The irony of this whole 650 page conversation is... even if chess were to get "solved" from a computational standpoint (and I don't think the arguments here for that happening are very well-formed) it would not make even the tiniest bit of difference when it comes to playing chess.
Hey its like Synchronized Swimming in the Olympics.
The irony of this whole 650 page conversation is... even if chess were to get "solved" from a computational standpoint (and I don't think the arguments here for that happening are very well-formed) it would not make even the tiniest bit of difference when it comes to playing chess.
Hey its like Synchronized Swimming in the Olympics.
I would argue rhythmic gymnastics...the regular gymnasts (engines) will do all the physical marvels, and the humans will work on artistry and beautiful form, that being all that is left to them. But yes, for the average person at 1000 rating, it's essentially meaningless.
It won't even be impactful in a world championship match. Paraphrasing Karparov: "I can assure you... even at the top level - world championship match - players make mistakes". Some cloud computer hard-cracking chess while performing a quintillion operations a second over 5 years is not going to then cause the world championship players to suddenly know all the moves and refutations of the game. The most likely result would just be.... "okay the Ruy Lopez is a draw". What do you do with that information? Nothing.
If anything what would probably happen is the computer would pronounce some opening refuted, most people would abandon it, a few years would pass... and next thing you know players are crushing the opposition with that opening. Case in point - the QID and alphazero.
It won't even be impactful in a world championship match. Paraphrasing Karparov: "I can assure you... even at the top level - world championship match - players make mistakes". Some cloud computer hard-cracking chess while performing a quintillion operations a second over 5 years is not going to then cause the world championship players to suddenly know all the moves and refutations of the game.
The point is, your average chess player (like your average Connect Four player), won't know or care if the game is solved. The top players, using engines for analysis, will obviously know they can't even begin to touch engine results or even comprehend them past a certain point, and that knowledge does have a psychological effect. It's already changed the game tremendously at the top levels, and it's telling that super GMs frequently lose games when they muddle up their engine prep.
it would not make even the tiniest bit of difference when it comes to playing chess.
i know right ?...e/o stopped playing checkers as soon as it got solved.
The irony of this whole 650 page conversation is... even if chess were to get "solved" from a computational standpoint (and I don't think the arguments here for that happening are very well-formed) it would not make even the tiniest bit of difference when it comes to playing chess.
yeah we cant have any real discussion here because two individuals have literaly fantasies hat we have to keep others from believing
If we could convert @Optimissed's ego and @tygxc's obstinacy into computational power, we could get it done by Friday.
I would agree regarding tygxc's obstinacy.
(perhaps tygxc was or is a trial defense lawyer)
but I don't agree regarding O's ego which is much too fragile and delicate to have any 'strength of conversion' or any strength at all.
And there were several blank pages at the end of this forum again.
Somebody has been muted.
@12940
"it'll need to be a very large algorithm"
++ No it is the same algorithm for going from 6 to 7 men endgame table base or 7 to 8 men.
A quantum computer has the posssibility to do it all in parallel.
80 years from now it will be done.
@12933
"the word logic by using it for something completely different"
logic = science that deals with the rules and tests of sound thinking and proof by reasoning - Merriam Webster
"identifying some general condition and using that to identify the move"
++ Yes, that is the core of chess.
'Chess is the art that expresses the science of logic' - Botvinnik, PhD. in engineering
"the relationship is an empirically based on many examples"
++ No, it is also based on logic and geometry.
"Your logic will instruct you to blunder"
++ No, logic is a way to narrow down candidate moves in play and in analysis.
In sharp positions calculation is inevitable, but those are the exception rather than the rule.
"my move draws if I can draw each of the positions reachable by a legal response to my move by my opponent." ++ That is brute force. No chess player of reasonable strength always calculates all legal responses. Logically derived candidate moves only, and not even that.
In a quiet position there is no need to calculate even 1 legal response, all goes by logic.
"In competitive chess, such thoroughness is usually impractical."
++ In chess analysis it is not useful either. It leads nowhere to calculate game trees of 1 a4, or 1 e4 e5 2 Ba6? After 1 e4 e5 2 Nf3 Nc6 there is no need to calculate 3 Ba6?, 3 Nd4?, 3 Nxe5?, 3 Ng5? 3 Nh4? they lose material without any kind of compensation,
that is to be dismissed right away without even 1 ply of calculation.
