Chess will never be solved, here's why

@avramtparra

"Your chess is insane." avramtparra said, as he slipped his feminine hand into 6673's pants and smirked. "Are you trying to mate me?" protests 6673, as avramtparra blushes, the boyish figure undressed before 6673. "Weak tempo play, avramtparra." The two kissed, deeply and passionately, and afterwards 6673 places his Rook into avramtparra open line.

     tygxc's proposal is to start with only a handful of "best" opening moves, consider only lines played by world-class correspondence players (who use multiple engines' analyses in their games), employ a few GMs to judge what alternative moves to consider (or ignore). It may well be possible to come to some conclusions within the time and budget parameters he and Sveshnikov propose.

     The question is whether most people would consider such a solution a definitive proof.

     Of course that would be enough "proof" for some people. Not for you or me or most others here.

@DesperateKingWalk

That appears to be correct. 

The number is 152.65 and it's not a mate score.  I've just seen it come up in Arena/SF15 against a move that can't win (lots in fact).

@6721

"Very little time, very little money, and very little effort."
++ 5 years, 3 million $, 3 engines & 3 grandmasters is a long time, much money and effort.

"it would have been done by now" ++ People struggle to get a 100,000 $ loan to build a house, nobody has decided to shed 3 million $ with no payback. 

Not "no payback", surely.

5 years and 3 million would probably be enough to complete the 8-piece database and the 9-piece one. At that point, only 99.9999999999999999 percent of chess is left unsolved. Huge improvement!

@6717

"White will win if both sides play perfectly" ++ No, when both sides play perfectly it is a draw.

"unless you go through the entire decision tree of possible moves you could never know"
++ It is beneficial to incorporate game knowledge into game solving.
If the 4 best moves cannot win for white, then the 16 worst moves cannot win either.

"what you mean by 146 moves exhausts all the moves"
There are 10^44 legal positions. 10^44 = 2^146.
Thus a forced win of 546 moves must contain a string of forced moves by the defending side.
146 choices between 2 moves yields more positions than chess has legal positions.

"8x8 chess is in EXP time"
++ To strongly solve chess 10^44 legal positions need to be visisted.
That is beyond present capability. It might become possible with a quantum computer.
To weakly solve chess 10^17 relevant positions need to be visited.
That can be done with 3 powerful computers running for 5 years.
Many here still fail to understand the difference between strongly solving and weakly solving.

"your method of pruning the data"
++ It is not my method, it is the same method that Schaeffer used to weakly solve Checkers.

@6731

"5 years and 3 million would probably be enough to complete the 8-piece database"
++ An 8 men endgame tablebase strongly solves all 8-men positions.
5 years and 3 million $ are to weakly solve Chess, i.e. calculate from the initial position towards a 7-men endgame table base draw or a prior 3-fold repetition.

Well, quite a big improvement I would say.

Problem is @tygxc doesn't plan to use them for that. He plans to use them to not solve chess.

With the 50 move and triple repetition rules in force? What a quick learner you are! (Probably even right if your'e talking only about two kings and six bishops on the same coloured squares.)

Presumably you won't be generating any with unequal material because you already know those are wins for the side with the extra pawn or more. That dispenses with the great majority of 8 man tablebases and all the 9 man tablebases. In fact you don't actually need any tablebases, you can just stop when you've got 31 men. That should leave you with plenty of spare capacity to concentrate on the serious business of not solving chess.

hi

Yes, someone wrote, "++ An 8 men endgame tablebase strongly solves all 8-men positions" a few posts back.

Some even fail to understand the difference between solving and not solving.

well maybe your right Optimissed..  but for the Human mind..  for a chess engine i'm sure will eventually figure out chess just like it did with checkers....

What are you talking about? It literally is Game Theory, both sides Black and White have perfect information therefore making impossible to know if its A Win, Loss or Drawn for any side. Its literally a theorem and its described in the quote I posted

@6732

What are you talking about? It literally is Game Theory, both sides Black and White have perfect information therefore making impossible to know if its A Win, Loss or Drawn for any side. Its literally a theorem and its described in the quote I posted. you deadass have to go through the whole decision tree because of this

Just read about it here, https://en.wikipedia.org/wiki/Perfect_information

@6742

"It literally is Game Theory"
++ Game theory is that each position including the initial posiytions is either a draw, a win, or a loss.

"both sides Black and White have perfect information therefore making impossible to know if its A Win, Loss or Drawn for any side." ++ It is quite possible for manty positions to determine if it is a draw, a win, or a loss. The initial position is a draw. 1 g4? or 1 e4 e5 2 Ba6? are losses for white. 1 e4 b5?, 1 e4 f5?, 1 d4 g5?, 1 d4 c5? are losses for black.

@6745

yeah from those positions it is, the question is to derive it from the start. Thats solving for all 8x8 chess

@6746
"the question is to derive it from the start. Thats solving for all 8x8 chess"
++ No that is not solving 8x8 chess.
Determining if the starting position is a draw, a win, or a loss is ultra-weakly solving chess. There is massive evidence from millions of human and engine games as well as the logical argument that a tempo is not enough to win to know that the initial position is a draw.

Weakly solving chess is figuring out how to draw from the starting position. It is thus the path from the initial position to other drawn positions until a known draw, like a 7-men endgame table base draw or a prior 3-fold repetition.

Strongly solving chess is figuring out for all legal positions if they are draw, a win, or a loss.

"Solving chess means finding an optimal strategy for the game of chess, that is, one by which one of the players (White or Black) can always force a victory, or either can force a draw (see solved game)" - Wikipedia

The only way to do that is to go through the entire position tree because you don't know if black's position is a starting loss. Because game theory says both sides have perfect information, and like you said about tempo advantage, you literally don't know if the starting position is a win loss or draw of white. I am not sure what you are saying, because what I am talking about is literally "solving chess", and this is coming from someone who took a class on weak and strong deterministic problems.

 I can see what you are saying about, "Strongly solving"chess, but if you think about its the same case, where you would need a quantum computer to just go through all the variations, and if you think about it, if you find the most optimal solution any other legal position would be a loss from which ever side changed from the given optimal solution and you wouldnt have to go through all those tree paths

Chess is a closed game and a perfect closed game always ends in a draw. Kind of like tic tac toe. Chess cannot be perfectly solved but that doesn't mean its unsolvable.