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

EndgameStudier

 It should be 100 moves, with castling or double checks also resetting the move count.

NihontoEurope

Hello!

I think Magnus Carlsen had a game where that rule was engaged. Pretty recently.

/Martin

 

MARattigan
NihontoEurope wrote:

Hello!

I think Magnus Carlsen had a game where that rule was engaged. Pretty recently.

/Martin

 

Thanks.

But I couldn't find any terminated by the 75 move rule in the chess-db database for Carlsen since 2017. So I'm still interested to find one.

Arisktotle
Numquam wrote:

You always need to prove that the algorithm gives the desired result. Some brute-force algorithm which simply goes through all possibilities is not that interesting, because you can't use it in practice. Defining a dead draw by using such an algorithm isn't any better than the definition in the article I linked.

What you really need is a smart algorithm. A human can often instantly see if a position is dead or not. So it should be possible to create an algorithm which finds 100% of all dead positions which appear in practice. If you can keep enough mating material on the board after the next 5 moves and the path to the king isn't blockaded, then you know in practice that the position is not dead. You don't need to find the checkmate. If there is insufficient mating material, then the algorithm should look for ways to restrict the movement of the king with his own pieces such that a checkmate is possible. Also note that the 75M and 5 REP rules are not included in the dead draw rule and it may not be possible to include these rules without going through all possible moves. It would be hard to prove that such a smart algorithm always finds a checkmate when a checkmate is possible, but it can be used in practice. Therefore it may be better to define a dead draw as a position in which that algorithm does not find a checkmate.

In fact, 2 sides of logical reasoning come together here. When playing a game on a human level, as much smartness should be used as can be found anywhere. But we are not discussing the intelligence of playing, but the use of rules we play by. Those rules should be as stupid as possible in order to avoid us wasting our brains on figuring out the legality of chess decisions instead of the performance power of our chess choices. We have massive problems in our western societies with enterprising people for no other reason than that they are incessantly frustrated by dealing with ambiguous and incomplete laws while the activities of their real interest are marginilized. Making bad laws should be a high crime.

The discussion we are waging is very similar to the discussion on relational databases 40 years ago. Those who liked it said they were great because of the transparency of concepts and those who disliked them argued the amount and size of table products would be impossible to manage computation wise. The concept people were right, as they almost always are. Why? Well, the fact that the relational model permits the human brain to conceive of enormous amounts of rows and colums and objects and attributes does not imply that they need to be generated on a computational level. Every clever shortcut can be and is used to reduce the number of manipulations required to get to the required result, as long as it conforms to the relational scheme - which is the human angle in the story. The simple relational model provides the grounding of all database activity, while the actual database engine may be as complex as you like

So it is with the dead rule. The proper definition of deadness should be as simple and stupid as possible, while the algorithmic implementation may be as intelligent as you describe - as long as it fully complies with the simple and stupid definition. The ultimate in chess stupidity is prove by move which requires no heuristics or understanding of meta-conclusions.

This makes it appear as if the discussion is completely redundant. When 2 roads - the stupid and intelligent one - lead to the same outcome why argue about it? Answer, when people get smart, they will make mistakes. Much of formal mathematics was designed to guard against overconfident intelligence. Nothing is more stupid than a formal mathematical proof, it's just an endless series of trivial inferences not even an idiot would dare to argue with. When people are confronted with it's conclusions - like the Gödel sentence - they often go mad trying to be intelligent about it. Some are even known to assert that the Gödel sentence is true, and they are unfortunately so numerous that attempts to lock them all up so far failed wink.png

The remaining elements are better presentable in a context of molecular moves which removes issues such as self-reference and paradoxicality regarding legal moves. Note that the simple approach of a game is dead as soon as checkmate is no longer possible does not provide you cover against the application of 75M and 5REP as factors in that evaluation. I know you believe that the distinction of basic and competition rules will help you out but that depends on sanitizing the definition of the legality of moves which in turn depends on its context in a molecular move.

I'll leave it here until I have a post on the molecular moves.

Arisktotle
SpiderUnicorn wrote:

Erm, @Arisktotle @Numquam and @MARattigan what is the conclusion of your debate?

