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

Thanks @narceleb. Do you know where I can find a record of the game?

An action picture from that tournament. The tension is rising, the excitement growing. Will she blow it or not?

Brilliant picture. Doubt if that was when the 75 move rule cut in though.

Understanding the dead rule

The handbook is and has always been a sort of semi-formal description of the rules. For instance, we had a discussion on the checkmate definition and I had to conceed that - for the first time - the definition was pretty solid. But in all the years before, players have been handing out checkmates and arbiters have been approving them without accurate FIDE laws. In the end, there is always a judge somewhere willing to explain to you that rule intention matters when there is space for interpretation or doubt.

Let me start by asserting that the dead rule should be in the competition section! Not because I say so but because that would reflect the reason for having that rule. They were sick of players attempting to win unwinnable games on the clock and thought this rule might help. Outside competition there is no real reason to have a dead rule except if you believe it is required to terminate a game which is not really true. There is no termination procedure connected to the dead rule so that would be an entirely different issue I will address below.

They did put it in the basic section because it is a technical rule - requiring no intervention by clock, players and arbiters - but functionality reigns over technicality and it was a wrong decision. Nobody in his right mind really believes that players need to know about the dead rule before they can play basic chess! Not that it matters greatly if you consider all rules equal and the rule split only there to provide a referential tool for looking up issues relating to basic chess and issues relating to competitions.

The first thing to consider is how to apply the dead rule to stalemate. It is completely out of the question to start a debate on the 5REP or 75M rules relating to deadness before making your mind up on stalemate. After all, stalemate was already there when deadness was introduced and it is an uncontentious part of the basic rule set and therefore applicable to all formats of the game. Recently, a position similar to this one was posted a number of times:

Is the position dead now or is it played to stalemate?

When you look at the dead rule technically, it is dead but if you look at the intention, there is no reason to terminate because the game will end before long anyway. Nobody will argue that the dead rule was necessary to put an end to all those endless forced stalemate sequences Frankly, I don't know FIDE's answer to this issue but I do know that the answer is required before we can move on to the 5REP and 75M issues of the same type.

There is one very interesting technical point to the answer though. With stalemate looming you can actually axiomatically prove that the position is dead by demonstrating that all moves lead to stalemate. Without a termination condition like stalemate, you need intelligent reasoning to determine a position is dead - something for which no simple move-based algorithm exists. And the same would be true for 5REP and 75M. Even if you do not like to apply deadness in conjunction with these competition rules, there is no denying that they provide a perfect backdrop for evaluating deadness. You actually need just one of them since both rules guarantee finite chess games. The stalemate rule by itself does not.

So, no definitive answers, but some interesting considerations!

Note: Obviously, I couldn't comment on the interaction with the half-dead rule without creating chaos. As the dead rule is almost - in games, not in problems - reduced to insignificance by the half-dead rule, every recourse to reason becomes questionable. It is extremely hard to make a point when no points are in stock.

No, I think there is a reason to terminate.

Comparable actually to the reason why perpetual check rule, which once existed, has been abolished.

There used to be a termination condition of "insufficient material to checkmate". For example, two bare kings can never check each other, let alone checkmate. Therefore, when two kings are left alone on empty chessboard, they are not required to play on until either player can claim 50 moves from last capture. The game is drawn immediately after the capture that leaves insufficient material to checkmate - no half-moves are to be played after all pieces are gone.

But the thing is, "insufficient material" is not the only condition that makes checkmate impossible. There are also some blocked positions, like:

There is not "insufficient material to checkmate" because the pawns might promote and mate... if they could be unblocked, and neither king has any legal way to reach opponent´ s pawns and unblock his own. It is also not a stalemate, because each king is free to move - on his own side, to no effect.

So what could be the language to bring it under a common denominator with "insufficient material"?

The rule adopted by Fide in 1997 is "impossibility of checkmate".

But this is overinclusive, because it also includes positions where reaching insufficient material or stalemate is forced. Like:

White has blundered to hang his queen, and black has only one legal move out of check - which would create insufficient material. Black does not have a legal choice to blunder into moving out of check and eventually getting properly mated.

Where should this game end? Here, with queen still on board, because the position is already dead? 1 half-move later, after there is insufficient material for mate? Or 101 half-moves later, after the kings have made no progress for 50 moves?

I suppose there are several issues here but let me concentrate on just one. I didn't say there is no reason to terminate; - more the opposite, see my paragraph on axiomatic proof. I only said that the "dead rule" itself contains no obvious trigger to terminate the game (as your diagrams show). Something else needs to happen like "flag falling", or a player claiming deadness (based on his private insight) or one of those other termination rules (5REP or 75M) becoming active. I guess we essentially agree on this.

