7 January 2008: An amazing King move

On one of the last days of 2007, the lowly rated Finnish player Tommi Luukkonen defeated a grandmaster with one of the most amazing moves of the year.

Kulaots - Luukkonen (2238), Rilton Cup, Stockholm, 27 December 2007
1.e4 c5 2.Nf3 e6 3.d4 cxd4 4.Nxd4 a6 5.Nc3 b5 6.Bd3 Qb6 7.Nf3 Nc6 8.O-O Qb8 9.e5 Nxe5 10.Nxe5 Qxe5 11.Qf3 Qb8 12.Bf4 Bb7 13.Qg3 d6 14.Nxb5 axb5 15.Bxb5+ Kd8 16.Rfd1 Bd5 17.c4 Nf6 18.a4 Ne4 19.Qe3 Bb7 20.Rd4 g5 21.Be5 Rg8 22.Rad1 f5 White's piece sacrifice was correct, and he had an easy win here with 23.Rxd6+ or 23.Bxd6, both based on a later Qxe6. 23.f3 Winning, too. 23...Ke7 24.Rxe4? But now he had to play 24.Rxd6 Nxd6 25.Bxd6+ Qxd6 26.Rxd6 Kxd6 27.Qb6+ etc. 24...dxe5 (Kf7!) 25.Rxe5 Be4 Here the computer gives 26.Rc5 as the easiest win, but it was difficult to see what was wrong with 26.Qd4
see diagram
26...Kf6! Stepping into a discovered check and sacrificing a piece at the same time. I have looked for, but not found, another example of such a move. It is the best move, but it should not be winning.
27.Rxe4+? The computer gives this series of typical computer moves as leading to equality: 27.Rxf5++ Kxf5 28.c5 Bc2 29.Qc3 Ra7 30.b4 Bb1 31.Rxb1 Qe5 32.Qd2 After 27.Rd5+ Kg6 28.fxe4 exd5 29.exf5+ Kh6 30.Rd3 g4 31.c5 however, White, a Rook down, is supposed to be somewhat better.
In the game, after 27...e5 28.Qd5 (Qf2!) fxe4 29.Qc6+ Kg7 30.Qxe4 Bc5+, White went steadily downhill: 31.Kf1 Ra7 32.Rd5 Bd4 33.c5 Rf8 34.b4 Rf4 35.Qc2 g4 36.Ke2 Raf7 37.Qb3 gxf3+ 38.gxf3 Rxf3 39.Qxf3 Rxf3 40.Kxf3 Qf8+ and White resigned.

Thanks to Kari Tikkanen for tipping me - the game is commented by Ülar Lauk (in Estonian) in his blog .

371. 24 December 2007: Christmas puzzle from Kuwait

Bader al-Hajiri, the Wizard from Kuwait, proposes this Christmas puzzle.

Proof game?
Bader al-Hajiri
Original, 2007

The position is legal, so a proof game can be constructed.

I'm offering a little prize for the shortest one, or a few prizes if more really good solutions come in at timkr, address xs4all.nl, before or on 4 January.

PS 7 January 2008: 21 proof games are in, from 19 solvers. In a few days, when I have the comments from Bader al-Hajiri, I will give the results.

370. 16 December 2007: Spotting progress in a tablebase endgame

Still on the subject of Karjakin - Shirov (see item 369 below), Kevin Begley quotes me from my story Perfection in Chess: "The 50-move rule would prevent winning these endgames over the board, but as they have never occurred in games, and probably never will, and couldn't be won by humans anyway, there is no need for a special 250- or 300- move rule."

"Karjakin-Shirov demonstrates that this ending can occur", writes Begley - in fact, as I now discovered, it was the third occurrence in around 2,4 million games. The other two, Boriss - Duschek, 1990 and Swathi - Laxman, 2006, were not in my database when I wrote that. More interestingly, both were theoretical wins, in 211 and 193 moves respectively, and in both a draw was agreed within 5 moves, when the positions were still winning.

The related RN vs. NN endgame has happened a little more often, and the results suggest (as did Karjakin - Shirov) that it is more difficult for the defender than for the attacker. In four of the seven times it arose it was a win, and in three of those four, it was won. The mate in 142 was won in 22 moves; the mate in 23 in 14 moves; the mate in 26 was drawn, and the mate in 44 was won in 16 moves. The three cases where it was a theoretical draw, all did end in draws.

"Two super GMs played this endgame remarkably well under considerable time pressure," Begley also wrote, "which suggests maybe some of the logic is not entirely beyond human understanding." I agree that R+B and R+N have good practical winning chances against NN, even within 50 moves - but here's a little test, from a much more complicated tablebase endgame.

These two positions, both with White to move, and presented in the alphabetical order of their filenames, are snapshots from Bourzutschky & Konoval's 517-move (Distance to Conversion) win in QN vs. RBN that I gave in item 316. If humans are to really understand anything about any tablebase win, it should at least be possible to recognize progress in such an ending - in this case: to tell in which diagram White is closer to the win, and, approximately, how much closer. The answer can be found on the linked page, but it might be fun to guess.

I will give the answer in a few days.

PS 8 January 2008: Four readers sent their guesses and their 2-2 vote is a nice illustration of the near-invisibility of progress in this type of endgame.
    White needs two hundred perfect moves to get from the diagram on the left to the one on the right - they are the positions after Black's 250th and 450th moves, respectively.
    Antonio Torrecillas made an educated guess that White "has made much progress in second diagram", based on the coordination of the black pieces (which is better in the first diagram); the coordination of the white pieces ("not a big difference"), the white King's activity ("in second diagram it is cooperating in the attack") and the safety of the black King where again, there is no great difference.
    Be that as it may, what to think of the diagram on the right? The black pieces seem to be cooperating reasonably, and the white King is even further away from joining the attack than in the first diagram - yet this is the position after Black's 500th move, when White is really close to the winning Rook capture at move 517.

PS 31 January 2008: Mark Thornton showed me another amazing example. The two positions below (again in the alphabetic order of their filenames) are from the 179-move tablebase endgame win KBBPKNN, found by Bourzutschky & Konoval. In both, Black is to move. One position arises at move 24, the other at move 124. But which is which?

I'll give the solution in a few days. I welcome guesses, but readers who want to know now, can play over the whole endgame at Gothic Chess.

PS 7 February 2008: No guesses this time, so here is my own. I thought the h3/e5 position had to be the one nearer the end because the pawn is free to advance and White's King is free to support it. Of course this obvious guess was wrong - which is precisely what makes this such a nice pair. In the position on the right, White must play a hundred perfect moves to reach the seemingly inferior position on the left.

369. 11 December 2007: Shirov in the face of Total Truth

After that Khanty-Mansiysk game Carlsen - Cheparinov, where the players blundered five half-points in a row in an endgame, a reader, Andras Roboz, wondered how special such series are, and what tablebase wisdom might reveal in this field.

Today's first rapid game in the semi-final play-off Karjakin - Shirov offered a unique opportunity to compare human playing strength with the Total Truth of the tablebases, when at move 49 the notorious RB vs. NN endgame arose.
    This once became famous as the tablebase ending with the longest shortest win (DTC; distance to conversion), of 223 moves; this was soon beaten by RN vs. NN with a DTC of 243 moves and a DTM (distance to mate) of 262 moves. (See Perfection in Chess, elsewhere on this site.) These days, the record length (DTC) for such endings is 517 moves, found in QN vs. RBN (see item 316 in this Diary.)
    Taking into account that most tablebase moves are beyond human comprehension, and that Shirov and Karjakin only had 10 seconds per move, they didn't do badly.

