Open chess diary 181-200

Chess is inexhaustable, but some parts of it are less inexhaustable than others. The endgame of King and two pawns versus King and one pawn, for instance, was thought to have been charted completely. The great man in this field was the Russian N.D. Grigoriev (1895 - 1938), who singlehandedly seemed to have shown all that was significant or beautiful with this material.
Grigoriev was one of Alekhine's early opponents (and was named by him as the loser in a famous game that was later exposed as a fake; see Alekhine's Five Queen Game, elsewhere on this site); he was champion of Moscow four times; 5th in the 1920 Soviet Championship; he was the chess editor for Izvestia, but his real fame rests on his endgame analysis, especially of small pawn-endgames. When in 1936 the French magazine La Stratégie held a composing tourney, with K+2P vs. K+P as a theme, Grigoriev won 10 of the 12 awards.
Even so, Grigoriev overlooked at least one little gem in this personal square millimeter of his, as the Dutch endgame-composer and expert Harold van der Heijden recently showed with this 'original, modest study by myself.'

White will sacrifice one pawn, and run with the other. But this is not possible immediately, because after 1.g5 fxg5, Black saves himself (and even wins) with a check on g3, and after 1.e5? fxe5, both sides promote. 1.Kh3 Kf3! 2.e5 also leads to mutual promotions and a draw, while 2.g5 is again losing.
1.Kh1! however, puts Black in Zugzwang. If he plays his King to the e-file, then after 2.e5 fxe5 3.g5, it is in the way of his e-pawn. After 1...Kg3 2.e5 fxe5 3.g5, White promotes with check, and 1...Kf3 loses to 2.e5 (not 2.g5? fxg5 3.e5? Kf2! and White is mated) 2...fxe5 3.g5 e4 4.Kg1! e3 5.Kf1 etc.
Black's best try, and the variation that gives this composition its charm, is 1...Kf1 2.e5! fxe5 3.g5 e4 4.g6 e3 5.g7 e2 6.g8Q e1Q Both have queened again, but now 7.Qg2 is a nice mate.

From 3-8 April last year, together with Rick Goetzee, I visited Leonid Yarosh, the man who composed the Babson Task, in Kazan in Tatarstan (Russia.) You can find some pictures on this page. The story I wrote about it, De man die de Babson Task maakte, is in Dutch.

For a player of my stature, it is not really done to publish your own games, but when you make a complete fool of yourself, that is an excuse. Here are two ridiculous ICC-wins that I scored on the same day, 30 December. (Compare items 170 and 176 in this Diary.)

TK - J**, ICC 5 0, 30 December 2002
1.e4 d6 2.d4 g6 3.Nc3 Bg7 4.Nf3 Nf6 5.Bc4 O-O 6.Qe2 c5 7.dxc5 dxc5 8.O-O Nc6 9.e5 Ne8 10.Rd1 Nd4 11.Nxd4 cxd4 12.e6 f5 13.Nb5 a6 14.Nxd4 Bxd4 15.Be3 Nd6 16.Rxd4 b5 (see diagram)
I've played well up to here, and any normal move must be winning. But I decided to go for an immediate mate - the first blunder. 17.Bh6 bxc4 18.Qe5 Rf6 I had missed that completely. However, my position had been so strong that even with a piece less, I had all the chances. 19.Rad1 Bb7 Better is Bxe6, but after 20.Bg5, White is probably winning. 20.Bg5 Blunder two: 22.Rxd6 wins immediately, but it escaped me that after exd6 23.e7 dxe5, I guard square d8 and Rxd8+ will mate Black quickly. 20...Rf8 21.Bh6 Rf6 22.Bg5 Blunder three, which is a repetition of blunder two. 22...Rf8 23.Bh6 Being half-glad to escape with a draw after hanging that Bishop, I clicked the draw-button, thinking we already had a threefold repetition. But apparently, as the game was not declared a draw automatically, we were half a move short of that. And in fact, Black declined the draw by breaking the repetition with a different move: 23...Qe8 That, however, was blunder four: 24.Qg7 mate.

I** - TK, ICC 5 0, 30 December 2002
1.d4 d5 2.c4 e6 3.Nc3 f5 4.Nf3 Nf6 5.Bf4 c6 6.e3 Bd6 7.Bxd6 Qxd6 8.Ne5 Nbd7 9.f4 O-O 10.Bd3 Ne4 11.O-O Nxe5 12.fxe5 Qe7 13.Bxe4 fxe4 14.Rxf8+ Qxf8 15.Qb3 Qe7 16.cxd5 cxd5 17.Rf1 Qg5 18.Nxd5 exd5 19.Qxd5+ Be6 20.Qxe6+ Kh8 (see diagram left) As so often in my favorite opening, the Stonewall, I am completely busted. But it's a little early to resign. 21.Rf4 h6 22.d5 Qh5 23.Rg4 Re8 24.Qg6 Easier is 24.Qd7 (Qxe5 25.Rxg7) 24...Qxg6 25.Rxg6 Rxe5 Now, after missing the best move a few times in a row, White has to think again. 26.Rd6 a5 27.g4 Kg8 28.Kf2 b5 29.Kg3 g5 30.Rd7 Kf8 31.h4 Ke8 32.Rd6 Ke7 33.Rxh6 Rxd5 34.hxg5 Rd2 35.Kf4 After Rh2 even I would have resigned. 35...Rxb2 36.Rh7+ Ke6 37.Kxe4 Rxa2 Now I had dangerous counterchances. 38.Rh6+ Kf7 39.Ke5 a4 40.Rf6+ Kg7 41.Rb6 Rb2 42.g6 a3 And now I was better. 43.Kf5 a2 44.Ra6 b4 45.Kg5 b3 46.Ra7+ Kg8

