Some time ago (this was originally a Diary entry for 26 September 2001) I received an email from Bader Al-Hajiri from Kuwait. He wanted to share an obsession with me that had had him in its grip for months.
Mate in 8
W. Shinkman, 1887
His story began with the well known gem on the left.
    The American Shinkman (1847 - 1933) was a genius of the early modern chessproblem. He composed a few thousand, but none of them has ever been as popular as this little joke. The solution, as it should be in a joke, is not difficult: 1.0-0-0 Kxa7 2.Rd8 Kxa6 3.Rd7 Kxa5 4.Rd6 Kxa4 5.Rd5 Kxa3 6.Rd4 Kxa2 7.Rd3 Ka1 8.Ra3 mate.
    However, there is a shadow over this classic, in the form of two anecdotes. In one, the problem was also solved when a mirrored version was given, (castling is then impossible) and in the other the solution was found by someone who thought castling was not allowed in problems. In fact, the sad dual solution 1.Kd2 (1.Ke2 in the mirrored column) with mate in 8, was first published in 1929. In 1975, Speckmann gave a version with Pa2 on c2; 1.0-0-0 then leads to mate one move sooner, with 7.Ra4.
    That, of course, was nowhere near the charm of the original and worse, that version too, was cooked, as Bader al-Hajiri showed: 1.c4 also leads to mate in 7: 1...Kxa7 2.0-0-0 Kxa6 Or 2...Kb8 3.Rd7 Ka8 (3...Kc8 4.Rb7 Kd8 5.a7 and 6.a8Q mate) 4.Rd8+ Ka7 5.c5 Kxa6 6.Rd7 Kxa5 7.Ra7 mate. Or 2...Ka8 3.Rd8+ Ka7 4.Kb2 Kxa6 5.Rd7 Kxa5 6.Rd6 Kxa4 7.Ra6 mate. 3.Rd7 Kxa5 4.Rd6 Kxa4 5.Kb2 Ka5 6.Ka2 (other moves work, too) 6...Ka4 7.Ra6 mate.

But this cook of the correction of a cook was not what fascinated Bader Al-Hajiri. He wondered whether Shinkman's original position, with the sextupled pawns on the a-file, was legal. It did seem so, because the white pawns must have captured 15 times and 15 black pieces are missing, but all those black men having gotten in the paths of the white pawns needs such a breathtakingly complicated history of black captures and/or promotions that to make sure it could really have happened, all that can really be done is to construct a game in which it has happened. And that is what Bader Al-Hajiri had done.
    In his first mail, he told me that he had not only wanted to construct a proof game, but that he had also yielded to the challenge to make it as short as possible. After a month of analysis he had found a 51-move game that did it, but after spending even more time, he had now achieved something far better - the game he proudly sent me was only 43 moves. He was convinced this was the minimum, but a few days later I again received a mail from him, in which he wrote: "I was absolutly wrong!! Can you believe it, I first found a shortest game in 43 moves, but now I found a game in 35 moves! And this is - finally - the minimum, I'm sure!"
    But, only a few days later, Al-Hajiri mailed me once more: "There is a proverb that says: 'the third time is immobile'! (I wonder of what that could be the translation - TK.) What do I say? I say that chess is fantasy, is ingenuity, is immortal. I am amazed by the art of chess! And I am amazed by its secrets! My friend, I first gave you the game of Shinkman's problem in 35 moves. I had analysed very deeply, and for a long time I tried to find a game in 34 moves, but I didn't manage. I thought it was impossible or very difficult to solve, and for this reason I sent you the game in 35 moves. But suddenly, I heard a voice inside me: 'You must try! It can be done in 34!' Then I analysed again and again, for many days, and finally I saw the IDEA, and I did it in 34 moves!! I couldn't believe my eyes!!! But it was really true! I thought I was dreaming!! Yes my friend, this is true! I finally did it, in 34 moves!! And here is the Immortal Game:
1.g4 e5 2.Nh3 Ba3 3.bxa3 h5 4.Bb2 hxg4 5.Bc3 Rh4 6.Bd4 exd4 7.Nc3 dxc3 8.dxc3 g3 9.Qd3 Rb4 10.Nf4 g5 11.h4 f5 12.h5 d5 13.h6 Bd7 14.h7 g2 15.h8B g1R!! 16.Bd4 Ba4 17.Rh4 Rg3 18.Bg2 gxf4 19.Be3 fxe3 20.Be4 fxe4 21.fxe3 exd3 22.exd3 c5 23.Rc4 dxc4 24.dxc4 b5!! This is the IDEA! 25.cxb4 Qa5 26.cxb5 Na6 27.bxa5 0-0-0!! 28.bxa6 Rd4 29.exd4 Rb3 30.cxb3 Ne7 The saviour! 31.bxa4 Nd5 32.dxc5 Nb6 33.cxb6 Kb8 34.bxa7+ Ka8.....Shinkman's problem; White to play and mate in 8 moves (two solutions).
    For this why I love chess too much!!"