Position after 49.Nxe4
Karjakin - Shirov
Khanty-Mansiysk, rapid, 11 December 2007

There followed 49...Ra2 50.Nf4+ Kh6 51.Ne6 Ra5+ 52.Kf4 Bd1 53.Nd4 Kg6 54.Nc3 Bh5 55.Ne4 Rd5 56.Ke3 Rd8 57.Nc6 Re8 58.Kd4 Ra8 59.Ne5+ Kg7 60.Nd6 Kf6 61.Ne4+ Ke6 62.Nc5+ Kf5 63.Nc6 Re8 64.Nb4 Rd8+ 65.Ke3 Be8 66.Nbd3 Bb5 67.Nb4 Rh8 68.Nbd3 Rh3+ 69.Kd4 Rh4+ 70.Ke3 Rc4 71.Nb3 Re4+ 72.Kd2 Rh4 73.Kc3 Ke4 74.Nd2+ Kd5 75.Nb4+ Kd6 76.Nc2 Rh3+ 77.Kb2 Kd5 78.Na3 Ba4 79.Nc2 Kc5 80.Kc1 Rd3 81.Ne1 Rc3+ 82.Kb2 Kb4 83.Nb1 Rb3+ 84.Kc1 Rh3 85.Nc2+ Kc4 86.Ne1 Rh1 87.Kd2 Rh2+ 88.Ke3 Bd1 89.Nd2+ Kd5 90.Nef3 Re2+ 91.Kf4 Re8 92.Kg3 Rf8 93.Kg2 and now Shirov showed his exasperation with 93...Bxf3+ - a draw was agreed 10 moves later.

In the graph, which gives Nalimov's Distance to Mate at each half-move, it can be seen that apart from Shirov's intentional blunder at the end, there was only one other move which threw away half a point: 56...Rd8, when White could have played Ne6 and Nf4+, exchanging the Bishop. (I have put draws, with their infinite distance to mate, at 300.)
    In the starting position, Shirov was 208 moves away from mate, assuming perfect play by both sides. And apart from that blunder, he did make steady progress, Karjakin defending well. Then on move 68 Karjakin blundered with 68.Nbd3 (going from a distance of 149 moves to 54), Shirov immediately repaying the compliment by letting him come back to 162 with Rh3+, instead of 68...Re8.
    Generally, Shirov played better, more or less staying at the shorter distances to the win that Karjakin allowed him. From move 49 to 70, he steadily progressed from 208 to 160, then he was around 110 for a few moves, jumping to 84 at move 76, and to 43 at move 77. He even managed to get under 30 - until 86.Ne1, when at only 25 moves, he was closest. But precisely at that point, he fell back to 131 in just two bad moves.
    Of course Shirov didn't know all that - but with the invisibility of his progress, the uncertainty of it, the possibility that his position wasn't winning at all, it must have felt like one of those dreams where you try to run, but with each step you float back, ever further from your goal.
    With 92.Kg3, Karjakin again allowed a huge shortcut: from 136 to a striking distance of 31, but at that point, it seems both players were through with this ending; Shirov's 92...Rf8 threw away 43 moves; 93.Kg2 threw back 35, and finally 93...Bxf3+ threw away the whole win.
    All in all, in his 45 moves in this endgame, Shirov played the best one 24 times; Karjakin scoring 19 out of 44 - 'best' meaning staying on the fastest road to victory or the slowest one to defeat.

I was afraid that after this torture, Shirov would be a broken man, but he showed great fighting spirit, winning the second rapid game, and qualifying for the finals.

PS: In these rapid games, the players do not have to write down their moves, and as a result, cannot claim 50-move-rule draws. It is therefore an academic matter that only at move 86, Shirov missed a (theoretical!) win within 50 moves of the last capture (49.Nxe4): 86...Kd4 87.Nc2+ Kd3 88.Ne1+ Ke2 89.Nc2 Rh1+ 90.Kb2 Kd3 91.Nb4+ Kc4 92.Nd5 Rh2+ 93.Ka3 Bc2 94.Ne3+ Kd3 95.Ng4 Rh5 96.Nf2+ Ke2 97.Nc3+ Kxf2 and mate in 13 move moves.

PS-1 14 December: Dennis Monokroussos makes an interesting point about this theoretical win inside the 50-move limit - "White can improve by playing a 'weaker' move."
    Indeed, as the Nalimov tablebases measure DTM (Distance to Mate), White can paradoxically save himself by allowing a quicker mate that includes a later conversion - because both beat the 50-move limit. After 94.Ne3+ Black mates in 17 and after 94.Nbc3 he mates in 15 - but that move draws, because this quicker mate is also a slower conversion: 94...Bd3 95.Ka4 Rh8 96.Ka3 ("another sub-optimal move") Rb8 97.Ka2 Kd4 98.Ka3 Bc4 and one white Knight would fall next move if they weren't both protected by the 50-move rule.

PS-2 14 December: In Hans Ree's weblog (in Dutch), I read that I was mistaken about the 50-move rule not being in effect in the Khanty-Mansiysk rapid games - it was. Of course, without a scoresheet, the players would have to gamble claiming it, risking a time penalty if the claim was wrong. This explains Shirov's 93...Bxf3+ - fearing he was running out of moves, he tried a new bunch of 50 - of inferior quality.

PS 16 December: Guy Haworth remarks: "It is very common for the DTC-minimaxing and DTM-minimaxing lines to diverge as the nearest event [in this case: mate or a capture - TK] approaches. The two sets of optimal moves typically start off by being identical, can then differ, and finally can become disjoint."

368. 6 December 2007: Silent castling brilliancy duplicated

Looking for something entirely different, I came across the unusual phenomenon of an unusual game that was duplicated twice - once with reversed colors.

Jukes - Pinch, cr 1974

Position after 15...Ke3

1.e4 d5 2.exd5 Nf6 3.Nc3 Nxd5 4.Bc4 Nb6 5.Bb3 Nc6 6.Nf3 e5 7.d3 Bg4 8.h3 Bh5 9.Nxe5 Bxd1 10.Bxf7+ Ke7 11.Bg5+ Kd6 12.Ne4+ Kxe5 13.f4+ Kd4 14.Rxd1 Qxg5 15.c3+ Ke3
see diagram
16.O-O! Qc5 17.Ng3 Nd4 18.Rf2 Ne2+ 19.Nxe2 Na4 20.Kf1 and Black resigned.

Ahn - Ruck, tt Belgium, 2007

Position after 16.Ke6

1.e4 e5 2.Nf3 Nc6 3.d4 exd4 4.Nxd4 Bc5 5.Nb3 Bb6 6.Nc3 Nf6 7.Bg5 h6 8.Bh4 d6 9.a4 Nxe4 10.Bxd8 Bxf2+ 11.Ke2 Bg4+ 12.Kd3 Ne5+ 13.Kxe4 f5+ 14.Kd5 Rxd8 15.Qxg4 c6+ 16.Ke6
see diagram
16...O-O! 17.Nd5 fxg4 18.Bd3 g6 19.Rhf1 Kg7 20.Nd4 Rfe8+ 21.Ne7 Bh4 22.Bxg6 Rxe7+ 23.Kf5 Rf8+ 24.Ke4 Nxg6+ 25.Kd3 Ne5+ 26.Kc3 Bf2 and White resigned.

The positions are identical, except for one a-pawn. Needless to say, the Queen's sacrifice wins in all variations - even recapturing the Queen (fxg4(5) in the diagrams) wins easily.

This silent castling brilliancy had already been duplicated.

Pirozhkov - Remizov, Moscow 1995
1.e4 Nf6 2.Nc3 d5 3.exd5 Nxd5 4.Bc4 Nb6 5.Bb3 Nc6 6.Nf3 e5 7.d3 Bg4 8.h3 Bh5 9.Nxe5 Bxd1 Better is 9...Qh4 10.Nf3 Qe7+, as was played in a few other games. 10.Bxf7+ Ke7 11.Bg5+ Kd6 12.Ne4+ Kxe5 13.f4+ Kd4 14.Rxd1 Qxg5 15.c3+ Ke3 16.O-O Qc5 and here White deviated with 17.Rd2 Nd4 18.Kh2 Qe7 19.Re1+ Kxf4 20.cxd4 Qh4 21.g3+ Qxg3+ 22.Nxg3 and Black resigned.

