Chess Will Never Be Solved. Why?

"I should repeat some hoping more to sink it in."
Repeating more - establishes more that it is misleading spam.
"I should repeat some hoping more to sink it in."
Not a new thing.
But it does get attention.  In that particular way - it is effective.
If it were to be compared with flat earth notions -
maybe about par with each other.
So that's the main dynamic - the fact that it gets attention.
A false positive.

there are 20 possible moves in the starting position

#159
"there are 20 possible moves in the starting position"
++ Only 4 of these are relevant: 1 e4, 1 d4, 1 c4, 1 Nf3.

"there are 20 possible moves in the starting position"
For each side.
But then it gets much tougher and more and more mathematically unmanageable.
Whereas upper bounds on the number of possible positions is relatively neat in comparison.
The math on types of positions starts off neater too.

Simply defining positions by the number of pieces on the board.
Could be called numerical classes of positions.   31 of them.
And its still neat and  manageable by computers - within each numerical class as to the subcategories of which types of pieces are on the board including how many of each.

So for two kings or for all 32 pieces - there's only one subcategory.
Itself in each case.
But for three pieces there's ten subcategories depending on what the extra piece is and its color.

For 31 pieces -   not so simple.
31 pieces means there's been a capture.
Which means promotion is enabled in that case.
The pawns can't get past each other unless something has been or is captured.
But computers could handle all those subcategories.
Not solve them.  But they could classify them.
That particular element of chess could be solved and is very likely solved a long time ago.
Why ? How?  Because the numbers and the math are not heavyweight on such classifications.

And yes - somebody will try to push for only 4 starting moves to be relevant and the other 16 irrelevant.
Will he get disciples? supporters? verbal pingpong from same?
Probably.
But that can also be posted around or ignored.

#161
"Simply defining positions by the number of pieces on the board."
++ Yes, this has been calculated:
32 men: 1.89 * 10^33 positions
31 men: 1.71 * 10^34 positions
30 men: 1.64 * 10^35 positions
29 men: 1.53 * 10^36 positions
28 men: 5.46 * 10^36 positions
27 men: 1.05 * 10^37 positions
26 men: 1.08 * 10^37 positions
25 men: 6.14 * 10^36 positions
24 men: 3.19 * 10^36 positions
23 men: 5.66 * 10^35 positions
Table 3, page 8
https://arxiv.org/pdf/2112.09386.pdf 

Interesting that they'd peak at 26 pieces.
Well that particular post isn't spam.

Plenty of games without those moves played by GMs in Informant.

If c4 is relevant - then f4 and b3 and g3 and a whole bunch of reversed openings are relevant too.  All 20 moves are relevant.
But this post by me now could be a mistake.  Its biting at the bait.
"Sir - we believe there's a Rolex watch orbiting Jupiter.
We'd like to know if you're a believer in that or an atheist or agnostic on that."
Correct reply:  "I'm none of the above.  I'm not even going to consider it."

#165
"Plenty of games without those moves played by GMs in Informant."
Yes, but not at the highest levels like in world championship matches or in ICCF correspondence.
Andersen has opened 1 a3.
Basman has opened 1 g4 and 1 h3.
Larsen has opened 1 f4.
Fischer has opened 1 b3 four times and won all 4.
See Figure 5 on page 17
https://arxiv.org/pdf/2111.09259.pdf 

The figure does not support your claim that 16 other first moves are irrelevant. AlphaZero used all of them in training.

If they lack theoretical interest, they do not get published in Informant. The periodical has clear standards.

#168
"The figure does not support your claim that 16 other first moves are irrelevant."
++ The figure supports that 1 e4, 1 d4, 1 c4, and 1 Nf3 are superior over the other 16 moves.
AlphaZero trained itself with no other input but the rules of chess. So it had to look at and dismiss the inferior lines.

"If they lack theoretical interest, they do not get published in Informant."
++ Oh yes, chess is a game and it is often beneficial to deliberately play inferior moves e.g. to avoid opening preparation. However to analyse chess those inferior moves are not relevant.
This game was published in Chess Informant, but its theoretical interest is questionable.
https://www.chessgames.com/perl/chessgame?gid=1068157 

"those inferior moves are not relevant"
And the Pope is a teenager.
It isn't even proven that the alternatives to the four favored moves are inferior.
Stats reporting they win less doesn't prove it.
But that's about as far as the conversation will go. Depth-wise.
Its like talking to Flat Earth believers. 