It was time for White to take the draw with Ra8+ etc., but after the horrible winning attempt 47.g7 it was I who was winning. (See diagram right)
Almost anything would make White resign now, most of all Rb1 48.Kg6 a1Q, or Rc2 48.Kg6 Rc8. But I played 47...Rf2, only to discover that after 48.Kg6 all my heroic resistance had been in vain - I was going to be mated. I clicked the resign button, but just before clicking the confirmation button, I saw that I still had 48...Rf8, and after 49.gxf8Q+ Kxf8 it slowly dawned upon me that this was even winning. Because after 50.Kf6 Ke8 51.Ke6 Kd8, 52.Kd6 Kc8 53.Kc6 Kb8 54.Rb7+ Ka8, White cannot stop the promotion. After 52.e4 b2 53.Rxa2 b1Q 54.Rd2+ Ke8 55.Rd4 Qg1 56.Rd5 Qxg4+ 57.Ke5 Ke7 58.Kd4 Ke6 59.Re5+ Kd6 60.Rd5+ Kc6, it hardly mattered that Black had a tablebase win - poor White overstepped the time limit.

PS 3 January: As Antoine Baudoux shows, this last win was even more ridiculous. After 52.Kd6 Kc8 53.Kc6 Kb8, White can stop the promotion: 54.Ra5! b2 55.Rb5+ Kc8 56.Rh5! Kd8 (Kb8 57.Rh8+ Ka7 58.Rh7+ Ka6 59.Rh8 Ka5 60.Kc5 etc.) 57.Kd6 Ke8 58.Ke6 Kf8 59.Kf6 Kg8 60.Rg5+ Kf8 61.Rh5 etc. Compare that other ridiculous win in item 170.

In the mid-eighties, Maurits Wind, a strong (around 2230, these days) player from The Hague and a member of one of Holland's oldest clubs, Discendo Discimus, became interested in the seemingly nonsensical Müller-Schulze Gambit; 1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6 4.Nxe5. Analyzing and playing it, he discovered its potential, and wrote about it in the DD-club magazine in 1991. This issue was later sent to the German R. Schlenker, who as a result also started playing and analyzing the gambit, and published his findings in 1993 in Randspringer, his own off-beat chess magazine. That article in turn was noticed by the Austrian chess programmer Steffen Jakob, who, from 1996 to 1998, let his crafty-clone Brause play the gambit on ICC and other servers, and who also coined the name Halloween Gambit - a name that stuck. When in 2000, I discovered thousands of highly entertaining Halloween games by Brause (and an increasing number of humans) in a database, I wrote a lengthy article about the history and the fun of the Halloween Gambit, A Breeze in the sleepy Four Knight's Game . And started playing it myself.
    And now the circle is round. Not only has the Halloween Gambit become a popular opening in internet blitz; that article also revived Wind's own interest. Analyzing the Halloween even more deeply, he will shortly publish an extensive article about it in Kaissiber, and, at the occasion of DD's 150-year jubilee, he organizes an (email) correspondence theme tourney around the Halloween, to test his belief that it can even be played against strong players who have time to think.
    This tournament will start in January; first Prize is Euro 500; prescribed moves are 1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6 4.Nxe5; 8-10 games will be played, and there are still a few free spots - players over FIDE 2100 can apply to Maurits Wind. He will give further details.
    As an illustration, Wind offers the following line he analyzed together with Kaissiber's Stefan Buecker.

1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6 4.Nxe5 Nxe5 5.d4 Ng6 6.e5 Qe7 7.Bb5 c6 8.O-O cxb5 9.Bg5 d5 10.Qd3 a6 11.Rae1 Be6 12.f4 h6 13.exf6 Qd6 14.Rxe6+ Qxe6 15.f5 Qd6 16.Re1+ Kd7 17.Bd2 b4 18.fxg6 bxc3 19.Qh3+ Kd8 20.Qxc3 fxg6 21.f7 Be7 22.Qb3 Kc8 23.Ba5 Qd7 24.Qb6 Kb8 25.Qxg6 Rc8 26.Bb6 "and Black is helpless".

White to play and draw
M. Zinar
Special Commended, Shakhmatyi v SSSR, 1984

Harold van der Heijden and I wonder whatever became of M. Zinar (no first name known) who, from 1974 to 1990, published around 300 studies, almost all king-and-pawn endings, and many of high quality. He was, perhaps still is, a worthy successor to N.D. Grigoriev - a modern Pawn Wizard.
In the diagram left, White loses his c-pawn, and stopping Black's h-pawn seems to give Black access to White's fortress: 1.Kg5 Kxc6 2.Kh6 Kd5 3.Kxh7 Ke5 4.Kg6 Kf4 and Kg3 etc. And after 1.Ke5 Kxc6 2.Ke4 Kd6 3.Kxe3 Ke5, Black wins easily with his outside passed pawn.
Only the stunning 1.Kg7, almost pushing the h-pawn into the race, works. 1...h5 After 1...Kxc6? 2.Kxh7 Kd5 3.Kg6 Ke4 4.Kg5 White is in time to stop Black - and even wins, but that is beside the point. 2.Kf6 Having given the passed pawn a free march, White now executes a Réti manoeuvre to catch it after all. 2...h4 If Black doesn't run, he loses. 3.Ke5 h3 4.Kd6 h2 5.c7 h1Q 6.c8Q and Black cannot escape a perpetual, e.g. 6...Qd1+ 7.Ke5 Qxe2 8.Qb8+ Kc6 9.Qc8+ Kb5 10.Qb7+ Ka4 11.Qa7+ Kb4 12.Qa3+ or even Qe7+ etc.: draw.

Anyone know anything about Zinar?

PS 6 December: Thanks to the readers who found out the following about Zinar. His first and father's name are Mikhail Afanasyevich, he was born in 1951, and in 1990 he pubished a book in Kiev titled Harmony in Pawn's studies.
The question remains: why did he stop composing in 1990?