PS 8 December: James Plaskett mentions that in his book Can you be a tactical chess genius?, he gives another duplicate of Jukes - Pinch. In De Barberis - Musso, Turin 1998, Black deviated after 16.0-0 with 16...Ne5, and had to resign after 17.Nxg5 Nxf7 18.Ne6
    I found yet another one; Vucinic - Durovic, Yugoslavia 1984, where Black simply allowed the threatening switchback mate: 16...Qh4 17.Rf3+ Ke2 18.Rd2+ Ke1 19.Rf1 mate.

PS 10 December: Clones keep coming in - I should have looked in my less serious databases, some of which have 8 copies of Anderssen - Kieseritzky and 13 of Morphy's Opera game, but are strong on this type of light game.

The oldest example is Perry - Willmatch, Cork 1917 (sometimes USA 1917, or cr 1917) where White deviated by already castling at move 15. There followed (15.0-0) Ke3 15.O-O Nd4 16.Rde1+ Ne2+ 17.Rxe2+ Kxe2 18.Bh5+ Ke3 19.Rf3+ Kd4 20.Bf7 Nd5 21.c3+ Nxc3 22.bxc3 mate.
    This game was entirely duplicated in Imbaud - Strumilo, cr 1922, quoted by Tartakower in his 1934 Bréviaire des échecs.

There are also two games where Black deviated with 14...Nb4 (instead of Qxg5); Rozentalis - Mikenas, USSR 1981 and Nicolaide - Birescu, 1987, both ending with 15.c3+ Ke3 16.O-O Nxd3 17.Ng3 and Black resigned.

Ahn - Ruck however, still seems to be the only duplication with colors reversed

Thanks to René Olthof and Joose Norri.

367. 5 November 2007: A mysterious brilliancy from Cuba

Eduardo Bauzá Mercére showed me a spectacular move from an old Cuban championship.

González - Rodríguez Águila, CUB-ch Havana (2) 1963
1.c4 e6 2.g3 Nf6 3.Bg2 c6 4.Nf3 d5 5.b3 Be7 6.Bb2 0-0 7.0-0 Qb6 8.Qc2 Nbd7 9.Nc3 Re8 10.d3 e5 11.cxd5 cxd5 12.Rac1 Bd6 13.Nd2 a6 14.Nxd5 Nxd5 15.Bxd5 Nf6 16.Nc4 Qd8 17.Bg2 Bb8 18.Nb6 Qxb6 19.Qxc8 Bd6 20.Qxb7 Qa5 21.Qxa8 Rxa8 22.Bxa8 h6 23.a4 Kh7 24.Bf3 h5 25.Rc6 Bb4 26.Rfc1 Bd2 27.R1c4 Bb4 28.Bxe5 Qxe5 29.Rxb4 Qa1+ 30.Kg2 Qa3 31.Rb7 Qb2 32.Rxf7 Kg6 33.Rb7 Qd4 34.a5 h4 35.Be4+ Kh5 36.Rxf6 Qxf6 37.Rb6 h3+ 38.Kxh3 Qxf2 39.g4+ Kg5 40.Rg6+ Kf4 41.Rxg7 Qxe2 42.Rf7+ Ke5 43.Rf5+ Ke6 44.g5 Qe3+ 45.Kh4 Qe1+ 46.Kh5 Qe2+ 47.Kg6 Qxh2 48.Rf6+ Ke7 49.Rxa6 Qf2 50.Rf6 Qb2 51.a6 Qxb3 52.a7 Qg8+ 53.Kh5 Qh8+ 54.Kg6 Qe8+ 55.Kf5 Qc8+
see diagram
Now, after Kf4 or Ke5 Black has Qc7+ winning the Pa7 with some drawing chances. The game is supposed to have ended with the brilliant unguarded guard 56.Re6+! and Black resigned. After 56...Qxe6+ 57.Kf4, Black has one more check and White promotes. But why not simply 56.Kg6? There was no threefold repetition in sight, and it's easy to see Black quickly runs out of checks in that case too, and White promotes and keeps his Rook.
    Furthermore, why did Black resign instead of grabbing the Rook with check, and maybe then resign?
    Finally, there is the bulletin, which gives 56.T5A, an impossible move. It is crossed out with T6R barely readable as a pencilled correction - the only correction in Bauzá Mercére's entire copy of this bulletin, which is full of uncorrected typos.

Was that brilliant Re6+ really played?

366. 1 November 2007: Some more discovered checks (and see the PS of 16 September 2011 below.)

In entry no. 343, I published a position constructed by Sampsa Lahtonen where 14 consecutive discovered checks were possible - with promoted pieces, in a legal position. He had beaten, I wrote, a 76 year old record by A. Lapierre with 13 such checks.

Now Olli Heimo informs me that Finnish composition GM Unto Heinonen has managed 15 (see below) but, sensationally (and causing a rather sheepish look on my face) that Joose Norri had remembered an old position he had once seen.

White to play
17 consecutive discovered checks (legal position, promoted men)
W. Frangen, Feenschach 1974

1.e6+ Nxf5+ 2.Ng3+ Nf3+ 3.Nd2+ Rb3+ 4.Nxa6+ Bxa5+ 5.Rxb3+ Rc6+ 6.Rxb1+ Ke5+ 7.exd7+ Ne7+ 8.Nge4+ Nxh2+ 9.Rf3+

So, the record in this domain is not 14 or 15, but 17. There seems to be no doubt as to the legality of the position.

Lahtonen's other record of 47 consecutive ordinary checks (see entry no. 351) still stands. I know he knows about Frangen's 17, and I would be surprised if he's not already working to get to 18.

And here is Heinonen's position:

White to play
15 consecutive discovered checks (legal position, promoted men)
Heinonen, Suomen tehtäväniekat 2007

1.Rg4+ Ng5+ 2.Ngxf4+ Bg6+ 3.Nd3+ Ke6+ 4.Rg7+ Nd7+ 5.Nxb6+ b3+ 6.Nxa4+ Rc6+ 7.Nec3+ Nfe4+ 8.Rgxg1+

PS 5 November 2007: Bader Al-Hajiri dispels all thoughts that the above two positions might not be legal, with proof games that can be played over on the left.

PS 7 February 2008: Alexey Khanyan managed a slightly less economic (28 vs. 27 pieces) position in which Frangen's number of 17 consecutive discovered checks is equalled. It combines some of Frangen's ideas with new ones.

White to play
17 consecutive discovered checks (legal position, promoted men)
Khanyan, original, 2008

1.Nbc5+ Rb3+ 2.Nxa6+ Bxa5+ 3.R7xb3+ Ke5+ 4.Nxd6+ Nxe3+ 5.Ne4+ Nxb8+ 6.Bxd8+ Nf5+ 7.Nxg3+ Rxe2+ 8.Rxb1+ Rb2+ 9.Nxf1+

PS 9 February 2008: Seeing the above position by Alexey Khanyan, Sampsa Lahtonen reminded me of a mail he sent me in October 2007, when he had been shown Frangen's record of 17 consecutive discovered checks. "I'm now trying to get higher than 17," he wrote. "That won't be easy but it doesn't look impossible either. I've already discovered some different ways to come up with 17, including one that leads to 18 with just one piece too many."

And now, February 2008, he showed me that position:

White to play
17 consecutive discovered checks (legal position, promoted men)
Lahtonen, original, 2007

1.Nbc5+ Rb3+ 2.Rb5+ Ke5+ 3.Nxd6+ Nxe3+ 4.Nde4+ Nxb8+ 5.Bxd8+ Nf5+ 6.Nxg3+ Nxe2+ 7.Nxb3+ Rc5+ 8.Rxb1+ Nc3+ 9.Nxf1+

Almost the same position as Khanyan's; many of the same checks. Which, joked Lahtonen, "proves that great minds do indeed think alike..."

There is some truth in that joke - the positions being so similar is not entirely a coincidence. Once you get to the limits of a construction task there is, as Lahtonen writes: "a limited supply of different ideas, and [...] this position, still having some Frangen ideas, was a logical one to come by."

