A Chess Mystery

White to move and mate in 2.

I did not compose this problem, but I can't tell you who did.  I mean I could, but then I would have to kill you.

Amy

Qa1, then h8# =D pawns are going down the board too lol. thnx for the puzzle =)

After 1. Qa1 then 1. ... O-O-O?

It's a good thing black can't castle then, since any previous move would have removed castling rights.

But if Black can castle, what was Black's previous move?

-EJ

OK, I'm giving it to you by a one-minute timestamp priority !

The puzzle is thought to date from 1859, composed by the redoubtable Sam Loyd, although if that date is correct, he would have been only 17 or 18 at that time.

Amy

Drat! I should learn to never leave this site!

- EJ

q c6+

Wow, looks like I got home at the exact right time!

So, then Black can not castle?  Otherwise, not a mate in 2.  

Yes, in a sense, this problem is a bit of a cheat, but it brings up an interesting metachessic point:  when presented with a chess problem, do we consider it as arrived at (from the legal starting position) by following all the rules of chess?

From what I've read, this problem by Loyd is the first in history to include an aspect of retrograde analysis, the type of chess problem made famous by Raymond Smullyan.

Can black castle to save himself from a mate in two?  If this position arose from an actual game, then no, as SirDavid and Aliyat-EJ indicated.  Why?  Because, since the black pawns are on their original squares, black's last move had to have been with the rook or the king.  Thus castling is not allowed.

Amy

But could black's last move have been a knight move, with that knight that moved now off the board, having been taken by the white queen or king?

Nevermind -- it is now white's turn, not black's, correct?

No, because if white had just captured the knight, it would be black's move.  But in the position shown it is white to move.

I think this is the solution (highlight):

1. Qf1 isn't that the answer?

Tajamoen black just moves the rook and you cant checkmate him in 2 =)

qixel i know, this puzzle is unsolvable =) because theree IS no solution =D

1. Qb7 mates in 5--there is no mate in 2

If black cannot castle, then how can moving the rook save him (or her) from mate in two?  The queen moves down to the corner, A1, and then whatever black does, queen to H8, check mate.  I'm not the brightest guy in the world, but by my count, that was two moves.

No, there is a mate in 2.

1. Qa1    Anything

2. Qh8#

By retrograde analysis, black cannot castle.

But the question remains:  Is it a proper assumption that chess problem positions must arise from actual games?  My answer to this is going to wait because I am working on an article about Raymond Smullyan that I hope will appear here soon.

Amy

I am quite sure chess problems do not have to have arisen in actual games.

Also I thought the assumption was castling is legal unless you can prove that it is not (ie King or rook have already moved)