In the following position, neither the white king nor the white rook have moved. The game now continues 4.Rg1 Rb8 5.Rh1 Ra8 6.Rg1 Rb8 7.Rh1 Ra8.
Has there been a triple occurrence? Discuss.
In the following position, neither the white king nor the white rook have moved. The game now continues 1.Rg1 Rf7 2.Rh1 Rf8 3.Rg1 Rf7 4.Rh1 Rf8.
Has there been a triple occurrence? Discuss.
Bill Smythe
No and No. These are old, moldy chestnuts.
I always thought the interesting case was something like this:
In the following position, neither the white king nor the white rook have moved. The game now continues 1.Kf1 Bc5 2.Ke1 Bb4+ 3.Kf1 Bc5 4.Ke1 Bb4+.
Could it be argued that the castling right was already effectively gone before the first Kf1 ?
.
No, it cannot be successfully argued. Geurt Gijssen has clarified this point a few times (on ChessCafe.com).
.
That’s what I thought, but I think it is a fun case to think about. Was Geurt Gijssen writing about USCF rules?
Is it true that, at one time, chess rules inadvertantly allowed one, in certain very peculiar situations, to castle vertically? Or is that a sort of chess urban legend?
I believe Bill’s point in his second example was that castling is impossible in the original position because of the Rook on f8, and is impossible on the third occurrence because the Rook has moved. However, I am reasonably confident that this is not a correct reading of the rules. The rules are quite clear that the possible moves of the pieces must all be the same, and that the (permanent) loss of the right to castle produces a different position.
I think that is what makes the case in 85549 fun. The possilbe moves are the same at the end and before the first Kf1. One could plausibly argue that, before the first Kf1, the right to castle is already lost forever. I know, I know. This argument doesn’t really work, but it sounds so plausible.
.
Geurt writes about FIDE rule only.
Kinda makes me wonder why Tim Just does not have a bi-monthly column in paper Chess Life similar to Geurt’s.
Yes, in the past the rules were worded sloppily, so that after e7e8/R White could supposedly castle with the Re8.
.
Some of you are overlooking something.
First of all, in Louis Blair’s example, the right to castle is already permanently lost even though neither the king nor the rook has moved. Therefore, when this position comes up twice more, there has been a triple occurrence.
In my first example, the right to castle is only temporarily lost, but after the third (and second) occurrence, the right is permanently lost, so there has not been a triple occurrence.
But please take another look at my second example:
(White to move. Neither the king nor the rook has previously moved.)
In this position, is the right to castle temporarily lost or permanently lost?
Bill Smythe
It seems to me that #85674 is essentially the same sort of situation as my #85549. I have to admit that I did not notice this when I first looked at the #85674 position.
Well … tongue in cheek … yes, you can castle vertically and neither the King nor the Rook have moved previously to castling.
Actually, there was chess problem in, I believe, one of the old CL&R magazines that had this as the solution. 
P.S. This WAS legal under Fide rules prior to 1972.
I’m sorry, but I must disagree. The “permanent loss of castling” is FORCED from that position, but it hasn’t OCCURED yet.
That was a humor piece. The position involved a pawn on e7 and a king on e1. White first plays e7-e8=R, then, on the next move, “castles” by moving the king to e3 and the rook to e2. The author dubbed this move O-O-O-O-O. The argument was that, since the king has never moved, and the rook has never moved (well, at least not since being a rook), it should be legal to castle by moving the king two squares toward the rook, and jumping the rook over.
That claim is quite a stretch, to say the least. It’s just part of the humor. For one thing, I’m pretty sure the FIDE rule said something about moving the king along the rank toward the rook. (But if FIDE decided to clarify the rule as a result of the article, more power to them.)
Bill Smythe
I must disagree with your disagreement.
USCF rule 14C (triple occurrence) says “… the position is considered the same if pieces of the same kind and color occupy the same squares and if the possible moves of all the pieces are the same …” (emphasis mine).
In the Louis Blair position, the possible moves (including future moves) of all the pieces are the same.
Bill Smythe
Perhaps I was seeing an incorrect distinction between the “immediate right to castle” and the “right to castle”. 14C says “including the right to castle”. Doesn’t a difference in having permanently lost the right to castle mean that there is difference in the position? The permanent right to castle hasn’t been lost until AFTER the diagramed position (see 8A3 where the permanent loss of the right to castle is defined).
Humor yes … and that’s why I said … tongue in cheek.
And yes, prior to 1972, there was no distinction between rank OR file for castling. Whether it was due to this little fantasy idea, I have no idea, just that the rule was not specific until then. Perhaps it should be a little clearer in saying that 0-0-0-0 was NOT illegal until then. 
I’d like to see the exact wording of the pre-1972 FIDE rule. If anybody has such, please post it here, verbatim, so that we can see what we’re talking about.
Bill Smythe
You know, I would also like to see it. I tried to find information about the exact wording of the rules prior to 1972 and have gotten nowhere. Couldn’t even find a used copy of the FIDE handbook from 30+ years ago.
But you made me do some more digging … actually, quite a lot. And now more than ever I would like to read the FIDE rules back as far as 1931. Pieter de Groot states clearly that it’s false. That even in the 1931 rules, it was clear that it must be on the same rank in order to castle.
xs4all.nl/~pijpersh/index.ht … er_71.html
But that’s what ‘urban legends’ are all about.
It goes this way:
Max Pam “discovered” a loop hole in the FIDE rules. Tim Krabbé composed a chess problem with this loop hole. Called it 0-0-0-0. The composition was published in 1972. Right after that, FIDE revised the rule to include the rank requirement.
I dug all around Krabbé’s website. (He has some really really interesting stuff.) But nowhere could I find out anything about this composition.
Perhaps one place to look would be The Complete Chessplayer by Fred Reinfeld. If I remember correctly, it was written in the fifties and contained a copy of the FIDE rules.
I own Krabbe’s book Chess Curiosities. He discusses the castling rules situation on p.8: (Referring to the “move” 0-0-0-0 from the problem, which may be viewed here) “… a new move apparently in accordance with the FIDE rules (art. 6.1): ‘The king is transferred from its original square, two squares toward the rook (here e3); then that rook toward which the king has moved is transferred over the king to the square immediately adjacent to the king (here e2).’”
He goes on to state that all the other stipulations are met and that the FIDE rules nowadays do not speak of ‘a rook’ but of ‘either rook on the same rank’.