     This is the difficulty with pruning positions or lines from the analysis. Can we be positive that there are no previously unexplored possibilities there that may contain a surprise. We're humans or present-day computers sufficiently knowledgeable the solution would be imminent but experts have been mistaken before as to the extent of our expertise.

Terrifically important game! Thanks.

Those are not Informant annotations, however. Too bad Keene was allowed to mess with Miles’ annotations. What Tony Miles had to say merits attention.

#172
"Those are not Informant annotations"
++ No, but this game was published in Chess Informant.

Because it is of theoretical interest.

#173

"A lot of moves look like blunders but if there's a faint possibility that they do in fact contain a forced win or forced equality, beyond the engine's search horizon, then it would be necessary to explore them."
++ That is why pruning is the job of the human good assistants.

"The algorithms in present search engines aren't capable of distingushing "normal blunders" from hidden brilliancies IN ALL CASES and so they can't possibly cope with such admittedly hypothetical lines we're discussing."
++ That is why the good assistants should do the pruning.
All present and future engine evaluation functions are inherently flawed.

"They're hypothetical in the sense that we don't know WHICH lines they may be and that's very much the point."
++ There is knowledge available for pruning. We know some endgames with opposite colored bishops or rook endings with pawns on one wing are draws even with > 7 men and not in the 7-men endgame table base. There is no need to calculate these further. We know from logic and from AlphaZero that 1 e4, 1 d4, 1 c4, and 1 Nf3 are superior the the 16 other possibilities. That is why it is not necessary to calculate the 16 other 1st moves. We know that 1 e4 e5 2 Ba6? is a loss for white, no need to calculate that until checkmate.

"The result is that the search has to be much wider than many, in fact, have assumed in this thread, to be sure that any solution for chess is a full one."
++ Solving chess is difficult enough as it is, there is no need to invent additional hurdles. To provide a solution is hard enough. A discussion if some solution is 100% or 99.9% or 99.798% is futile. It may be in order after a solution is provided.
Personally I would consider chess solved if black can prove a draw after both 1 e4 and 1 d4.

"I'm not discussing it in terms of "strong" or "weak". In fact, those ideas and confused and they contribute to the general lack of understanding of the subject matter."
++ The definitions of ultra-weak, weak, and strong are clear.
This use of strong and weak is usual in mathematics. Unusual definitions cause confusion.

"It means that a so-called "weak" solution is dependent on the achievement of a semi-strong one, in any case."
++ No, not at all. 'Semi-strong' does not exist. The weak solution only depends on the 7-men endgame table base, on brute force engine calculation, and on some human knowledge beneficial for pruning.

"Anyone who doesn't yet understand, that cannot be capable of understanding the problems involved, because understanding it is an elementary necessity."
++ Apparently you do not understand.

"It's also why I think that there'll never be a breakthrough unless chess is *mathematically* solved ... ie rendered into an immensely complex mathematical equation or set of equations, which can quantitively identify points of imbalance in chess games."
++ No, not at all. All present and future engine evaluation functions are inherently flawed.
Now you advocate something you erroneously accused me of. There exists no genaral mathematical function where you can input any position i.e. FEN and get the output draw / win / loss. For some positions it is immediately clear, e.g. the initial position is a draw equal material and equal position, and 1 e4 e5 2 Ba6? is a loss for white. For most positions only calculation towards the 7-men endgame tablebase can decide it.

"the normal routes would take millions of years of computing, even at the fastest speeds"
++ No, as I have pointed out Sveshnikov was right and it takes 5 years on modern computers.

"we can safely assume that chess will never be solved."
++ No, we cannot. This is a total show of ignorance. You can disagree on the number of legal, sensible, reachable, and relevant positions. It is stupid to argue: 'I do not know, I do not understand, so it must be millions of years, let us say never.'

#175
"Because it is of theoretical interest."
++ Not even Miles thought 1 e4 a6 to be superior over 1 e4 e5.

#180
"A million human grandmasters?"
++ I did not say a million. Like 3 human grandmasters or ICCF grandmasters should suffice. Sveshnikov said 'good assistants' hence more than one. They should select and set up the starting positions for analysis and they should terminate engine calculation when the outcome is without doubt so as to save resources. They should prune like 1 a4 or 1 e4 e5 2 Ba6.

"I'm afraid you don't grasp what's involved." ++ I grasp that more than you.

"Why don't you try to contribute positively?" ++ I always contribute positively even if you call my positive contributions 'drivel' and the like.