Chess will never be solved, here's why

Elroch wrote: Not only "could", pretty sure that has been done. I think all of Euclid has been mechanised. Here is a paper about mechanical verification of book I.

Oh, please

Which is what my post suggests.  Something else.
And 'wasn't' is past tense.

Hi !
Who said 'rely' ?
By the way - welcome to chess.com. 
You having joined the site yesterday.

Regarding the 50 move rule - it can be considered an adjunct to chess -
like the issue of what is 'mating material' when somebody's flag falls in a sudden death game.
If the side whose flag didn't fall has an apawn or hpawn in a dead drawn position - then that side wins.  For example.

In a situation where the 50 moves is counting - the exact number of moves left dictates strategy.
Especially for the defending side.
The rule could determine whether the position is a win or a draw anyway though.  With or without strategy.
But that should be Skipped by the initial computer bank and Skipped as initial goal.
It should be skipped as Secondary goal too.
Should be low on the priority list.  Lowest?  Not necessarily.

///////////////////////////////

In a correct way to go at the hypothetical project -
I say - mating material and 50 move rule and repetitions of position are skipped.  (a possibility of a repetition of position  - if there's already been a 2-fold repeat - could again determine strategy.  But could also determine whether its a draw or win because a defending side could perhaps force a final repetition)
But that should also be skipped in the initial stages too.
(nobody else in these forums has confirmed yet that he/she understands why.)

In other words - such a project would be done in Stages.
Multiple computers or banks of computers assigned.
So the first computer bank's job is to generate positions - not games.
Arrangements of pieces which do provide for either side to move.
(Otherwise they'd be the so-called 'diagrams'.  Not good enough.)

Does that mean there's already a legality issue?
In some positions - it would be impossible for either side to move.
For example white's g-knight at f3.  Every other piece at its home square.
Would it be possible for it to be white's move? I don't know.
Is there a sequence of moves that would bring that about ?
Should these issues of 'could it have got there legally?' be left out in the initial stages of the project ?
That's harder. 
It depends on how positions to be solved are to be generated.

Somehow - simplicity has to be maintained - but without the ridiculous premises being pushed over and over again by @tygxc .
That's not personal.  Simply addresses his posts.

Approach - define the initial goal for the initial computer bank.
Assign other relevant tasks to backup computer banks.

Suggestion for that initial goal:
A computer bank proceeds with position generation.
It then classifies every position it generates into categories and then refers each category to the appropriate secondary computer bank.
Of which there might be several.
And like the issue of whose move it is - castling and en passant should be provided for but not in a way that confounds and cripples the initial goals.
Other computer banks get those positions -
further down the Goals priority list.

Many ways to set up those Categories.
Part of solving - classifications of positions.
Links up Hugely with actual chessplaying.

But both the position generation and the classifications would be designed to fit with and greatly enhance the efficiency and effectiveness of the different computer algorithms assigned to each category.
That would be Top Priority.

Without such stages - the whole project degenerates.

But @tygxc is probably not. He has only to state something to consider it proven.

I don't know about AZ, but I just downloaded SF15 and it seems that you can't even know when it is certain. 

I wanted to see at what level it became out of it's depth in KNNKP so I let it play a mate in 36 position and it drew by repetition after 14 moves. The following image shows SF15 announcing mate in 38 (full) moves from the initial position while simultaneously showing a search depth on the same line of only 33 half moves! It lasted for a few seconds then it changed its mind to 0.80 and after another second or so to 0.00.

Incidentally we seem to again show minimax pathology at work.

I first ran the kibbitzer until the top line showed depth 48 and captured the top line, which is inserted into the game log. Notice this follows the game until Black's move 3 when it contains 3...Kg6 instead of 3...Kh5 which I played. The kibbitzed move would slip 3 moves on the move I played which is accurate. However in the later version (bottom) which has reached only depth 33 it is showing 3...Kh5.  

"But @tygxc is probably not. He has only to state something to consider it proven."
That's correct.  And then he repeats it over and over.  As if repetition is proof.
And again - without being personal - its similiar to the positions of flat-earthers.
They simply make assertions. And therefore/thereby consider them correct.
In tygxc's case he'll mention Sveshnikov or den Herik and also then take a position that mentioning their names is additional 'proof'.
The flat-earthers are on the rise in numbers.
Especially among millenials.
There may be a corresponding and growing cult that believes that anything can be done with computers.

Hi all

Please join this club

https://www.chess.com/club/wizarding-chess-1

#3071
"He has only to state something to consider it proven."
++ I am the only one who provides proofs and references.
Some people do not understand the proofs and then deny it being a proof.
All other people state something without any proof at all.
Chess is already strongly solved for 7 men or less.
There is no point in rehashing 5 men positions.

