Toldsted skrev:
Ng3 x h1
Board with missing corner rook
Sort:
Can you please explain?
I mean, how do we know that this move has been definitely played? Thank you!
No, it's Nh3 x Rg1. It has something to do with the fact that knights cannot lose tempi (plural for tempo). Parity or something like that. For the position to be "Black to move", White must have changed the tempo once. Knights cannot lose tempi, so the White rook at h1 will have to lose one tempo by moving to g1 to provide the said tempo.
Had the final position been "White to play", the capture was made on h1 (Ng3 x Rh1).
Jun 9, 2019
0
#6
There are some Morphy games with the rook off the board. Why to make it even or fair when playing a lesser opponent. Or just for fun.
Thank you but I don't get this tempo thing 
Also, why does the rook have to be captured by the knight? Aren't there any other possibilities?
thomasgalle wrote:
Thank you but I don't get this tempo thing Also, why does the rook have to be captured by the knight? Aren't there any other possibilities?
The tempo thing is not an easy concept to understand. It requires some counting to realise that the capture must have been made on g1.
The pawns have yet to move. The only piece which can jump past the pawns is the knight, so only the knight is capable of capturing the White rook.
But we cannot make sure ANY previous black move. The puzzle has no solution!
Logic of "tempo" here.
A knight necessarily changes colour each move.
Therefore, a knight needs two, or a larger even number of moves to return to where it was before.
If pawns have not moved then the players can bring out their knights, but bringing knights out AND returning them to starting position would take an even number of moves - for both. After all knights are back, it must still be white to move.
A rook usually can lose tempo. Like, move a1-c1, then c1-b1, then b1-a1 - 3 moves to go back to previous position.
But if pawns have not moved then the rook also cannot lose tempo - bishop cannot get out, and with knight out of the way, the only moves allowed are e. g. h1-g1 and back g1-h1 - still 2 moves to return.
If it is White to move and the rook is gone, then the only way rook could disappear (other than giving rook odds) is for either black knight (but unknown which) to get out, travel across the board and capture the rook. If it is White to move then both must have made an even number of moves. White knights must have left to allow black to move, and departure of g knight may have allowed h rook to move, but in order for it to be white to move, the rooks must have made even number of moves and therefore been captured at h1. Thus we know one black move: Ng3-xh1
If it is black to move then White must have made an odd number of moves. White rook therefore must have made an odd number of moves and ended up captured on g1. We then know one white move: Rh1-g1.
But we do NOT know any black move, as originally asked. Because Nf3-xg1 and Nh3-xg1 are both possible, and we have no way of knowing which. Thus, the puzzle has no solution.
jsaepuru wrote:
But we cannot make sure ANY previous black move. The puzzle has no solution!
Logic of "tempo" here.
A knight necessarily changes colour each move.
Therefore, a knight needs two, or a larger even number of moves to return to where it was before.
If pawns have not moved then the players can bring out their knights, but bringing knights out AND returning them to starting position would take an even number of moves - for both. After all knights are back, it must still be white to move.
A rook usually can lose tempo. Like, move a1-c1, then c1-b1, then b1-a1 - 3 moves to go back to previous position.
But if pawns have not moved then the rook also cannot lose tempo - bishop cannot get out, and with knight out of the way, the only moves allowed are e. g. h1-g1 and back g1-h1 - still 2 moves to return.
If it is White to move and the rook is gone, then the only way rook could disappear (other than giving rook odds) is for either black knight (but unknown which) to get out, travel across the board and capture the rook. If it is White to move then both must have made an even number of moves. White knights must have left to allow black to move, and departure of g knight may have allowed h rook to move, but in order for it to be white to move, the rooks must have made even number of moves and therefore been captured at h1. Thus we know one black move: Ng3-xh1
If it is black to move then White must have made an odd number of moves. White rook therefore must have made an odd number of moves and ended up captured on g1. We then know one white move: Rh1-g1.
But we do NOT know any black move, as originally asked. Because Nf3-xg1 and Nh3-xg1 are both possible, and we have no way of knowing which. Thus, the puzzle has no solution.
Howsoever we try Ng3xh1 case, we cannot get to the final position with Black having the move.
A tempo has to be lost somehow. It does not matter which knight does the capture.
EDIT: I did not read carefully the text that you wrote that the last part. I presume the last move should be called Nxg1. In other words, the last move was "a capture of the White rook on g1 by a Black knight".
EDIT 2: I just realised, the reason why it HAS to be Nh3xg1 is because Nf3+ is check! This means that, because the White king is smothered, a capture of the Black knight must be performed, so a knight at f3 cannot proceed to g1. This was what I had considered while attempting the moves.
jsaepuru wrote:
But we cannot make sure ANY previous black move. The puzzle has no solution!
Logic of "tempo" here.
A knight necessarily changes colour each move.
Therefore, a knight needs two, or a larger even number of moves to return to where it was before.
If pawns have not moved then the players can bring out their knights, but bringing knights out AND returning them to starting position would take an even number of moves - for both. After all knights are back, it must still be white to move.
A rook usually can lose tempo. Like, move a1-c1, then c1-b1, then b1-a1 - 3 moves to go back to previous position.
But if pawns have not moved then the rook also cannot lose tempo - bishop cannot get out, and with knight out of the way, the only moves allowed are e. g. h1-g1 and back g1-h1 - still 2 moves to return.
If it is White to move and the rook is gone, then the only way rook could disappear (other than giving rook odds) is for either black knight (but unknown which) to get out, travel across the board and capture the rook. If it is White to move then both must have made an even number of moves. White knights must have left to allow black to move, and departure of g knight may have allowed h rook to move, but in order for it to be white to move, the rooks must have made even number of moves and therefore been captured at h1. Thus we know one black move: Ng3-xh1
If it is black to move then White must have made an odd number of moves. White rook therefore must have made an odd number of moves and ended up captured on g1. We then know one white move: Rh1-g1.
But we do NOT know any black move, as originally asked. Because Nf3-xg1 and Nh3-xg1 are both possible, and we have no way of knowing which. Thus, the puzzle has no solution.
@MARattigan and @EndgameStudier are two of the players on Chess.com who probably know this tempo and parity concept well, much better than I do,
Hello to everyone! I am new in chess. I was given a puzzle with an (alleged) real game where the board is set up exactly as at the beginning, with the only exception that the white corner rook h1 is missing and it's the blacks' turn to play. The puzzle is asking for one previous move which we know for sure that was played by the blacks. Any clue??