PS 5 August 2007: Thanks to Siegfried Hornecker who advised me of the following post by Sergiy Didukh on the MatPlus.net forum, dated 31 July 2007:

"In 1995 Mikhail Afanasievich Zinar moved from Crimea to a little village Gvozdavka-2 in Odessa region (300km from Odessa). He taught labour discipline at school. In 1997 S.N.Tkachenko came to see him and M.Zinar told him he abandoned composition because of financial difficulties. Those were really hard times for Ukrainian people and especially for teachers. They were paid 10-20 dollars a month. A little later M. Zinar prematurely died."

On this forum you can find some more discussion about Zinar and a few of his compositions

PS 17 January 2008: René Olthof advises me that the Arves site announces Zinar as the judge of an international composing tourney for pawn studies in Ukraine. So, he is alive and hopefully well. In fact, it had already been mentioned on that MatPlus forum that the report of his death had been erroneous.

Inspired by item 184, Paul van Linde suggests a new record: maximum number of legal moves in game positions. His opening bid is this very famous game:

Karpov - Kasparov, 16th match game, Moscow 1985
1.e4 c5 2.Nf3 e6 3.d4 cxd4 4.Nxd4 Nc6 5.Nb5 d6 6.c4 Nf6 7.N1c3 a6 8.Na3 d5 9.exd5 exd5 10.cxd5 Nb4 11.Be2 Bc5 12.O-O O-O 13.Bf3 Bf5 14.Bg5 Re8 15.Qd2 b5 16.Rad1 Nd3 17.Nab1 h6 18.Bh4 b4 19.Na4 Bd6 20.Bg3 Rc8 21.b3 g5 22.Bxd6 Qxd6 23.g3 Nd7 24.Bg2 Qf6 25.a3 a5 26.axb4 axb4 27.Qa2 Bg6 28.d6 g4 29.Qd2 Kg7 30.f3 Qxd6 31.fxg4 Qd4+ 32.Kh1 (see diagram)
In this position, Black has 62 legal moves.
32...Nf6 33.Rf4 Ne4 34.Qxd3 Nf2+ 35.Rxf2 Bxd3 36.Rfd2 Qe3 37.Rxd3 Rc1 38.Nb2 Qf2 39.Nd2 Rxd1+ 40.Nxd1 Re1+ and White resigned.

There should be plenty of positions with far over 62. Only serious and verifiable tournament games, please.

The record for constructed positions is perhaps still 143, as in this composition by H.F.L. Meyer from 1880.

PS 1 December: Of course I knew I had tapped a rich reservoir here. To start with, Olaf Teschke sent me three games where the above record of 62 was beaten; the greatest mobility being 68 moves.

Abramian - Kadirova, Cannes 1997
1.e4 c5 2.Nf3 e6 3.d4 cxd4 4.Nxd4 a6 5.Bd3 Nf6 6.O-O Nc6 7.Be3 Be7 8.c4 O-O 9.Nc3 h6 10.Kh1 Qc7 11.f4 b6 12.e5 Ne8 13.Nxc6 dxc6 14.Qg4 Bb7 15.Rf3 Rd8 16.Bc2 c5 17.Rh3 Kh8 18.f5 Kg8 19.f6 Qxe5 20.fxe7 Nf6 (see diagram) White can play 68 moves here. 21.Qh4 Rfe8 22.Bxh6 Rd7 23.Bxg7 Qh5 24.Bxf6 Qxh4 25.Rxh4 and Black resigned.
It is a little bit unfortunate that this game is from a 'Girls under 10' (world) championship, but it is a "serious and verifiable tournament game". And White is certainly a sharp player. To refrain from 21.exf8Q+ (Fritz plays Qh4, too) and to sacrifice a Bishop instead of capturing a Rook next move, even if that is not the best, shows promise.
But: like in item 184, it can be argued that the different promotions are not really different moves. Perhaps there should be separate records here too; in any case, Karpov - Kasparov, with 62, still is the record for 'promotionless mobility'.

Martin Stichlberger sent me a number of constructed records in this domain. There

are records for legal and illegal positions, promotions being possible or not, and moves of the two sides together, or just of one side. Restricting myself to 'legal positions; one side', that leaves two records. In the diagram on the left, (J. Ban, 1960) Meyer's record was broken with one move - White has 144 moves there. In the diagram on the right, (W. Cross, 1967), where promotions are not possible, White has 109 moves. (But it strikes me that if we do not count underpromotions, Ban still has 111 moves, two more than Cross.)

PS 6 December: Olaf Teschke sent me two more games with mobility records.

Vaganian - Pigott, London 1975, simul
1.d4 Nf6 2.c4 c5 3.d5 e6 4.Nc3 exd5 5.cxd5 d6 6.e4 g6 7.f4 Bg7 8.Bb5+ Nbd7 9.e5 dxe5 10.fxe5 Nh5 11.e6 fxe6 12.Nf3 exd5 13.O-O O-O 14.Qxd5+ Kh8 15.Bg5 Ndf6 16.Qxc5 Bf5 17.Rad1 Qa5 18.Nd4 Rac8 19.Qe7 Bg4 20.Rd2 a6 (see diagram) White has 69 moves here. 21.h3 axb5 22.hxg4 Nxg4 23.Rxf8+ Rxf8 24.Qe6 Nhf6 25.Bxf6 Nxf6 26.Ndxb5 Qb4 27.Qd6 Qh4 28.Re2 Ng4 29.g3 Rf1+ and White resigned.
Perhaps a minor blemish: this was only a simul.

