# This position is not dead.

Jirakon
Posers and Puzzles 09 Mar '08 06:03
1. 09 Mar '08 06:03
Black

White

It's White's move. Can White castle?
2. 09 Mar '08 06:511 edit
I cannot see what black's last move could have been...
Yes i do actually. It must have been Kf3e3, and the white move before that was Re4e3.
So one of the rooks was promoted.
3. SwissGambit
Caninus Interruptus
09 Mar '08 07:31
Originally posted by Jirakon
Black
[fen]8/8/2Q1N3/3B4/2PP4/3Nk2P/2P5/B2RK2R[/fen]
White

It's White's move. Can White castle?
Since castling is assumed legal unless otherwise provable, it's not necessary to stipulate whether the position is alive or dead. And I'm not sure why Rd1 is needed even with that stip.
4. 09 Mar '08 09:15
The question is: is it provable?
5. 09 Mar '08 09:471 edit
It is easy to find ways to arrive at this position without white having to play with either his king or his h-rook. So, it cannot be proven that castling is illegal, which means it is legal for the purpose of this puzzle.

edit. I assume the rook on d1 is put there just make it a bit harder to find the last few moves (a.o. the need to have a rook promoted)
6. 09 Mar '08 16:581 edit
castling is assumed legal unless otherwise provable

Not in this problem. Either prove that it's legal, or prove that it's not. No assuming.

In fact, I don't even have to tell you that the position is not dead. You should be able to deduce that from the position itself.
7. SwissGambit
Caninus Interruptus
09 Mar '08 17:29
Originally posted by Jirakon
castling is assumed legal unless otherwise provable

Not in this problem. Either prove that it's legal, or prove that it's not. No assuming.

In fact, I don't even have to tell you that the position is not dead. You should be able to deduce that from the position itself.
You do, however, need to stipulate that the FIDE law about "dead" positions will be employed. This is usually indicated by the condition "Dead Reckoning".

Am I right in thinking that Rd1 is unneeded?
8. SwissGambit
Caninus Interruptus
09 Mar '08 17:33
Originally posted by Dejection
The question is: is it provable?
It is, given that the problem is "Dead Reckoning". This indicates the use of a FIDE rule that states that a position with zero possibility of checkmate is immediately drawn.

If White cannot castle, then Black could not have played KxR on the last move, because the position would already have been dead. Therefore, DR proves White can castle.
9. 09 Mar '08 17:34
Am I right in thinking that Rd1 is unneeded?

Nope.

And maybe I could make the problem more interesting like this: Prove that this position is not dead, then prove whether or not White can castle.

And isn't dead reckoning necessarily applied? It is a rule of chess, after all.
10. 09 Mar '08 17:363 edits
You beat my post by less than a minute!

And if the d1 rook weren't there, there would be the possibility of 1 c3 Kxd3, so castling would not be provable.

A.G.Buchanan composed it.
11. 09 Mar '08 17:37
Originally posted by Jirakon
castling is assumed legal unless otherwise provable

Not in this problem. Either prove that it's legal, or prove that it's not. No assuming.

In fact, I don't even have to tell you that the position is not dead. You should be able to deduce that from the position itself.
It is possible to construct games arriving at this position both with white having lost his castling rights and not having lost them. Since that means we can't prove that castling is illegal, the conclusion is that it is legal: YES white can castle.

ps. a rook on d1 or not does not change this, it only makes the retraction a tiny bit more complicated in my opinion
12. 09 Mar '08 17:39
It is possible to construct games arriving at this position both with white having lost his castling rights and not having lost them.

That's not true. The fact is that the position would be illegal if White had lost his castling rights. And the d1 rook is needed for the reason stated above.
13. SwissGambit
Caninus Interruptus
09 Mar '08 17:41
Originally posted by Jirakon
And isn't dead reckoning necessarily applied? It is a rule of chess, after all.
It cannot be applied to chess problems by default, because it would ruin certain genres of problems, like self-stalemates. The current consensus is that the rule only applies if it is expressly stated by the composer.
14. 09 Mar '08 17:491 edit
The current consensus is that the rule only applies if it is expressly stated by the composer.

Okay. I'll remember that next time.
Although it makes more sense to me, instead of never applying unless stated, to always apply it unless stated that it's not applied. But if that's the consensus, I won't argue.
15. 09 Mar '08 18:03
Originally posted by Jirakon
It is possible to construct games arriving at this position both with white having lost his castling rights and not having lost them.

That's not true. The fact is that the position would be illegal if White had lost his castling rights. And the d1 rook is needed for the reason stated above.
Look through these games

a) white retains castling rights:
1. a3 a5 2. d4 Ra6 3. Nf3 d5 4. e4 Nf6 5. e5 Bg4 6. exf6 exf6 7. h3 Bd6 8. Ne5 O-O 9. f3 Re8 10. fxg4 Rc6 11. b3 Rc4 12. bxc4 Na6 13. Bd3 Nb4 14. axb4 Qd7 15. bxa5 Re7 16. Qe2 f5 17. Nxd7 fxg4 18. Nf6+ Kf8 19. Nxh7+ Ke8 20. Nf6+ Kd8 21. Nxd5 b6 22. axb6 f5 23. Qxe7+ Kc8 24. Qxd6 f4 25. Qc6 Kd8 26. Nxf4 Ke7 27. Be4 g6 28. Nxg6+ Kf7 29. bxc7 Kg7 30. Ne5 Kf8 31. Nxg4 Kg7 32. Nf6 Kf7 33. Nd2 Kg7 34. Nb3 Kf7 35. Nc5 Kg7 36. Nd3 Kf7 37. Ng4 Kg7 38. Nf6 Kf7 39. Bd5+ Kg6 40. c8=R Kf5 41. Bb2 Kg5 42. Rd1 Kh4 43. Re8 Kg3 44. e4+
Kxg2 45. Ng5+ Kg3 46. Re4 Kg2 47. Ba1 Kg3 48. Ne6 Kf3 49. Re3+ Kxe3

b) white lost castling rights:
1. a3 a5 2. d4 Ra6 3. Nf3 d5 4. e4 Nf6 5. e5 Bg4 6. exf6 exf6 7. h3 Bd6 8. Ne5 O-O 9. f3 Re8 10. fxg4 Rc6 11. b3 Rc4 12. bxc4 Na6 13. Bd3 Nb4 14. axb4 Qd7 15. bxa5 Re7 16. Qe2 f5 17. Nxd7 fxg4 18. Nf6+ Kf8 19. Nxh7+ Ke8 20. Nf6+ Kd8 21. Nxd5 b6 22. axb6 f5 23. Qxe7+ Kc8 24. Qxd6 f4 25. Qc6 Kd8 26. Nxf4 Ke7 27. Be4 g6 28. Nxg6+ Kf7 29. bxc7 Kg7 30. Ne5 Kf8 31. Nxg4 Kg7 32. Nf6 Kf7 33. Nd2 Kg7 34. Nb3 Kf7 35. Nc5 Kg7 36. Nd3 Kf7 37. Ng4 Kg7 38. Nf6 Kf7 39. Bd5+ Kg6 40. c8=R Kf5 41. Bb2 Kg5 42. Rd1 Kh4 43. Re8 Kg3 44. e4+
Kxg2 45. Ng5+ Kg3 46. Kf1 Kh4 47. Ke2 Kg3 48. Ke1 Kh4 49. Ba1 Kg3 50. Re5 Kh4 51. Ne6 Kg3 52. Re4 Kf3 53. Re3+ Kxe3

Both give (I hope I didn't mistype) the same end position. This means that we cannot prove that castling is impossible. The convention therefor is that castling is allowed.

You could easily construct games like the above without a rook on d1, in fact it is a lot easier, in both cases.