Likewise 3 Bb5 and 3 Bc4 are logically superior to 3 Bd3 or 3 Be2.
'chess is a generalised trade' - Botvinnik, 1968
Material is a value in chess: 1 pawn is enough to win.
You can sacrifice material, but only for something in return.
All materially unbalanced positions are sharp, need calculation.
Interesting enough the empirical N = B = 3 P, R = 5 P, Q = 9 P all contain odd number factors.
Thus only positions with an even number of men can be materially balanced, while all positions with an odd number of men are materially unbalanced and thus sharp, i.e. in need of calculation.
"constructing a weak solution"
'Next to brute-force methods it is often beneficial to incorporate knowledge-based
methods in game-solving programs' Games solved: Now and in the future
You are obsessed with brute force methods only.
Connect Four has been weakly solved by brute force by Allen as well as knowledge-based by Allis before it got strongly solved by Tromp.
Checkers is more calculative than Chess, Checkers is more like a Chess pawn endgame.
That is why knowledge-based methods are better suited to Chess than to Checkers.
@12933
"the word logic by using it for something completely different"
logic = science that deals with the rules and tests of sound thinking and proof by reasoning - Merriam Webster
"identifying some general condition and using that to identify the move"
++ Yes, that is the core of chess.
'Chess is the art that expresses the science of logic' - Botvinnik, PhD. in engineering
"the relationship is an empirically based on many examples"
++ No, it is also based on logic and geometry.
"Your logic will instruct you to blunder"
++ No, logic is a way to narrow down candidate moves in play and in analysis.
In sharp positions calculation is inevitable, but those are the exception rather than the rule.
"my move draws if I can draw each of the positions reachable by a legal response to my move by my opponent." ++ That is brute force. No chess player of reasonable strength always calculates all legal responses. Logically derived candidate moves only, and not even that.
In a quiet position there is no need to calculate even 1 legal response, all goes by logic.
"In competitive chess, such thoroughness is usually impractical."
++ In chess analysis it is not useful either. It leads nowhere to calculate game trees of 1 a4, or 1 e4 e5 2 Ba6? After 1 e4 e5 2 Nf3 Nc6 there is no need to calculate 3 Ba6?, 3 Nd4?, 3 Nxe5?, 3 Ng5? 3 Nh4? they lose material without any kind of compensation,
that is to be dismissed right away without even 1 ply of calculation.
Likewise 3 Bb5 and 3 Bc4 are logically superior to 3 Bd3 or 3 Be2.
'chess is a generalised trade' - Botvinnik, 1968
Material is a value in chess: 1 pawn is enough to win.
You can sacrifice material, but only for something in return.
All materially unbalanced positions are sharp, need calculation.
Interesting enough the empirical N = B = 3 P, R = 5 P, Q = 9 P all contain odd number factors.
Thus only positions with an even number of men can be materially balanced, while all positions with an odd number of men are materially unbalanced and thus sharp, i.e. in need of calculation.
"constructing a weak solution"
'Next to brute-force methods it is often beneficial to incorporate knowledge-based
methods in game-solving programs' Games solved: Now and in the future
You are obsessed with brute force methods only.
Connect Four has been weakly solved by brute force by Allen as well as knowledge-based by Allis before it got strongly solved by Tromp.
Checkers is more calculative than Chess, Checkers is more like a Chess pawn endgame.
That is why knowledge-based methods are better suited to Chess than to Checkers.
Btw chessbrahs said they can't take rapid chessplayers seriously I'll send a YouTube clip of it and actually it is one of the reasons I typically don't play rapid ill send once the videos out xd
@12940
"it'll need to be a very large algorithm"
++ No it is the same algorithm for going from 6 to 7 men endgame table base or 7 to 8 men.
A quantum computer has the posssibility to do it all in parallel.
80 years from now it will be done.
