Chess will never be solved, here's why

you'll notice how tygxc did completely ignore the possibilities of improvement you presented, just as i predicted, and instead tried to deflect by redefining "improvement" when you were obviously exclusively talking about reducing errors.

Chess has mathematical patterns but the solution isn't mathematical.

Well, i feel its some kinda STEM. donchu ? and maybe even some kinda STEAM. a/w we needta start by saying how much (lets take e.g.) pawns are worth. e/o says uhhh....1 !! thats sooo dopey. before a game begins a d-pawn is probably worth more than a a-pawn right ? well, how much more ? well thats gonna be difficult. Accurately evaluating ea & every piece/pawn relative to ea & every accurately evaluated square it controls &/or occupies (or CAN !) needsta happen...PER PLY. our effort is to solve chess right ?...not guess as to how many legal positions there are...hello-hello-hello. iows thinking like a computer isnt gonna get u a/w. trust me.

@Maets-Nori
from your post:
"So "spotting errors" depends mainly upon the infallibility of the engines involved."
I would put it differently.
Fallibility. Not infallibility. With that second term suggesting 'binary'.
Fallibility. Degree of fallibility.
Infallibility of the engines never established. Might never be.
'Spotting errors' depends on degree of fallibility.

And - lets see if EE deletes all his posts here every night as he does in other forums.

Hmm, yes I did notice that he redefined improvement.

Even giving him the benefit of the doubt, though, I find it quite a leap in logic to conclude that no more chess improvements (regarding the moves themselves) can ever be found ... It suggests that engines have maxed out.

And yes, I know tygxc would clarify this by saying "engines + humans" - though I remain unconvinced that the humans have any positive impact on the games, at this point ...

Come to think of it .. if we were to declare that engines perform better with humans than without - then wouldn't that, in itself, imply that engines have room for improvement?

And around we go, with the circular debate ...

A good catch, I agree. That would have been a better way to phrase it.

I appreciate the correction.

the 'determining which moves are optimal' problem and 'constructing a very long mate chain' problems, while connected, are separate.

for 'determining which moves are optimal' - the retrograde analysis is not done ON the moves to be analyzed, but has been done BEFOREHAND in order to create the tablebase that is used, and referred to, in determining which moves are optimal.

it's not that a reverse tree has been created by a singular end checkmate position and traced backto the move to be analyzed, it's that a reverse tree has been created for EVERY SINGLE POSSIBLE end position, and ALL possible trees have ALREADY been traced to positions on the tablebase.

thus from reverse game trees to all positions, the true game tree has been created for your specific position.

Ive seen it for years, tygxc just asserts that some unjustified statement is absolute fact, people question and rebut it, tygxc ignores the rebuttal and introduces a new deflection with a new unjustified statement, each time refusing to address the actual content and instead making the merry go round. there isnt even a true 'cycle', eventually tygxc just stops responding when someone lands on a core fallacy that he makes (for example he just did that with elroch) and then starts talking to someone else and acts like he's facing all comers.

tygxc doesnt answer me anymore because ive seen the cycle too many times, he calls me a troll for the fact that i repeatedly point out his core fallacies. But i don't mind that. I'm here to make sure that tygxc doesn't mislead people, and I have been very successful at that.

I appreciate your appreciation.
And T and 'Washi' in the climate 'hoax' forum continue to be comparable.
That forum by the way - like this one recently got 'moderator intervention'.
It needed it.
Washi's ridiculous claims are as preposterous as T's.
Washi not as 'serene' as T but perhaps doesn't entertain a notion that he 'knows better than everybody'.
And as pointed out by another here - Washi has far more backers than T does.

Oh ok, I figured there was another component to it.

Yes, that's it. Basically, you generate all the positions that are "mate on the board", then all the positions that are "mate in 1", and then work up to "mate in n-ply" with optimal play (the even ones are "mated in k moves".

Each time you generate information about a position by retrograde analysis (a single reversed move leading from a solved position to another position), that tells you what the move achieves when played forwards from the other position (because you know what the position it reaches achieves).

Hope that doesn't make things less clear.

A point worth emphasising is that when you generate a position by a backward move, it tells you exactly what the move played forward achieves but it doesn't tell you anything about the other moves.

HOWEVER, because you are working backwards from all shorter mates, the first time you reach a position tells you what is optimal if the move is a winning one. That is because if there was a faster mate, the position would already have been reached.

The same does not apply for losing moves that lead from a new position. Eg if you generate a new position and you have a move that leads to a position where the opponent has mate in 2, you know you can at least stave off mate for 2 moves. But later on in the generation of the tablebase another backward move might provide a different move for you that leads to a position where you get mated in 5.

Better still, a move might be generated where YOU have a mate in 6 (despite the position being first reached with a (backward) losing move.

The final thing is that when you have no more new backwards moves left to play from the full set of positions with winning or losing moves that you have generated, there are loads of legal moves left in the positions. In fact there can even be positions that you have never reached. All the unevaluated moves are drawing moves by default. In the case where the position has never even been reached, all the legal moves are drawing moves.

Do you think we'll have 8 pieces before 2030? Will we ever hit 10 pieces?

Yes, we will have 8 pieces this decade.

highly disagree with the comparison, for the reasons u mentioned and more, the climate change debacle has actual stakes, and movements behind each side. people have investments in those positions and many personal political repercussions for themselves and for how they see the world with each statement, and a culture behind their positions. there's literally nobody on tygxc's side. theres no debate of the topics that tygxc claims. in fact, that's been one of the bigger difficulties in getting sources against tygxc, because tygxc's "interpretations" of various definitions are so completely ludicrous that nobody would even think to clarify against them.

571 pages to 729 pages. WHAT?

A certain poster's posts have finally become visible again. Longest mute I have ever seen enforced all the way, actually. Two months? Actually, 70 days (or more).

No.I think it was someone else who suggested they were all colluding. Probably the 'p' person.

no, check the comment he was responding to. this is exclusively tygxc. playerafar isnt even in this conversation.

https://www.chess.com/forum/view/general/chess-will-never-be-solved-heres-why?page=726#comment-105653469

tygxc has actually made this claim several times. tygxc genuinely believes that the only explanation for chess not being a draw despite the iccf games drawing, isnt that the ICCF games all missed the winning line, but they all; not only secretly know the winning line, but also colluded to not use it.

you might recall, that tygxc uses the draws to claim that the games are perfec- and now hes using his claims of perfection to justify it circular.

But this person (if it is just one) needs to account for about 1 in 5 posts in the whole discussion! Seems a bit implausible. Maybe multiple simultaneous unmutes?