Chess will never be solved, here's why

Here's a peach of muddled thinking I missed earlier, but found looking for the link for the above post!

So, how exactly do you determine a move by the opponent to be "non-viable" (as opposed to a deep brilliancy, say) WITHOUT ANALYSING IT?

I'm not 'sure'.  
Elroch said something about top GM's struggling against computers.
I then tried to research it and found that Nakamura could not beat the computers even with white and spotted a pawn - and did in fact lose one of those games.
But that was Seven Years ago !  
So I am therefore thinking that Elroch is correct on this.
Apparently he has done some other research on these kinds of matches happening since then.
The computers and chess software are stronger now than then.
However much knight odds might be a 'won position' that doesn't mean that GM's spotted that knight advantage are going to beat today's engines.  And the engines aren't getting 'weaker'.  They're going in the other directon.
But would GM's be willing to play such matches?
Have they done so?
Could be Embarassing. 

'Instructions'.  Imaginary authority.  Figures.  Consistent with the pattern.

Another random nonsense "The strong solution becomes u;tra-weak [sic]"

No. An ultra-weak solution involves means determining the value of a game without exhibiting an optimal strategy. A strong solution is the opposite extreme - indicating optimal play in every legal position. Neither does a weak solution (exhibiting optimal strategies) ever turn into an ultra-weak solution (not exhibiting optimal strategies).

I am not sure why you come out with these things, @Optimissed.

At this point, know one is at the argument, we just don't follow the math and the trigonometry (Pronounced= chicken-on-a-tree) That, was a meme.

Alright, let me get to my point, the 10% of active brain cannot acknowledge and cannot come to the point of understand  the point and are of a king. I am in 3rd grade don't judge me.

again the pattern - 'hard guy's' total failure to follow his own advice
and we got a 'strawman' too - 'people's right to express themselves' that he confused with his 'instructions' and his delusion that he 'thinks faster and more accurately'.
Plus he finally got a supporter (fortunately not the forum's opening poster)  and he now wants to 'pronounce victory' immediately and is now filled with much hope.  

In situations like this - there's often an issue:
Does the person believe his own nonsense?
'psychotic'?  use of that term ... projecting again ?
His feelings of 'superiority' are remarkable ...
suggesting that the three paranoias are worth mentioning ...
paranoid schizophrenia (probably not) - classical paranoia (a rarer thing but probably not the case here)
and paranoid personality disorder.
'delusions less common in this disorder but when present are transient and not well organized'
Bingo !   Fits his pattern.  Consistently.  

Can all of this be 'reconciled' with the forum topic ?
Maybe - maybe not.
If the 'translation' of the forum title is something like:
"you really do have time for this topic ?  Why?  How?"

Issue: can progress be made with such discussion ?
Some. 
Overview of the game.   
Relationships of chess with modern technology.
Individual views of same. 
But not that much progress on those.
But some progress made on 'exposing' too.

The forum subject is partly about 'perceptions' of 'solved' as they might pertain to chess.
And perceptions are more about psychology than they are about maths and physical sciences.  (except for brain physiology perhaps?  No because the human brain is not 'solved' either.)
So - the psychology of perception is relevant here.  
(constant occurrence - when various psychologies are mentioned - often the reaction will be the behaviours mentioned instead of discussing the behaviours. )

#1072
There is a massive difference between a weak and a strong solution.

A strong solution i.e. a 32-men table base would require to visit 10^36 positions, that would take 10^27 seconds = 3 billion billion years on a 10^9 nodes/s cloud engine and require at least 10^36 bits of storage draw / no draw = 10^12 yottabyte. That is not feasible.

A weak solution can toss away huge numbers of positions that are nor reachable and are thus not relevant because each pawn move and each capture are irreversible. That leaves about 10^18 relevant positions.

A weak solution can also prune away the openings to look at by a factor 10. If 1 e4 e5 draws then for a weak solution it is not necessary to establish if 1 c5 draws as well or not.
That leaves 10^17 positions, doable on one 10^9 nodes/second cloud engine in 10^8 seconds = 3 years.

The evaluation draw/win/loss does not come from the evaluation function, but from the 7-men table base once it is hit or from a previous draw by repetition of positions or stalemate.

As calculated 3 verification passes retracting all white moves lead to less than 1 error per 10^20 positions so that is enough to prove for the 10^17 positions.

10^17 positions are reached from 28 moves deep at 4 branches per move. That is enough too: as we see from ICCF or human grandmaster play games do not last more than 28 moves beyond the point where they deviate from their predecessor to reaching a 7-men table base.