Ribeiro - Novikov, Yerevan Olympiad, 1996
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Qxd4 a6 5.Be3 Nf6 6.Nc3 Nbd7 7.O-O-O e5 8.Qd2 b5 9.Nh4 g6 10.f3 Be7 11.g4 Nb6 12.g5 Nfd7 13.Ng2 Bb7 14.h4 h6 15.gxh6 b4 16.Nd5 Bxd5 17.exd5 Nf6 18.h5 Nxh5 19.Qf2 Nd7 20.f4 Nhf6 21.Be2 Qa5 22.Kb1 Rb8 23.fxe5 Ne4 24.Qf4 Nc3+ 25.bxc3 bxc3+ 26.Ka1 Rb4 27.e6 Ne5 28.exf7+ Kd7 29.Rd4 Rhb8 30.h7 Rb2 31.Ra4 Qxd5 32.Bg4+ Kc7 33.h8Q g5 (see diagram) And here, White has 71 moves. There followed 34.Qhxe5 and Black resigned.
But this 71-move record has two minor blemishes - it includes 3 underpromotions, and perhaps there should be a separate record for positions with promoted pieces.

PS 27 April 2003: But see item 212 in this Diary.

Mig Greengard, who has a keen eye for the bizarre in chess, called my attention to the following game, played in the Bled Olympiad.

Ivanchuk - Radulski, Bled 2002 ol, 9th round
1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.O-O Be7 6.Re1 b5 7.Bb3 d6 8.a4 Na5 9.Ba2 b4 10.c3 c5 11.d4 cxd4 12.cxd4 Bg4 13.Nbd2 Bh5 14.Nb3 Nxb3 15.Qxb3 Qb8 16.Nh4 O-O 17.Qg3 Bg6 18.Nxg6 hxg6 19.dxe5 dxe5 20.Bb3 Kh8 21.Qh4+ Kg8 22.Bg5 Rd8 23.Re3 Qd6 24.Rh3 Kf8 25.Qh8+ Ng8 26.Rh7 Bf6 27.Bh6 Ke7 28.Bxg7 (see diagram) and Black resigned.

This made Mig think of another game:

Spassov - Perez, Manresa 1996
1.Nf3 Nf6 2.c4 c6 3.Nc3 d5 4.e3 e6 5.Qc2 Nbd7 6.a3 Bd6 7.b4 O-O 8.Bb2 a6 9.Be2 b5 10.c5 Bb8 11.Nd4 Bb7 12.g4 g6 13.g5 Ne8 14.h4 e5 15.Nf3 Ng7 16.O-O-O Ne6 17.Rdg1 Qe7 18.h5 e4 19.Nxe4 dxe4 20.Qc3 f6 21.hxg6 hxg6 22.gxf6 Rxf6 23.Rxg6+ Rxg6 24.Qh8+ Kf7 25.Rh7+ Ng7 26.Bxg7 (see diagram) 26...Ke6 27.Nd4+ Kd5 28.Qg8+ Re6 29.Bg4 and Black resigned.

And it made me think of one of the craziest games I've ever seen:

Kostina - Te, Moscow ch women, 1978
1.e4 c5 2.Nf3 Nc6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 d6 6.Bg5 e6 7.Qd3 a6 8.f4 Be7 9.O-O-O Bd7 10.Nb3 b5 11.Bxf6 gxf6 12.Be2 b4 13.Nb1 a5 14.Bh5 Qb6 15.N1d2 a4 16.Nc4 Qc7 17.Nbd2 b3 18.axb3 Nb4 19.Qd4 axb3 20.c3 Nc2 21.Qd3 Ra1+ 22.Nb1 Qa7 23.Ncd2 Qa2 24.Rdf1 Ba4 25.f5 d5 26.fxe6 Ba3 (see diagram) 27.exf7+ Kf8 28.exd5 Bxb2+ 29.Kd1 Na3 30.Qe4 Rxb1+ 31.Ke2 Bb5+ 32.Kf3 Rxf1+ 33.Rxf1 Bxc3 34.Qe8+ Bxe8 35.fxe8Q+ Kg7 36.Qe7+ Kh6 37.Qe3+ Kxh5 38.Qxc3 b2 39.g4+ Kg6 40.Kg3 Rf8 41.Ne4 Qc4 42.Rxf6+ Kg7 43.Qxb2 Rxf6 44.Qxf6+ Kg8 45.Qd8+ Kf7 46.Nd6+ and Black resigned.

The grandmaster title earned at 12 by Sergey Karjakin (see picture; thanks to kasparovchess), makes one wonder about the development of this record.
In 1950, when FIDE had better things to do than to make chessplayers piss in public, and first instituted an official grandmaster title, it nominated 27 players. The youngest among them was Bronstein, 26 - and since then the "youngest grandmasters ever" have been:

Spassky, 18 (1955)
Fischer, 15 (1958)
Judit Polgar, 15 (1991)
Leko, 14 (1994)
Bacrot, 14 (1997)
Ponomariov, 14 (1997)
Bu Xiangzhi, 13 (1999)
Karjakin, 12 (2002)

Of course, today there are not 27 but over 800 grandmasters - an effect produced by inflation, propaganda but also, and not in the least, by the general rise in the standard of play. In any case, beating Xiangzhi's record with 1 year and 3 months, Karjakin made the greatest leap toward birth since Fischer.
    An obvious question is who earned the grandmaster title at the latest age. That first batch of 1950 included 85-year old Jacques Mieses, but he was (like Bernstein, Duras, Rubinstein and a few others) some sort of an honorary grandmaster. There have been many honorary grandmasters later - Carlos Torre was given the title in 1977 on the grounds of a career that had ended in 1926.
    When this became a thread in rec.games.chess.misc, a good candidate was found in the Latvian Janis Klovans, who became a grandmaster in 1997 at 62 by winning the first senior world championship (he won it again in 1999 and 2001.) He had been a strong master for over 40 years, and was certainly of grandmaster strength, but - his title was an automatic result of his championship. And even if he shares this little blemish with Fischer (who automatically became a grandmaster when he qualified in the 1958 Interzonal), the question remains: what is the greatest age at which a player became a grandmaster in the regular way?
    A first bid was Edmar Mednis who was 43 when he received the title in 1980 - but the Dutch chess journalist Johan Hut beats this by pointing at Mark Tseitlin. This former Soviet master, who now lives in Israel, had 'IM' next to his name in 1995, and 'g' now. Having been born in september 1943, this means Tseitlin became a grandmaster when he was at least 51.
    But how old exactly?

