Chess will never be solved, here's why

Sort:
Avatar of DiogenesDue
playerafar wrote:

Only read the part with the snipped quote but without the 'snip' message.
Somebody (not I) needing to resort to that - hoping to bait ...
but again - he's failing.
The posts about 'gatekeeping jargon' stand.
And its a subject that could be discussed more.
But by those willing and able - which are many.
-----------------------------
Also from the internet:
"Bill Nye is a great example of someone who intentionally avoids gatekeeping jargon. His whole communication style is built around translating complex ideas into everyday language so people feel invited into science rather than pushed out of it."
-----------------
Plus - the people who prefer the gatekeeping jargon - including the jargon that is extra-misleading - tend (not all) to be the same people who discredit science by their trolling against nonconformity and by their trolling against those not conforming.
They're not doing anything for the science side by so doing.
We'll see a lot of defensiveness about their real motivations too. And projection.

There's no "we" to speak of, and you are still avoiding posting what the AI(s) actually told you about the term "perfect information". The "jargon" garbage is is purely deflection, because you haven't got the character or integrity to post what an AI says about chess being a game of perfect information straight up. It's pitiable, but par for the course.

Avatar of OctopusOnSteroids
playerafar wrote:
OctopusOnSteroids wrote:
playerafar wrote:
OctopusOnSteroids wrote:

@playerafar

Okay but question.. if you say chess shouldn't fit the classification "game of perfect info".. how do you distinguish chess from poker to illustrate one game technically disables some information about future outcomes from the players and thusperfect solution not possible even in theory... while one is just too complex for humans but all is available?

For a technical definition like "game of perfect information" its pretty problematic to base it on limitations of human cognition, since we'll never know where exactly that line is

Earlier - I responded to the 'how do you distinguish?' part with the word 'easily'.
Now I'll respond to the second paragraph.
Who based it on 'limitations of human cognition'? Who did that?
And as to 'never know' - then such limits would become moot in the context anyway in a scenario x centuries from now if and when chess is thoroughly solved ...
which would depend on humanity surviving its own activities of self-destruction which include nuclear arsenals - manmade climate change - mass pollution - disinformation and other things.

You did, and I'll quote and show where.

Start of quote/

'Perfect information' is actually inappropriate with chess ...

Here's why - because in so many gazillions of positions 'perfect play' isn't known./end of quote

"Known" refers to a cognitive state. Thus.. perfect information is inappropriate in your opinion because of human cognitive limits in processing this information... You prefer to use this as a condition for 'game of perfect information' rather than the fact that the deterministic information exists.

And yes you did say "easily" earlier, which is hardly an alternative definition for something you call inappropriate..

Perfect play isn't known because that's the state of affairs.
The context.
'limits of human cognition' is your idea. And something different.
I didn't say 'limits of human cognition'. Your idea entirely.
So your accusation is unfounded.
----------
And 'easily' stands. 
Because poker is easily distinguishable from chess.
You don't need 'gatekeeping jargon' to do so.
See my previous post about the subject of 'gatekeeping jargon'.
I don't believe 'D' is capable of addressing those points objectively.
But you might be.
Most people would be.

We still got some missing pieces since the existing definition of 'perfect information' serves some purpose(s).... so in theory chess can be solved and the output would be a deterministic map which provides an outcome for any given chess move, in advance, with certainty... same cant be said about poker. Why is that.. whats the property that chess has that poker is missing? Cant be called 'game of perfect information' since its inappropriate according to you, fine, but how would you describe it instead in more appropriate way..?

Avatar of AG120502

According to Wikipedia, chess fits the definition of a perfect information game. I don’t really see the problem here. I know Wikipedia isn’t the most trusted source in discussions like these, but I can’t think of a single person in the field of game theory who would define chess not to be one.

As for the distinction between data and information, I don’t really see how it applies here. One still has the facts about something, which is how the dictionary defines it. And the definition used in game theory says nothing about useful information.

