This line of reasoning has some superficial plausibility, but Gijssen covered it in a recent comment:
[quopte]Regarding the first position, I agree that after 1…Qg3+, White has no possibilities to castle at the moment or in the future. Nevertheless, the Rules Committee proposed and the Presidential Board decided to establish a uniform rule that “a player loses his right to castle with his own move and not by an opponent’s move.”
[/quote]
Anyway, back to the original questions that began this thread –
Everybody seems to agree that, if the right to castle is temporarily lost in one position, and permanently lost in another, then the two positions are not identical (for triple-occurrence purposes). In fact, I don’t think any reasonable person would disagree.
There is, however, disagreement on what constitutes the permanent loss of the right to castle (for triple-occurrence purposes). There are two reasonable possibilities:
A. The right to castle is permanently lost (for triple-occurrence purposes) if either the king or the rook has moved from its original square.
– or –
B. The right to castle is permanently lost (for triple-occurrence purposes) if there does not exist a sequence of legal moves leading to legal castling.
LBLAIR asks the question, and seems (maybe) to favor definition A. GENE_M and TANSTAAFL definitely are in the camp of definition A. SLOAN (and RFEDITOR, until just now) do not weigh in on this issue, but seem to be overlooking something about the second example I presented. LBLAIR finally noticed this something, and observes that it casts his example and my second example as two instances of essentially the same situation.
In supporting definition A, TANSTAAFL points out that rule 8A3, which defines the permanent loss of the right to castle, is essentially the same as definition A.
This is not, however, a strong argument in favor of definition A in the triple-occurrence rule (14C). In fact, 14C never talks about permanent vs temporary loss of the right to castle at all. Rather, it 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”. Here “possible moves” includes “future possible moves”.
Then 14C attempts to clarify itself, but succeeds only in muddying the waters, by adding the phrase “including the right to castle”. In implying that situation X includes situation Y, when in fact some elements of situation Y are outside situation X, this phrase in 14C worsens clarity rather than improving it.
In addition to the above, definition B is more in keeping with the spirit of 14C. If all the pieces are on the same squares and all possible future moves are the same, why shouldn’t the situation be considered the same?
In LBLAIR’s example:
(neither K nor R has moved)
– and in my second example:
(neither K nor R has moved)
– there is no possibility of future castling (everybody please convince yourselves of this in the second example), so any repetition of either position after the king or rook has moved does not reduce future move possibilities. This, in my opinion, makes definition B more reasonable.
Bill Smythe
I’m not going to weigh in on which rule would be “better,” but I think it’s clear from Gijssen’s comment that the FIDE Rules Committee explicitly considered and rejected (a statement equivalent to) your definition B.
In the second example (in Bill Smythe’s previous post), the white pawn COULD move, right? So the right to castle HASN’T been lost by EITHER definition in the original position.
I don’t believe definition B is unambiguously ruled out by the rulebook, but definition A is the way I’ve always heard the rules interpreted. Perhaps we should look at some earlier editions of the rulebook for some additional guidance on the intent of the current rule. (on the theory that IF the rule used to be more clear on this point AND if there’s no clear evidence that we’ve intentionally changed the rule then maybe the older definition is correct…)
Incorrect. (I made the same mistake, at first.) Consider what has to happen after the pawn move.
SLOAN weighed in on this issue years ago - the last time you instigated this very same discussion.
Those who refuse to remember the past are like goldfish - every trip around the bowl is a new and wonderful experience.
Every new trip around can be a new and wonderful experience, if it picks up new and wonderful contributors to an interesting topic.
Bill Smythe
Oh? Then please show me a sequence of legal moves leading to legal castling (definition B).
Good idea.
I just checked editions 4 (Goichberg), 3 (Redman), 2 and 1 (Morrison), as well as the Handbook (Harkness) and its update, the Rulebook (Harkness). (I don’t have the earlier Blue Book (Harkness).) The wording has not changed in all this time, i.e. the current ambiguity has been there since Noah was a boy. Editions 2 and 1 even include a “FIDE interpretation” which also glosses over this same ambiguity.
Apparently, FIDE eventually recognized the ambiguity, and recently decided to resolve it in favor of definition A, but USCF has not followed suit.
I find definition B much more satisfying, and more in accordance with the spirit of the triple-occurrence rule, in that the same situation must recur before a draw claim can be considered valid.
Nevertheless, I fully understand the practical problems associated with definition B. For one thing, definition B would be virtually impossible to implement in a computer program such as Fritz, or in an online chess-playing service such as ICC, because the program would probably have to generate all legal move sequences starting with the position in question.
The same problem could be faced by a human TD, because there could be a “problem-like” position where it is difficult to determine if there is a legal castling sequence. Indeed, my second example comes close to being such a position, as evidenced by the fact that only LBLAIR (so far) has posted a correct answer, and only on his second try, at that.
So, I can readily understand the sentiment in favor of definition A, even though, to us purists, definition B may be preferable.
Bill Smythe
There’s a difference between the right to castle and the ability to castle.
For example, you have the right to remain silent.