"AlphaZero outputs probabilities." ++ AlphaZero autoplay yields results.

"Capablanca is less reliable than AlphaZero."
++ Capablanca was human and provided human insight

You repeated this many times, but it's plainly false. Several posters have proven or disproven concepts and provided references. You did not prove your ideas because the "proofs" are flawed.

To start, you use a supposedly known but not proven game-theoretic value of the game (a draw), to "deduce" (after other fallacies and mistakes) an error rate per move. I asked you to give a reference of a scientific work where the authors claim they know something which is not either proven, or at least never disproven by a history of experiments. You provided only a statement: "provability is a higher degree of truth", which is unreferenced, out of the context, ambiguous (if not completely wrong) and unfocused (it is about provability and truth, not about proofs and knowledge). This is a very fragile premise to start with.

#3075

"You repeated this many times, but it's plainly false." ++ It is not. That is what you say. Prove.

"Several posters have proven or disproven concepts and provided references."
++ Like what? So far none have proven or disproven anything let alone provided references.
Most exclaim like you "false", "flawed", "fallacies", "mistakes", "fragile" etc. and that is it.

"You did not prove your ideas because the "proofs" are flawed." ++ They are not. That is what you say. So I make a statement, provide proof and / or reference and you can just say it is flawed

"To start, you use a supposedly known but not proven game-theoretic value of the game (a draw)" ++ I even provided evidence for that.
A) Expert opinions of world champions
B) AlphaZero autoplay with reference paper high draw rate and more with more time and even if stalemate is a win
C) ICCF WC results: high draw rate, rising over the years, even with 7 men win claims allowed > 50 moves without capture or pawn move
D) TCEC results high draw rate even with imposed slightly unbalanced openings
E) Human world championship matches based on cloud engine preparation

"to "deduce" (after other fallacies and mistakes - ++ that is what you say, proof?) an error rate per move. Yes, that is right from AlphaZero autoplay follows an error rate per move
I thus have proven that 4 white candidate moves suffice at 17 s/move for 1 error in 10^20 positions
That is consistent with opening knowledge like the reference from AlphaZero I provided and also consistent with Carlsen saying in an interview he considers 3 candidate moves.

"I asked you to give a reference of a scientific work where the authors claim they know something which is not either proven, or at least never disproven by a history of experiments. You provided only a statement: "provability is a higher degree of truth", which is unreferenced" ++ I told you that is from a feature article in Scientific American around 1977. I do not have that issue any more.

"This is a very fragile premise to start with." ++ That is not a starting premise.
Starting premises are:
The count of legal positions by Tromp 10^44.
I proved none of the sampled positions are sensible i.e. can occur in a game with > 50% accuracy. Mainly because of the multiple underpromotions to pieces not previously captured, each such underpromotion being a blunder.
I presented the more relevant Gourion count of 10^37 and I pointed out that even the vast majority of these are not sensible, thus arriving at an estimate of 10^32 sensible positions.
I pointed out that the vast majority of these positions is not reachable in the course of the solving process and made reference to solving Losing Chess and Checkers. Thus I estimate 10^19 reachable positions.
I pointed out that of the reachable positions the majority is not relevant for weakly solving chess.
I gave examples of non relevant positions and pointed to a reference indicating that using chess knowledge is allowed in weakly solving chess.
I thus arrived at an estimate of 10^17 legal, sensible, reachable, and relevant positions.
I then pointed out that from this follows that 3 cloud engines can weakly solve chess in 5 years, like Sveshnikov predicted.
I also analysed ICCF WC results and pointed out 99% of the drawn games are already ideal games with optimal moves from both sides, i.e. perfect play. I thus pointed out we already have over 1000 ideal games with optimal moves i.e. perfect play.
I also showed that ICCF WC games end on average in 39 moves.
I also showed that the 50-moves rule is never invoked and thus can be disregarded.

Nobody has responded yet to the point that if you know three pieces checking a King at once is illegal - 
you don't need 'retrograde analysis' to know that four or more pieces so checking is illegal too.
A kind of 'loud silence' in response to that one.
4 pieces checking - is 'illegal on the board'.  No need to computer it to show it 'couldn't have got there.
Illegal on the board versus 'couldn't have got there'.
'Response' wouldn't include spamming about Sveshnikov and phony computer speed figures and cosmetic reductions in the number of digits in numbers of positions to be solved though.
Is there a 'real discussion' in the posts ?
Yes I think it exists. 
Is it like 'Cherenkov radiation' and it has to be Discovered ? 
 