According to agreed upon understanding of deadness, your first and second diagram are drawn as they stand. But they traditionally came under the heading of "intelligent understanding"  of deadness and fell outside axiomatic provability. Since 2017 however, in competitive environments, the automatic 75M rule and the automatic 5REP rule guarantee finite future draws which can be proved by "move analysis" (which is axiomatic). Provided you accept they apply in conjunction with deadness, see Numquam's posts and my "anticipated stalemate" diagram.

Note: Insufficient material has always been a failed concept. Glad it disappeared - except in USCF which is not my side of the ocean.

One of the problems with the dead position rule is that it is self referent. When it says "neither player can checkmate the opponents king", the assumption is that this means with the rules in effect for the game, but that includes the dead position rule. This can give the arbiter some knotty problems.

Suppose the following position appears with a ply count of 147. Only the mainline produces a win for White within the 75 move rule.

Suppose White touches the g5 knight intending to play variation A. After playing the g5 knight there is no possible win within the 75 move rule, so the arbiter could assume according to 

4.2.2 Any other physical contact with a piece, except for clearly accidental contact, shall be
considered to be intent.
4.3 Except as provided in Article 4.2, if the player having the move touches on the
chessboard, with the intention of moving or capturing:
4.3.1 one or more of his own pieces, he must move the first piece touched that can be moved

that the position is dead as soon as White touched the piece. But if the position is dead then White cannot make any move with the g5 knight because the game is terminated. Therefore by

4.5 If none of the pieces touched in accordance with Article 4.3 or Article 4.4 can be moved
or captured, the player may make any legal move.

the arbiter must direct white to move a different piece. Suppose then he touches his king with the intention of playing variation B the same occurs. 

The arbiter would therefore direct White to play the d2 knight. 

But in that case the arbiter would have effectively guided White to the win. 

The main problem though is if the position is dead after White first touched the knight then it's not dead, whereas if the position is not dead then it's dead.

This is probably another situation in which software interfaces do not arrive at the correct result (whatever that may be), but then they don't usually implement "touch move" in the first place. 

Edit: The same applies here. If White touches the rook can he argue that the rook can't be played because the position is already dead if he must play the rook, so he should be allowed to play something else by 4.5?

But if you just excise the last 25 moves your "proof" then proves it's 25.

Not much of a proof really.

(It shouldn't take you more than ten moves to mate with a queen from any position by the way.)

I can answer these issues by referring to semi-formality and rule intention - as I think FIDE would do - but that would be a bit cheap. Also I recognize the issues as similar to the ones I encountered in the retro-field and which forced me to develop generic retrograde logics. So I might have some answers for you.

First the "touched piece issue". In my theory, the progress of a game should not be treated through a time continuum but as a discrete, ordered particle stream, The basic particle is atomic, corresponding to a full legal move. There are also subatomic particles corresponding to elements such as "touched pieces", "incomplete moves", "illegal moves" and microphases in retro-move evaluations. Note that illegal moves are also incomplete because they require correction. There are also molecular particles corresponding to groups of moves, such as "a legal series of moves" or "a justification sequence for an a posteriori move" (retro) or "the 2 justification moves following checkmate or stalemate". A molecular move is always legal or illegal or something else in its entirety. All full moves in it are assigned the same legality status. Note that the full moves in a molecular move appear internally atomic but are not always contextually legal - inevitable paradoxality connected to the molecular theme (to be explained later).

The entities of "atomic move", "subatomic move" and "molecular move" are of completely different types and governed by completely different rules. For instance, article 4 doesn't look anything like any of the other articles in the handbook. There are also differences and similarities between atomic moves, as there are differences  and similarities between subatomic particles and between structures of molecular moves.

There is a default understanding of positions/diagrams and moves in relation to these concepts. For lack of additional information you are always to assume that they align with games on full legal move boundaries. This is not self-evident. There is a retro-variant (MDR = minimum deviation of the rules) which permits you to publish diagrams half-way of an atomic move, for instance when you've taken an enemy piece off the board but not yet moved your own piece to its square. Before continuing the game normally, you must first complete or correct the underway move. Interestingly, I composed an MDR problem very close to your diagram example with deadness but different on an essential point.

