Chess will never be solved, here's why

@7133

"the computers and programs of the year 2100 will far outclass those of the present"
++ Yes, they will make fewer mistakes for the same time per move.
That does not change anything, it will only go faster.

"punch large holes in the analyses that the Sveshnikov five year plan"
++ Because the plan of Sveshnikov to 'bring all openings to technical endgames' depends on the 7-men endgame table base and you cannot punch holes into that.

"unanalyzed lines the human experts judge to be unworthy of consideration"
++ 1 e4 e5 2 Ba6? is unworthy of consideration and will stay so.

"possibilities that exist following opening moves that the plan will never examine"
++ If 1 e4 & 1 d4 are calculated to 7-men endgame table base draws, then 1 a4 is not relevant.
If 1 Nf3 is calculated to a 7-men endgame table base draw, then 1 Nh3 is not relevant.
Likewise 1 f3 and 1 g4? are not relevant.
If the best moves cannot win for white, then the worst move cannot win for white either.

@7139

"not a way to solve chess"
++ It is a way to weakly solve Chess, just like done for Checkers, Losing Chess, Connect Four

"person involved in peer-reviewed research on the subject"
++ Prof. van den Herik, authority on the subject: 'it is beneficial to incorporate game knowledge into game solving'. Schaeffer used best first heuristic and pruning in solving Checkers.
Allis solved Connect Four with knowledge rules only.
There is no reason why weakly solving Chess should be subject to more stringent restrictions than weakly solving Checkers, Losing Chess, Connect Four, Nine Men's Morris etc.
Chess is 1000 times more complex than Checkers to weakly solve.
You need not invent additional complications.

How long would it take to solve chess with an unlimited supply of stubby pencils? Sorry about your insanity tygxc. I do okay living with mine.

@7146

"Chess is not checker!"
++ No, Chess is not Checkers. Chess is 1000 times more complicated to weakly solve than Checkers: 10^17 relevant positions instead of 10^14.

"1. Could not play a game of chess.
2 Could not beat any human at chess.
3. Could not give a refutation on any chess positions.
4. Could not play a perfect game of chess.
5. Could not give any kind of analysis."
++ None of these are part of weakly solving Chess.
Please re-read the definition of weakly solving a game.
'weakly solved means that for the initial position a strategy has been determined to achieve the game-theoretic value against any opposition'
A strategy can be a set of moves, like Checkers, or a set of rules like Connect Four, or a combination of both.

"And no more useful then a opening book of moves we already have as best."
++ Oh yes, as it gives a path from the initial position to a 7-men endgame table base no matter what white tries. It will never end with 'white is slightly better' or +=, or +0.50, it will only end in 7-men endgame table base draws.

"Your opening book can give no more answers, or give any other moves."
++ All relevant white tries are included.

"Chess is not a forcing game like checkers"
++ Nine Men's Morris is not a forcing game either and has been solved to a draw as well.

Your stance is that the line is unworthy of consideration but you still can't win it as Black against Stockfish.

If you actually had a solution you might even be able to win it against Stockfish.

That's what solutions are for.

That's just one of the reasons why your proposal is to spend 5 years not solving chess.

And 4.82 x 10^44 is the estimated number of positions in basic rules chess, which is a drop in the ocean compared with the number of positions in competition rules chess.

The difference is huge for a strong solution of chess, but not for a weak solution which merely has to achieve the optimal result of the starting position, not to successfully take advantage of blunders by the opponent.

The reason is that if you have been following an optimal strategy in a game, you need not fear a repetition of a position you have already been in changing the value of the position, regardless of the value of the opening position. If the value of the opening position was a win to you, your strategy will never permit the opponent to return to a position (if it did, they could go round in circles so your strategy does not win). If the value of the opening position is a draw, repetitions do no harm to getting the theoretical result.