PS 15 October 2001: This was not the last of Al-Hajiri's discoveries in the Shinkman problem. A friend he showed the position, gave him a "very old book without cover." In this book, he read: "Shinkman originally intended it as a retro-analytical exercise showing the position after Black's 34th move including seven consecutive Black captures, so it was not originally intended as a sound 8-mover."
    Not only, then, had 34 moves been the minimum that Shinkman himself had arrived at, his solution had this wonderful 7-move capturing series in it. However, the game itself was not given, so all Bader Al-Hajiri could do, was to accept this new challenge - and he managed to solve it again.
1.Nh3 f5 2.Na3 h5 3.g4 hxg4 4.Nf4 Rh3 5.Nc4 Ra3 6.bxa3 e5 7.Bb2 d5 8.Bd4 Bd7 9.h4 Bb4 10.h5 Ba4 11.h6 g3 12.h7 g2 13.h8B g1R 14.Rh3 g5 15.Rc3 Rg3 16.Bg2 exd4 17.Be4 dxc3 18.Bd4 gxf4 19.Be3 fxe3 20.dxc3 fxe4 21. Qd3 exd3 22.fxe3 dxc4 23.exd3 c5 24.dxc4 b5 25.cxb4 Qa5 26.cxb5 Na6 27.bxa5 0-0-0 28.bxa6 Rd4 29.exd4 Rb3 30.cxb3 Ne7 31.bxa4 Nd5 32.dxc5 Nb6 33.cxb6 Kb8 34.bxa7+ Ka8 The captures are in moves 16-22.

PS 19 September 2007: Bader Al-Hajiri found a witty way to correct the original Shinkman problem:

Mate in 8 (Chess960)
Bader Al-Hajiri (after Shinkman), Kuwait, 2007
dedicated to Tim Krabbé

1.0-0-0! Kxa7 2.Rd8 Kxa6 3.Rd7 Kxa5 4.Rd6 Kxa4 5.Rd5 Kxa3 6.Rd4 Kxa2 7.Rd3 Ka1 8.Ra3 mate.

The idea is that this problem does not have the dual solution 1.Kd2. The initial position was as in the diagram on the right; after 1.0-0-0, the pieces end up on the same squares as in normal Queen's side castling.

Al-Hajiri also offers a shortest proof game: 1.Nd3 e5 2.Nc3 Ba3 3.bxa3 h5 4.g4 hxg4 5.Bg2 Rh4 6.Bb2 g3 7.Qb1 Rb4 8.Ne4 f5 9.Bd4 d5 10.Qb3 c5 11.Qc4 Nf6 12.Bf1 Qa5 13.h4 Bd7 14.h5 Na6 15.h6 0-0-0 16.h7 Ba4 17.h8N g5 18.Ng6 b5 19.Ngf4 dxc4 20.Nd5 exd4 21.Ndc3 dxc3 22.Rh8 fxe4 23.Rh1 exd3 24.exd3 g4 25.dxc3 g2 26.dxc4 g3 27.cxb4 gxf1N 28.cxb5 g2 29.bxa5 gxh1B 30.bxa6 Ne3 31.fxe3 Rd4 32.exd4 Bd5 33.dxc5 Bdb3 34.cxb3 Nd5 35.bxa4 Nb6 36.cxb6 Kb8 37.bxa7+ Ka8, pointing to the five consecutive Black pawn captures from move 19 to 23.

Al-Hajiri adds that there are many intitial chess960 positions that would allow 1.0-0-0 as a unique solution, but he liked this one best, also because only the King and the Ng1 have swapped places.

© Tim Krabbé, 2001

Top of the page | Main chess page | Main page