Seems to me a definition that would exclude all games that involve any action at all that is not forced on the players would not be terribly useful. Presumably @Optimissed disagrees.
Perfect information was explained to @Optimissed well before he was muted by chess.com for three months at the end of last year.
Which doesn't say it's impossible.
The counter argument is like saying the only way to solve Fermat's last theorem is by calculating a^n+b^n for all combinations of nonzero integers a, b and n>2 and checking if the result is c^n for some integer c. There are obviously far too many integers for this to be pracicable even on the fastest current computers, so Fermat's last theorem will never be solved.
No, I am simply saying it's somewhat pointless to discuss a premise were not a single step has been taken towards proving it is viable.
But if everyone takes the same view, not a single step will ever be taken towards proving it is viable.
[General aside: still catching up with some posts, but no reason to take these later replies as re-opening something if the person I am replying to considers it closed.]
I disagree. Many innovators would prefer to bring something to the table first in order to get something going... "I was interesting in this, so I did that, and the results were this, has anyone tried anything similar, what did you find?", etc..
Regarding the purpose of AI as it pertains to computer projects -
idea: the AI doesn't process the projects itself - it assists heavily in writing complicated code and then computers execute that code.
Could or does AI execute its own code that it has generated?
I don't know. My first guess is no. My second is rarely.
The answer is yes, Stockfish evaluation functions is 100% generated by the AI. The only thing it was taught was the rules of chess.
That is why Alphazero, and Leela Chess Zero. Are named Zero. Everything was machined learned, and generated by the self teaching AI. (Stockfish uses Lc0 AI learned Evaluation)
"Mathematical rules can be abstracted from chess positions."
But these mathematical rules, either generated by a AI, or code written by a human. Can not solve chess.
That is why all chess engines either being AI generated, or code written by a human. Still have to do a search of the game tree of chess.
Because any evaluation code is just an approximation of the true evaluation. And a search gets you closer to the true evaluation of the position up to the effective search horizon.
If we had a code that could do what these people suggest. No search of the game tree would ever be needed.
But here is the problem. Chess is a 100% tactical game. And can not be solved with just some mathematical rules. That are just an approximation, or also know as a guess. Of the real true evaluation of the position.
And there are only 3 true evaluations of any chess position. White wins, Black wins, or draw with perfect play.
Does that 'yes' answer mean that Stockfish is 'AI'?
AI in the general sense has been around for a long time.
One could say any operating computer and its software is 'AI'.
So Stockfish software inside a computer is 'AI' in that sense.
But what about AI in the more recent sense?
Things like chatgpt and copilot?
Stockfish is in that group?
You can have a conversation with Stockfish like you would with those AI's?
You can verbally tell Stockfish to set up programs?
I guess I'll see the reply next time.
Yes, Stockfish is a AI chess engine. As already stated.
Stockfish play was learned by the computer itself, without human intervention. And it does talk, and the chess speaks for itself.
And BTW, ChatGPT, and copilot sucks at chess. And can not play real chess.
In my opinion, anyone who talks with playerafar has no judgement and isn't worth talking to. Is it because there's no-one better to talk to, so you talk to a 60 year old or whatever, with an emotional age of 8 and a mental age of 13? Oh well, obviously it's up to you and it's your problem.
I am good with that. As I do not want to talk with anyone who thinks it is a good idea to taunt a fake IQ claim.
Bro This IS TOO MUCH STOP BRAINROTTING PEOPLE BRO
I'm not entirely sure what Dubrovnik's talking about either. All I know is that he's deliberately helping a troll to feel self-important and is encouraging him. He would much prefer to talk to people who haven't a clue than to people who do have a clue. And there are not many of those here. Hence, I don't see any possible profit from talking here any further. Just wanted him to know that.
I am glad to see Optimissed finding new friends that he can agree and get along with, and whose arguments he can relate to.
what the !@#$%^are you people talking about - please speak clearly on point in a forum debate about whatever it is you are talking about-- hate and trolls. and troll haters seem to go hand and hand here now
Due to appallingly weak moderation by Chess.com, which has been going on for many years. I know personally very many people who have left C.c due to their incompetence.
