Does anyone know of a game that was drawn under the mandatory 75 move rule?

Right! But read post #182!

hmm thats wierd

Yes , I discussed that earlier. The officials may overrule the actual game outcome. In China they would be sent to a "re-education" camp, in the west they wander free. But actually, chances are that FIDE permits these rule violations as part of the freedom for organizations to determine their own competition rules. Chess sites, inclusing chess.com, have been breaking the rules for years with regard to dead positions. This is to some degree excusable since detecting dead positions can be highly complicated. But seeing a fiive-fold repetition is very simple for an interface, be it rapid, blitz or bullet - so missing it should never happen.

@danielbaechli

I agree with your preferences.

There must have been many real chess games where a player won a KBNK endgame for example after his opponent failed to claim under the fifty move rule. (You don't see many recorded games where that endgame is drawn under the rule and play hasn't actually continued continued well beyond fifty moves.)

The triple repetition rule even more so, because the repetitions can be separated by many moves and difficult to keep track of.

The automatic draw action in these cases is performing a function that is the responsibility of the player in real chess, hence rendering unfair assistance.

I think most chess programs do the same thing.

I don't think it's deliberate policy; just lazy programming. (The XBoard protocol supports draw claims (almost).)   

@danielbaechlie Martin Stahl's replies to your queries are simply ridiculous. What he says is: "We don't listen to FIDE, we listen to USCF" but of course he won't admit to that openly.

Also, do not confuse discussions on "insufficient material" with those on 75M/50M/3R/5R. The latter four are clearcut and very simple for engines to implement and execute. The "insufficient material" issue is difficult as it replaces "dead positions" which are indeed difficult to evaluate. FIDE should launch a project to develop a fast algorithm which captures at least 99.9999% of the death cases and web-sites should implement it. The remaining few (still many) are mostly in the professional composition domain and could be ignored in game play. 

this is true

I don't agree.

The term "legal move" is defined in the rules and can be a move by either colour. The term series is not defined, therefore must be taken in its normal English sense. A series of legal moves is hence just one defined legal move, as defined, after another (necessarily a finite number of times because no position is defined after an infinite sequence).

So from the following White to play position

 the series e8=N, Nc7+ (both with White pieces) would be a series of legal moves.

The point is that the players cannot play this sequence under the rules in force for the game (whether basic or competition), because of art 1.2 "The player with the light-coloured pieces (White) makes the first move, then the players move alternately, with the player with the dark-coloured pieces (Black) making the next move" and art 4 "The act of moving the pieces" which latter attempts, at least, to prohibit a player making a move with an opponent's piece. That the players can play such a sequence is therefore an essential inclusion.

Without it I think the definition doesn't work anyway. For example from this White to play position

Nc7+ is a sequence of (one) legal moves. But is it checkmate?

It's checkmate if White has achieved the objective in art 1.4 "The objective of each player is to place the opponent’s king ‘under attack’ in such a way that the opponent has no legal move. ... The player who achieves this goal is said to have ‘checkmated’ the opponent’s king and to have won the game."

The phrase "has no legal move" here must be taken to mean, "has the move", meaning his opponent has made his move  (which used to be a definition) but can play  no legal move (so he doesn't have for example the legal move Ne8 with the White knight after Nc7+). After a sequence of moves that are not assumed to be made by the players, nobody actually has the move.

The rule 5.2.2 without reference to a played sequence (i.e. ignoring any rules not directly or implicitly referred to in 5.2.2, which only apply to the rule as it stands via the phrase "... player can checkmate" implying can checkmate under (all) the rules of the game, would mean something completely different.  But then without including some phrase such as "can checkmate ignoring rule 5.2.2" (which refers to the label for the rule rather than it's content) the rule as it stands becomes self referent.

Yes, it's about "legal moves". There are 2 articles in the laws (don't have them at hand but you know them) which clarify that a move is only legal in a context. Many moves are not legal when the king is in check and others not when e.g. a king or rook moved or a pawn did not move. So legality is not a private property of a move, it concerns legality in the context of the game. Which implies by concept that you cannot ratify 2 legal white moves in a row or identify a legal move outside the game.

The move you describe I would describe as a well-formed move as is common in math. It looks OK but to evaluate legality you should look at the overall board position and the game state.

Unfortunately FIDE didn't make an effort to define a well-formed move or just a common chess move outside the legality range. I certainly agree there is a function for 2 types of "chess move qualities".

@Arisktotle

... For instance, you agree that you can't castle when you moved your rook, but you are OK with the idea of castling after a wild article 5.2.2 series of 6 white moves leading to a position which could never be reached with castling right from the game start in a legal game. Do you really believe FIDE intended that?

--- No, I think that's why it's worded, "... neither player can checkmate ... with any series of legal moves" as opposed to your formulation. I think a related rule did appear along your lines in one set of rules but reverted back, probably for that reason. 

Note that FIDE has no command over the games or the rules after a game is terminated (it may still have command over the tournament rules). Termination matters because absolutely nothing in the game rules applies outside it's domain, including article 5.2.2. That is a meta-rule which applies to all systems and no system can modify it.

--- We appear to be almost in agreement. Rules that say the game immediately terminates imply that any subsequent actions are not part of the game. The FIDE definition of legal move remains the same for the duration of the laws that are published, however, irrespective of whether anyone is actually playing or not. It's simply a definition.

Yes, you take the rules literally and I don't. Simply because I don't believe that FIDE paid attention to all the details. For instance, you agree that one can castle when one has not moved his rook or king. Next you are OK with the idea of castling after a wild article 5.2.2 series of 6 white moves leading to a position which could never be reached with castling right from the game start in a legal game. Do you really believe FIDE intended that?

