Edit: I did not read the column closely enough. Ken’s discussion of how to claim if a threefold repetition has already happened via an opponent’s move is correct. IM Gurevich’s procedure is correct if the claimant is actually making the move that causes the repetition.
It is simple to state that there is no 10-move limit and you go back to the point of any illegal move or position even if that means being on move 65 and backing up to move one to correct the starting position.
I will push back against this argument. First, it is almost always the case that when the game starts with the pieces incorrectly, it is because one or both players have the king and queen swapped. This even happens (very infrequently) with higher rated players, as IM Gurevich stated in the article. It’s the easiest mistake to make when setting up the pieces because the rooks, knights, and bishops are placed on the board with left/right symmetry, but there is only one king and one queen. (And, almost inevitably, the problem is “solved” by the player just swapping the king and the queen and continuing the game.) However, when this happens with experienced players, it is highly unlikely that the mistake will not be noticed within the first few moves of the game.
With all due respect, the other hypothetical situations (three rooks and one knight, nine pawns) are simply too ridiculous to merit any serious consideration. If any of those ridiculous situations were to affect castling, I would be fine with leaving the situation to the TD’s discretion (applying the sound judgment and absolute objectivity called for in rule 1A).
Second, the description of castling in both the Official Rules of Chess and the FIDE Laws of Chess is written in a manner that works just fine if the king starts the game on the d1 or d8 square. (Side rant: Why in the name of Caissa is it necessary for the Official Rules of Chess to have its own text for how the pieces move instead of using the text from the FIDE Laws of Chess?) Looking simply at the text defining the castling move, there is no a priori reason castling is disallowed. While I am among the loudest proponents of reducing the length of the rulebook (having described the text as “blathering” for over 100 pages), I am even more opposed to having ambiguity in the rules.
Third, it requires very little text to resolve this ambiguity. In the FIDE Laws of Chess Appendix A (Rapidplay rules, analogous to US Chess “quick” rules), this text appears:
Not only does this address the case of the game starting with the king and queen swapped, it addresses all the ridiculous hypothetical situations offered as reasons the rulebook would bloat to 2000 pages if we tried to address all situations. (This text is in the rapidplay rules instead of the main rules because the main rules do not limit correction of illegal moves (including incorrect initial position) to ten moves. Interestingly, the draft of the revised Laws of Chess to be considered in Baku introduce a ten move limit for correcting illegal moves but do not lift this text into the main rules, creating ambiguity.)
I think it is reasonable to spend two sentences to eliminate ambiguity. I’m confident we can easily squeeze two sentences out of the blather in exchange for this. 
In fact, we could partially pay for this by eliminating blitz rule 3a:
(Yes, that’s another unnecessary difference from FIDE rules.) Or, if is decided that castling in this case should be allowed, we can just move this rule to the main rules, keeping the word count the same.
You’re making the assumption that each player set up their own pieces. If sets are provided, players from the previous game may have set the board back up incorrectly. It’s also possible that black arrived early and set up both sides and reversed white’s king and queen. Should white be penalized for failing to notice black’s mistake? One could argue that he made a mistake in not noticing. Personally, I would argue both players share some responsibility because each failed to make sure the board was set up correctly before the game started.
Whichever way is better (allowing castling or not), I am in 100% agreement with Ken that it shouldn’t be ambiguous.
Considering this magazine is geared towards kids, I can’t help but feel like the errors were very minor and that this is shooting a fly with a cannon. I understand where rules purists may differ, but consider the audience here.
As far as a target audience goes are you talking about parents who will use anything they can to get a ruling that will go in their kid’s favor regardless of what the rules actually say?
I agree that the above version clarifies the situation admirably. Unfortunately (as you pointed out), within FIDE this is only a Rapidplay rule, even though FIDE is now adopting a ten-move rule for all of its events. This rule should be moved into the general rules.
Even then, the rule may not avoid all ambiguity, at least not in a way that would make sense in all situations. What if, for example, the pieces are set up incorrectly, then on move 6 one of the players castles (illegally, i.e. short on the queenside or long on the kingside), and neither player notices until move 12? Here the illegal castling was caught within 10 moves, but the initial incorrect setup was not. So apparently the game would then revert to move 6, with castling disallowed, but the initial K-Q reversal would not be corrected.