Those who have left entirely or comment no more in the forums, due to online trolling, to my knowledge include people in the medical profession, at least one nationally known U.S. politician, a Greek chess I.M. called "Pfren" ( can't spell his real name) who didn't get on with me at first although we became good friends. Many more too, including a lot of good people who don't like having their respectfully voiced opinions closed down by people such as we have to put up with in this thread.
By very many, you mean a handful of posters that are not really gone because they always pop back up like bad pennies. Case in point, somebody talking about daffodil harvests today on another thread.
You're a bit gullible when your ego is being catered to. There is no "nationally known US politician" among your friends here. They claimed to be a US ambassador appointed by Eisenhower. The reality is quite a bit more mundane:
https://culturalvistas.org/our-programs/professional-study-tours/citizen-ambassadors-program
The Citizen Ambassador Program was first established in 1956 by the U.S. State Department during the Eisenhower Administration. Its purpose was to promote exchanges between private citizens of different cultures, allowing them to gain a better understanding of each other and their professions through the sharing of ideas and experiences.
So, essentially, the business equivalent of a foreign exchange student. An "ambassador" in name only, not appointed by Eisenhower but merely participating in a basic program with very low barrier to entry, established during the Eisenhower administration.
Pfren in all his years here never showed a single sign of your narrative and was definitely not the type of person that liked people associating themselves with him purely for gaining extra credibility. Of course, we can't ask him now, and he won't show up to disabuse you here, and that's probably the key to your chosen reference in this case.
Don't know which doctor you are referring to, but chances are good that is also bunk.
Optimissed is either just arguing for the sake of it or doesn't get what "perfect information" actually means. In chess, both players can see everything on the board at all times (as well as the prior moves), so there’s nothing hidden—unlike games like poker where you don’t know your opponent’s cards. Just because chess is insanely complex and no one can fully calculate it doesn’t mean it isn’t a perfect information game. If he actually knows this but still argues the opposite, then he’s either messing around or trying to sound deep for no reason. Same with the AI search claim—AI changes how results show up, but it obviously doesn’t make searches impossible. If he’s as smart as he says, he knows that’s not true, which makes it seem like he just likes stirring things up.
You catch on quickly Alyra.
Opto's manners are terrible and then he complains about the results he gets.
"he’s either messing around or trying to sound deep"
Try both. Some of the kids believe him.
And some of the people who like to spout disinformation play along with him.
You've probably jaundiced her view.
...
The very minor relevance to this discussion is that chess is a game of perfect information in the strongest sense, and all theorems about such games apply. ,,,
You don't say what you mean by "the strongest sense". All previous actions known to all players at any stage? That would apply to flip flop flue bang (aka rock-paper-scissors I think).
I agree the relevance is very minor. I don't see that perfect information in any sense is a prerequisite for solvability in any sense. For example with a suitably amended set of chess rules both chess and Kriegspiel would be solved in each of the senses defined in the Wikipaedia article, though Kriegspiel would not match any of the definitions of a game of perfect information.
(Although the games would be solved according to the definitions, it is possible that they would not be considered solved (by person or persons unknown) or it would be unknown whether they would be considered by the same to be solved according to the later paragraph:
Given the rules of any two-person game with a finite number of positions, one can always trivially construct a minimax algorithm that would exhaustively traverse the game tree. However, since for many non-trivial games such an algorithm would require an infeasible amount of time to generate a move in a given position, a game is not considered to be solved weakly or strongly unless the algorithm can be run by existing hardware in a reasonable time. ...
That would depend on what is meant by "a reasonable time" (how long is a piece of string?) and in the case of (amended) chess at least the lack of knowledge of how long a minimax algorithm would actually take translates to an uncertainty as to whether the relevant persons consider the game to be weakly or strongly solved. They would apparently accept it as an ultra-weak solution.)
what's the point in running in baseball, your quest begins and ends on the same base
Don't know, we haven't got around to solve that yet. Still doing chess...
You seem to be quite a good chess player but you're only 16 so you have no understanding of much. Keep up the good work though. We're on your side. After all, you have to live in this new world longer than me!
Ah, thank you, Optimissed, for bestowing upon me the knowledge that my 16-year-old brain is simply incapable of understanding much. Clearly, my ability to play chess well is just a happy accident, and any thoughts I have about AI, search engines, or, well, anything must be immediately disregarded.
But I appreciate your support—especially since you’ve already accepted that the world is doomed, but hey, I have to live in it, so best of luck to me! (Lol) Truly inspiring words of encouragement.
