Chess will never be solved, here's why

@9358

"There is no such thing as 'weakly solved'."
++ There is, e.g. Schaeffer's proof how to draw Checkers.
Please read Prof. van den Herik's scientific paper Games solved: Now and in the future.
https://www.sciencedirect.com/science/article/pii/S0004370201001527

@9357

"Once 7 men are reached castling rights are forfeited. I never heard of such a rule."
++ It is no rule, it is a practical observation. You cannot construct a reasonable game that reaches a 7-men position with castling rights intact. Players castle, or move a king, or the rooks.

"there is no proof chess is a draw with best play"
++ I presented an inductive proof as well as a deductive proof. Ii took care to explain in simple terms, but If you still do not understand either, then you do not need to be ashamed.

"There's a high percentage of GM draws. But - that doesn't mean no mistake was made."
++ At the 2024 Toronto Candidates' Tournament they made 1.1 error / game average.
However, an engine at 3 minutes / move is much stronger than a grandmaster at 3 minutes / move. An ICCF (grand)master with engine at 5 days/move average is much stronger than an engine at 3 minutes / move. At this year's ICCF World Championship Finals they have reached no mistakes as proven.

"It can mean nobody made a big enough mistake" ++ There are no small or big mistakes.
Per GM Hübner there are only errors (?) that change the game state from a draw to a loss, or back from a win to a draw, or blunders or double mistakes (??) that change a win to a loss.

Yeah, you go ahead and let us know when these "proofs" are accepted by anyone that matters.

@9355

"The above are all human approximations."
++ AlphaZero reached about the same, with no other human input but the Laws of Chess.

"This debate is still in exactly the same state it was years ago."
++ No. New fact: 104 draws out of 104 games in the ICCF World Championship Finals.

"Chess is not proven a draw with best play"
++ It is. I presented an inductive as well as a deductive proof.

"nor indeed can humans or current engines determine best play"
++ It can be determined from the results of the ICCF World Championship Finals.

"A solution is not forthcoming in our lifetimes" ++ It is already there.

"tackling the issue (10^43 positions to traverse)"
++ This mistake keeps coming up. To strongly solve Chess, i.e. a 32-men table base needs to traverse 10^44 legal positions and that is indeed beyond current technology.
Weakly solving Chess, as Schaeffer did for Checkers needs only to traverse 10^17 positions.

"arbitrarily shave away 20+ orders of magnitude"
++ Not arbitrarily, it is the vital difference between strongly and weakly solving a game.

"Claiming to "know" chess is a draw" ++ 'Chess is a draw' - Fischer

@9354

"you dont get to say something is absurd and invalidate it. thats not how proof works."
++ That is how many proofs work in mathematics: reductio ad absurdum.
Why would there only be odd numbers of errors and never an even number?
Why would all errors come in pairs when 17 different entities play?

I could understand errors sometimes in pairs for 1 entity autoplay, e.g. in a drawing position black fails to recognise a winning Greek Gift Sacrifice for white and erroneously allows it. White also fails to recognise it and erroneously does not play it. Play goes on in a drawing position.
But all errors coming in pairs and never a single error makes no sense.

@9363

"Fischer either said this as a joke or assumed it." ++ He said so after a disappointing draw.
"He never proved that statement."
++ He knew it, he did not have to prove. He was a player, not a mathematician.

@9353

"The tablebase project is an attempt to solve chess." ++ Table base = strongly solving.

"these correspondence games mean very little" ++ They mean a lot: perfect games.

"Imagine some player from the future (with technology from the year 2100) play games against these guys. Would these guys manage to score even a single draw?"
++ Yes, they would reach 100% draws just the same.

@9367

"you need proof to say chess is a draw"
++ Fischer and many othert greats said so. They are free to pronounce expert opinions.
They have no obligation to provide proof.
I have here presented two proofs: an inductive as well as a deductive one.

- Your made-up 10^17 number in no way corresponds to what would actually be considered a weak solution by definition:

Provide an algorithm that secures a win for one player, or a draw for either, against any possible moves by the opponent, from the beginning of the game.

That's the definition. You are not meeting it.

- Fischer didn't know either, and neither does any GM, or any human being that has ever lived.

- ICCF finals are based on engine play, which is flawed play and will remain flawed play, as engines that still improve with every single release amply demonstrate. The human factor is all but meaningless in ICCF play at this point. All they are doing is running multiple engines and pretending to make significant choices in choosing lines...they are basically administrative assistants to the engines at this point.

He believes in the method of "proof by 100 examples", and he also thinks the stated opinions of good human chess players are proofs.

No sensible discussion is possible, unfortunately, since all points are ignored.

I am sure you will agree, there is a big difference between practical chess heuristics and solving chess!

Facepalm emojis were designed for anyone who thinks there is a precise, quantifiable relationship between tempi and material and between material and results.

@9372

"10^17 number in no way corresponds"
++ I provided rational arguments to arrive at this estimate. The true number may be a bit higher or lower, as will be clear after Chess is officially weakly solved.
Schaeffer was also slightly off when he first estimated the number to weakly solve Checkers.

"what would actually be considered a weak solution by definition"
'weakly solved means that for the initial position a strategy has been determined to achieve the game theoretic value against any opposition' - Prof. van den Herik

"Fischer didn't know either" ++ He knew. "neither does any GM" ++ Most know.
"or any human being" ++ Many know.

"engines that still improve with every single release"
++ For a given time per move they make fewer and fewer errors.
Now at 5 days/move and jockeyed by an ICCF (grand)master they have reached 0 error.

"The human factor is all but meaningless in ICCF play at this point." ++ No it is very important.

"All they are doing is running multiple engines"
++ You are wrong. Sign up for an ICCF World Championship qualifier and try to qualify.

"they are basically administrative assistants"++ They are ICCF (grand)masters for a reason.
They play the ICCF World Championship Finals because they qualified.

@9375

"basic maths in a flawed way"
++ What flaw?
Capablanca's statement is a verbose way to say 1 pawn = 3 tempi and 1 pawn wins.

@9376

"there is a big difference between practical chess heuristics and solving chess"
++ Yes, but heuristics like 1 pawn = 3 tempi and 1 pawn wins (ceteris paribus) can be seen as theorems derived from the Laws of Chess after centuries of playing and analysing and thinking by great minds.

"there is a precise, quantifiable relationship between tempi and material and between material and results" ++ There is. Any master has learned how to convert an advantage of a pawn. As Capablanca asserts you can grab a pawn and lose 2 tempi as black, or lose 3 tempi as white.

@9381

"All your statements have no complete proof"
++ I am the only one on this thread who proves statements.

""All your statements have no complete proof"
++ I am the only one on this thread who proves statements."

give one example lmfao.

10^17 claim - no proof

engine games were optimal claim - no proof.

already proven to be a draw - no proof.

1 tempo is worth less than a pawn - no proof.

i love how tygxc downvotes posts that show his idiocy most spectacularly, like does he not think that the rest of us dont know its him downvoting?

""there is a precise, quantifiable relationship between tempi and material and between material and results" ++ There is. Any master has learned how to convert an advantage of a pawn. As Capablanca asserts you can grab a pawn and lose 2 tempi as black, or lose 3 tempi as white."

thats not proof, we are asking for proof.

hey tygxc, you still havent come forward about missing your calcualtions on cloud computing by a factor of over 100 million.

its not even that, hes also assuming that they are talking in the direct context that tygxc wants them to be speaking in. ah yes, fischer complaining about a loss is him making a general statement that chess has an absolute peer reviewed proof of a draw. lmfaooooo