It’s this kind of situation that makes me uncomfortable with any attempt to create a one-size-fits-all rule to deal with illegal moves. Sooner or later, a position is going to arise where the solution prescribed by the rule just doesn’t seem right.
This rule dates way back to the WBCA blitz rules from 1980 or thereabouts. The only difference is, the WBCA version had an exclamation point at the end! Somehow, this rule eventually made its way into the USCF blitz rules. I agree, let’s get rid of it.
Well, maybe because one of those FIDE rules defines the knight move as being to one of the closest squares that is not on the same rank, file, or diagonal as the departure square. What person, who does not already know how the knight moves, would be able to decipher that wording correctly?
Bill Smythe
I’m sorry, Mike, but this Pandora’s Box was opened by IM Gurevich’s article, and you cannot now put the lid back on. An article by a titled player in an official US Chess publication is going to be taken as authoritative by the tens of thousands of kids and their parents who read Chess Life for Kids.
The article was well intentioned; Daniel was simply trying to write an interesting column showcasing some of the stranger rules of our game. The problem is that when you discuss rules that are complex (how to claim three fold repetition of position) or unsettled (castling allowed if King and Queen reversed?) and you are not yourself a strong TD you risk treading on quicksand. The problems and ambiguities in this article need to be addressed soon. If not they will likely create further problems down the road when people rely on this misinformation.
I applaud Mr. Ballou for taking the initiative to write his letter to the editor, and I hope both that its text will be included in the next edition of Chess Life for Kids, and that his offer to have the Rules Committee review for accuracy any rules related content in a proposed article of a US Chess publication will be accepted.
Actually, the fact that the magazine is geared toward kids is, in my opinion, the precise reason the errors need to be corrected. Many of the readers of Chess Life for Kids have little or no tournament experience, and are therefore most likely to be influenced in their understanding of the rules by this article. And some of the scenarios described in the article are more likely to occur in scholastic tournaments than in adult tournaments. Kids are probably also more likely to assume that anything they read about USCF rules in a USCF magazine must be correct.
Bob
Having a king and queen reversed at the beginning of a game is guaranteed to change “the normal conduct of the game.” Allowing castling in this situation doesn’t significantly add to the change. In fact, it could be argued that making castling illegal for one player changes it even more.
True. This only happens when more than 10 moves have elapsed since the error was made.
Bob
I think you’ve been overly hard on Daniel, for while he was wrong about the rule, he was correct about the possible consequences of using it. And there is a weirdness about the USCF rules in this case, even though it is not exactly as he explained it: The USCF rules allow both players to stop recording when only one is under time pressure. But it makes no sense for the player who is not under time pressure to take advantage of this in a non-sudden death time control, since (a) he has no need to try to save time (unlike his opponent), and (b) failing to record is likely to leave him ineligible to make a time forfeit claim against his opponent at the precise time when he is likely to want to make such a claim.
You failed to mention what is, in my opinion, his most serious error: A player does not lose his right to claim a triple occurrence of position draw simply because he moves his piece before making the claim (as Rule 14C4 clearly states). The only way he can lose the right to make the claim is if, after moving his piece, he presses his clock button, thereby completing his move (since only a player who is on move may make the claim).
Bob
I have wondered this myself! And as the rules are currently written, I would have to think that castling is still permitted precisely because it does not say it isn’t permitted.
In Fischer Random, castling is still possible despite the odd piece placement at the beginning of a game. I’ve thought it might be funny to collude with a friend and play a fischer random game in a regular rated event- as long as we blitz out our first ten moves the TD wouldn’t be able to do anything and would have to submit it for rating.
Of course, this could easily be addressed with a clarification of the rule or adopting the FIDE laws as a catchall for instances that fall outside the scope of those things expressly enumerated in the USCF rules.
I might add that the USCF Rulebook could benefit from more “Catchall” clauses which are common in legal statutes to encompass unforeseeable situations that the legislature would like to provide courts guidance on - or at least a hook on which a judge can hang his hat when making a ruling. As a TD, I can’t tell you how many times I wish there were a catchall clause at the end of certain sections in the Rulebook.