Compared to Khanyan's position, Lahtonen's position has two reasons of existence. It has, like Frangen's, only 27 pieces - and of course, it has room for that 18th check. Add a Black Rd1 and Bd2, and a final 9...Bf4+ is possible. (Note that this is not possible in Khanyan's position, as his Rc2 crosses d2 (twice.))

Alas, adding Rd1 and Bd2 also robs the position of its legality. Thirteen promotions would then need an explanation, when only two white pawns and one black one have disappeared. And one PxP capture only yields three possible promotions, as in 1.e4 d5 2.exd5 which enables two promotions on d8, and one on e1.

That 18th discovered check is still elusive.

PS 16 September 2011: In some sports, to have new records, measuring had to go from fifths of a second to tenths to hundreths. Perhaps something like that has happened in this realm, too. In the positions below, Alexey Khanyan equals two records for consecutive discovered checks, but with one piece less in both cases.

In this position without promoted men, there is a series of 11 discovered checks, as many as in the 81-year old record by O. Stocchi (see items 125 and 308 in this Diary) but with one piece less, 17 vs. 18.

White to play
11 consecutive discovered checks
Alexey Khanyan, original, 2011

1.f8B+ Kd6+ 2.Nge5+ Ne6+ 3.Bf4+ Nfd4+ 4.Ng6+ Nxf4+ 5.Ncxe7+ bxa6+ 6.Nc6+

And in the following position, Khanyan equals the 17 consecutive discovered checks by Frangen and Lahtonen in this item, but again he does it with one piece less; 26 vs. 27.

White to play
17 consecutive discovered checks (legal position, promoted men)
Alexey Khanyan, original, 2011

1.Nf5+ B4d6+ 2.Ned4+ Ngxf7+ 3.Rg5+ Nfe4+ 4.Rf3+ Nb3+ 5.Rcxc1+ Nc6+ 6.Nfxd6+ Nfe5+ 7.N4f5+ Ned2+ 8.Nc4+ Nd7+ 9.Nbd6+

365. 29 October 2007: Patterns, paths, Volkswagen logo

Over the past months, several readers showed me some strange phenomenon they had noticed in a game of their own or somebody else's, wondering whether something like that had ever happened before. It invariably had, and had been surpassed - by far.

Michael Spencer had played a blitz game in which, after 25 moves, he still had a Bishop on c8, hemmed in by pawns on b7 and d7. Had something like that ever occurred in a real tournament game?

White to play
Belson - Opsahl
Toronto, ch-Canada, 1936

After more than 70 moves, the Bishop is caught on its home square by the enemy King: 70.c6 bxc6 71.Bxe7 Kxe7 72.Rxc8 and White won.

And in top-level chess:

Black to play
Bronstein - Boleslavsky
Riga, ch-USSR, 1958

White's last move was 42.Kf2-e1, after which Black resigned.

PS 1 November: Steve Wrinn suggests naming this Bishop the Steinitz-Bishop, as Steinitz often had them in one cramped variation of the Evans Gambit he stubbornly defended. He scored ½ out of 5 in those games, but the draw was the game with his worst-ever c8-Bishop.

White to play
Chigorin - Steinitz
17th match game, World Championship, Havana 1889

Here, 24.b6 buried that Bishop forever, it would seem, but Chigorin didn't guard it well, and after breaking loose with 40...b5, it became the strongest piece on the board. White had some problems holding a draw, but that was the result at move 70 - the game can be played over on the left.

Spencer also wondered what the latest move would be where a player still had both of his central pawns on their original squares.

White to play
Schmitt - Panzalovic
Altensteig 1993

Here, after Black's 61st move Ne1-f3+, a draw was agreed.

Dennis Breuker had observed that in a game l'Ami-Nijboer from this year's Dutch championship, Black had played 11...Bc8-e6; 12...Be6-c4; 13...Bc4-a6 and 16...Ba6-c8, completing a Bishop's tour. He wondered if something like that had ever happened before.

Such tours are quite common, as I discovered with CQL. There are thousands of them, and even many hundreds consecutive tours, of all sizes and shapes, everywhere on the board, and by all of the pieces.

Just a little one that finished a game.

White to play
Crouch - Stevenson
Aviemore 1997

34.Bxg5 Kg7 35.Bf4 Rd8 36.Be5+ Kh6 37.Bf6 and Black resigned.

Compare the Rook's tours in item 354 below.

Jeff Caveney showed me an ICC game in which he'd had a five pawn chain, later adding a game Rogers-Paunovic, Belgrade 1988, where both players had such a chain. Caveney wondered how common such chains are, and if there had ever been a chain of 6 pawns.
    5-pawn chains occur once in around a thousand games, and I found thirteen 6-pawn chains; one in 170,000 games. Here's a recent one from top-level chess.

Judit Polgar - Bacrot, Bastia 1999
1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.O-O b5 6.Bb3 Bc5 7.c3 d6 8.d4 Bb6 9.Be3 O-O 10.Nbd2 Bb7 11.Re1 Re8 12.d5 Ne7 13.Bxb6 cxb6 14.Bc2 Qd7 15.Bd3 Rec8 16.Rc1 Rc7 17.b4 g6 18.Qe2 Rac8 19.c4 Nexd5 20.exd5 Nxd5 21.Be4 f5 22.Bxd5+ Bxd5 23.Qd3 Be6 24.Ng5 bxc4 25.Qh3 Qe7 26.Nxe6 Qxe6 27.Qc3 b5 28.f4 e4 29.Nf3 Qf7 30.Nd4 d5 31.Rcd1 Re8 32.Kf2 Qg7 33.h4 Rce7 34.Re3 Rd7 35.h5 Qh6 36.Ne2 Qxh5 37.Qf6 Qh6 38.Qc6 Red8 39.Qe6+ Kh8 40.Rh3 Qg7 41.Nd4 Rd6 42.Qe5 Qxe5 43.fxe5 Rb6 44.Ke3 Kg7 45.Ne2 Re6 46.Nd4 Rxe5 47.Nc6 d4+ 48.Kf4 Red5 49.Nxd8 Rxd8 50.Ra3 d3 51.Rxa6 c3 52.Rc6 c2
see diagram
53.Rc7+ Kf6 54.Rc6+ Kf7 55.Rc7+ Kf6 56.Rc6+ Kf7 57.Rc7+ Ke6 58.Rh1 d2 59.Rxc2 d1Q 60.Rxd1 Rxd1 61.Rc6+ Rd6 62.Rc5 Rd2 63.Rxb5 Rxg2 64.a4 g5+ 65.Ke3 Rg3+ 66.Kd4 Rd3+ 67.Kc4 Ra3 68.a5 g4 69.Rb8 g3 70.Rg8 Ke5 71.Kb5 f4 72.a6 e3 73.Kb6 Ke4 74.a7 e2 75.b5 and White resigned.

That had already been surpassed in this very strange game.

Cherniaev - Colure, New York open 1993
1.e4 c5 2.Nf3 Nc6 3.Bb5 Qc7 4.O-O e6 5.Re1 Nge7 6.c3 a6 7.Ba4 b5 8.Bc2 d5 9.a4 b4 10.d4 c4 11.Nbd2 Bb7 12.e5 b3 13.Bb1 g6 14.Nf1 h6 15.h3 O-O-O 16.g4 Bg7 17.Ng3 Rdg8 18.Nh4 Kb8 19.Ng2 Bc8 20.f4 Bf8 21.Qf3 Qd8 22.Re2 Bd7 23.a5 Nc8 24.f5 gxf5 25.gxf5 Be7 26.f6 Bf8 27.Nh5 Kb7 28.Kh2 N8a7 29.Bf4 Nb5 30.Rd2 Nxa5 31.Ne3 Nc6 32.Ng4 a5 33.Qe3 Qa8 34.Nxh6 Bxh6 35.Bxh6 a4 36.Ng3 a3 37.Bg7 a2
see diagram
38.Bxh8 axb1Q 39.Rxb1 Rxh8 40.Rg1 Rh4 41.Ne2 Qh8 42.Rg7 Qh5 43.Nf4 Qf5 44.Rf2 Be8 45.Qg3 Rh6 46.Rg8 Bd7 47.Qg7 Rh7 48.Qf8 Nc7 49.Rg7 Be8 50.Rxh7 Qxh7 51.Qg7 Qe4 52.Qg2 Qh7 53.Qg3 Nb5 54.Rd2 Qe4 55.Ne2 Nb8 56.h4 Nd7 57.Nf4 Nf8 58.Re2 Qf5 59.Qg5 Qb1 60.Nh5 Ng6 61.Ng3 Nxc3 62.bxc3 Qa1 63.h5 b2 64.hxg6 fxg6 65.f7 Bxf7 66.Qe7+ Ka6 67.Qd6+ Ka7 68.Qc7+ Ka6 69.Qc6+ Ka7 70.Rf2 and Black resigned.

