Chess will never be solved, here's why

A very long time as in billions of years of computer time. So it won't happen, ever. The World could be Putinned next Thursday.

@6531

"chess can be solved" ++ Yes

"it will take a very long time and dedication" ++ 5 years

"it would be far too difficult for a human to ever memorize all the right moves"
++ A human can memorize 10,000 perfect games with optimal play from both sides.
The solution might also be condensed into a set of rules that are easier to remember.

@6525

"the goal in solving chess is to discover whether the game can be won by force by either player from the starting position against all possible counterplay"

++ No. The goal of ultra-weakly solving chess is to determine the game-theoretic value of Chess i.e. the outcome if both players play optimally. For all practical purpose Chess is already ultra-weakly solved and we know the game-theoretic value is a draw.

The goal of weakly solving Chess is to determine how black can draw.

"vast numbers of possible lines are proven worse"
++ It is proven that 1 e4 e5 2 Ba6? is worse than 2 Nf3.
It is proven that 1 a4 is no better than 1 e4 or 1 d4.
It is proven that 1 Nh3 is no better than 1 Nf3.
Here is a paper that proves it with no other input but the Laws of Chess:
https://arxiv.org/pdf/2111.09259.pdf 

@6516

"Sveshnikov's is a glib comment, not one arrived at quantitatively."

Sveshnikov probably arrived at it quantitatively. He held a MSc and almost a PhD in engineering.
I have arrived at it quantitatively. I try to explain in a different way.

Let us assume we calculate all w legal white moves, all w legal black responses and so on all the way up to checkmate or a draw. Then we calculate all 10^44 legal positions and we strongly solve chess to a 32-men table base. That would take 10^27 years.

Let us now restrict ourselves to weakly solving chess: finding a strategy i.e. one strategy for black to draw: against all possible w white moves 1 black response that draws. w*1 = Sqrt (w*w). Thus that would need to calculate Sqrt (10^44) = 10^22 positions and take 300,000 years.

Now observe that the vast majority of the 10^44 legal positions contain multiple underpromotions to pieces not previously captured. Underpromoting instead of queening is like blundering a piece. Let us first restrict all promotions to pieces already previously captured. There are 10^37 such positions.
Now note that positions with 3 or 4 queens do occur in perfect games with optimal play from both sides. So multiply by 10 to include such positions with 3 or 4 queens. 10 * 10^37 = 10^38.
Now the same reasoning as above: Sqrt (10^38) = 10^19. That would take 316 years.

Now note that black tries to achieve the draw. So white tries to oppose to the draw.
Moves that lose for white like 1 e4 e5 2 Ba6? do not even try to oppose to the draw.
Thus such moves can be pruned away. That has to be done judiciously, hence grandmasters are needed. Some losses of a piece might be sacrifices. Grandmasters must decide if there is any compensation of any kind. If yes, then calculate. If no, then prune.
Likewise some endgames e.g. with opposite colored bishops are known draws. The computer can calculate a long time until a 3-fold repetition, while the humans see it is not possible to win in any way. Also here some opposite colored bishop endgames can be won. Here the grandmasters judge. If there is any chance to win, then calculate. If there is no chance to win, then adjudicate a draw so as to save calculation time.
So this human incorporation of knowledge reduces another 2 orders of magnitude to 10^17 relevant positions, solvable in 5 years.

I only see 24. Not that it matters, but if I'm right and you're wrong..... THAT would definitely be a record....or at least a first.

  I am sure we are both right about our observations. That is a very odd discrepancy.

Upon reflection, I was looking at reactions rather than downvotes, so I think you were being quite gentlemanly not to point that out. 

I'm sure it would do better if it weren't borderline unintelligible.

Imagine a paradigm of 100 Kasparovs, solving two Ruy Lopezes each, blah blah blah, chess will never be solved QED.

As an aside, I'm a little tired of having to catch my own spelling mistakes... I rely on the auto thing at this point.

No, there can be no way to move forward unless and until computers can produce a facsimile of the same level of positional understanding, which you forlornly hope that the GMs will supply.

I had 41 yesterday and today, 43.

I was looking at "thumbs down" reactions rather than down votes.  I now see the error of my ways.

 Now I understand. A little ambiguity.

A perfect engine should be able to convert any advantage to a win

,

my brother be like.

With perfect play, chess will probably end in a draw. Going by that standard, this may be a perfect game of chess: