I once heard that Bobby Fischer pawn promoted to a rook, and then on a subsequent move, castled with that rook. I have been searching for the move history of this game, if it really did happen, but I haven't been able to find it on the Internet. Am I just searching for a rumor, or did this really happen? Where can I find the move history for this game if it really did happen?
5 Answers
There was a time when the FIDE rules didn't specify that a king and rook need to be on the same rank in order to castle. This meant that, assuming the other requirements for castling were met, it was legal for White to castle with a rook on e8 (or Black with a rook on e1), provided that rook had never moved (which could only happen if it was a promoted pawn). This interpretation was never intended, of course, and the wording has since been fixed.
I've never heard of Fischer (or any other player) taking advantage of this glitch in the rules, but it was used in at least one chess problem.
-
2It reminds of a trick I once heard of: In a blitz game, you just play h4, Rh3, Re3 and sacrifice it somewhere … later in the game you castle short … with the rook from a1 on the board to the right of you! ;-) Commented Dec 25, 2014 at 16:53
Not Fischer. But there is the infamous Tim Krabbe vertical castling problem:
http://www.futilitycloset.com/2009/12/11/outside-the-box/
Although it appears this was already illegal as of the problem's publication in 1972.
-
4Tim Krabbé credits the "vertical castling" idea to Max Pam. chess.stackexchange.com/a/18100– Rosie FCommented Apr 19, 2019 at 7:19
By FIDE rule 3.8.b, it would not be valid as the rook would have to move.
-
Someone asks a rules question, and I get a downvote for citing the authoritative rules reference? Commented Dec 26, 2014 at 14:35
-
Countered the downvote - it seemed unfair to downvote an answer that is technically correct & cites the official rules.– user1108Commented Dec 29, 2014 at 11:01
-
I give +1 for this too. Although it does not mention anything about Fischer making such a fantastic castling. I've never heard of such a story. Although if I did, I would most probably believe it. Cheers.– user2001Commented Jan 1, 2015 at 1:42
-
10Rule 3.8.b does not disqualify castling in this situation. If the pawn is promoted to a rook on the appropriate square and that rook has not moved, then 3.8.b is satisfied. The other answers indicate that the rules have been updated to specify that castling must occur on the player's first rank. This is specified in 3.8.a. It would be nice to know when this rule was updated to specify castling must occur on the player's first rank. Commented May 12, 2015 at 15:34
I ran a search on the games in GM Robert Huebner's World Champion Fischer CD looking for a game where one side promoted a pawn to a rook. Chessbase reports 'no games found.' There were 47 games where someone promoted to a queen, but no games where someone promoted to a bishop or knight. This CD has close to 1000 of Fischer's games.
In the 8th game of Mikhail Tal's match with Bent Larsen at Eersel in 1969, Tal as White checked Black's king, sending it to d8. Later in the game he could have won by advancing his a-pawn, but became worried that Black could develop an attack by castling(!), and so focused his play on preventing this from happening. After the game, a fan asked Tal why he hadn't pushed the a-pawn, and it was only then that Tal realized his mistake. You can't castle once your king or rook have moved.
This won't have happened in a tournament because it is against the rules but could have happened in a friendly blitz game. Given Fischer's character I think it's extremely unlikely even in a blitz game. The promoted rook would have had to make its way back to one of the corner squares and I don't think Fischer would be able to even consider castling with it.