There is the connected question of whether a single pawn has ever travelled the path of such a chain. Not very often; there have been twelve such pawns. Nine captured their way from b2 to g7 (or mirrors); two from a2 to f7 or mirrors, and one from c2 to h7. No pawn has ever extended such a path by promoting after a further capture, let alone in the matching direction.

The fastest of these pawns crossed the board within 8 moves; almost an Excelsior march. It's a pity it happened after that game was essentially over; White must have been in time trouble.

Willim - Franz, Germany 1999
1.c4 e5 2.g3 Nc6 3.Bg2 g6 4.Nc3 Bg7 5.Rb1 d6 6.b4 f5 7.b5 Nce7 8.e3 Nf6 9.Nge2 O-O 10.d3 g5 11.f4 gxf4 12.exf4 Ng6 13.O-O h5 14.Nd5 h4 15.Nxf6+ Bxf6 16.Be3 Qe8 17.Qd2 Kh7 18.Rbe1 Qf7 19.Nc3 Qg7 20.Nd5 Bd8 21.Qd1 Rh8 22.g4 h3 23.Bf3 fxg4 24.Be4 Kg8 25.f5 Nh4 26.Kh1 Ng2 27.Rg1 g3 28.f6 Bxf6 29.Nxf6+ Qxf6 30.Bd5+ Be6 31.Qg4+ Kf7 32.Rgf1 Nf4
see diagram
33.hxg3 Rag8 34.Bxe6+ Qxe6 35.Qxe6+ Kxe6 36.gxf4 Rg2 37.Rf2 Rg3 38.fxe5 Rhg8 39.exd6 Rg2 40.dxc7 Rxf2 41.Bxf2+ Kd7 and Black resigned.

Today, Dennis Monkroussos called my attention to the V-sign Ivanchuk created (although for him, it was really a /\ sign) the in the second play-off blitz game against Leko.

White to play
Leko - Ivanchuk
Match, 2nd tiebreak blitz game, Mukachevo 2007

After Black's last move, 47...Ra1-d1, White resigned.

Have similar V-signs ever occurred? Obviously, depending how many other pieces you allow, they are common fare, happening once in a hundred games. Ivanchuk's V-sign only has five extra pieces, which is pretty rare, but I found two games with only four.

After 61.Rg4-g7
Parr - Sommerville
Blackpool 1971

Six extra pieces indeed, but after 61...Kxe5 62.Ne4 Bxf6 63.Nxf6 that was down to four, while the V-sign was alwas kept up. After 63...Kxf6 64.Rxa7 a draw was agreed.

Almost as soon as I posted this, Wijnand Engelkes reminded me of an old famous V-sign. My search hadn't included such crowded V-signs, but I guess this one is so famous and beautiful I must add it.

Capablanca - Treybal, Karlsbad 1929
1.d4 d5 2.c4 c6 3.Nf3 e6 4.Bg5 Be7 5.Bxe7 Qxe7 6.Nbd2 f5 7.e3 Nd7 8.Bd3 Nh6 9.O-O O-O 10.Qc2 g6 11.Rab1 Nf6 12.Ne5 Nf7 13.f4 Bd7 14.Ndf3 Rfd8 15.b4 Be8 16.Rfc1 a6 17.Qf2 Nxe5 18.Nxe5 Nd7 19.Nf3 Rdc8 20.c5 Nf6 21.a4 Ng4 22.Qe1 Nh6 23.h3 Nf7 24.g4 Bd7 25.Rc2 Kh8 26.Rg2 Rg8 27.g5 Qd8 28.h4 Kg7 29.h5 Rh8 30.Rh2 Qc7 31.Qc3 Qd8 32.Kf2 Qc7 33.Rbh1 Rag8 34.Qa1 Rb8 35.Qa3 Rbg8 36.b5 axb5 37.h6+ Kf8 38.axb5 Ke7 39.b6 Qb8 40.Ra1 Rc8 41.Qb4 Rhd8 42.Ra7 Kf8 43.Rh1 Be8 44.Rha1 Kg8 45.R1a4 Kf8 46.Qa3 Kg8 47.Kg3 Bd7 48.Kh4 Kh8 49.Qa1 Kg8 50.Kg3 Kf8 51.Kg2 Be8 52.Nd2 Bd7 53.Nb3 Re8 54.Na5 Nd8
see diagram
The Sign had already been there since move 42 and now White finished the game: 55.Ba6 bxa6 56.Rxd7 Re7 57.Rxd8+ Rxd8 58.Nxc6 and Black resigned.

As Engelkes remarks, the Black W-sign completes a Volkswagen logo.

PS 1 November: Brian Karen (naisortep on ICC) shows me a crowded but really amazing V-sign, a wedge really, from an early game of Petrosian's I had never seen before.

Petrosian - Petrovsky, USSR under 18 championship, Leningrad 1946 1.d4 Nf6 2.c4 e6 3.Nf3 Bb4+ 4.Bd2 Qe7 5.g3 Nc6 6.Bg2 Bxd2+ 7.Nbxd2 d5 8.O-O O-O 9.Re1 Rd8 10.e3 Bd7 11.Rc1 Rac8 12.a3 Na5 13.Ne5 c6 14.c5 b5 15.f4 Be8 16.Qc2 Nd7 17.e4 f6 18.exd5 fxe5 19.d6 Qf8 20.dxe5 g6 21.b4
see diagram
21...Nb7 22.Nb3 Qf7 23.Nd4 Nf8 24.a4 a6 25.Ra1 Qg7 26.Qe4 Ra8 27.axb5 cxb5 28.Qxb7 and Black resigned.

I found a few other Wedges - one occurs after 50...d3 in the game Polgar - Bacrot already given above.

364. 29 October 2007: A move to go berserk

It's always fun to try to recognize games from board positions in pictures, or descriptions in books or magazines. Now we have YouTube, too. This morning's Chess Today quoted John Saunders' blogspot in which he said the clip below is his "favourite chess video clip of all time."

That raised the question: what was Kasparov's move and what did Anand do to make him go berserk? It wasn't hard to find.

Anand - Kasparov, blitz, Geneva 1996
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Be3 Ng4 7.Bg5 h6 8.Bh4 g5 9.Bg3 Bg7 10.Be2 h5 11.Bxg4 Bxg4 12.f3 Bd7 13.Bf2 Nc6 14.Qd2 Ne5 15.O-O g4 16.f4 Nc4 17.Qe2 Rc8 18.b3 Na3 19.Nd5 e6 20.Nb4 Qa5 21.Qe1 h4 22.Be3 h3 23.g3 Nb5 24.Rd1 Nc3 25.Nd3 Qc7 26.Rc1 Nxe4 27.f5 e5 28.f6 Nxf6 29.Nf5 Bxf5 30.Rxf5 Qc6 31.Qe2 Qe4 32.Rf2 Nd5 33.Re1
see diagram
Black is completely winning. But: 33...Qxe3?? 34.Qxg4
There followed: 34...O-O 35.Rxe3 Nxe3 36.Qxh3 Nxc2 37.Qd7 Nd4 38.Qxb7 a5 39.Kg2 Rc3 40.Nb2 Nc2 41.Nc4 d5 42.Nd6 Ne3+ 43.Kh3 f5 44.Qd7 f4 45.Qe6+ Kh7 46.Nf7 Rxf7 47.Qxf7 Rc6 48.gxf4 Rf6 49.Qc7 e4 50.f5 d4 51.Qe7 Rh6+ 52.Kg3 Nd1 53.Rf4 e3 54.Rg4 and Kasparov resigned.