A chess model and algorithm is a very large thing. I have written a chess model myself... it is much more complex than an RSA algorithm... you have to code all the parameters for the game itself - it's not just a simple math equation you're coding - you need a concept of a board, a concept of a piece, logic for what castling is, logic for not moving off the edge of the board, logic for en passant, logic for promoting a pawn, logic for moving a pawn twice on its first move, logic for pawns not capturing forward, logic for the king in check... You all of that need that before you can even begin to place pieces on the board and run any kind of algorithm... the largest quantum computer *in the world* today has 1121 qubits. You will not be able to code a chess game with 1121 bits. I don't know how many bits you'll need, but it's going to be alot more than that. Qubits are in very short supply, as the number of qubits grows linearly the system complexity grows exponentially.... I am skeptical that quantum computer is going to scale the way you're imagining it will, or in the way traditional computers have. At least, there is no evidence yet that it will, and there is reason to be skeptical.
Bottom line is you just have a hope, an opinion... which you're entitled to have, but my point is your opinion is not fact.
Btw chessbrahs said they can't take rapid chessplayers seriously I'll send a YouTube clip of it and actually it is one of the reasons I typically don't play rapid [...]
You'll note that the Chessbrahs are not exactly feared in OTB play among their 2500-ish peers. The most robust and elegant chess games require a time control that does not really inflict significant time pressure.
@12953
"you have to code all the parameters for the game itself"
++ All chess engines have the game itself coded.
If you translate that code to Python, then you can run it on a quantum computer.
The point is than 8 men position FEN8 leads to known win/draw/loss 7 men position FEN7 can be executed in parallel.
You could derive 8 men Qubits from 7 men Qubits in one step.
"Qubits are in very short supply" ++ For now, yes. Just like RAM was in very short supply.
Dude, regular computers can process gigabits worth of data in seconds, the largest quantum computer in the world can handle 1121 bits and that's for the entire system.
The chess model is not going to be represented by 8 qubits, I'm sorry to break it to you.
"you have to code all the parameters for the game itself"
++ All chess engines have the game itself coded.
If you translate that code to Python, then you can run it on a quantum computer.
The point is than 8 men position FEN8 leads to known win/draw/loss 7 men position FEN7 can be executed in parallel.
You could derive 8 men Qubits from 7 men Qubits in one step.
"Qubits are in very short supply" ++ For now, yes. Just like RAM was in very short supply.
No, you cannot...the quantum computing version of Python has a limited instruction set...we've been through all this before.
@12954
"Chessbrahs are not exactly feared" ++ Eric Hansen 2610
"The most robust and elegant chess games require a time control that does not really inflict significant time pressure." ++ 5 days / move ICCF, 3 min/move classical, but Magnus Carlsen no longer plays the Classical World Championship, only the Rapid and Blitz World Championships.
@12940
"it'll need to be a very large algorithm"
++ No it is the same algorithm for going from 6 to 7 men endgame table base or 7 to 8 men.
A quantum computer has the posssibility to do it all in parallel.
80 years from now it will be done.
A chess model and algorithm is a very large thing. I have written a chess model myself... it is much more complex than an RSA algorithm... you have to code all the parameters for the game itself - it's not just a simple math equation you're coding - you need a concept of a board, a concept of a piece, logic for what castling is, logic for not moving off the edge of the board. You need that before you can even begin to place pieces on the board and run any kind of algorithm... Qubits are in very short supply, as the number of qubits grows linearly the system complexity grows exponentially.... I am skeptical that quantum computer is going to scale the way you're imagining it will, or in the way traditional computers have. At least, there is no evidence yet that it will, and there is reason to be skeptical.
Bottom line is you just have a hope, an opinion... which you're entitled to have, but my point is your opinion is not fact.
weve been trying to inform tygxc that his ramblings do not count as facts for years. he's beyond hope. ive literally told him that i had my rebuttals to his claims personally verified by mathematicians and his response is to ignore me and keep posting the same things.
Lol, just for specifying what the type of ONE piece is you must specify in bits the color (2 possible options) + the type of piece (6 possible) > so you need 4 bits for that... at least. You can have 32 pieces on the board... so you're already at 128 bits at a minimum and we're probably underestimating, because you're probably gonna need some metadata too. You've used up 12% of the available bits and you have not even BEGUN to code anything in the chess model, or the game, or anything at all... all you have done is just given the pieces a name...
The irony of this whole 650 page conversation is... even if chess were to get "solved" from a computational standpoint (and I don't think the arguments here for that happening are very well-formed) it would not make even the tiniest bit of difference when it comes to playing chess.