At 16 Alyra is already demonstrating clearly more understanding than Opto is.
And very quickly too. WIth Opto complaining that he was talked back to.
But its mostly phony behaviour by Opto. That's the part that's usually missed.
Regarding 'perfect information' in formal game theory the application of the term is not hard to grasp in the context of that game theory where that term is used.
A simple comparison with poker will suffice.
But the term 'perfect information' is not a good term for general discussion purposes.
As the recent round of discussion proves.
--------------------
I've mentioned this before about when jargon is used.
And 'jargon' that uses common everyday words with broad meanings becomes worse. Being applied to special contextes.
'Game theoretic value'. 'Weak solution'. (is worse)
If you've got a position where one player has just forced a won position for himself and he will win if he doesn't make a big enough mistake from there -
then that's called 'a won position'.
In some cases its 'a book win'.
In some cases its a checkmate forced in a particular number of moves which in a way is both a minimum and a maximum.
In some cases the winning player will take more moves to checkmate than that particular number of moves.
If its still a forced mate then its still 'strong'.
The term 'weak solution' for all these various kinds of solved positions is a very poor construction of jargon.
People defending the jargon or labels will say 'Hey its recognized terminology within game theory. Go look it up if you don't understand it.'
Repeat - 'weak solution' is a very poorly constructed jargon.
A won position is a won position. Its strong. Not weak.
Both the position and the terminology 'won position'.
Stands.
-----------------------
'Perfect information' is possibly even worse.
Its not necessary. They are still won positions.
A won position doesn't need the jargon 'perfect information' at all - to be discussed. The term is inane in the context.
-------------------------
But there's still 'trouble' when that number 5 x 10^44 is discussed.
The number of possible chess positions formulated by John Tromp in the year 2000.
I got that date from AI just now - not google search. Chatgpt specifically.
Its much easier and faster and better with the AI. Many types of questions.
(Perhaps a particular person (not I) will now again complain bitterly about the mention of AI. He will want to personalize that too.)
----------------------------------
Here's the 'trouble' with that huge number of 5 x 10^44 possible positions ...
Say by simply stopping the solving of any position once it becomes a forced win ...
that would reduce the number of positions to be solved to 1 five-thousandth of the number from John Tromp.
That looks like a very big cutdown. Right?
'Hey the computer project gets to skip about 4999 positions out of every 5000 ?? That's alot'
No it isn't.
Because then you've still got a number that's just too big.
10^41 positions to be solved.
That's still 100,000 trillion trillion trillion positions.
To be solved.
Say the project computers could solve a trillion positions per second!
(they can't - nowhere near that fast)
Then to solve all those positions you're still looking at
31 million trillion years. (yes I used the AI again - I made a mistake that it didn't catch but then I did and it corrected too.)
31 million trillion years - even after cutting the positions needing to be solved to one five-thousandth of their previous amount!
By stopping the processing each time the computer gets to a won position.
Understanding of the nature of the task!
--------------
That has some similiarity to the way humanity goes at the study of the game.
...
The very minor relevance to this discussion is that chess is a game of perfect information in the strongest sense, and all theorems about such games apply. ,,,
You don't say what you mean by "the strongest sense".
If condition A implies condition B, A is stronger than B. To put it another way the set of entities with property A is a subset of those with property B.
Maths is about abstraction, and one aspect of abstraction is weakening conditions (axioms), with the idea that a weaker axiom suffices for some results.
...
The very minor relevance to this discussion is that chess is a game of perfect information in the strongest sense, and all theorems about such games apply. ,,,
You don't say what you mean by "the strongest sense".
If condition A implies condition B, A is stronger than B. To put it another way the set of entities with property A is a subset of those with property B.
Maths is about abstraction, and one aspect of abstraction is weakening conditions (axioms), with the idea that a weaker axiom suffices for some results.
I like the part about the subsets.
'weaker axioms' looks good.
------------------------
'weakening conditions (axioms)' .
Could be qualified some more. That particular phrase.
No complaints. Just commenting.
If 5x posts are made within 30 seconds by an user one would hope some spam alarm should be ringing. Let's measure some mod reaction time.
The time to get chess solved is never
The time to make up your mind about people , on the other hand....