Avatar of Optimissed
DiogenesDue wrote:
playerafar wrote:

Only read the part with the snipped quote but without the 'snip' message.
Somebody (not I) needing to resort to that - hoping to bait ...
but again - he's failing.
The posts about 'gatekeeping jargon' stand.
And its a subject that could be discussed more.
But by those willing and able - which are many.
-----------------------------
Also from the internet:
"Bill Nye is a great example of someone who intentionally avoids gatekeeping jargon. His whole communication style is built around translating complex ideas into everyday language so people feel invited into science rather than pushed out of it."
-----------------
Plus - the people who prefer the gatekeeping jargon - including the jargon that is extra-misleading - tend (not all) to be the same people who discredit science by their trolling against nonconformity and by their trolling against those not conforming.
They're not doing anything for the science side by so doing.
We'll see a lot of defensiveness about their real motivations too. And projection.

There's no "we" to speak of, and you are still avoiding posting what the AI(s) actually told you about the term "perfect information". The "jargon" garbage is is purely deflection, because you haven't got the character or integrity to post what an AI says about chess being a game of perfect information straight up. It's pitiable, but par for the course.

A suggestion there that you skipped to the last line because his post was hard going. I won't disagree with that.

However, this fatuous claim that "there is no we" .... divide and attempted rule. It means you're saying that no-one agrees with him. If his posts were a bit shorter and more to the point, I'm sure there would be more agreeing with him and therefore more agreement too .... or maybe the same people agreeing who would then agree more. But even so, those reading posts are the "we". So are you saying no-one reads posts on chess.com?

Avatar of crazedrat1000

It would be a perfect data game but not a perfect information one, as optimised pointed out earlier. It’s not complicated, tbh. Game theory is a model and models are incomplete, they have yet to make distinctions such as these to my knowledge, but if they were to do so they’d be capable of modeling some games more fully. 

Avatar of Optimissed
AG120502 wrote:

According to Wikipedia, chess fits the definition of a perfect information game. I don’t really see the problem here. I know Wikipedia isn’t the most trusted source in discussions like these, but I can’t think of a single person in the field of game theory who would define chess not to be one.

As for the distinction between data and information, I don’t really see how it applies here. One still has the facts about something, which is how the dictionary defines it. And the definition used in game theory says nothing about useful information.

Well, I do see the problem and I'm normally seen as quite a bright guy. I know there are those trying to detract from that but it's just from necessity really ... their necessity.

An analogy could be wartime codebreaking. Without the Enigma Machine, the code received from intercepts say in North Africa between Rommel and his commanding generals would simply have been data ... useless data. After analysis by Enigma it became useful information which enabled the Allies to defend Tobruk by making sure they always managed to sink the supply ships coming from Italy. That was the turning point of the war.

I really don't think that looking at the original board position and knowing how the pieces moves "solves chess". In fact, we know it doesn't. Information is only information in relation to its proposed uses. The price of gold in Timbuktu wouldn't help a gardener grow rhubarb round his compost heap. The proposed use for chess info is "being able to always make good moves". If the initial position doesn't enable us to solve chess then it isn't full information enabling that end.

Basically the Wiki definers, who are simply people like you and me, are wrong. Simple as.

Avatar of DiogenesDue
crazedrat1000 wrote:

It would be a perfect data game but not a perfect information one, as optimised pointed out earlier. It’s not complicated, tbh. Game theory is a model and models are incomplete, they have yet to make distinctions such as these to my knowledge, but if they were to do so they’d be capable of modeling some games more fully. 

...and now it's 3 navel-gazers telling everyone that game theory isn't a valid field of study. All 3 of them without a single applicable credential to their name.

Avatar of Elroch
playerafar wrote:
mpaetz wrote:
Optimissed wrote:

We can define "game state" as the real game situation as to whether it's a forced win for one side or not. It isn't important that we cannot determine the game state since it's an abstract theoretical concept only.

If "we cannot determine the game state" we cannot know in any move (or series of moves) alters the game state, making it impossible to classify moves as "perfect play".

Exactly.
Which means in turn that chess is Not a game of 'perfect information' ...

No it does not, and chess is the archetypal game of perfect information.

Avatar of Elroch
playerafar wrote:
DiogenesDue wrote:
playerafar wrote:
mpaetz wrote:

If "we cannot determine the game state" we cannot know in any move (or series of moves) alters the game state, making it impossible to classify moves as "perfect play".