We are like the prosecutor and defense attorney in a court case, likely to disagree from beginning to end. Judgement will only be passed by the authoritative body in chess FIDE - if they ever catch up on the discussion.

Note that my goal is in the future. The issues of game chess, problem chess, retrograde chess and archetypal chess (such as chess960) can only be resolved in their entirety. It requires a more conceptual and generic approach to a number of rules which I cannot all discuss here.

 

danielbaechli
MARattigan wrote:

Just that.

Just that:

Topalov vs Nakamura, 2016
Yudasin vs Erenburg, 2017
Firouzja vs Demchenko, 2019

All drawn after exactly seventy-five moves each without capture and pawn move.

Srimurugan108

That's terrific 

Arisktotle
danielbaechli wrote:
MARattigan wrote:

Just that.

Just that:

Topalov vs Nakamura, 2016
Yudasin vs Erenburg, 2017
Firouzja vs Demchenko, 2019

All drawn after exactly seventy-five moves each without capture and pawn move.

Surprising! They were obviously fooling around ignoring repetition and 50M claimables but still they somehow decided to go thru the motions. I don't know about the time incrementals at that stage. if really small they might have decided to sit it out knowing that the interface would interfere to save them.

danielbaechli
Arisktotle wrote:

[...]
Surprising! They were obviously fooling around ignoring repetition and 50M claimables but still they somehow decided to go thru the motions. I don't know about the time incrementals at that stage. if really small they might have decided to sit it out knowing that the interface would interfere to save them.

All games with two seconds increments per move as it seems:

Topalov vs Nakamura, 2016 - guess 5+2
Yudasin vs Erenburg, 2017 - guess 15+2
Firouzja vs Demchenko, 2019 - guess 15+2 or 10+2

 

MARattigan
danielbaechli wrote:
MARattigan wrote:

Just that.

Just that:

Topalov vs Nakamura, 2016
Yudasin vs Erenburg, 2017
Firouzja vs Demchenko, 2019

All drawn after exactly seventy-five moves each without capture and pawn move.

Thanks for that.

For reasons previously mentioned I think the first one actually terminated in a dead position half a move before the 75 move rule was reached.

Arisktotle

Yep and it appears that Andrew Buchanan agrees to that view as well. He has done a lot of work on dead reckoning - a term he avoids these days - mainly in the composition field. The situation is way more complicated there due to the factor of uncertain pasts and the composition conventions.

It's still possible though that FIDE one day decides that it was all a big misunderstanding and that the 75M termination was supposed to be invisibe to the DP (dead position) rule. Just as the 5R and 75M rule explicitly exclude mutual visibility by their conditions.

Interestingly, the Topalov Nakamura game almost should have terminated 2 half moves earlier - for instance if the white rook had been on a8.

 

lincolnpeters

I didn't even know that existed!

danielbaechli
MARattigan wrote:

[---]

Thanks for that.

For reasons previously mentioned I think the first one actually terminated in a dead position half a move before the 75 move rule was reached.

I guess that by the introduction of the fivefold repetition rule and the seventy-five-move rule 2014, there was simply no awareness that the introduction of these two rules had an impact on the dead position.

 

The rule "5.2.2 The game is drawn when a position has arisen in which neither player can checkmate the opponent's king with any series of legal moves." could have been changed in 2014 to "... with any series of legal moves, disregarding 9.6.1 (fivefold repetition) and 9.6.2 (75-move rule)". Such an addition would have left dead position unchanged and would have avoided the intricated questions which arise today. 

 

The game Guijarro vs Antipov at the World Blitz Championship 2015 is an example for a game continued beyond 75 moves without capture or pawn move. 137. Kg2 hit this limit. Are the remaining moves illegal? Nobody would say so, but already this question can not be easily put away. Surely playing beyond checkmate if undetected is not ok, and the 75-move rule is at the level of checkmate regarding game termination.

 

The game ended in a draw. But if one player would have won, this would lead to another discussion, if the result is valid.

chesschesskid

wow thsi is really cool!

i cant believe games actually ended like that

but what about five-fold repetition?

jsaepuru

The game Guijarro vs Antipov at the World Blitz Championship 2015 is an example for a game continued beyond 75 moves without capture or pawn move. 137. Kg2 hit this limit. Are the remaining moves illegal? Nobody would say so, but already this question can not be easily put away. Surely playing beyond checkmate if undetected is not ok, and the 75-move rule is at the level of checkmate regarding game termination.

 

The game ended in a draw. But if one player would have won, this would lead to another discussion, if the result is valid.

How common are precedents where players play through mutually overlooked checkmate and terminate game with a different result (draw, or checkmated player winning)?

Arisktotle
danielbaechli wrote:

I guess that by the introduction of the fivefold repetition rule and the seventy-five-move rule 2014, there was simply no awareness that the introduction of these two rules had an impact on the dead position.

 

The rule "5.2.2 The game is drawn when a position has arisen in which neither player can checkmate the opponent's king with any series of legal moves." could have been changed in 2014 to "... with any series of legal moves, disregarding 9.6.1 (fivefold repetition) and 9.6.2 (75-move rule)". Such an addition would have left dead position unchanged and would have avoided the intricated questions which arise today. 

 

The game Guijarro vs Antipov at the World Blitz Championship 2015 is an example for a game continued beyond 75 moves without capture or pawn move. 137. Kg2 hit this limit. Are the remaining moves illegal? Nobody would say so, but already this question can not be easily put away. Surely playing beyond checkmate if undetected is not ok, and the 75-move rule is at the level of checkmate regarding game termination.

 

The game ended in a draw. But if one player would have won, this would lead to another discussion, if the result is valid.

Actually, these are precisely the questions we discussed in the foregoing posts of this thread - some of which I already discussed 20 years ago because composers had automatic draw conventions before the new FIDE game rules. The discussion in compositions is more complex as it involves the status of the "draw-rule-modifying" conventions.

My views in short are;

Yes, I am sure FIDE overlooked the interaction of 75M/5R with the dead rule or they would have clarified it. That does not necessarily imply they want it different. I defined the options as "soft DRexit" (your proposal) and "hard DRexit" (current FIDE laws). DRexit = "Dead Reckoning exit" after Andrew Buchanan's introduction of the term "Dead Reckoning" for the evaluation of "deadness".

A game terminates when a dead position is reached. After any sort of termination, game rules no longer apply. There are no longer legal or illegal moves, there are no longer players, there is no running game. So the game phase after a dead position does not exist - it's just backroom analysis even when played in the theater. Which should mean you can no longer be checkmated, resign or lose by arbiter intervention. [Btw, the fact that no legal moves exist after terminations implies that the FIDE definition for dead positions is circular - another thing which is extensively discussed in this thread]

However there is a slight complication. There is an environment for chess games which involves for instance the signing of a scoresheet and protest arbitration. If you sign the scoresheet with a loss, you still lose the game (I suppose) while arbiter decisions may overrule game outcomes simply because the arbiters don't understand the rules or have no consensus on them.

You're welcome to place the issues on the FIDE table. I hear that is very hard!

Arisktotle

On the circularity of 5.2.2: I just realized that article 5.2.2. is not that bad. With some good will, it could be read as non-circular but nobody reads it that way. A better formulation would be: 

5.2.2:  "The game is drawn when a position has arisen in which no series of legal moves exists ending in checkmate."

This is a non-algorithmic definition. It doesn't tell you how to find out whether or not a position is dead by playing legal moves. For an algorithmic move-based definition you need to introduce a "shadow type" for the legal move which you might call the "analytical move" or "imaginary move". And you need a criterion to stop searching before infinity.

SmallppJake

lol I just found this out in a 2/1 bullet game when I had a queen and a king vs a king ,but only had 1 sec. I thought it would be smart to keep pre moving around my king so that I would get more time. Thought I was a genius untill I got a draw lol

Arisktotle

I don't know how the chess.com interface is currently programmed but if it follows the FIDE rules then it will only terminate the game long after your opponent could have claimed a draw. So you were lucky to even get that far! By the way, it might have terminated your game by repetitions if you didn't take care of changing overall board positions.

Moon482

few weeks ago I observed game of top players (somewhat 2700+) here at Chess.com . 
time control was 3 minute without increment.
One of player had around 10 seconds, other one had plenty of time
First guys chaotically moved peaces to reach move limit, other tried to void position repetition and put opponent into state of thinking.
but game was finally drawn after 50 or 75 moves