(my MDR problem described - can't show it without elaborate explanations): Picture a diagram which is illegal unless you assume some "minimally illegal"move is underway - an "article 4 diagram". In fact, there is a choice amongst several "minimally illegal" moves preceding this diagram. One of these however, would lead to a forced correction move and stalemate! Since it was forced, the position was already dead and therefore the correction move impossible! The difference with your diagram is of course that the history in mine is unknown and can be completed in more than one way. What our examples have in common is that they both say you can't start on a move in a dead position, not even on a subatomic part of it like "touching". But your examples goes one step further. You also state that the death evaluation continues during the subatomic phases of a move. Here we part.

Considering that the entities atomic move and subatomic actions are governed by completely different rule sets, I see no reason to claim that the dead rule should apply to subatomic phases. It refers to "positions" and "legal moves" but nothing subatomic at any place. Clearly, the view that death-evaluation should align with full moves is better justifiable in the presented discrete view of game events than in one of a time continuum.

There is a practical point related to this as well. An arbiter need not interfere in a touch-move situation, neither does he in partial move situations. After all, letting go of all pieces does not complete your move unless it is a full legal move. When it is not, your opponent may simply wait for you to make the correction or completion prescribed by article 4 - until we have an atomic move. No arbitration anywhere. Only if you push the clock or your flag falls, arbitration needs to take place. That's why the clock plays an essential role in detecting the termination of an illegal move. Though I give this as a practical point, it is easy to project it as a basic concept. Arbitration is on full moves, full-move positions and clock aborts only. Subatomic operation is on account of the players - as long as it abides by article 4 and subatomic information is made available to the arbiter when he is required to intervene. An onboard camera would do nicely.

Edit 11/7: Note: I originally thought you were discussing the situation where white touched and let go of Ng5 but, upon rereading, I now assume white simply started on a legal move. It leaves unchanged what I wrote about incomplete and illegal moves which are more complex cases of similar incidents. The argument for all is that there is no need and no advantage in perpetuating deadness evaluations - either by machine or by human arbiter - during un unfinished move. Besides the reasons already given, it is also hard to complete the scoresheet with the going-ons. You can't write down "touches" or "half moves" as game actions resulting in formal terminations. And what are their FEN/PGN notations?

To be continued, can't do all this in one sweep.

Need more time for the remainder. Will follow.

@Arisktotle

Some comments on the treatise so far:

I can answer these issues by referring to semi-formality and rule intention - as I think FIDE would do ...

Most game rules are semi formal. I don't think that is a problem, but if the rule makers have to fall back on "intention" to explain a point, that is simply an admission that they didn't write them right (even if the wrote them by rote). What they are then implicitly saying is that the rules don't say what they were intended to say. I don't even know what FIDE's intention is in the above example, if indeed they have an intention.

There is a default understanding of positions/diagrams and moves in relation to these concepts. For lack of additional information you are always to assume that they align with games on full legal move boundaries.

I question that.

There is nothing stated in the laws (I'll start using "laws" as FIDE do - rugby union players get excited about which word is used, but mainly it's less typing than "rules") to the effect that chess is a sequence of discrete steps running from move termination to move termination. Indeed, if a player resigns or his flag falls he cannot then complete his move. (The FIDE laws define two games - neglecting the appendices - depending on whether the laws are taken to be those under Basic Rules and Competition Rules or those under Basic Rules only. The point about flag falling doesn't necessarily apply in the latter case, but even then there is usually some understanding of a termination of the game due to time constraints that would not usually apply only on termination of a move.)

I think the laws of any game, chess included, are intended to apply for the duration of the game (which is normally fixed by the rules). If the laws are properly written the prescribed starting position(s) (more on "position" later) do not contradict the laws. and that must remain true until some event occurs that is referred to in the laws. In those cases the events must comply with the laws. The points at which the laws must be enforced are then naturally determined by the events referred to in the laws.

In the case of chess (basic variant) these events would be completion of a move, offer of a draw, acceptance of a draw, resignation, touching a piece, ceasing to touch a piece. These are the events to which I would say application of the rules applies. Different games would have different events (which may not form a discrete series) but I would say the way any written laws should be interpreted are the same in each.  

There is a practical point related to this as well. An arbiter need not interfere in a touch-move situation, neither does he in partial move situations.   

According to the current arbiter's handbook it is the arbiter's responsibility to enforce touch move. It doesn't say that he should wait until the clock is pressed by the player engaged in illegal activity in this respect. Obviously if the arbiter is present then the most convenient point would be when the player touches a second piece with intent to move. In fact it is stated here

http://arbiters.fide.com/images/stories/downloads/2018/Arbiters-Manual-2018-v1.pdf

If an arbiter observes a violation of Article 4, he must always intervene
immediately. He should not wait for a claim to be submitted by a player.