363. 22 October 2007: Longest decisive game? (+ PS 26 October)

The above positions occurred, from left to right, after moves 25...c6, 75.h4 and 120...Kg8 in the game Smith - Burwick, Swedish Team Championship, 21 October 2007. The game ended when White was mated on move 196. Time controls, according to André Nilsson who sent me this game, were 2 hours for 40 moves, and 1 hour for the rest. Some rest - even if the position wasn't very challenging until 121.Nb5.

But is it the longest decisive game ever, as Nilsson thinks it might be? On my records page, I give a game Stepak - Mashian, Israel 1980 as such, which White won after 193 moves. But I also mention an Internet game TINTIRITO (T. Petrosian jr.) - Daki (Milanovic), Dos Hermanos 2005, which Black won after 199 moves. That was only "2+8" blitz, but during the last so many moves, TINTIRITO and Daki must have had more time than Smith & Burwick who finished with 3 vs. 16 seconds - too little for White to resign which he could safely have done after move 178. (But see the PS to this entry.)

Anyway, here is the game. With 148.c6, White could have won, Smith writes on www.schackforum.se

Smith - Burwick, Swedish Team Championship, 21 October 2007
1.e4 g6 2.d4 Bg7 3.Nc3 d6 4.f4 a6 5.Nf3 e6 6.Bd3 Ne7 7.Be3 Nd7 8.Qd2 f5 9.Rg1 Nf6 10.e5 Ng4 11.h3 Nxe3 12.Qxe3 d5 13.g4 Bh6 14.Ne2 Bd7 15.b3 Bb5 16.c4 Bc6 17.c5 Bb5 18.g5 Bxd3 19.Qxd3 Bf8 20.Kf2 Qd7 21.b4 h6 22.a4 h5 23.b5 Kf7 24.Rgb1 a5 25.b6 c6 26.Nd2 Bg7 27.Nb3 Ra6 28.Nc3 Rb8 29.Ra2 Bf8 30.Kg3 Qd8 31.Rc2 Rc8 32.Rbc1 Qe8 33.Kg2 Rb8 34.Nb1 Qd7 35.Na3 Qe8 36.Kf3 Bg7 37.Rb1 Bf8 38.Kg2 Bg7 39.Qc3 Rba8 40.Nd2 Rb8 41.Rcb2 Bf8 42.Qd3 Qd7 43.Rg1 Bg7 44.Nab1 Rd8 45.Nc3 Rb8 46.Rgb1 Raa8 47.Qf3 Rh8 48.Qd1 Rhb8 49.Ne2 Ra6 50.Ng3 Rh8 51.Qe2 Raa8 52.Qe1 Ra6 53.Rc2 Rb8 54.Nf3 Rh8 55.Qe2 Raa8 56.Nh4 Qd8 57.Kf1 Qd7 58.Ke1 Qd8 59.Kd1 Qd7 60.Qd3 Qd8 61.Kc1 Qd7 62.Rb3 Qd8 63.Kb1 Qd7 64.Qc3 Qe8 65.Rbb2 Qd7 66.Ne2 Rhb8 67.Ka2 Qe8 68.Nc1 Ra6 69.Ng2 Bf8 70.Ne3 Qd7 71.Re2 Kg8 72.Nc2 Qe8 73.Na1 Qd7 74.Nab3 Rba8 75.h4 Nc8 76.Nd3 Ne7 77.Re1 Qe8 78.Reb1 Qd7 79.Ndc1 Qe8 80.Ne2 Qf7 81.Qd3 Qe8 82.Nc3 Rb8 83.Ra1 Kh7 84.Kb1 Raa8 85.Rba2 Nc8 86.Qd2 Ra6 87.Kc2 Ne7 88.Ne2 Rba8 89.Ng3 Kg8 90.Nh1 Qf7 91.Nf2 Qe8 92.Nd3 Qd7 93.Qc3 Qe8 94.Nd2 Kf7 95.Qb3 Rb8 96.Nb1 Raa8 97.Nc3 Kg8 98.Qb2 Qd7 99.Nc1 Qe8 100.Nb3 Kf7 101.Qc1 Kg8 102.Qe1 Ra6 103.Nb1 Rba8 104.Na3 Kg7 105.Qe2 Rb8 106.Qd2 Rba8 107.Qd3 Rb8 108.Rb1 Kf7 109.Qe2 Kg8 110.Qh2 Kf7 111.Qd2 Rba8 112.Rc1 Kg7 113.Qd3 Rb8 114.Rd1 Kg8 115.Qd2 Rba8 116.Qg2 Rb8 117.Rda1 Kf7 118.Qe2 Kg8 119.Kb2 Kf7 120.Kc3 Kg8 121.Nb5 cxb5 122.axb5 Raa8 123.Rxa5 Rxa5 124.Rxa5 Nc8 125.Qa2 Qd7 126.Kb2 Be7 127.Qa4 Kf7 128.Nc1 Qe8 129.Nd3 Qd8 130.Qb3 Kg7 131.Ra4 Kf7 132.Ra2 Kg7 133.Ra1 Kf7 134.Ra4 Kg7 135.Nb4 Kf7 136.Ka2 Kg7 137.Qc3 Kf7 138.Nd3 Qe8 139.Ra5 Kf8 140.Nc1 Kf7 141.Nb3 Kf8 142.Kb1 Kf7 143.Qd2 Kf8 144.Qa2 Kf7 145.Kc2 Kg7 146.Ra8 Rxa8 147.Qxa8 Qd7 148.Qa4 Kf7 149.Na5 Bd8 150.Qb4 Nxb6 151.cxb6 Bxb6 152.Nb3 Qc7+ 153.Kd3 Qd7 154.Nc5 Qc8 155.Na4 Bd8 156.Nc5 Qa8 157.Na4 Be7 158.Qb3 Qa5 159.Nc3 Qa1 160.Qd1 Qb2 161.Qc2 Qa3 162.Qa2 Qc1 163.Qa8 Qxf4 164.Qxb7 Qf3+ 165.Kc2 Qf2+ 166.Kd3 Qf3+ 167.Kd2 Qf4+ 168.Kc2 Qxd4 169.Qc7 f4 170.b6 f3 171.b7 f2 172.b8Q f1Q 173.Qcc8 Qdd3+ 174.Kb3 Qc4+ 175.Qxc4 Qxc4+ 176.Kb2 d4 177.Ne4 Qe2+ 178.Kb3 Qxe4 179.Qc7 Qe3+ 180.Ka2 Qc3 181.Qd7 d3 182.Qb7 d2 183.Qb1 Qa5+ 184.Kb2 Qb4+ 185.Kc2 Qxb1+ 186.Kxd2 Qb2+ 187.Kd3 Qxe5 188.Kc4 Qg3 189.Kd4 Qxh4+ 190.Kd3 Qxg5 191.Ke2 h4 192.Kf3 h3 193.Kf2 h2 194.Ke2 h1Q 195.Kf2 Qf4+ 196.Ke2 Qhf1 mate.

PS 26 October: The question whether Smith - Burwick was the longest decisive game ever, had a resounding answer that same day, 21 October, when Kosteniuk, with Black, beat Fressinet in 237 moves in a rapid (25' + 10'') tournament in Villandry, France. It's not clear why Fressinet, who was down in the dreaded R+B vs. R ending, never claimed a draw but allowed it to go on for 116 moves. This created the further coincidence that Kosteniuk finally won by obtaining the "Philidor position" - in a tournament that was part of a "Chess and Music Festival" in memory of Philidor. On the eve of this game, a descendant of Philidor's opened a conference on the great player / composer, and some Philidor music was performed.