From some AI folks:
""It's not currently possible to solve chess within a practical time frame, and it's unlikely that it ever will be.
Explanation
The complexity of chess is so great that it's not possible for modern computers to solve it.
In 1950, Claude Shannon calculated that it would take 1090 years for a computer to make its first move in chess.
Computer analysis has solved chess for almost every position with up to seven pieces.
However, solving all eight-piece positions is still ongoing.
The complexity of the game and its infinite possibilities make it unlikely...."
Got to make an awfully big dent in that 5 x 10^44 number of positions.
No indication yet of any such dent.
Nor the Shannon number. That's even more hopeless.
...
The very minor relevance to this discussion is that chess is a game of perfect information in the strongest sense, and all theorems about such games apply. ,,,
You don't say what you mean by "the strongest sense".
If condition A implies condition B, A is stronger than B. To put it another way the set of entities with property A is a subset of those with property B.
Maths is about abstraction, and one aspect of abstraction is weakening conditions (axioms), with the idea that a weaker axiom suffices for some results.
Yes, I didn't phrase that very well did I?
What I intended to say was, "You don't say what you mean by " a game of perfect information in the strongest sense"."
The Wikipaedia article I referred to earlier defines perfect information only for sequential games, with the definition
In game theory, a sequential game is defined as a game where one player selects their action before others, and subsequent players are informed of that choice before making their own decisions.
and also states
Games with simultaneous moves are generally not considered games of perfect information.
In the FIDE laws the term "move" can reasonably be taken to mean only one of the moves of the pieces as described in art.3, but that cannot be the meaning in the latter quote, because games in general need not even have pieces. I read the term "move" there to mean one of the actions referred to in the former quote.
So chess under any ruleset I have seen is not a game of perfect information according to the Wikipaedia definitions, because simultaneous actions; resignations, agreed draws or piece moves are allowed. The games are not sequential as defined.
You have stated that chess (of an unspecified flavour) is a game of perfect information in the strongest sense, hence the implied question, "What is your definition of a game of perfect information in the strongest sense"?
I do note that the Wikipaedia article also states
Chess is an example of a game with perfect information, as each player can see all the pieces on the board at all times ...
but they're obviously confused. They have presumably failed to realise that the games they do describe under the link "Chess" are not sequential.
As I said earlier the topic doesn't seem to be particularly relevant because perfect information is not a necessary attribute of a solvable game.
You make a really interesting point about the definition of perfect information and whether chess truly fits the bill. The whole simultaneous vs. sequential move debate is a tricky one, but I’d still argue that chess qualifies as a game of perfect information in the strongest sense.
I'm assuming by, " perfect information in the strongest sense", you mean conforming with all constraints analogous to those in your next sentence.
At any moment, both players can see the entire board, know every move that’s been played, and don’t have to deal with hidden information or randomness.
Here you are presumably taking "move" to mean action, rather than the FIDE definition, so the players know at any moment whether a draw offer has been made etc.
The statement that they don't have to deal with hidden information or randomness is questionable.
If a player offers a draw under FIDE basic rules chess, then sees he has a mate in one, he doesn't know if his opponent will accept it or not before he finishes playing the mate. That makes the difference between a win or a draw unless his opponent accepts the draw simultaneously with the mate, when it makes the difference between a win or a simultaneous win and draw.
If the players simultaneously resign under chess.com chess then only one will win, but which depends on the state of the electronics between the players and is random.
Sure, things like resignations or draw offers could technically happen at the same time, but they don’t really change the structure of the game itself. Moves still happen in a clear turn based order.
The moves of the pieces happen in a clear turn based order, but the actions affecting possible continuations need not. If you talk of a game of perfect information, the phrase would appear to apply to games other than chess and those actions must also be counted as moves. Also in chess even though it's in conflict with the use of "move" in the FIDE laws.
Even in formal game theory, chess is treated as a sequential game.
As it is in the Wikipaedia article, but that is in error.
It does however at least specify the rules of the games it is talking about (it says they're FIDE rules but differs in significant ways).
I've seen many game theory articles that talk about chess, but no others that actually say what the rules are of the game they talk about. E.g. Tromp talks about just the number of legal chess positions and apparently assumes FIDE basic rules since 2017, but I don't believe he says that anywhere. The number would be much higher pre 2017 or with the 50/75M and 3/5R rules in effect assuming positions with different possible continuations are not counted as the same.
