Three-fold repetition and en-passant

Btw, in case of „castling“ the rules speak of „rights“: a king or rook had castling rights

So, even if you never be able to castle (see position with that Pe2 in front of the King which can never be removed without giving up castling rights) the theoretical right to castle counts.

Question: 3x repetition?

Same position, White cannot castle in all three positions. In 1&2 he has the theoretical right to...

1. e4 c6 2. Nf3 Qa5 3. Bb5 Qxb5 4. Nc3 Qa6 5. Nb1 Qb5 6. Rg1 Qa6 7. Rh1 Qb5

 

The rule as written, at least as presented here, doesn't talk about rights. It talks about moves. Those are two different things.

So, in #43 either player can claim 3x times repetition successfully?

The rook had moved, in all three positions White couldn’t castle. 2x temporarily, 1x forever.

 

 

With the rule as written, yes, I believe both players could have claimed 3 fold repetition.

And for the record, although I believe the rule as written should be interpreted as I stated, I think the rule should be clarified to make absolutely clear what was intended.

Most arbiters I know would decide against. The possibilities to move do not affect only the very next move. 

FIDE talks of „right“ explicitly.

I sent an email to FIDE with this exact question, and they responded that in their opinion (yeah, opinion, go figure) the rule should be interpreted like it was earlier in this thread. In other words, the rule only looks if right-to-castle or right-to-en-passant-capture exists, not whether the move can actually be done or whether it's impossible because of something else.

I suppose that if FIDE officials interpret their own rules like that, that's the correct way of interpreting it. At least in FIDE tournaments.

OK, so I went looking for the full rule in the FIDE handbook. The FULL rule is below. I agree that the FULL rule (as opposed to what was presented earlier) is now clear, and that rights to castle or en-passant do matter regardless of the actual ability to make the move. Thus the sequence BL4D3RUNN3R presented does NOT qualify.

I think that the FIDE rules are slightly ambiguous in this regard. When they say "a pawn could have been captured en passant" it's not 100% clear whether it's talking about an opposing pawn having just moved two squares to its side, or whether it's talking about actually being able to do the capture. But their own interpretation is the former. It doesn't matter whether the capture can actually be made or not, it still makes the position different.

The FIDE rule is clear in specifying how to deal with the special situations of e.p. and castling. It's pretty obvious when you remember that the point of the 3-fold rule is to avoid endless repetition, so in deciding whether a position is actually repeated, you should consider whether there's any real, practical difference between the positions.

Note that the rule refers to castling rights but not e.p. rights for a good reason. Castling rights are preserved or lost regardless of the legality of the castling move in any particular situation. Suppose the game starts with 1.h3 Nf6 2.Rh2 Ng8 3.Rh1. Because 0-0 right is lost, there's a real, practical difference between the positions and there's NO repetition, regardless of the fact that 0-0 is illegal in both positions (thanks to the pieces on f1/g1).

In e.p., however, there's no such thing as preserving the "rights" to make the capture because the move has to be made immediately. Therefore unlike castling, there's no distinction between the "rights" to make an e.p. capture and its legality, and it's just confusing matters to speak of e.p. "rights" as if they were analogous to castling rights when they are not. An e.p. capture is simply legal/possible or illegal/impossible. That means in the second diagram of the OP, when the potential e.p. capturer is pinned, e.p. is illegal at first and illegal later and there's no real, practical difference between the positions, i.e. the position IS repeated for the 3-fold rule.

The #12 position posted by BL4D3RUNN3R is interesting - in such a situation there's no practical difference between the positions because White will never be able to castling anyway, with or without castling rights. However, it seems that the FIDE rule says "The castling rights are lost only after the king or rook is moved" specifically to deal with such a weird situation. Regardless of how 0-0 is not possible in the future, castling rights are lost ONLY after the K or R has moved. That means a real change has occurred if a K or R moves and returns to its original square, and the position is NOT regarded as repeated.

I read one time a prominent Grandmaster - I believe was Korchnoi, but not positive on that - claimed a draw by 3-time repetition (unsuccessfully) , unaware that according to the rules, the position is considered different, if one of the players was allowed to castle in the first incident, but not the remaining.

He must have been embarrassed.

