@dwl1945
I believe that is called “Blackburne’s Law.”
Yes and he’s dead now. A cautionary tale.
White to move and win.
Bill Smythe
Can White castle kingside?
The standard rule in chess problems is, I believe, that castling is assumed legal unless it can be proven that it is illegal.
Hint #2: I posted this problem in partial response to a post by Bill Brock earlier in this thread.
Bill Smythe
It would be cool to get to this position via some fireworks (e.g., White’s last move was a capture on f8)
There is also some retrograde analysis potential: could we construct a similar position in which we could prove that White must have moved one of the existing pieces?
Kolty told a tall tale on his old public TV show:
Steinitz (?–the story would be better with Blackburne, but that’s the way I remember it) plays a patzer on a special chess set: each piece is a decanter of booze, with volume proportional to the piece’s value. The game begins something like 1.e4 e5 2.Qh5 Nc6 3.Qxf7+!! (Steinitz is required to drink a tasty little shot) 3…Kxf7 (patzer is required to drink a large amount of brain poison), and the Black king, with no general to offer sensible direction to the army, does not survive for long.
I heard that the story was about Em Lasker while steaming to America.
Another version of the story is told, with comments, at chesscafe.com/text/bruce118.pdf
Hmm, that could present a problem for the usual “chess problem convention” that castling is assumed legal unless proved otherwise.
With just three white pieces (Ra1, Ke1, Rh1), in a position where we could somehow prove that black’s last move was not a capture, we would then know that white’s last move was with one of the three pieces, and hence, castling on at least one side would be illegal now. But we would have no way of knowing which! Conundrum in the world of problem conventions.
Bill Smythe
That would be amusing…we’d probably need many more White pieces. Ideally, the winning move would be either 0-0 or 0-0-0, and we’d be able to construct two different proof games, one in which the a-rook had moved, and one in which the h-rook had moved.
(Duh: only a quantum Black king could be on both the f-file and the d-file, so one of the decisive moves would have to be a non-check.)
I assume by now everybody has figured out that, in the position I diagrammed, O-O is the only winning move.
OK, everybody, here’s your next challenge. Come up with a position, white to move, with white rooks on a1 and h1, white king on e1, where it can be proved that O-O and O-O-O cannot both be legal, yet where either could be.
Bill Smythe
I have a challenge. How about we have a discussion about the actual topic?
IMO the chess problem above is a good explanatory example of #5 in your original post and keeps within the spirit of the thread.
It is indeed axiomatic that rules are made to be broken. (Apologies to all the cliche-haters — please don’t be snarky or snippy).
Apologies for my contribution to thread drift…but it’s an interesting concrete question of zero practical use.
But “rules” have their limits. John Nunn, in Understanding Chess Middlegames:
Rules of thumb are best learned to be forgotten during concrete analysis…
With only the king and both rooks on the board, obviously White had to have moved one of them back to it’s home square.
As we said in math class when the proof became tedious, “trivially obvious to the casual observer”.
However, there might have been a 4th piece which White last moved and Black then captured.
Knights can’t lose tempi, but rooks can, and knights can choose which square they wish to be captured on. So this position can be created with either side to move. In this proof game game, we know that with Black to move, White has lost castling rights (with which rook? or with the king? that we don’t know), but we can’t say anything about Black’s rights.
[Date “2012.08.27”]
[White “Proof Game”]
[Black “Who can castle?”]
- Nf3 Nf6 2. Nc3 Nc6 3. Na4 Na5 4. Nh4 Nh5 5. Ng6 Nb3 6. Nxf8 Nxc1 7. Nb6 Ng3 8. Nxc8 Nxf1 9. Ne6 Ne3 10. Nxd8 Nxd1 11. Nc6 Nb3 12. Nd4 Nxd4 13. Nb6 Nc3 14. Nd5 Nc6 15. Nxc3 Nd4 16. Nb5 Nxb5 17. Rb1 Nd4 18. Rc1 Nb3 19. Rd1 Na1 20. Rxa1 *
“5. Rules about how to think” - one of the most interesting (and most difficult!) tactics books I’ve ever seen is Paata Gaprindashvili’s Imagination in Chess.
Gaprindashvili discusses ways of making your combinations work - essentially “how to revise your idea when your original idea doesn’t work.” One well-know rule of thumb along this line: when a multi-move forcing combination doesn’t work, try inverting the move sequence.
Bill B, you didn’t answer the question of Bill S. He wanted to know if you could construct such a position with white to move, not with black to move.
I don’t think of numbers 3-5 from the original post as being rules at all. They sound more like advice about how to use the rules in 1 & 2.
That’s not to say your breakdown might not serve some purpose… perhaps if you clarified what you’re trying to accomplish with these classifications, it might help guide the conversation.