Fressinet - Kosteniuk, Villandry rapid 25' + 10'', 21 October 2007
1.e4 c5 2.Nf3 d6 3.Bb5+ Nc6 4.O-O Bd7 5.Re1 Nf6 6.c3 a6 7.Ba4 b5 8.Bc2 e5 9.h3 Be7 10.d4 O-O 11.d5 Na5 12.Nbd2 g6 13.b4 Nb7 14.a4 Qc7 15.Nf1 Nh5 16.Bh6 Ng7 17.Ng3 f6 18.Qd2 Rfc8 19.Ra2 a5 20.Rea1 axb4 21.cxb4 bxa4 22.Bxa4 Bxa4 23.Rxa4 Rxa4 24.Rxa4 cxb4 25.Rxb4 Nc5 26.Be3 Qd7 27.Rc4 Rb8 28.Bxc5 dxc5 29.Nf1 Ne8 30.Qc2 Qb5 31.N3d2 Qb2 32.Qa4 Kf7 33.Rc2 Qb4 34.Qd7 Nd6 35.Ne3 Kf8 36.Kh2 Rd8 37.Qg4 Qd4 38.Nec4 f5 39.Qe2 Nxe4 40.Nxe4 Qxe4 41.Qxe4 fxe4 42.Rd2 Bg5 43.Rd1 Rb8 44.Kg1 Rb4 45.Nxe5 c4 46.Nc6 Rb2 47.Nd4 c3 48.Re1 c2 49.Nxc2 Rxc2 50.Rxe4 Rd2 51.g3 Be7 52.Rf4+ Kg7 53.Kg2 Bc5 54.h4 Rxd5 55.Rc4 Kf6 56.Rc2 Kf5 57.Ra2 Bb4 58.Rb2 Bc3 59.Rc2 Rd3 60.Kf1 Bd4 61.Kg2 Rb3 62.Rc7 h6 63.Rf7+ Ke6 64.Rf4 Bb6 65.Rf8 Rb2 66.Rf3 g5 67.hxg5 hxg5 68.Rf8 Bc5 69.Rf3 Ra2 70.Kg1 Ra7 71.Kg2 Ke5 72.Kh3 Ra4 73.Kg2 Ra2 74.Rf7 Ke4 75.Rf3 Bd4 76.Rf8 Kd3 77.Rf5 Ke4 78.Rxg5 Rxf2+ 79.Kh3 Rf8 80.Rg4+ Ke3 81.Rf4 Rg8 82.Rg4 Bg7 83.Kg2 Kd3 84.Rg6 Ke4 85.Rg5 Kd4 86.Kg1 Kd3 87.Kg2 Ke4 88.Rg4+ Kf5 89.Rf4+ Kg5 90.Kf3 Ra8 91.Rg4+ Kf6 92.Rf4+ Ke6 93.Kg2 Be5 94.Rf3 Rh8 95.Kg1 Kd5 96.Kg2 Ke4 97.Rb3 Rc8 98.Kh3 Rc2 99.Kg4 Rg2 100.Ra3 Rg1 101.Rb3 Bd6 102.Rc3 Kd4 103.Rb3 Bc5 104.Rf3 Bb4 105.Rf4+ Kc3 106.Rf3+ Kd2 107.Kf5 Ke2 108.Rb3 Bd2 109.Ke4 Ra1 110.Kf5 Be3 111.Kg4 Ra4+ 112.Kh3 Ra7 113.Rb5 Rf7 114.Kg4 Rg7+ 115.Kh3 Bf2 116.Rb3 Kf1 117.Rb1+ Be1 118.Rb3 Rh7+ 119.Kg4 Kg2 120.Kf4 Rf7+ 121.Ke4 Bxg3 122.Rb2+ Kh3 123.Rb3 Kg4 124.Kd5 Bf4 125.Ke6 Rf8 126.Rb1 Re8+ 127.Kd7 Re5 128.Rd1 Kf5 129.Kc6 Re2 130.Rd8 Rc2+ 131.Kb5 Ke6 132.Rd4 Bd6 133.Rc4 Rb2+ 134.Kc6 Rb8 135.Rh4 Rc8+ 136.Kb7 Rc1 137.Kb6 Bf8 138.Kb5 Rc5+ 139.Kb6 Rc1 140.Kb5 Bd6 141.Rc4 Ra1 142.Kc6 Be5 143.Rc2 Ra8 144.Rc4 Rd8 145.Rc5 Bd4 146.Rc4 Rd6+ 147.Kc7 Rd7+ 148.Kc6 Be5 149.Rc5 Rd6+ 150.Kb5 Rd8 151.Kc6 Bd4 152.Rc4 Ke5 153.Kc7 Rh8 154.Kc6 Rh6+ 155.Kb7 Kd5 156.Rc7 Rb6+ 157.Kc8 Be5 158.Rb7 Rh6 159.Kd7 Bd6 160.Rb5+ Bc5 161.Rb7 Re6 162.Rc7 Bb6 163.Rb7 Rh6 164.Kc8 Ba5 165.Kd7 Kc5 166.Kc8 Rd6 167.Rd7 Rg6 168.Rb7 Bd2 169.Kd7 Bf4 170.Kc8 Kd5 171.Kd7 Rg7+ 172.Kc8 Rg1 173.Kd7 Bd6 174.Rb5+ Bc5 175.Rb7 Rd1 176.Kc8 Ke6 177.Kc7 Rc1 178.Kc6 Bd6+ 179.Kb5 Kd5 180.Rb6 Bc5 181.Rb7 Kd6 182.Ka6 Ra1+ 183.Kb5 Kd5 184.Rd7+ Bd6 185.Rb7 Rh1 186.Rb6 Bc7 187.Rb7 Rc1 188.Ka6 Kd6 189.Kb5 Rb1+ 190.Ka6 Ra1+ 191.Kb5 Bb8 192.Kb6 Bc7+ 193.Kb5 Ra5+ 194.Kb4 Rh5 195.Kc4 Kc6 196.Rb3 Rh4+ 197.Kd3 Ba5 198.Rb8 Kc5 199.Rg8 Rd4+ 200.Ke3 Bd2+ 201.Kf3 Rd7 202.Rg4 Kd5 203.Ke2 Ke5 204.Kf3 Bh6 205.Ra4 Rd5 206.Rh4 Bg5 207.Ra4 Rd3+ 208.Ke2 Re3+ 209.Kf2 Rb3 210.Ke2 Be3 211.Ra8 Bd4 212.Re8+ Kf4 213.Kd2 Be5 214.Kc2 Rc3+ 215.Kd2 Rc7 216.Kd3 Rd7+ 217.Kc4 Ke4 218.Kb5 Kd5 219.Rc8 Rb7+ 220.Ka6 Rb1 221.Rc2 Bd6 222.Rc8 Bc5 223.Rd8+ Kc4 224.Rc8 Rb2 225.Rc6 Kd5 226.Rc8 Rb6+ 227.Ka7 Kd6 228.Rb8 Rb1+ 229.Ka8 Ra1+ 230.Kb7 Ra7+ 231.Kc8 Rh7 232.Rb1 Kc6 233.Rd1 Be3 234.Kd8 Bc5 235.Kc8 Ra7 236.Kb8 Ra2 237.Rc1 Rh2 and White resigned.

Thanks to Wiebe Cnossen for drawing my attention to this.

362. 20 October 2007: A wonderful Steel King

Today's Chess Today has a wonderful fighting game, a true Steel King, from the open tournament in Barcelona. It is deeply analyzed by Mikhail Golubev - I'll just take one or two remarks from that analysis.