This would also be the usual point the objection would be raised in a game played under Basic Rules. 

Note that illegal moves are also incomplete because they require correction. 

My view would be that there are no illegal moves in a chess game. In fact there are no illegitimate moves, by which I mean moves that contravene any of the rules (to distinguish from the FIDE defined term "illegal move"). The game ceases to be a chess game at the point any action is taken which contravenes the rules. This arguably wouldn't include moving a piece not first touched because

appears to say that the violation is claimable and play can continue as part of a legitimate chess game. I would say the FIDE laws are strictly speaking inconsistent on this point

There are also molecular particles corresponding to groups of moves, such as "a legal series of moves" ...

The FIDE laws refer only to "series of legal moves", never "legal series of moves". The term "legal move" is defined in art.3.10.1, so "series of legal moves" should be interpreted as a series of these.

(prematurely posted - I'll post a continuation)

On post #92:

I see we operate from an entirely different paradigm. The primary difference is that I consider the conceptual explanation of rules almost completely absent while necessary to interpret rules where not formally complete or accurate. Therefore, a number of the concepts I added myself. I will never try to resolve rule issues by merely relying on what they seem to say as such is wholly insufficient. Zooming in on rules and formulations never resolves anything (except in math).

In my paradigm, I start from the problem side, which is that the rules do not appear to resolve certain issues - like the ones you raise - and therefore disqualify the rules as they stand. Subsequently, I introduce missing concepts which I believe constitue the basis of intended rules and interpretations. No rules are random axioms, all are based on archetypal concepts, certainly in chess which is a very simple game. These concepts yield rules and intentions in the interpretations of those rules. Example: nobody knows how a bishop moves on a 10x10 board based on the exact rules of standard chess. But everyone knows intuitively that a bishop follows the archetypal law of physics that an object not hindered by an external force moves in a straight line (until stopped by a barrier) - Newton's first law of motion. The only way to ever solve any serious rule question is to first go back to concepts, ask the question for the concepts, and then project the answer back into the rules. Nothing else works. In casu, the bishop will move as unhindered to the borders on a 10x10 board as it does on an 8x8 board. You will argue that this example is not in standard chess but it shares the characteristic of your questions that it relates to a blank space in the rules. Therefore I will never claim that your answers or interpretation are impossible, only that more natural concepts exist leading to different answers. Concepts are inductive, they cannot be deduced of proofed; they can be recognized as archetypals (in your mind) and they can be agreed upon - or not.

Example: An archetypal concept of basic chess is that time does not exist. There is only change and not all change is of the same type and on the same level and governed by the same rules. The type of change that rules apply to is either clearly defined or understood by concept. The dead rule is defined in terms of positions and moves. As recognized by everyone (including problemists where this understanding is vital), a position is always on a full move boundary and a move is always a full move. Which makes clear that deadness is conceptually about full move changes. It doesn't literally say so in the rules, but who cares? When you sit back and relax for 5 minutes you will see that this is the concept and this is what is intended. If you can't, it is because you refuse to step outside the logics of zooming in and deduction which are fruitless.

Indeed, illegal moves do not exist in a chess game - on a full move level. But they exist on a subatomic move level. Now, subatomic moves do not exist in the laws. True, but they exist on the concept level associated with article 4; everybody reading it, inevitably thinks "partial", "incomplete" or "illegal" moves. Possibly "move" is an incorrect term here but it doesn't matter as long as you think the right concept which is "subatomic move".

The fact that an arbiter enforces the correction or continuation of a "touch move" says nothing about the point of intervention. It would be extremely funny when a player touching a piece - and apparently not completing his move - would receive a warning from an arbiter. After all there is no reason to assume he will go against the rules by moving another piece until he has done so and pushed his clock. That is the proper intervention point for arbitration. This is also the point where it is legitimate to speak of an "illegal subatomic move". The player has created a contradiction: (1) by pushing the clock he says "I completed a legal move" while (2) no legal move was executed on the board. And contradictions cannot be permitted, forcing the arbiter to intervene.

Molecular moves are only a concept which is very effective in generating rules for all the scenario's it applies to. The distinction between "legal series of moves" or "series of legal moves" is not relevant in my story, as it clearly only refers to situations where the status of the whole series is questionable (considering self reference or other things). There is always something fishy going on or we wouldn't be referring to the move series at all.