"3 billion billion years on a 10^9 nodes/s cloud engine and require at least 10^36 bits of storage draw / no draw = 10^12 yottabyte. That is not feasible."
Now you're talking.
Three billion billion years on a 'Cloud' computer.

And I like that term 'yottabyte'. 
It could be the name of a new ice cream sandwich at fast food places.  

Three billion billion years - that's more years than the number of fatal plutonium doses in the bloodstream from one gram of plutonium?
Plutonium is the number one poison apparently.  In theory technically.
I read somewhere that it only takes 10 to the minus 12 of a gram of plutonium  in the blood to kill a person.
Comparing 10 to the 12th power - with three billion billion years ...
that's 3 x 10 to the 18th power.  Still less than the Avogadro number. 
But that's 'enough' years.  
Just 10 to the 18th.  Years.  Without the three.  Hey that might be more than enough time to travel to the Andromeda galaxy.  Or even to our nearest other star.  Alpha Centauri.  By conventional travel.  Also known as Betelgeuse and Rigel I believe.

I am not sure if that was because you didn't know that there was already a key concept called "ultra-weak solution", or that you are ignored the requirement not to redefine terms that already have an accepted definition in a technical discussion (because doing so would obviously cause confusion every time the term was used). Also you didn't "redefine" the term anywhere anywhere other people can see.

If you have a new concept (this is not clear) you need a new term for it.

Where in that link exactly?

I looked through the Wiki page a bit.
Work on 8 pieces onboard began in 2021.
And the article reminded me of something I had found out on my own about 30 years ago -
that there are ten types of squares on the board. 
But that doesn't work for pawns though.  Nor for castling.
 
The ten square types can be defined by the lengths of their diagonals.
I ran across this while I was determining how many kinds of knight move there are on the board.  21 in fact.  
But the article did seem to say that the tablebases take en passant into account but they seem to skirt around castling.
https://en.wikipedia.org/wiki/Endgame_tablebase]]

Yes, there are 10 equivalence classes of squares under the action of the 8-fold symmetry group of the chessboard (ignoring the directionality that only affects pawns and castling).

4 of them have 4 squares (the corners of the 4 nested squares of width 8, 6, 4 and 2), the other 6 have 8 squares (the sides of the nested squares)

Hi @Elroch !
Yes - 'nested' squares.  As it happens there are a total of 204 large and basic and 'nested' squares on the board.
But unlike the ten square types - I doubt the 204 figure could be of any usefulness in helping players visualize the board and pieces and piece motions and controls of squares.

Regarding the ten square types - there's different ways to define them.
Does it have any application to chessplaying?
Yes. Helping to mentally visualize the board.

Four types of squares on the two long diagonals
Three more square types of the ten on the four 7-square diagonals adjacent to the long diagonals.
(they and the long diagonals share the four central squares) 
two more square types on the four bishop 'original' six-square diagonals (three altogether - they again share. They share the eight squares like d3 and c5 
(which are the only 8 squares where a knight can move directly between the same square types.)
And finally - one more and final type of the ten on the Royal squares and their corresponding squares on the wings.  

Each of the ten sets of squares has a uniqueness that helps to visualize them on the board.
The eight squares like c2 and b3 I call the 'catseye' squares.
The only squares that are not on an edge and also not on a central file nor central rank nor central diagonal. 
Can be compared to the Union Jack flag !
Those 8 squares are a knight's only route to or from the corners too.

Weakly solving chess = produce at least one complete ideal game (all moves start to end) with proof that each move is optimal for the player making it.

Consider the following game

It ends in a draw. It is a candidate ideal game. Proof that each move is optimal consists in retracting all of white moves starting from move 36 backwards and determine that alternative white moves produce nothing more than a draw either. 3 verification runs of retractions suffice.

I think its invalid right away.
How are you going to prove that e5 for black and Nf3 for white are 'optimal' ?
Or 1) e4 either.
There are arguments that 1) e4 is sub-optimal for white.
1) d4 is solider.  The e4 move loses the option to protect the diagonal leading into f2 - by e3.  Plus the e4 pawn is unprotected and can become a target.  It is also not available to provide control of d4.
Also - a pawn at d5 can become more powerful while less vulnerable than a pawn at e5.  
And - if memory serves - 1) Nf3 has the lowest loss percentage.  Yes that might have changed.  I read it somewhere and I don't have a link for it.

Yes: only 26% black wins against 1. Nf3

Only move that has done better is 1. a4 (but in only 34 master games, and not significant)