(edit: I think you agree with this) Note that FIDE has no command over the games or the rules after a game is terminated (it may still have command over the tournament rules). Termination matters because absolutely nothing in the game rules applies outside its domain, including article 5.2.2. That is a meta-rule which applies to all systems and no system can modify it.

Interestingly, all these issues come back on the composition level. Without a concept of what a "legal move" is, it is impossible to construct appropriate composition rules. Since the whole past is uncertain, many legal moves are uncertain and there is no way you can resolve that with the FIDE handbook without a conceptual understanding of its terminology.

(edit: you provided a great example yourself!) Note: If you permit black and white to play one-colour move series then you might miss dead draws. Many draws depend on "who is on move" and some on the "the rhytm of alternating black and white moves" and have different outcomes when just playing on. That is not acceptable. But in my opinion FIDE never intended that whatever it says in the rules.

I just looked at your diagrams and the text going along with it. If I get it right you started with taking the rules literally ("series of moves") and end up with problems you attempt to solve by interpreting rues about checkmate and other things.

My conclusion is different. You started on the wrong foot by assuming that the moves in a "series of legal moves" were any different from the series of moves in a game and you therefore ended in an absurd place. I am quite willing to bet all my possessions that FIDE never intended one-colour move series in 5.2.2. It is just anti-chess.

How the heck did you get your reply in before my comment??? My comment was poorly formulated and I edited it but you have the oldest version here - probably before I wrote it

What I tried to say is what appeared in different wordings in other comments as well which is "I don't think the move series should end up in places where you can only arrive (without 5.2.2) when you play these one-colour move series!". I emphasized that in my previous comment!

The FIDE rules don't say that you can play one-colour move series. Article 1.2 and article 4 are there to ensure legal moves are not played out of turn. Rules that refer to possible continuations are phrased, "the player can checkmate ..." etc. with the assumption that those rules are followed.

It's only if you talk about disembodied series of moves that such considerations come into play. Hence my objection to your wording.

This is slightly different from my disagreement with @Numquam in that he doesn't assume those rules are followed in possible continuations (at least for 5.2.2) but will only allow rules that are explicitly or implicitly (i.e. definition of terms) referred to in the rule being applied and, in the case of 5.2.2, this doesn't include the rules I referred to. In my view this doesn't leave a playable game by which the players could continue, so, "neither player can possibly checkmate the opponent’s king", would always be true and on Numquam's view I believe all positions would be dead under the rule.  

OK, but a definition for what? My opinion is that the function of the definition is to describe what you can play in a particular game position (context) and I don't see many places which put that in doubt. In fact, article 5.2.2. is the only one that is clear to me and I contribute that to poor formulation. And then I do not even place doubt on the "series of legal moves" which is OK to me.

You don't catch me napping!

I didn't intend to use "play" here in the game sense but in the analytical sense! To demonstrate the result you will have to execute these moves in some way. Note that in my theory I agree that the moves in a "death analysis" have a different status from the ones you play in a game. I call them "analytical" or "imaginary". The difference between us is that I see nothing special in the "series of legal moves" from the move series in a game. The other difference is that I associate legality with playability and therefore reject the legal moves in an analytical series as a FIDE misstep. For me a move is either "legal" or "analytical", for you a move can be both "analytical" (not playable) and "legal".

The practical problem is this. In communication with chess players and problemists, they always appear to assume that a legal move is a playable move. This also shows up in articles and comments on diagrams with castling or e.p. rights: "It is legal" to castle, or "not legal" to play e.p.. Nobody ever uses the term "playable" because it is not part of our chess vocabulary (except with a different meaning). By limiting legality to playability, you capture at least the primary function of legality which is so vital in communication.

Arts 3.1.2-3.9.2 constitute definitions of piece moves mostly independent of context. Arts. 3.7.4.1-3.7.4.2 restrict the e.p move to particular contexts, so the e.p. move is not defined outside of those. Similarly 3.8.2.1-3.8.2.2.2 restrict the castling move to particular contexts. 3.1.2-3.1.3 and 3.9.1 are definitions not of moves but of terms used in other rules. Arts 3.1, 3.1.1 and 3.9.2 are rules of play rather than definitions and 3.10.1 and 3.10.2 define legal and illegal moves (as moves conforming with 3.1-3.9 or not).

Section 3 doesn't actually say what you can play in a particular game position. That is determined by the rules of play in art. 4 The act of moving the pieces. In addition art. 1.2 is a rule of play that says somebody must move something (only applicable before the game is terminated of course).

Arts. 1.2 and 3 make no mention of the colour of the piece moved, but art 1.2 specifies the player to move and art. 4 the colour of the piece he must move. It is the combination of arts.1.2, 3 and 4 that tells you what you can play in a particular position  Art 3. by itself doesn't say you may or must play any of the moves defined.

I asume you meant to say 5.2.2 is the only one that is unclear to you. I don't find it unclear apart from the self reference. What is it you don't like.

As I said before a reference to a series of legal moves without specifying in some way that they are played  under some meaningful set of rules, while not being unclear to me, would clearly allow wild sequences of moves by the same coloured pieces, because 3.10 which defines legal moves doesn't restrict the moves to be any particular colour.

@Arisktotle

"The difference between us is that I see nothing special in the "series of legal moves" from the move series in a game."

The difference I see is that in a game arts 1.2 and 4 are in effect. It is clear what is meant by the result of a series of legal moves. But according to the FIDE laws such a sequence need not alternate the colour of the pieces moved and there is no player associated with the moves in the sequence unless you explicitly specify it in some way.

I don't think an argument on the grounds that FIDE should have said something they didn't is valid.

Btw. if somebody says it is illegal to castle that fact would normally follow from the definition of "legal move". If they mean it's illegal for White to castle with the black king however, it wouldn't follow.