More about molecular moves in my follow-up post

Irregularities

To explain my statement, " My view would be that there are no illegal moves in a chess game":

The laws covering various games often include provisions for irregularites (i.e. events which while violating the laws nevertheless occur when a game is in progress). These will often allow for continuation of the game, possibly with some penalties. I believe football, for example, has a law that you must not clog an opponent when the referee is looking, but the game is normally allowed to continue in such a case. Article 7 in the FIDE laws contains such rules.

My view is that the game runs to the point of the irregularity and then from the continuation if that is allowed.

In the case of touch move we have:

and

which seem to paradoxically allow illegality. If a player must move the first piece touched under 4.3 what can his opponent claim against under 4.8?

As an aside, from my reading of the FIDE documentation if, in the following position, White first touches the queen with intent to move it and then touches another piece, then in a tournament game the arbiter should demand he play the queen, whereas in a friendly Black has the option of ignoring the transgression. Would that be other people's understanding?

On #94:

There is an conceptual illogicality in article 4. While it is in the basic rule, set, it once refers to the arbiter who's presence is commonly only guaranteed in competition environments. Which immediately raises questions on the article 4 actions of the arbiter in various competition forms such as rapid and blitz chess. It is possible to write article 4 without arbitration, simply by demanding that a full legally executed move must be completed before the move turn goes to the other player. The procedure only matters mildly in the basic rules since the players must agree anyway. If one refuses to correct, what can you do?

The way I read article 4.8 is that after the opponent his pushed his clock, the other player or the arbiter may intervene by demanding a correction of the opponents move, since it wasn't executed in accordance with the rules. In effect an illegal subatomic move was executed, demanding correction. A problem can arise when there is no clock. What is the intervention point? That's is why I consider a clock indispensible - not to keep time but to terminate the move!

@MARattigan, if you interpret the rules like I do, you wouldn't have all these problems. There is a difference between the 'act of moving pieces' and a 'move'. The 'act of moving pieces' is how you communicate that you are playing a particular move. Clearly this is different for online chess, but most of us still call online chess chess. I have a different view on this than Arisktotle, the 'act of moving pieces' is not part of a 'move'. I prefer to use mathematical terms to define these things. I just googled it and I found a nice link where it is defined in mathematical terms: http://home.planet.nl/~narcis45/Chess/Chess%20Math%20Definition.pdf

Excluding castling a move can be seen as a pair (a,b), where a and b are squares on the board such that a is not equal to b and there is a piece on square a. You move a piece from one square to another square after all. The set of legal moves is a subset of the set of moves. A precise definition of legal moves is given in that link. Note also that the dead draw rule as defined there does not include the 75M rule or 5 REP.

The move is terminated irrespective of clocks whether or not the clocks are mandatory (but under 7.5.1 an illegal move only is terminated by pressing the clock in games where Tournament Rules are in force). 

Most interesting and valuable! Do not thiink I disagree with it or have a different viewpoint - I only have a much broader scope including fairies and retrograde chess. The article you refer to implements precisely the theory based on atomic moves which are full moves only. It defines positions, moves and the dead rule only in terms of full (valid) moves. Subatomic moves (actions as I also refer to them) do not exist. That gets you a simple theory which works.

However, when attempting to address MARattigans suggestions of deadness after a "touch move" I cannot start by denying that such a thing exists. Math does that with the greatest of ease. After all, "touch move" is discussed at length in article 4 of the handbook. So I have to introduce a different category of objects (you would call actions) and explain how they are governed by different rules and why deadness shouldn't apply to them. Also, as described in my posts, there is a fairy form making use of the separation of actions inside a valid chess move. They even publish diagrams with half-way moves. Can't really get around that, can I? 

Yes, yes, yes! I already explained the clock is only needed to identify the termination of illegal moves. Don't expect me to repeat that in every sentence on the subject  Formally, as a player may not be aware he played an illegal move, one should require him to push the clock after every move he believes completed; until that point he may still correct it in accordance with article 4.

Note that I am not quoting or even interpreting FIDE rules on the clock. I have argued conceptually that a clock is needed in the basic rules as a termination device to prevent the players waiting forever. After all, one player played a move which places his king in check (unaware of it), the other player waits patiently until the move is retracted and replaced by a legal move. Of course, he will not disturb his opponent to complain as he might already be aware of the situation and thinking about a replacement move. Pushing the clock clears up the matter!

What I mean is that if I read a 'series of legal moves',then I do not include the 'act of moving' in that series. So in my opinion there is no such thing as a subatomic move. A series of actions is not a move. I prefer to use a mathematical definition like in that article. So the rules which apply to these actions like touching a piece do not apply to a series of moves.