In practical play, just because two players might agree to a draw at the same time doesn’t mean there's any hidden information or uncertainty affecting decision making. The Wikipedia article’s claim that games with simultaneous moves don’t count as perfect information seems to be aimed more at situations like rock-paper-scissors, where you don’t know what the other person is picking in real time. That’s not the case with chess. Even if two players happen to make an action at the same time, the game state is still fully visible and unchanged.
The game state will not normally be unchanged.
It will be fully visible exactly as it is in rock-paper-scissors ("flip flop flu bang" when I learnt it). All previous actions (i.e. none if it's a single shot game - mine continued until a win) are also visible in that game.
I also totally agree with you that perfect information isn’t a requirement for a game to be solvable, though it does make it easier in some cases. Chess’s solvability has way more to do with the sheer number of possible positions than whether it meets a strict definition of perfect information.
I would agree that perfect information has little to do with it.
Whether chess is or is not solvable depends on your definition of "solvable" and the version of chess. There are also multiple incompatible versions of the different flavours of "solvable" in game theory.
Because the outcome of some simultaneous events gives results under the FIDE laws that do not correspond with the win/draw/loss that is normally assumed in competitive two player games (e.g. simultaneous resignations result in a win for both players in 2018 or (usually) a loss for both players in 2023, a draw agreement simultaneous with a mate move gives a draw where one player wins ... there are of course more) an ordered reciprocal set of payoffs for each player at game termination is not defined. That is FIDE chess of any flavour is not a zero sum game. So far as I know, there is no definition of "solved" in game theory literature that applies to such a game, so I think it can be firmly asserted that FIDE chess according to any of their sets of laws going back a long time will never be solved according to any current definitions of the term.
The situation is different in the case of computer chess. The rules are embodied in programs and all events will normally be serialised. The payoffs may be only win/draw/loss in reciprocal order and the arbiter actions if any will usually be fixed and clocks may not be implemented. In that case, so long as the program actions are fully known (e.g. accurate documentation) the the game is usually already solved by the minimax algorithm in all senses according to the definitions in Wikipaedia but may not be considered solved according to the later caveat
Given the rules of any two-person game with a finite number of positions, one can always trivially construct a minimax algorithm that would exhaustively traverse the game tree. However, since for many non-trivial games such an algorithm would require an infeasible amount of time to generate a move in a given position, a game is not considered to be solved weakly or strongly unless the algorithm can be run by existing hardware in a reasonable time. Many algorithms rely on a huge pre-generated database and are effectively nothing more.
(Make of that what you will. It obviously depends on how long is reasonable, but whether it is so considered apparently depends on the actual time it would take to complete, which is currently unknown for any game that would be recognised as chess (it could fall within what is taken to be reasonable). That suggests that the relevant group don't know whether they consider it solved. It presumably should read, "... unless the algorithm has been run by something and the results are accessible in a reasonable time". But at any rate if you have a wonderful proof of a strategy for winning that doesn't involve running the minimax algorithm game theorists will not consider the game solved, at least not if they're Wikipaedia editors.)
And in tournament chess. Draw offers can not be made at the same time.
Yes they can, just so long as one or both of the players has said "hail Mary" and given solemn consideration to art.11.5.
A player wishing to offer a draw shall do so after having made a move on the chessboard and before pressing his/her clock. An offer at any other time during play is still valid but Article 11.5 must be considered. No conditions can be attached to the offer. In both cases the offer cannot be withdrawn and remains valid until the opponent accepts it, rejects it orally, rejects it by touching a piece with the intention of moving or capturing it, or the game is concluded in some other way.
In chess, you can offer a draw only when it is your turn to move before making your move. According to the FIDE Laws of Chess (Article ), the correct procedure is to offer the draw before making your move and then complete your move on the board. The opponent can then either accept, decline, or ignore the offer. If they continue playing, the offer is considered declined.
That is the correct procedure, but the laws allow you to be incorrect with some minor precautions.
Edit: Just noticed I forgot to undo the chess.com editor's vandalism in my least paragraph. Should make more sense now. Also added a some paras regarding the solvability of chess.
This post will never be solved because its like politics and religion.
look at how many responses and its still the same rhetoric and worthless conjecture.
I concur