 

For example, I have provided proof and reference that about 10¹⁴ nodes have been searched to solve checkers, not 10⁷ like you repeated for many posts; I have also provided a paper that states clearly how the authors interpret van den Herik definition of "weak solution", while you didn't; MARattigan has showed the pathology in Stockfish. But I have not to convince you, only those who are interested in the topic, who are invited to read from the beginning.

Your "evidence" is not a proof, otherwise chess would be already ultra-weakly solved. Just your word that you read the statement "provability is a higher degree of truth" on Scientific American proves nothing, sorry.

All your idea to cut off lines at early stage of the search is based on the supposed error rate per move, which in turn is "deduced" from the supposed game value. The real "accuracy" (not that computed by engines or by you) will be known (i.e. by proof) only after the solution. If it was really known, the game would be already weakly solved. Again, you use "known" but not proven things to prove other things, so all your idea of "solution" is based on that unreferenced an ambiguous statement.

@haiaku
But I have not to convince you, only those who are interested in the topic, who are invited to read from the beginning.
Regardless of 'convince' - that post is good - and numerous other good posts you've made.
The real discussion does happen some of the time.
It is not undiscovered 'Cherenkov radiation'.
😁^✅ 👋 🎵🎸

Well, of course some proofs may be not as convincing as I think, but it's undeniable that not only @tygxc have provided references, as everyone can see just reading previous posts. @tygxc thinks (and hopes) nobody ever read previous posts, so he repeats that no one has been able to find real problems in his ideas.

#3079

"I have provided proof and reference that about 10¹⁴ nodes have been searched to solve checkers, not 10⁷ like you repeated for many posts"
++ I originally wrote 10^9 as written in the first paper of Schaeffer. Indeed, his later paper said 10^7 nodes in the tree with each 10^7 nodes.

"I have also provided a paper that states clearly how the authors interpret van den Herik definition of "weak solution", while you didn't"
++ It is an interpretation. I still think 'any opposition' implies an act of opposing.

"MARattigan has showed the pathology in Stockfish."
++ He showed a pathology in KNN vs. KP which is a 5-men position known to be troublesome to engines and strongly solved. I invited him to provide an example for a relevant 7-men position like KRPP vs. KRP, but he has not found one so far.

"Your "evidence" is not a proof, otherwise chess would be already ultra-weakly solved."
++ personally I consider chess already ultra-weakly solved on the basis of the evidence I provided

"Just your word that you read the statement "provability is a higher degree of truth" on Scientific American proves nothing, sorry." ++ That is not my problem. I did not make it up and I read it there and then. If you do not believe that, then that is your problem.

"The real "accuracy" (not that computed by engines or by you) will be known (i.e. by proof) only after the solution."
++ Yes, that is true. It will all be known after it is done. The question now is if it is feasible. Likewise Schaeffer first wrote about checkers 10^9 nodes needed investigation and he ended up with 10^7 * 10^7

"If it was really known, the game would be already weakly solved."
++ Yes, that is true, but before somebody starts to do it, he needs to know if it is feasible.

"Again, you use "known" but not proven things to prove other things"
++ To weakly solve chess we must know if it is feasible. If we can only prove feasibility by solving chess, then it is chicken or egg. First the feasibility must be established even approximately.

#3081
"@tygxc thinks (and hopes) nobody ever read previous posts"
++ I hope people read, but I think people do not read or do not understand

"so he repeats that no one has been able to find real problems in his ideas."
++ I repeat hoping people read and understand.
The challenge is not to find problems, the challenge is to find solutions.

tygxc's postings remind me of arguments for a flat earth.

10⁷ is the top of the proof tree, not the nodes searched; 10¹⁴ are the nodes searched. You had even conceded that.

The special meaning of "act of opposing" (a limited subset of the legal moves that an opponent can play) that you think is correct, is not shared by game theorists. I have provided one reference specifically about van den Herik definition, because you use that, but other equivalent definitions are interpreted the same way. You, instead, did not provide any scientific paper supporting your personal interpretation (and common dictionaries are generally not sufficient for technical definitions). Using a personal interpretation of a definition is equivalent to use a personal definition. But that was not even the point. You said that only you provide references: that is clearly not true.

Exactly, personally, but no scientist announced that chess has been ultra-weakly solved. What is evident, is that your "evidence" is not a proof.

Again, you said you provided proofs and references. Previously, you said that we can use "known" but unproven things to prove something else and you based your claim on that statement, which is unreferenced; therefore it was a problem in your idea, not mine. But now you say that you think chess is ultra-weakly solved, so you don't even need that statement, right?