PS 12 November: Paul van Linde tells me that Valery Grechihin, born 6 September 1937, was a GM on FIDE's ratinglist for October 2002, while on the July 1996 list, as published in Informator, he was an IM. This means he was at least 58 when he became a grandmaster.

PS 7 January 2003: Grechihin's record for becoming a grandmaster turns out to be 60. Johan Hut found that he was an IM on FIDE's rating list of 1 January 1998, and a GM on the list for 1 July 1998.

PS 14 March 2007: Åge Trulssen notifies me that 55-year old Leif Øgaard has just become Norway's ninth grandmaster. That is not a record, but the fact that one of the necessary norms was obtained 25 years ago, might be. (At some point FIDE dropped the rule that these norms have to be made within 5 years.)

Immediately after uploading the previous item, when I took my nightcap of having a look at ICC, and typed my alias for observing the highest rated game, I saw this.

Relange - YARDBIRD, ICC 5 0, 22 October 2002
White is the French GM Eloi Relange, Black the German/American GM Eric Lobron.
1.e4 c6 2.Nf3 d5 3.Nc3 Bg4 4.h3 Bh5 5.d3 e6 6.Be2 Bd6 7.O-O Nd7 8.exd5 exd5 9.Bg5 Ne7 10.Nh4 Bxe2 11.Qxe2 O-O? 12.Nf5 Nxf5!? Black understandably doesn't fancy the ordeals of 12...Nf6 13.Bxf6 or 12...f6 13.Qe6+, and decides to play on with two minor pieces against the Queen. 13.Bxd8 Raxd8 14.Qg4 g6 15.Rfe1 h5 16.Qd1 h4 17.Qd2 Kg7 18.Ne2 Rfe8 19.Qa5 a6 20.Rad1 Nf6 21.Qd2 Rd7 22.a4 Rde7 23.a5 d4 24.c4 c5 25.Kf1 Nh5 26.Nc1 Bf4 The compensation has grown and grown, and now 27.Qc2 would even lose: 27...Nhg3+! 28.fxg3 Be3! and White is helpless, e.g. 29.Ne2 Re6 30.gxh4 Ng3+ 31.Nxg3 Rf6+ 32.Qf2 Rxf2+ and Black wins. 27.Rxe7 Rxe7 28.Qc2 Bh2 Preparing a unique perpetual. White's last winning chance would have been 29.Ne2. 29.Re1 (see diagram) 29...Nfg3+! 30.fxg3 Nxg3+ 31.Kf2 Nh1+! 32.Kf1 After 32.Kf3 Rxe1, Black must be winning, but why not 32.Rxh1? Because White is mated: 32...Bg3+ 33.Kf3 (K-1 Re1 mate) Re3+ 34.Kg4 f5+ 35.Kg5 Bf4+! 36.Kxf4 (36.Kxh4 Kf6 and Bg5 mate) 36...Kf6 and mate next move by g5. They played a tempo around here; would they both have seen this mate, and especially Bf4+? After 32...Ng3+ 33.Kf2 Nh1+ 34.Kf1 Ng3+ 35.Kf2, a draw was agreed.

Once more a readers game, this time from Vienna, between two club players with ratings of around 1900.

Pichler - Hoza, Vienna 1991
1.e4 e6 2.d4 d5 3.Nc3 Bb4 4.e5 c5 5.a3 Bxc3+ 6.bxc3 Qc7 7.Nf3 Ne7 8.Bd3 Nbc6 9.a4 c4 10.Be2 f6 11.Bf4 Ng6 12.Bg3 f5 13.Nh4 Nxh4 14.Bxh4 O-O 15.f4 Qa5 16.Qd2 b5 17.O-O bxa4 18.Rfb1 La6 19.g4 Rfe8 20.gxf5 exf5 21.Bf3 Rab8 22.Qg2 Rxb1+ 23.Rxb1 Rb8 24.Bxd5+ Kh8 25.Re1 Nxd4 Despair, but clever despair. 26.cxd4 a3 27.Kh1 a2 (see diagram)
    White, with his extra piece, had underestimated the danger of Blacks passed pawn, but believed that with 28.Rg1, he was in time; Qc7 is met by 29.Bf6 and mate. However, the Queen sacrifice 28...Qxd5 woke him from his dreams; 29.Qxd5 Bb7 and Black wins. "After a few minutes of painful pondering, I resigned," wrote Pichler.
    Things became even more painful when, the opponent having left, some club members had a look at the position. Could White have prevented the Queen sacrifice? "And then, suddenly, somebody saw it. Instead of allowing the Queen sacrifice, I could have made an at least equally beautiful Queen sacrifice myself!" In fact, after 28.Qxg7+, Black is mated: 28...Kxg7 29.Rg1+ Kh6 30.Bg5+ Kg7 31.Bf6++ Kh6 (Kh5 32.Bf7+ Kh6 33.Bg7 mate) 32.Bg7+ Kh5 33.Bf7+ Kh4 34.Bf6+ Kh3 35.Rg3 mate.
    And now Pichler wondered whether this had ever happened before - a move missing a Queen sacrifice, and thereby allowing one. My first thought was: quite often, probably, but to my surprise, I have not been able to find even one example.
    Anybody?

