Chess will never be solved, here's why

'Words and concepts' versus 'math and computer math and logic and logic math'.
The two groups of things constantly collide in this subject of Chess Not Solved.
But that Collission pertains to many other things too. Even in law.
-----------------------
@Elroch it is a mistake to think that because a classification of errors seems to be 'complete' that it is therefore 'adequate'.
That doesn't follow.
Respectfully - I think you've got this 'wrong'.
--------------------------
And the result is Advantage to tygxc.
He'll exploit other people's terms too.
Like 'game-theoretic value'.
Another booby-trap semantic in a huge minefield of such booby-traps.
Idea: keep it generic. 'game-theoretic value' is brand-name.
Its like a science person using the word 'believe' in a roomful of science deniers.
Causing them to yell 'enemy on periscope. battle stations!'

I play from afar

You are far out.

Yes, so are u.

More like a sanity test, really.

Elroch your post there shows that the point went right by you.
Attempts to classify errors can fail in at least two ways - one is from a programming standpoint and all the issues and vicissitudes of same.
Including 'playing for a win' as opposed to 'playing for a draw'.
Even a linear determination of 'advantage' is suspect.
---------------------------------------
But the other failure is in discussion including discussion with persons like tygxc.
Example: Martin - that's MarAttigan trying to tell tygxc he 'Can't' do something. That's 'advantage to tygxc.'
--------------------------------------------
He'll look at your post and think 'Yes! Victory again! Elroch's post is 'gobbledygook' !!'
Elroch your 'methods' are effective with people like Optimissed ... but with tygxc you're dealing with something different. A disinformationer.
It isn't about me.
You trying to make it as about me is more proof you don't get it.
But - and on the other hand - you're doing fine.
------------------------------------
tygxc will continue with 'nodes per second' and 'square root of number of possible positions' and '114 draws proves'. All invalid positions.
Could there be an analogy to describe what he's doing?
A park chess player sits reading. With board and pieces and clock set up in front of him.
He's got white. Has already moved 1) g4.
Another player comes up.
'Can I get a game? Why have you moved already? You're supposed to let black start your clock first.'
Reply: 'Lets play. 5 minutes each. There I've put the pawn back. Sit down.'
New player sits down. 'And why 1) g4 ?? Isn't that the worst move?'
Reply: 'Probably. But I play it and win with it. And also with 1) Na3 and other bad first moves.'
------------------------------------------------
they play. white wins and continues to win all games with other bad first moves with white or black.
New player: 'I don't get it. Okay you won with bad moves. Why don't you play better first moves?'
Reply: 'You're right. You don't get it. Lets try a different idea. I'll take off my b-knight every game before starting. Lets see what happens.'
Result: Hero continues to win every game even playing a knight down.
Other players come and the same thing happens.
They complain. They ruminate. They commiserate.
Hero: 'You guys don't get it about tactics.'
---------------------------------------
tygxc is 'Hero'.
His message to all is that he can keep repeating the same invalid claims to all comers and nobody will beat him.
In park chess - 'play the winner' means you have to get whoever's up off his seat by beating him if you want to play somebody else. Beat not draw.
Nobody is going to 'get tygxc up off his seat'.
The analogy holds up remarkably well.

I don't think they do that here. That's more of a Lichess thing ...

d4 d5 e4
The Blackmar-Diemer.
White does OK.
https://www.365chess.com/opening.php?m=6&n=300&ms=d4.d5.e4.dxe4.Nc3&ns=7.8.75.132.300#search_block_opening_explorerLow percentage of draws.
Pertains to forum subject?
Yes.
At the same site I compared d4 d5 c4.
Doesn't win as much for white as the Blackmar.
40% compared with 42% winning for Blackmar Diemer with white.
Point: dogma about trying to 'arbitrate' one opening move as better than another.
'Solved' ... g4 e5 f3 Qh4 checkmate.
The 'bad news'. White didn't have to play f3 which 'falls on his sword'.

It's quite remarkable the amount of pretentious drivel someone can write about such an unimportant subject.

yo guys completely unrelated to the topic i beat an irl person with over 500 elo more than me