The feasibility of a weak solution is already proven. You want to prove the feasibility of your particular project, which is to get the solution in five years, but the problem is that your solution would be invalid. That's because your algorithm is supposed to cutoff lines on the basis of the "calculated" error rate per move, which is in turn based on the game-theoretic value, which is not proven to be a draw. If the solution "proved" that the game is indeed a draw, that "solution" would be a circular process: the game value is a draw, therefore the error rate is that, therefore those lines can be cut off, therefore... the game value is a draw. That is not acceptable as a proof, it's a fallacy. On the other hand, if the solution proved that the game value is not a draw, it would contradict the premise, and therefore the whole process would be inconsistent. Your "solution" might be feasible, but not valid, anyway. That's why no one attempted such a thing, it's not just a matter of money.

Nobody wasted money on false logic projects based on Sveshnikov saying something.
Figures.  Anybody who has that kind of money for big projects knows how to manage it much better than that.
That's why they've got it.
And so with @tygxc doing the pushing with the false logic -
well this isn't personal but refers to his posts -
that money is Safe !!

#3085

"10⁷ is the top of the proof tree, not the nodes searched; 10¹⁴ are the nodes searched."
++ Yes, I agree with that. My previous figure of 10^9 was from the elder paper by Schaeffer.

"The special meaning of "act of opposing" (a limited subset of the legal moves that an opponent can play) that you think is correct, is not shared by game theorists."
++ Van den Herik was/is one of the leading game theorists and approved using game knowledge to weakly solve a game as beneficial in his paper.

"known but unproven things"
If something is proven then it is true.
If something is not (yet) proven, then that does not mean it is not true.
There are things that are not yet proven but yet believed to be true based on available evidence.
Examples in mathematics are the Riemann Hypothesis and the Goldbach Conjecture and were Fermat's Last Theorem and the Four Color Theorem. Mathematicians work/worked on proving these, not on proving the contrary.

"The feasibility of a weak solution is already proven."
++ No, not formally. Strongly solving chess is possible, but not feasible. 10^44 nanoseconds and 10^44 bits of storage are not feasible with present technology. For weakly solving Sveshnikov said 5 years and I provided evidence that 5 years are plausible indeed.

"That's because your algorithm is supposed to cutoff lines on the basis of the "calculated" error rate per move"
++ Yes, to weakly solve chess it is essential to use game knowledge so as to reduce the number of nodes to be searched. 10^17 nodes is still a formidable number. Losing Chess needed 10^9 nodes, Checkers 10^14 nodes.

"On the other hand, if the solution proved that the game value is not a draw"
++ Nobody right in his mind believes chess to be a forced win. It would contradict all evidence.
A) AlphaZero autoplay with more draws with longer calculation time and even if stalemate is a win
B) ICCF WC with more draws in more recent years despite 7-men table base win claims > 50 moves without capture or pawn move allowed
C) Classical World Championships with nearly all draws in recent years
D) TCEC with high draw rate despite imposed unbalanced openings
E) Checkers (a draw) needing more nodes 10^14 than Losing Chess (a win) 10^9 despite being a simpler game.

"it's not just a matter of money."
++ It is a matter of money. 5 years to rent 3 cloud engines plus the assistants is a lot of money.

As usual, it is difficult to understand whether you do not understand what people write, or you just try to avoid issues.

I did not question the sequence of words van den Herik uses to define a weak solution, but the particular and personal meaning you give to those words. It is almost impossible that you misinterpreted me. As for the rest, some posts ago:

 Your citation is again out of the context. If you read carefully, In that section four methods are presented and they are all general methods, which can be applied to many games, not only chess. Those heuristics do not cut off lines arbitrarily. Alpha-beta pruning cuts off lines too, but when it is used to solve a game, that happens when the value of some positions has already been determined after a TBH (Table Base Hit) or equivalent (3-fold repetition, etc.); in that case and under some conditions, some lines can be indeed cut off, because it is proven that one or the other player would have a better path to choose. You, as usual, overgeneralize the idea.

Yes, but things believed to be true although not proven are conjectures. If they are used as premises, the result is a conjecture too, not a proof, therefore no solution.

We already know how to proceed, the problem is time. You aim to reduce time, but to do that you not only use conjectures to prove something else, which is already illogical; your method would use the conjectured game-theoretic value of chess to ultimately prove the game-theoretic value of chess (which is part of the solution): that is just a fallacious circular reasoning.

This is an argument from incredulity. We have not to believe anything, we have just to find out, but, for what above, your method would give only an approximation of a conjecture, at best.

I did not say it is not a matter of money. I said it is not just a matter of money, if a project to solve chess has not started yet (at least AFAIK).