Nakamura - Fluvia, Barcelona, 1st Round, 18 October 2007
1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.Bg5 Ne4 5.Bh4 Nxc3 6.bxc3 dxc4 7.e3 Be6 8.Qb1 Qd5 9.a4 Nc6 10.Nf3 Bh6 11.Be2 g5 12.Bg3 g4 13.Nh4 f5 14.O-O O-O-O 15.Bd1 Qa5 16.Bc2 Rhf8 17.Qb2 Bg5 18.Rab1 b6 19.Rbe1 Bxh4 20.Bxh4 Bd5 21.Bg3 h5 22.f3 gxf3 23.gxf3 f4 24.Bxf4 Nxd4 25.cxd4 Rg8+ 26.Kf2 Qd2+ 27.Re2 Rg2+ 28.Kxg2 Qxe2+ 29.Rf2 Rg8+ 30.Bg3 Bxf3+ 31.Kh3 Bg4+ 32.Kh4 Qxe3 33.Qb5 Qxd4 34.Qe5 Qd8 35.Rd2 Bd7 36.Rd4 e6+ 37.Kxh5 Be8+ 38.Kh6 Qe7 39.Rf4 Bf7
see diagram
40.Qf6? 40.Rxc4! (Qf8+ 41.Kh7) was immediately winning. 40...Qf8+ 41.Kh7 Kb8 Or 41...Rg6 and now 42.Bxg6 Bg8+ 43.Kh8 Bf7+ etc. with an immediate draw, or 42.Qxg6 Bxg6+ 43.Kxg6 when it's not clear if White has any advantage. 42.Rf2 e5? Now 42...Rg6 was forced, with the same possibilities as on the move before. 43.Qxe5 and Black resigned.

PS 24 October: This winning King's march was Nakamura's first round game. In the second, he followed it up with a brilliant winning King's chase.

Krasenkow - Nakamura, Barcelona, 2nd Round, 19 October 2007
1.Nf3 Nf6 2.c4 e6 3.g3 d5 4.Bg2 Be7 5.O-O O-O 6.b3 a5 7.Nc3 c6 8.d4 Nbd7 9.Qc2 b6 10.e4 Ba6 11.Nd2 c5 12.exd5 cxd4 13.Nb5 exd5 14.Nxd4 Rc8 15.Re1 b5 16.Bb2 Re8 17.Qd1 bxc4 18.bxc4 Qb6 19.Rb1 dxc4 He must have seen the Queen's sacrifice here. 20.Nc6 Rxc6 21.Bxf6
see diagram
21...Qxf2+ 22.Kxf2 Bc5+ 23.Kf3 Rxf6+ 24.Kg4 Ne5+ 25.Kg5 Rg6+ 26.Kh5 f6 27.Rxe5 Rxe5+ 28.Kh4 Bc8 and White resigned.

Thanks to Chess Today and Mark Thornton.

361. 16 October 2007: Sleepless in Tokyo

In 1982, in Ken Whyld's famous column Quotes and Queries in the British Chess Magazine, he published a construction task record by M. Montgomery - a mutual stalemate in 19 moves; 38 ply to be precise. A year later, Evgenyi Gik in his book Shakhmatyi i Matematika, brought that down to 18½ moves, or 37 ply.

A few months ago, this caught the attention of the Italian composer Enzo Minerva. For his work he often has to travel to Japan, and during the inevitable jetlagged nights there, he set out to beat this record. Gik's 37 ply stalemate had 22 pieces, and Minerva soon managed 37 ply with 15, and later with 13 pieces. But he agrees that with this sort of task, less might not be more. Also, Gik's stalemate was symmetrical, which may be what he aimed for.

Then, during a new trip with new sleepless nights in Tokyo, Minerva achieved 36 ply, becoming the sole holder of this record.

White to play
Minerva, mutual stalemate in 36 ply
l'Unità, August 2007

1.c4 d5 2.Qb3 Bh3 3.gxh3 f5 4.Qxb7 Kf7 5.Qxa7 Kg6 6.f3 c5 7.Qxe7 Rxa2 8.Kf2 Rxb2 9.Qxg7+ Kh5 10.Qxg8 Rxb1 11.Rxb1 Kh4 12.Qxh8 h5 13.Qh6 Bxh6 14.Rxb8 Be3+ 15.dxe3 Qxb8 16.Kg2 Qf4 17.exf4 d4 18.Be3 dxe3 stalemate.

The previous records:

White to play
Montgomery, mutual stalemate in 38 ply
British Chess Magazine, 1982

1.e4 h6 2.Qg4 d5 3.Qxc8 Nd7 4.Qxa8 Ndf6 5.Qxb7 Nh7 6.Qxc7 Qxc7 7.Ke2 Qxc2 8.Kf3 Qxc1 9.Kg4 Qxf1 10.a4 Qxg1 11.d4 Qxh2 12.f4 Qxh1 13.Na3 Qxa1 14.Kh5 Qxa3 15.bxa3 a5 16.g4 g5 17.f5 f6 18.e5 Kf7 19.e6+ Kg7 stalemate. The final position has three pieces more than Minerva's.

White to play
Gik, mutual stalemate in 37 ply
Shakhmatyi i Matematika, 1983

1.e4 d5 2.e5 d4 3.c3 f6 4.Qf3 Kf7 5.Qxb7 Qd5 6.Kd1 Qxg2 7.Kc2 Qxf1 8.Qxc8 Qxg1 9.Qxb8 Rxb8 10.Rxg1 Rb3 11.Rg6 Ra3 12.Rh6 gxh6 13.bxa3 Kg7 14.Kb2 d3 15.e6 a5 16.h4 a4 17.h5 c5 18.f4 c4 19.f5 stalemate

Thanks to Harold van der Heijden and René Olthof for bringing this to my attention. The history of this record was described by Minerva himself in the Italian problemist's magazine Best Problems - the story (in Italian) can be found on page 93. I took the picture from there; thanks to Joost de Heer for that link.

PS 9 May 2008: Alexey Khanyan shows me a symmetrical mutual stalemate game of 22 moves that he composed. It was first published in 2002 in a book Nyeobychnyye Shakhmaty (Unusual Chess) by Evgenyi Gik.

1.a4 a5 2.b4 b5 3.axb5 axb4 4.Nc3 Nc6 5.bxc6 bxc3 6.Ra4 Ra5 7.Rf4 Rf5 8.e4 e5 9.exf5 exf4 10.h4 h5 11.Qg4 Qg5 12.hxg5 hxg4 13.Nf3 Nf6 14.gxf6 gxf3 15.Be2 Be7 16.fxe7 fxe2 17.f6 f3 18.d4 d5 19.Rh3 Rh6 20.Bxh6 Bxh3 21.gxh3 gxh6 22.h4 h5 stalemate, see diagram.

PS 11 May 2008: Almost immediately, Sampsa Lahtonen sent me two games in which he manages the same trick in one move less, both resulting in the diagram.

1.a4 a5 2.b4 b5 3.axb5 axb4 4.Nc3 Nc6 5.bxc6 bxc3 6.Ra4 Ra5 7.Re4 Re5 8.d4 d5 9.dxe5 dxe4 10.Nf3 Nf6 11.exf6 exf3 12.h4 h5 13.Rh3 Rh6 14.Bxh6 Bxh3 15.gxh3 gxh6 16.e4 e5 17.Qd4 Qd5 18.exd5 exd4 19.d6 d3 20.Be2 Be7 21.dxe7 dxe2 stalemate.

1.a4 a5 2.b4 b5 3.axb5 axb4 4.Nc3 Nc6 5.bxc6 bxc3 6.Ra4 Ra5 7.h4 h5 8.Rh3 Rh6 9.Rf4 Rf5 10.e4 e5 11.exf5 exf4 12.f6 f3 13.d4 d5 14.Bc4 Bc5 15.dxc5 dxc4 16.Qd3 Qd6 17.cxd6 cxd3 18.Ne2 Ne7 19.dxe7 dxe2 20.Bxh6 Bxh3 21.gxh3 gxh6 stalemate.

Lahtonen: "I mark it as "probable but not certain" that this is the quickest."

© Tim Krabbé 2007, 2008

Top | Main chess page | Main page | Contact me |