Raymond Smullyan retrograde analysis puzzles

For those of you who don't know, Raymond Smullyan wrote some interesting chess puzzles that are not at all like normal puzzles. Instead of looking for the best move, or a mate in a certain number of moves, you are given a chess position and must devine what happened in the PAST. Take a look at the board below:

On what square was the white queen captured? Using nothing but logic and the rules of chess, you can actually figure this out.

Sure, these puzzles won't make you any better at chess, but they sure are fun!

Check out my page at http://www.philipbrocoum.com/?page_id=228 for some more puzzles that I came up with in the same vein as Raymond Smullyan's other puzzles, as well as the solutions.

Philip

The piece captured on c3 is the Black's black-sqaured bishop, and in order for it to get out, White's queen's knight is captured on f6.

Now to have a capture on e6 only the queen is possible since black-sqaured bishops can't go on white squares

I used the same reasoning as xie1995 and came to the same conclusion.

And for the other puzzles on that page:

- White cannot castle. The queen must have been captured on e6 before the c8-bishop was captured on d3. So white must have played f2-f3 and Ke1-f2-g3/Ke3 to play Qd1-e1-f2. The bishops on f1 and c1 were just captured by black knights from e3/g3 and b3/d3.

- If we label the machines A, B and C, then first tell A to use green for B and red for C. Then ask A "Which of the machines is broken?". If it answers B, then choose C. Then C cannot be the broken one, because then A was lying and thus the broken one. Similarly, if it answers C, choose B.

Nice puzzles :)

Phobetor- Your solution to the broken machines problem must be wrong. The machines can only say yes or no. Besides which, your solution would be flawed even if they could give different answers, I think. (I don't need to go into that.)

I changed the solution, since I missed that it doesn't answer all questions but only yes/no questions. Still I'm not sure if it is right. Can you tell the machine which light to use for what? In other words:

--- Can you ask machine A: "Is machine B broken? Use green for yes, red for no."? Or is that not allowed? ---

Not allowed.

Actually, white's queen must have been captured on e6: before the queen and bishop got out, white captured black's bishop on c3, meaning that black captured on f6 before the c3 capture, and therefore it must have been the Nb1 captured on f6. Since white's only other missing piece, the Bc1, could never have been captured on e6 (it's on the other color), it must have been the queen captured on e6.

Interesting. I checked out his site (http://www.philipbrocoum.com/?page_id=228) and in particular this composition:

What piece does not belong on the board?

(In other words, remove one unit to make the position legal).

Am I the only one who does not find this "mind-numbingly difficult to solve"? I got it in a few seconds.

Yes, that is the first key thing to notice. It establishes that the position is illegal. It also gives a strong hint of which type of unit (and of which color) to remove to make it legal.

Yes, but you can't remove the Bc3, because then the White K would be in an impossible to create double check (regard both rooks).

I would remove the Pg2.  This pawn should not be there, because it advanced to g5, captured a Black N on h6 (the only Black piece that is gone) and then promoted to a dark square bishop on h8.  This bishop then went to c3 via g7-h6-e3-d4-c3.  White's original dark square bishop was captured by something.  There is no other way for white to create a dark square bishop by only making one capture.

EDIT: Made text white to remove spoiler

That's what I got. The problem is a bit 'over-composed'. He put Qa3 and Be3 there to stop the removal of Pb2 and Pd2, but they are already stopped because the last move for black was Rh1(or g1)xBf1 (but see edit!), and there is no way a Rook could have gotten in there with Pe2/f2/g2/h2 and Bf1 all at home.

Edit: oh, snap; it could have been Rh1xNf1. Maybe he should have moved a5 to b6 (so that wN had to be captured there) and withdraw the Qa3 and Be3 to use this more subtle means of ruling out -Pb2 or -Pd2

What a necro. I got that same answer too (in 5 minutes, but not "mind-numbingly difficult".) And that IS a nice change to make, BigDoggProblem. Adds a very subtle element of reasoning in.

One thing is indeed 100% certain in Smullyan's retrograde puzzles: all his hypothetical players must be complete patzers to obtain such positions. And it is unlikely for such complete patzers to avoid illegal moves and even set up correctly the starting position sometimes - this leaves retrograde analysis to be pure academic.

Thankfully problems are not limited to what is likely in a competitive game of chess.

Correct.

But puzzles not resembling real game are ugly and worthless.

Well, you're entitled to your ignorance.

I don't think many retrograde puzzles resemble real games.

You call it ignorance, I call it aestheticism.

I call it bizarro-aestheticism.

The definition is:

: devotion to or emphasis on beauty or the cultivation of the arts

If your view had prevailed, hundreds of masterful compositions would have never been made. Your view opposes cultivation of the art, and severely limits the possible beauty.

They're not supposed to.