By the way, there's a variant to the en-passant situation I described in my original post, which does not involve pinning the pawn:

 

I could ask someone to explain this for me, but I think I'd rather give up chess... It makes my head hurt.😵

DjVortex wrote: In the context of the three-fold repetition, a position is considered the same, and I quote, "if and only if the same player has the move, pieces of the same kind and colour occupy the same squares and the possible moves of all the pieces of both players are the same."This means that, for instance, possibility of en-passant capture vs. no such possibility makes the positions different, and thus wouldn't count against the 3-fold repetition rule. Example:But what happens if the en-passant is actually not a legal move because the king is in check, or would be put in check if the capture were done? Is the possibility-of-en-passant still considered as making the position different?

The FIDE rules are ambiguous as the scenarios presented in this thread demonstrate.

FIDE state in the main rule 9.2.2 that positions are only the same if "....the possible moves of all the pieces....are the same" If they had left the rule there, I suggest there would be no ambiguity. An arbiter would simply look to see if an ep. capture/castling  was possible in the 3 repeated positions. If the answer was no in all three positions he could rule 3-fold rep had occurred.

But FIDE goes on to try and clarify in 9.2.2.1 and 9.2.2.2 :

9.2.2.1 says positions are not the same if at the start of the sequence a pawn "could have been captured ep." This adds no clarification at all. What is the difference between a "possible" move (9.2.2) and "could have been captured" (9.2.2.1). None as far as I can see.

9.2.2.2 introduces the concept of castling "rights".  It specifically states that these rights are only lost after a K or R has actually moved. This does add clarification to the main rule as it differentiates between positions where the King is unable to castle on a specific board position, and unable to castle for the rest of the game (at least with a specific R if the latter had moved).

There is nothing to imply that ep. "rights" should be treated in the same fashion, although DjVortex's reply from FIDE to the question suggests that they think it does !

 

I think FIDE should define the concept of "right to en-passant capture" more formally, perhaps in a manner that makes it independent from whether that capture can actually be done or not (ie. if the move cannot be done because of something else, usually because of the pawn being hard-pinned, or the player being in check). This "right to en-passant capture" could simply be a kind of "state" in which the game is at that specific position (which happens when a pawn has moved two squares, to the side of an opposing pawn) that doesn't care about whether the subsequent capture can be legally made or not because of something else. They could then specify more unambiguously the concept of "repeated position" and the three-fold repetition rule in this respect.

That would just needlessly complicate a clear and unambiguous rule. Once again, the e.p. part of the rule says plainly, "Thus positions are not the same if: 9.2.2.1 at the start of the sequence a pawn could have been captured en passant" The distinction is between CAN and CANNOT, which you can phrase in other ways, e.g. possible vs impossible, legal vs illegal, playable vs unplayable, has rights vs has no rights. A pawn that was immobilised by pin COULD NOT HAVE made an e.p. capture, therefore the positions are in fact the same.

Unlike the castling part of the rule, for en passant there's no distinction between "rights" and "playable". If the FIDE representative you contacted actually agreed with you and made this silly distinction for e.p. that doesn't exist in the FIDE rule book, I'd be very surprised. Can you copy & paste their email that supposedly refers to the e.p. cases where the capture is unplayable (due to pin or check) and yet the "rights" bizarrely exist?

 

If as you claim 9.2.2 is unambiguous and 9.2.2.1 just says the same thing as 9.2.2, how does saying the same unambiguous thing twice make it ambiguous?

 

Easy......

9.2.2 is unambiguous insofar as it uses the words "....the possible moves of all the pieces....are the same"

The rules then attempt to clarify by saying :

"Thus positions are not the same if:

9.2.2.1 "at the start of the sequence a pawn could have been captured en passant"

9.2.2.2 "a king had castling rights with a rook that has not been moved, but forfeited these after moving. The castling rights are lost only after the king or rook is moved."

In attempting to clarify the ep. rule they use a different term ("could have been captured" instead of "possible move"). While I cannot see any difference, the fact that they have attempted to clarify and used a different term adds ambiguity, not clarity.

In clarifying the castling rule, they again introduce a new concept, "castling rights" which is not mentioned in the main rule. While the definition is clear, the fact that it is not mentioned in the main rule again adds ambiguity.

 

The whole section 9.2.2 through 9.2.2.2 is poorly written as a whole.