Van Leeuwen (2170) - Van der Linden (2200), tt Belgium, 6 october 2002
After 1.e4 d5 2.exd5 Qxd5 3.d4 e5 4.Nf3 Bg4 5.Be2 exd4 6.Nc3 Bb4 7.O-O Bxc3 8.bxc3 Nc6 9.cxd4 Nge7 10.c4 Qe4 11.d5 Ne5 12.Nxe5 Bxe2 13.Qa4+ Kf8 14.Nd7+ Kg8 15.Re1 Ng6 16.Qb4 Nh4 17.f3 Qd4+ 18.Kh1 Qf2 19.Rg1, the position in the diagram arose, with Black to play. He played the obvious 19...Nxf3 and White, who did not see how to meet the mating threat without a heavy loss of material, resigned.
He could have saved the day.
Thanks to Valeer Maes for this nice example of an ultimate blunder. Compare item 188 in this diary.

As a young boy, I sometimes dreamed of having a magic brew I could sprinkle on newspaper photographs, to see what happened next - cars would start moving, people walking. It would be great for chess books and magazines, too. A diagram with the initial position would be enough; the magic would make it come alive and show you the whole game.
Palview is a (freeware!) program that makes this dream come true. The chessboard below contains all the diagrams that follow - use your mouse on the symbols or on the moves themselves.

Black to play and not draw
Conrady, Hayworth 2002

1...Qd5+ 2.Kc8 Qe6+ 3.Kb8 Qe5+ 4.Ka7 Qa5+ 5.Kb7 Qd5+ 6.Ka6 Qd3+ 7.Ka7 Qe3+ 8.Ka8 Qa3+ 9.Kb8 Qd6+ 10.Ka7 Qa3+ 11.Kb6 Qe3+ 12.Ka5 Qc5+ 13.Ka4 Qa7+ 14.Kb5 Qb7+ 15.Kc4 Qc6+ 16.Kb3 Qf3+ 17.Kb4 Qb7+ 18.Ka3 Qa7+ 19.Ra4 Qc5+ 20.Rab4 Qa7+ 21.Kb3 Qe3+ 22.Kc4 Qe4+ 23.Kc5 Qe5+ 24.Kc6 Qe6+ 25.Kb7 Qd5+ 26.Kb6 Qd8+ 27.Kc5 Qc7+ 28.Kd5 Qd7+ 29.Ke4 Qc6+ 30.Kf4 Qf6+ 31.Kg3 Qg6+ 32.Kf3 Qc6+ 33.Re4 Qc3+ 34.Re3 Qc6+ 35.Kf4 Qf6+ 36.Ke4 Qc6+ 37.Kd3 Qd5+ 38.Ke2 Qc4+ 39.Rd3 Qe4+ 40.Kd2 Qf4+ 41.Kc3 Qc7+ 42.Kd4 Qd6+ 43.Ke3 Qe5+ 44.Kf2 Qc5+ 45.Ke2 Qc2+ 46.Ke3 Qc5+ 47.Rd4 Qc3+ 48.Ke4 Qc6+ 49.Rd5 Qc2+ 50.Kd4 Qf2+ 51.Kc3 Qe3+ 52.Kc4 Qe2+ 53.Kc5 Qe3+ 54.Kc6 Qe6+ 55.Rd6 Qc8+ 56.Kb5 Qb7+ 57.Kc5 Qc7+ 58.Kd5 Qa5+ 59.Kc6 Qc3+ 60.Kb7 Qf3+ 61.Rc6 and White wins.

Thanks to Helmut Conrady and Guy Haworth for this tablebase position which is the longest shortest win with this material. During the enigmatic "DTC (Distance-to-Conversion)-minimaxing play", Black has a record 60 consecutive checks (and one spite check if he wants) before the Rooks start doing their work.

188. 1 October: A tricky Queen's ending

For the last 25 moves or so, in Pedersen (2266) - Bergström (2306), Eurpean Club Championship, Halkidiki, 28 september 2002 (see diagram; White to play), White had been trying to win with his extra pawn, finally in a Queen's ending. Now he saw his chance: 54.Qe5, followed by Qh8 mate or a won pawn ending.
Black saw it too, and resigned. Both missed the not too hidden mate in 3 with 54...Qxe5 55.fxe5 f4 and g5 mate.
Thanks to Johan Østergaard for the tip.

For more resignations in won positions, see items 12, 150, 182 and 186 in this Diary, and The ultimate blunder, elsewhere on this site.

187. 20 September: The case for cheating

Regulary, lengthy threads develop on rec.games.chess.misc about 'computer cheaters' on ICC and other servers, with postings fuming with anger, accusing specific players, suggesting detection methods, ways to stop it.
    I've never understood what all the fuss is about. An inescapable fact of Internet chess is that you can't see your opponent - you play a chessmove-producing entity you know nothing about. This is also true of correspondence chess where the use of computers is fully accepted - and what else is Internet chess but very fast correspondence chess?
    I fail to see how I am cheated when I play a 2600-rated entity that produces 2600-calibre moves. Or how I'm cheated when my opponent pretends to be something he's not, when I didn't know what he was in the first place. Of course, many players, myself included, dislike playing computers. But if an entity beats you too often, or you don't like its moves, you can stop playing it. It's true that you might have to play it in ICC's 5-minute category where pairings are automatic - but if it really matters that much to you; avoid that category. As it is, the 5-minute system, with its constantly changing partners, protects you from playing computerlike humans too often. Because let's be reasonable - only a very small minority of ICC-players deny themselves the pleasure of playing chess, to create ratings that have no meaning even to themselves.
    The hunt for computer cheaters is a refusal to acknowledge that Internet chess cannot be like over-the-board chess. The attempted remedies are worse than the disease. ICC's program Blitzin installs spyware on your computer, without telling you so during installation. The witchhunt encourages paranoia, bad sportsmanship ("you computercheating bum you, you use a computer, Bc5 gave you away") and worst of all, slander.
    Recently, I was disheartened to see that even Roman Parparov, one of r.g.m.c.'s most interesting and serious posters, stooped to paranoia and slander when he posted the following game in which he had White.

