Friday, December 25, 2009

The twelve puzzles of Christmas

Harley, 1926

White to play and mate in two moves


Jonathan B said...

I'm not much of a puzzler usually but I'm going to try to solve as many of these as I can.

This took me a few minutes ... I think I got the key principle pretty quickly but it took me a while to make it work and then the right move suddenly popped into my head.

What I liked about it was finding more problems to solve once I found the right idea.

Are we supposed to be putting the solution in our contributions or will you be publishing this later in the series EJH?

Anonymous said...

Two-movers are about my limit. This one was nice. I'd exhausted most of White's plausible options before I hit upon the winner – a move that would have been counter-intuitive in a game situation (not that this position resembles a game situation).

By the way, did anyone see Leonard Barden's puzzle in yesterday's Guardian, and are the kings supposed to be on d1 and d8?


ejh said...

JB - do give the solution if you think you've got it, yes.

James - no, do tell.

Anonymous said...

1. Qc2 (threat is Qc3).
1. ... Na4 2. Ra4
1. ... d4 2. Bd2

Mike G

Anonymous said...

1 Ra5 seems to work, and is certainly thematic.


Jonathan B said...

Well I thought 1. Ra5 with the idea

1 ... Kxa5
2. Bd2 mate

1. ... Nxa5
2. Ba3 mate

1. ... Nd6
2. Bd2 mate

1. ... e3
2. Rb5 mate

Jonathan B said...

I don't think there's a mate in one after

1. Qc2 Nc4(!)

Anonymous said...

Barden's challenge is to construct a game starting with 1.e4 and ending with 5...Nxh1 mate. He says this is a tough puzzle which has defeated three world champions.

After some thought I found the right way to restrict the white king: 1.e4 Nf6 2.f3 Nxe4 3.Qe2 Ng3 4.Qxe7+ Qxe7+ 5.Kf2 Nxh1 mate.

However, Barden's diagram of the starting position shows the kings and queens interposed: white king d1, queen e1, black king d8, queen e8. Barden doesn't comment on this. Could it be another Grauniad misprint? Or does the challenge really start with the white king on d1? If so, I
haven't solved it yet.


ejh said...

You mean this?

In the diagram I've got, the kings are in the conventional positions, so I wonder if somebody's got it wrong.

ejh said...

In our puzzle, by the way, 1.Ra5 is indeed correct. The puzzle is from Pick Of The Best Chess Problems, complied by BP Barnes, Elliott Right Way, 1976.

Anonymous said...

Thanks for the link. It's clearly the same puzzle, except that Barden is more specific about the last move (the knight must capture on h1), and his diagram has kings and queens interposed.

I'm inclined to think Barden's diagram is wrong and the king should be on e1. First, because both Barden and Friedel (in the link) refer to the puzzle defeating three world champions, and Friedel's text contains a number of failed attempts all of which require the king to be on e1. Second, because if there is a mate with the king on d1, the opening move 1.e4 and the last move 5...Nxh1, then there is a mirror mate with the king on e1, the first move d4 and the last move 5...Nxa1. It is clearly more satisfying to start off from the original position.

There may well be a mate with the king on the wrong starting square, taking a move longer to reach f2 or g3, but since I no longer believe the starting position is correct, I'm less inclined to search for it.

I don't understand why so many people have found this problem so difficult. The white king has to end up on f2 or g3, taking three or four moves in the process (e4 is given and f3 is necessary), so he has at most one or two other moves. Black must spend four moves getting his knight to h1 so he has at most one other move. It doesn't take long to realise that the white king can't end up on g3, because it would have four escape squares (f4, g4, h4, h3). Black can only cover two of these (...e5 or ...g5 cover f4 and h4, ...d6 or ...d5 cover g4 and h3) and White only has time to cover one of them (Qg4, Nh3).

So the king must be mated on f2. Now we need to cut off three squares on the e-file (e1, e2, e3). Again it doesn't take long to see that the black queen must guard these squares along the file, therefore both sides' e-pawns must disappear. The black knight takes care of White's e4 pawn en route to the corner, the white queen uses up the two spare moves in capturing the black pawn on e7, and Black's one spare move is to recapture with the queen on e7 and get on to the (now) open e-file. And that's all there is to it.

OK, I've hijacked this post, but I hope people found this interesting. I'm looking forward to Barden's explanation.


Anonymous said...

James. The original puzzle was not Nxh1#. It was 5 NXR# without stipulating which rook, which square it was on, and even which colour gave mate. Therefore it was a much harder puzzle when Kasparov failed to solve it.

an ordinary chessplayer said...

Jonathan B:

What is this key principle you mentioned? I thought it was mate?

Like you "the right move suddenly popped into my head". Actually my first tries were, in the order I looked at them: Qb3+, Bd2+, Ba3+, Ra6!?, Qd4+. Then some dithering (Qc2, K-moves, etc.) before looking at Bd2+ and Ba3+ again. While considering Ba3+ I concluded "it has to be a rook move". (Notice the Ba3 and Ra6 sequence first time through.) Then pop.

So just curious -- did anybody else find the solution directly while considering Ba3+?