# How is this position possible in a normal game of chess?

A friend of mine gave me this problem. It goes like this.

In a normal game of chess, the white player checks the black's king with a rook and a bishop as in the following image.

``````8/8/8/8/8/7k/8/3K1B1R b - - 0 1
``````

Now the image seems a bit awkward. In the last move before the check, the white player will have moved either his rook or his bishop (but not both). Suppose that in the last move the white rook is moved, so the black king and white bishop would already be at the same position as they are now. So the black king was already checked by the bishop and hence it was impossible for the black king to remain in that position. Similarly, if the white bishop is moved in the last move, the black king would already be checked by the rook and hence it would be impossible for the king to stay in that position.

But it is assured that this position can be reached without any breach of rules. Does anyone know how this can be done?

• IIRC, this is the very first puzzle in Raymond Smullyan's classic "The Chess Mysteries of Sherlock Holmes." If you like this sort of retrograde analysis (plus some tricks!), I highly highly recommend everything by Smullyan. – Quuxplusone Aug 30 '17 at 20:34
• In the original puzzle, it is recounted that this position was merely observed by Dr. Watson in his club; he didn't see the play that led up to it. It is significant that when Watson drew the position for Holmes, he didn't place the usual "A–H" and "1–8" annotations along the edges of the board (because he didn't know which player was playing White). You didn't need to be given the information that it is currently Black to move; that's deducible from the fact that Black is in check. – Quuxplusone Aug 30 '17 at 20:37
• @Quuxplusone I have read about Smullyan in my Discrete Maths book but I didn't know that he used logic in chess. – Faiq Irfan Aug 31 '17 at 13:00

Double check is only possible by using discovered check. So either the rook check or the bishop check was discovered by moving something in between on the previous move.

I don't see how that's possible with the rook check, but it is possible with the bishop -- if the board is shown with black at the bottom, contrary to what is usually done. Then there could have been a pawn in between the black king and the white bishop, it could have captured something on h8 and promoted to a rook.

E.g.,

``````[FEN "3K1B1q/6P1/7k/8/8/8/8/8 w KQkq - 0 0"]

1.gxh8=R+
``````
• Would there be any advantage to promoting to rook instead of queen? – jpmc26 Aug 31 '17 at 4:02
• @jpmc26: it makes the puzzle a bit harder as people are more likely to think of promotion if it's a queen. – RemcoGerlich Aug 31 '17 at 6:41
• @jpmc26 Maybe the nearest queen was far away, and white just wanted to finish the game quickly and couldn't bother to rise from their seat. – JiK Aug 31 '17 at 8:34
• There are a few (very few) positions where under-promoting to a rook is the best move as a queen allows stalemate. en.wikipedia.org/wiki/Saavedra_position is the most famous. – TheMathemagician Sep 5 '17 at 15:53
• How do you create a chess game like that for other users to see? – Mr Pie Jul 13 '18 at 9:48

In the first position which you've given, there could have been a white pawn on b7 and a black piece on a8 prior to that with white to move. Then by capturing the black piece with the white pawn and promoting it to a rook, the position shown would have been reached, with black to move. The board is typically shown from the white perspective with the pawns moving up the board.