romm (2039) - X (2588), ICC 5-minute, 2001
1.e4 c5 2.Nc3 g6 3.Nge2 Nc6 4.d4 cxd4 5.Nxd4 Bg7 6.Be3 Nf6 7.Nxc6 bxc6 8.e5 Ng8 9.f4 Nh6 10.Be2 d6 11.exd6 exd6 12.Bd4 Nf5 13.Bxg7 Nxg7 14.Qd4 O-O 15.Ne4 (see diagram) 15...Ne6 16.Nf6+ Kg7 17.Nh5+ Kh6 18.Qf6 Qxf6 19.Nxf6 Kg7 20.Ne4 d5 21.Nc5 Nxc5 and White resigned.
    'Allowing a move like 16.Nf6+,' wrote Parparov, who also gave X's handle, 'is what no normal human player would do. And playing 16....Kg7 was also extremely suspicious. (It) forced me to submit a complaint.'
    But not to accuse X publicly, not with such scant evidence. Any experienced player knows that this Queen-Knight battery is almost never as strong as it looks, especially when g7 is guarded, and the Queen is under attack - discovered and even double checks tend to be harmless under those circumstances.
    Fritz, by the way, does not play Ne6, but the human 15...d5.
    And slander is not warranted by a few stupid people pinching a few stupid rating points.

Pilnik - Stahlberg, Beverwijk 1963
1.e4 e6 2.d4 d5 3.Nc3 Nf6 4.Bg5 dxe4 5.Nxe4 Nbd7 6.Nf3 Be7 7.Nxf6+ Bxf6 8.Bxf6 Qxf6 9.Bd3 O-O 10.O-O e5 11.Re1 exd4 12.Nxd4 g6 13.Nb5 Qb6 14.a4 Nc5 15.a5 Qc6 16.Bc4 a6 17.Nc3 Be6 18.Bd5 Bxd5 19.Nxd5 Qd6 20.Qd4 (see diagram) 20...Ne6 21.Nf6+ Kh8 22.Qh4 Kg7 23.Ng4 h5 24.Qf6+ Kh7 25.Rad1 Qf4 26.Rxe6 Qxg4 27.Rd7 Kg8 28.Ree7 Rae8 29.h3 Rxe7 30.hxg4 Rxd7 31.gxh5 Rd1+ 32.Kh2 Rd6 33.Qe7 gxh5 34.Qxc7 Rfd8 35.Qxb7 Rg6 36.b4 Kg7 37.b5 Rdd6 38.b6 Rd2 39.Qe4 Re6 40.Qf4 Rd7 41.c4 Kf8 42.c5 and Black resigned.

186. 18 September: Ultimate Blunder & King Hunt

Thanks to Dolf Wissmann and Rein Halbersma for sending me two interesting games.

Wissmann - De Groot, St. Jacobi Parochie, 1985 (ratings around 2000)
1.e4 e5 2.f4 exf4 3.Nf3 h6 4.d4 g5 5.Bc4 d6 6.O-O Ne7 7.Nc3 Bg7 8.g3 Bh3 9.Rf2 Nbc6 10.gxf4 g4 11.Be3 d5 12.exd5 Nb4 13.Ne5 Nbxd5 14.Qd2 Kf8 15.f5 Bxe5 16.dxe5 Nxe3 17.Qxe3 Nc6 18.e6 Qd4 19.e7+ Ke8 20.Bxf7+ Kxf7 21.Qe6+ Ke8 (see diagram) Here Qg6+ wins easily, but Wissmann (a top problem solver,these days) played 22.f6, with a formidable threat that made Black resign.
Only later did he see that Black's resignation was an ultimate blunder: 22...Qxf2+ 23.Kxf2 g3+ and Black wins. "I was so nasty as to call De Groot and tell him this," writes Wissmann. The explanation of their mutual blindness must be the very visibly congested h3-e6 diagonal.

Eppinga - Van der Wal, Groningen, year unknown
White was a former junior champion of the Netherlands, Black a former world champion of draughts (1982) who died in 1996 at the age of 39. In chess, he was a FIDE master with a top rating of 2245. He beat several IMs, but here, he is brilliantly beaten himself.
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Bc4 e6 7.O-O b5 8.Bb3 b4 9.Na4 Nxe4 10.Re1 d5 11.Qf3 Bb7 12.Nxe6 Starting a sacrificial attack that never stops. 12...fxe6 13.Qh5+ Kd7 14.Rxe4 dxe4 15.Bg5 Qc7 16.Qf7+ Kc6 17.Qxe6+ Bd6 18.Qxe4+ Kb5 19.c4+ Ka5 20.Qf5+ Be5 21.a3 Nc6 22.axb4+ Kxb4 23.Bd2+ Bc3 24.Qc5+ Kxb3 25.Ra3+ Kc2 26.Rxc3+ Kd1 (See diagram.) A chase over the whole board has ended. 27.Rc1+ Ke2 28.Qe3 mate.

185. 28 August: Excavations

Courtesy ChessBase From their discovery of a "chess piece" in an excavation of a 5th or 6th century AD Byzantine palace in southern Albania, archeologists concluded that chess was played there at the time. This would change the European and the general history of the game, as there is no proof that chess was played anywhere before AD 600. However, the fact that it was a single piece, and might therefore just as well have been some trinket or toy, makes me think of a much older dating of chess.
How do we know chess was played in Sumeria as far back as 3000 BC? Excavations did not turn up any chess pieces or boards, which proves the Sumerians played blindfold chess.

184. 6 August: How many moves can you kill?

A witty problem I came across.

