Construct your own retro!

Puzzle: Find a legal chess position in which we can conclude that at least one of the players has castled, but we can't know which.

Position after move 18 from White. I'm not sure, but I don't think you can tell who castled. However, I'm afraid you also can't conclude that one site has actually castled.

Originally posted by crazyblue
Position after move 18 from White. I'm not sure, but I don't think you can tell who castled. However, I'm afraid you also can't conclude that one site has actually castled.

I think I can reach that position after white's 14th move (you can add a few extra ones if you want!), without castling, so I don't think it solves the problem.

1. Nf3 Nf6 2. Ne5 Ne4 3. Ng6 Ng3 4. Nxf8 Nxf1 5. Ng6 Ng3 6. Rf1 Rg8 7. Rg1 Rh8 8. Rh1 Nxh1 9. Nxh8 Ng3 10. Ng6 Ne4 11. Ne5 Nf6 12. Nf3 Kf8 13. Kf1 Kg8 14. Kg1

(you can omit the R moves, but then it's after black's move)

No indeed. For instance: 1.Nf3 Nf6 2.Nh4 Nh5 3.Ng6 Ng3 4.Nxh8 Nxh1 5.Ng6 Ng3 6.Nxf8 Nxf1 7.Ng6 Ng3 8.Nh4 Nh5 9.Nf3 Nf6 10.Kf1 Kf8 11.Kg1 Kg8. No castling needed.

I think they both castled.

But that doesn't solve the puzzle, does it?

Is this even possible? Any pawn moves will result in the possibility of king/rook bypassing each other. Even without pawn moves, unless there is something I am missing (maybe if a restriction on the number of moves is issued) it is either impossible to tell if anyone has castled, or it is clear which side has castled.
So, I guess if there is to be a solution it should impose restriction on the number of moves needed to reach the position.

There is a solution without imposing restriction on the number of moves needed to reach the position.

Originally posted by Mephisto2
[fen]rnbq1rkn/pppppppp/8/8/8/8/PPPPPPPP/RNBQ1RKN w - - 0 11[/fen]

I think they both castled.

But that doesn't solve the puzzle, does it?

I don't know how to post a board but, starting from this position, try removing all 4 pieces on the Q side from both sides.

Originally posted by sugiezd
I don't know how to post a board but, starting from this position, try removing all 4 pieces on the Q side from both sides.

Sorry, that doesn't work either.

I'm pretty sure I just found the solution while thinking about it in the bathroom 😀

Here's it in theory:

- Both sides only make knight moves (and king move/s of course). This is important because knights can't make tempo moves.

- The Bishops and Rooks and Kingside will be captured and the Kings will be on g1/g8

- If both sides castle that will be 2 halfmoves for the king to reach their destination square. If neither of them castle it will be 4 halfmoves. And now comes the keypoint: If only one of them castled it will be 3 halfmoves.

- Now remember that knights cant do tempo moves.

Conclusion: If in a given position you don't know who's turn it is, you won't be able to tell which side castled. You will only be able to tell (from the position of the knights towards each other) that the kings made 3 halfmoves, which means 1 side castled and 1 didn't.

Correct?

Originally posted by crazyblue
I'm pretty sure I just found the solution while thinking about it in the bathroom 😀

Here's it in theory:

- Both sides only make knight moves (and king move/s of course). This is important because knights can't make tempo moves.

- The Bishops and Rooks and Kingside will be captured and the Kings will be on g1/g8

- If both sides castle that wil er) that the kings made 3 halfmoves, which means 1 side castled and 1 didn't.

Correct?

What about rook moves to compensate that?
The more I think about it, the more it seems that this is not a matter of odd/even number of moves since that is easy to get around. I am still clueless as to what else we can use.

Originally posted by ilywrin
What about rook moves to compensate that?
The more I think about it, the more it seems that this is not a matter of odd/even number of moves since that is easy to get around. I am still clueless as to what else we can use.

Yech, forgot about those. If both sides castle, its not possible, but if both sides do not castle, you can even the halfmoves to 4 and that's it for my solution. And even more, once the king is on g1/g8 and rook+bishop are gone, the queen has two squares to odd/even the moves.

One other idea as to what we can use....maybe something with an en passant move? This would mean that white just opened his position 1 move ago (for example with f4 or so), while blacks position has been opened with the e-pawn before. but i really dont see how that would help either.

Or maybe something with promotion (it matters which pieces the pawn has taking on the way to promotion then) and if he promotes somewhere near king it might have prevented castling or so....oh well 🙂

Hm, maybe the f-pawns can be exploited. 3 captures for one side, 2 for the other if they get promoted on the e-file. Can this justify a castling? Need to think some more...

I give up on this one. Whats the solution? You can also pm me if you don't want to tell yet.

Originally posted by crazyblue
I give up on this one. Whats the solution? You can also pm me if you don't want to tell yet.

Seconded.