Exactly.
Which means in turn that chess is Not a game of 'perfect information' ...

Exactly wrong, you mean. The definition of a game of perfect information has nothing to with perfect play being known or unknown.

Translation: 'perfect information' is a phrase made up of common words -

This reveals a fundamental lack of understanding about the use of language in a precise subject such as game theory. The use of the two words in the whole of the rest of the world is of not the slightest relevance. All that matters is what the DEFINITION of a game of perfect information is. "game of perfect information" is just a label for that definition. (Actually "game" itself has a definition, and "of perfect information" is a defined property of games, but this is going to confuse you further..

To repeat a tip you have been given before: when you feel inclined to comment on the language used in a precise subject you are not expert in, don't. You arein a worse position than going to France and complaining about the effect of the acute accent on pronunciation.

and 'gatekeeping jargon'

An irrelevant term to forget and never use again. The terms used in subjects like mathematics and computer science are NOT to exclude people. They are for people to communicate. And that includes anyone who has the capability to understand a definition.

Avatar of playerafar

'Gatekeeping' and 'gatekeeping jargon' are both realities on the internet.
There is no reason to pretend they don't exist.
Stands.

Avatar of Optimissed
Elroch wrote:
playerafar wrote:
mpaetz wrote:
Optimissed wrote:

We can define "game state" as the real game situation as to whether it's a forced win for one side or not. It isn't important that we cannot determine the game state since it's an abstract theoretical concept only.

If "we cannot determine the game state" we cannot know in any move (or series of moves) alters the game state, making it impossible to classify moves as "perfect play".

Exactly.
Which means in turn that chess is Not a game of 'perfect information' ...

No it does not, and chess is the archetypal game of perfect information.

No. that's a ridiculous claim, since many people disagree and since it's very easy to give, say, noughts and crosses as a game of perfect information which can be analysed clearly and accurately. So if we should wish to use words according to their true meanings, we wouldn't use "archetypal" in the context of something which is unclear and is a target of reasonable doubt and disagreement.

Avatar of Optimissed

It's correct, though, to cite chess as being *seen by some** as an archetypal game of p,i,

Avatar of Optimissed
Elroch wrote:
playerafar wrote:
DiogenesDue wrote:
playerafar wrote:
mpaetz wrote:

If "we cannot determine the game state" we cannot know in any move (or series of moves) alters the game state, making it impossible to classify moves as "perfect play".

Exactly.
Which means in turn that chess is Not a game of 'perfect information' ...

Exactly wrong, you mean. The definition of a game of perfect information has nothing to with perfect play being known or unknown.

Translation: 'perfect information' is a phrase made up of common words -

This reveals a fundamental lack of understanding about the use of language in a precise subject such as game theory. The use of the two words in the whole of the rest of the world is of not the slightest relevance. All that matters is what the DEFINITION of a game of perfect information is. "game of perfect information" is just a label for that definition. (Actually "game" itself has a definition, and "of perfect information" is a defined property of games, but this is going to confuse you further..

To repeat a tip you have been given before: when you feel inclined to comment on the language used in a precise subject you are not expert in, don't. You arein a worse position than going to France and complaining about the effect of the acute accent on pronunciation.

and 'gatekeeping jargon'

An irrelevant term to forget and never use again. The terms used in subjects like mathematics and computer science are NOT to exclude people. They are for people to communicate. And that includes anyone who has the capability to understand a definition.

Even if the definition is misleading and/or inaccurate? It's an easy mistake to make, to confuse data with information. But then the game theorists who were asked to help solve chess made such a mess of it that it really is best to junk ALL their ideas and start again.