Mate in 3
K. Wenda
1st Honorable Mention, Shakhmatyi v SSSR, 1963

    If it were Black's move, he would have to open the b-, c- or d-file, and White could mate on the 7th rank. However, White is in Zugzwang. Queen and Rooks must clearly stay in ambush, and all waiting moves with the Bishop have their fault; on b4, c5 and d6, it cuts off a line of attack, (1.Bb4? bxc1Q; 1.Bc5? cxd1Q or even a3; 1.Bd6? dxc1Q 2.Qd4 Qh1 and White is one move too late); on e7, it obstructs the 7th rank and on f8, it loses control over f8.
    Still, the paradoxical 1.Be7 is the key move, and now in 3 variations, White always re-opens the 7th rank, while closing a file:
1...dxc1Q 2.Qd7 Qd1 3.Bd6 mate
1...cxb1Q 2.Rc7 Qc1 3.Bc5 mate
1...bxc1Q 2.Rb7 Qb1 3,Bb4 mate
And of course, 1...a3 2.Bxa3, and now Black is in Zugzwang - it is mate next move.

The possibility 1.Rxc2, which stalemates, and kills all 16 moves Black would have had, gave me the idea of a construction task: what is the maximum number of moves that can be killed with 1 move?
    My opening bid is the diagram on the left, where Qxd4 stalemates, when Black to move would have had 47 moves; 25 by the Queen, 14 by the Rook, and 8 by the pawns.
    Clearly, many variations of this setting are possible. In some of them, a few pawns could be added to give Black an en passant capture, making it 48. But I feel the en passant capture is a move that can only be killed by the side not making it.

PS: Hardly had I posted this article, or Rob van der Plas sent a position where White kills 191 moves (see diagram right), because after Qxb7, Black is mate. Nice, but I forgot a few stipulations, I'm afraid: legal position, no promoted men, and White must stalemate.

PPS: Within minutes of each other, Dirk Hoppe and again Rob van der Plas sent me true new killmove records.

Dirk Hoppe; Bxd5 stalemate

Rob van der Plas; Bxd5 stalemate

On the left, Dirk Hoppe reaches 48, while Van der Plas, with a comparable setting, and the same killing move, hits the 50 mark.

PS 7 August: I had quite a lot of reactions to this construction task. Joshua Green sent this (diagram right) improvement over Van der Plas' position: he needs one piece less, and gives the white King a role.
    Several readers objected to counting a promotion as four different moves. Even if the difference between promotions led to some of the most beautiful problems ever composed (see Diagram of the Century ), there is a point to this objection - the four promotions of a pawn are

W.A. Shinkman; Bxe3 stalemate

the same movement. Perhaps there should be a separate record for promotionless move-killing. This record turns out to have been higher than Hoppe's 48 for a long time already - the great Shinkman (1847 - 1933) having shown such a position where 49 moves are killed. (See diagram left) Thanks to Martin Stichlberger for showing it to me. He found it in Chernev's The Chess Companion, where no source is given. Strangely, this construction task is not treated in the bible for this kind of thing, Dawson's Ultimate Themes from 1938. It has quite a few positions by Shinkman, but not this one.

PS 9 August: Piet Peelen suggests a mutual move-killing construction task, and makes the opening bid on the right. With QxQ, both sides can kill 31 moves, for a total of 62.

PS 11 August: After several attempts, the record in the mutual move-killing construction task, is now at 70, again by Dirk Hoppe. In his position (diagram right) QxQ kills 35 moves each.

PS 2 January 2003: To see what the Canadian/Romanian problem composer Cornel Pacurar did with this theme in miniature settings, see this page on his site. For positions with 8-10 men (and a more economic version of Shinkman's record), see this page.

183. 3 August: Kubbel's dreams

In the story Dream combination on this site, I give the combination on the left, which once occurred to me in a dream. Black to move, wins with 1...Qh3+ 2.Kxh3 Nf4+ etc. I couldn't immediately think of other examples of this simple combination but recently, I saw two beautiful studies by Leonid Kubbel, where it is used to great effect - and which, of course, I may have seen before that dream.

White to play and win, Leonid Kubbel, Listok SKP 1921
1.Qe4+ Kb8 2.Rb6+ Bxb6 Or Kc8 3.Qb7+ Kd7 4.Ne5+ Ke7 5.Qxc7+ and wins. 3.Ka6 Rd7 Or Rd5 4.Qxd5 Qc8+ 5.Kxb6 and wins. And now a game of three-cushion billiards: 4.Qa8+ Kxa8 5.Nxb6+ Kb8 6.Nxd7+, followed by 7.Nxf8 and wins.

White to play and win, Leonid Kubbel, Shakmatny Listok 1924
I like this one even better. 1.Qa2+ Now Black must watch out for checks by the Ne2. 1... Kb4 2.Qb2+ Kc4 3.Qc2+ Kb4 4.Kb2! Threatening Qb3 mate as well as Qc5+ with mate next move. 4...Qd5 The only move. But: 5.Qa4+! Kxa4 And there the Knight goes again: 6.Nc3+ Kb4 7.Nxd5+ Kc(a)4 (or Kb5 8.Nc7+ etc.) 8.Nb6+ followed by 9.Nxa8, and White wins.

182. 1 August: The latest in ultimate blunders

In a game Yetman - Brollini, Tucson 2001, White to play, White resigned because he thought he had to exchange Queens and would remain two pieces down. Perhaps he saw 35.Bb7+, but not that after 35...Kd7 (35...Kd8 36. Rxb8+ and mate), he had the further check 36.Bc8+!, which would have won;
36...Kxc8 37.Qc6+ Kd8 38.Qc7+ and mate, or:
36...Rxc8 37.Qb7+ and mate, or:
36...Ke8 37.Bd7+! and mate in a few moves, or:
36...Kd8 37.Qxe3 Bxe3 38.Bxe6 etc.

181. 14 July 2002: The litte guys

Browsing Brad Darrach's Bobby Fischer vs. the Rest of the World (why was it never reprinted?) I found a detail I had forgotten. One of Fischer's demands for the match against Karpov in 1975 was that there had to be a special pair of tiny pawns for the drawing ceremony - they could be concealed better.