A member - a good member - told tygxc that he tygxc 'can't' do something.
Which is what tygxc likes to hear.
Then tygxc will 'demonstrate' that he's going to do exactly what he 'can't' do.
Year in year out.
As to how the various members 'go at tygxc' I'm not going to talk about that the way the O-person does.
-------------------------
Instead - I'll put it this way:
When countering disinformation on the net one can 'micro-manage' one's interference with that disinfo person ... or instead 'look at the bigger picture' and proceed that way ... or concentrate on all points in between.
Or do all three.
----------------------------
If there's inattention to any of the three - then advantage to the disinformer.
tygxc knows about disinformation tactics.
He uses those tactics constantly.
For others to counter those tactics (and to counter the tactics of climate science deniers too) then something well beyond just 'being right' is in order.
tygxc knows he's wrong.
That's the point.
He's 'serving notice' like this 'Yes spreading disinformation. Now what are you going to do about it?'
-----------------
Reminds me of a scene in a movie where Clint Eastwood boxing coach is telling young Hilary fighter between rounds ... (Hilary fighter is losing) about her opponent ...
'She's a better fighter than you. Now what are you going to do about it??'
Hilary comes out swinging and puts her opponent down.
-----------------------------------------
But that isn't available here.
There's no 'knockout'.
So perhaps the best is to just keep refuting tygxc's disinformation.
Year in year out.
------------------------
Result: things come out about the nature of chess.
For example the business of 'playing for a win' versus 'playing for a draw'.
In lower levels of the game it usually becomes obvious at some point which one a player must be doing. In other words not how to make such decisions but when. Its usually obvious. Although it may be unexpected when it happens - that its time.
But in GM play but moreso with supercomputers playing against each other and the programming of such computers ... that's a whole different ball game.

Seems like DTM/DTM50 valuation. Optimal play is probably most often used to mean optimality according to one of those objectives.

Another common meaning of "optimal" might be play that is likely to maximise your expected result (counting 1 you win, 0 you draw, -1 you lose). That may depend on an estimation of your opponent and is not necessarily the same. E.g. I recommend not playing the lone king accurately in KBNK here.

That post by Elroch that MarAttigan just quoted - is 'better'.
But there can be much missed.
Like the computer just doesn't know whether a position is winning or losing but then what about the determination 'play for a win' versus 'play for a draw'?
-------------------------------
Idea - the computer is 'at the mercy' of the human programmer there.
And the human programmer can't possibly think with the depth of the computer ... so they're 'at each other's mercy - subject to each other's foibles and failures' and result - the computer can't properly address the issue of 'playing for the win' when its supposed to - including because of the Grim Reaper. Chess isn't Solved.
----------------------------
No need to be grim about that though.
Chess isn't solved. Can't be solved with today's technology.
tygxc pretends that others are Grim about that reality - and that they 'don't like' the idea of chess being solved.
Corresponds to climate science deniers falsely maintaining that people who accept science realities are 'alamists'.
Similiar tactics.

@MARattigan's first comment is of course correct. I was thinking of the distance to mate numbers in a tablebase when I wrote the definition.

By contrast the second comment is a very context-dependent concept. It is probably best thought of as being a statistical notion, including on a model of opponent behaviour. For example, to maximise the expectation against a stronger player may involve aiming for the most drawish (and simplest) positions, while maximising it against a weaker player would involve the exact opposite.

yap

'distance to mate' is a nice term.
But in any situation where there's no such 'distance' found because it doesn't exist or not yet -'then what'? 
if there's a 'distance to mate' then that means the position is solved.
Sometimes I've seen Stockfish give huge evaluation numbers on positions like +50. 
But with no 'distance to mate'.
Does that mean the computer's 'not sure'?
A player can often be sure he's got a win - with aboslutely no 'distance to mate' number.
But can Computers be 'sure' in that situation?
I would say No.
Computers do something called 'number- crunching'.

@Elroch . Hi!
Yes ... definition of a winning position. Obviously.
But there's other points.
Like computers often show 'winning' evaluation numbers when there's no 'distance to mate'.
Plus people can know they have a win. Again with no 'distance to mate'.
So I'm suggesting that that 'distance to mate' doesn't mean much in this context.
Regarding tablebases - the computer might assign a maximum distance to a mate position that has more than 7 or 8 pieces on board. In other words its not a 'tablebase' situation. Yes. Again obvious. Just pointing that out.
---------------------------
Regarding tygxc's tactics with 'nodes per second' and 'taking the square root' - well your methods and the methods of others here may be more than enough to keep refuting his posts on those.
But he's really keeping his jaws clamped on that 114 draws bone.
Yes his claims are invalid to anybody who really thinks about it.
----------------------------------------
My inclination is to think that those computers weren't programmed to play for the win properly and instead were programmed with an opposite extreme.
Whether by design or by 'incidence' or anywhere in between.
Gets back to the issue of classifying mistakes.
Play that is very deeply refutable - but is beyond the horizon of another similiar computer that is programmed similiarly - to refute that play so the computer fails to refute it. Which is itself an even deeper 'error'.
What happens if two recently deaf people are stranded on a desert island?
They can't hear themselves speak so the speech of both of them becomes more and more garbled. 'Draw'.
With two similiar and extremely strong computers its not progressively more 'garbled' - the win is just 'unintelligible' to both of them.
Note five days per move.