Avatar of playerafar
Optimissed wrote:

It's correct, though, to cite chess as being *seen by some** as an archetypal game of p,i,

Hi Roger.
A long time ago - here - I think a member suggested that with 'perfect information' (very misleading) we'd expect to know the result.
And somebody else had a very sharp reply to that.
But - removing the word 'expect' and simply instead noting that both the actual 'game state' and the 'perfect play' are both unknowns in a gigantic number of positions - 
then 'perfect information' is already instantly invalid. 
That's not only objectively true though. Its also subjectively true.
But in the subjective - there's a lot more variance.
happy

Avatar of playerafar

An example of a 'perfect information' game would be tic tac toe.
There's probably a list of such games.
Chess isn't one of them.
But might be someday ... they'll have to have a pretty big computer though ...
happy

Avatar of Optimissed

Incidentally, I just now gave AI a chance. The question was "data or information?" It backs my judgement completely.

AI Overview 
 
Data is raw, unorganized facts (like numbers or words) that lack inherent meaning, while information is that same data processed, organized, and put into context to become meaningful and useful for decision-making, like turning individual test scores (data) into an average class score (information). Essentially, information is data with context, making it valuable, whereas data is the fundamental building block. 
Data
Definition: Raw, isolated facts, figures, symbols, or observations

Avatar of Optimissed

Having said that, its examples are REALLY bad!

Avatar of playerafar

Conversation from the internet:
My question:
"In ten lines or less please tell me something about how 'gatekeeping' works on the internet. 'Gatekeeping' in its negative figurative sense."
AI response: 
"Here’s a tight, ten‑lines‑or‑less take on negative gatekeeping online:
It happens when people claim the “right” to decide who belongs in a community or fandom.
Gatekeepers police knowledge, taste, identity, or experience to exclude others.
They often demand obscure expertise as an entry fee, even when it’s unnecessary.
Social media amplifies this by rewarding performative authority and call‑outs.
It can create hostile spaces where newcomers feel unwelcome or intimidated.
Gatekeeping usually protects status, not quality or authenticity.
At its worst, it reinforces bias—who “counts” often mirrors existing power structures."
-----------------
I then continued - somewhat more specifically - like this:
My question:
"I've seen three people who constantly push 'its the cogent people versus the crackpots' - one of them constantly keeps making new accounts on the website for trolling purposes - another constantly accuses people of that which he does and gets caught doing so - and the third one backs the two of them up - enabling them. Please comment in 12 lines or less."
AI response: 
"That trio is basically running a textbook gatekeeping ecosystem:
The “new‑account troll” manufactures noise and disruption to unsettle others.
The “projection guy” flips accusations to keep everyone on the defensive.
The “enabler” provides social cover so the first two seem more legitimate.
Together, they create the illusion of a coherent in‑group policing “cogency.”
The rhetoric of “cogent vs. crackpot” becomes a shield for bad‑faith behavior.
It’s less about standards and more about maintaining control of the narrative.
Their dynamic works only if others accept the frame they’re imposing.
The healthiest move is to refuse that frame entirely and engage on your own terms."
-----------------------

Avatar of playerafar

The conversation continued at length ... (was chess.com mentioned?)
with the AI maintaining that such 'engines' are quite common and in addition to Reddit 
typical sites include:
Discord servers — especially topic‑focused ones where a few mods dominate the culture.
Facebook Groups — admins often curate reality by removing dissent and elevating loyalists.
Quora Spaces — owners can shape the narrative and block critics.
Stack Exchange communities — smaller sites sometimes develop tight in‑groups that police tone or “expertise.”
Special‑interest forums (science, gaming, philosophy, tech) — many run on old forum software where OPs or moderators have broad control.
Fandom wikis and forums — cliques often enforce their own “canon” and ostracize dissenters.
Niche hobby sites — anywhere with user‑run threads or rooms can develop these micro‑power structures.
---------------
I also asked again about the specific ingredient of one of the three constantly making new accounts - and it stated that's common - with the other two backing up that person and the three all essentially backing each other.
The AI then also mentioned some game sites where the pattern occurs.

Avatar of Optimissed

Even so, if such a thing were to exist here, which is highly unlikely, no mention should be made of it, since it would cause offence to others. In particular, correct etiquette has it that the word "cabal" must never be mentioned. Remember that whenever accusations are made, the same accusations will be directed back at you; and quite rightly. Therefore it's best to never mention anything like that and stick to topic. Naturally, the moderators would take correct action if such a thing were to occur.