@7150

"What path to a 7 man position"
++ For example this one: in 57 moves from the initial position to a 7-men endgame table base draw https://www.iccf.com/game?id=1164344

"you would not need to use other peoples work that does solve chess with 7 man or less"
++ Schaeffer had to create his own table base for Checkers. For Chess it already is there.

"why would you even need the table base" ++ Because only the table base provides the exact evaluation draw / win / loss. Nothing else can. 

"you can not force me as the other player to play one of your book moves"
++ All reasonable white moves are covered, black draws.

"Chess is not a forcing game like checkers."
++ Nine Men's Morris is no forcing game either and has been solved to a draw.

"1000"

@7156

"a game where every move by one side is forced"
++ 'For example' means there are many perfect games.
In ICCF WC Finals alone there are > 1000 perfect games.
Nobody said any moves were forced.
If white plays differently, then black draws differently.
How is the subject of weakly solving Chess.

@7157

"1000"
++ 10^17 / 10^14 = 1000.

10^15 / 10^14 =10

So I can fail to solve chess more quickly than you.

@7160

Again explained

Positions                                                             Strongly solving          Weakly solving
All legal moves                                                   10^44                          10^22
Underpromotions to captured pieces only        10^38                          10^19
Reasonable moves only                                      10^34                          10^17

The difference between weakly and strongly solving is a square root
w^d = Sqrt (w^(2d))
w = width, d = depth

10^44 comes from the Tromp paper
10^38 comes from the Gourion paper multiplied by 10 to include 3-4 queens
Reasonable comes from the random sample of 10,000 Gourion positions
If you doubt that, then take any of the 10,000 randomly sampled FEN and try to construct a reasonable game from the initial position that leads to it. You will fail to do so.

You can reduce it even further by not including minor pieces.

You could solve it for the dark squares (just don't include any of the light ones).

Then solve it for the light squares.

Put the two solutions together and you've solved chess.

QED (quite easily done)

It's plainly obvious that if you do not have an algorithm that can solve any position at a very few passes at most (and you don't have such an algorithm) then thinking you can solve chess by positions only is a glaring mistake. tygxc is far from being the only one here who's making that mistake. People arguing against him also seem to believe that it can be solved by assessing positions, rather than games.

Since you can't solve by assessing positions, every position has to be analysed at least in all its reasonable continuations. Again, there's a problem with the extremists on the other side. If you don't accept that 1. e4 e5 2. Ba6 is definitely lost for white, you have no business arguing here and expecting to be taken seriously, because you do not accept the realities. You propose no method of determining what is a reasonable continuation and what is not. That approach makes chess insoluble in any case, because that is just a tangle of lines, with no way to assess them at all.

Therefore, if continuations must be assessed, it makes sense to approach it game by game, which also has the benefit of cutting out illegal positions, which seems to be an enormous stumbling block to what is already an insanely wrong approach: of thinking you can do it by positions, rather than games.

For years now, I've been ( quietly ?! ) confident that there's nobody, pretty much, arguing on these and similar threads, who has any serious hope of coming up with workable proposals. This indicates that they don't understand the problems involved and so, naturally, they will fall back upon the incorrect approaches previously made, by so-called professionals in the field: and all the insanely confusing definitions and other claptrap that comes with their approach. The result is that anyone coming upon all this for the first time and believing them becomes just as confused as they are.

Anyone reading this and believing that they disagree with it really ought to reread it until they understand what I'm talking about.

With a bit of help from me and you, @tygxc should be able to solve chess with less computation than checkers. 

Fair enough, I was referring to the current FIDE rules.

You aren't taking the 50-move rule into account. If you shuffle your king in the position you mentioned, you'd at some point be unable to mate within the 50-move rule, thus changing the evaluation from "white mates" to "draw". Hence my hypothetical algorithm would force mate in less that 50 moves in your given position to prevent a draw, which changes the evaluation. Using the website, you similarly have to take the 50-move rule into account.

No, a consistent result does not indicate perfect play, but perfect play requires a consistent result.

MAR cannot get away from his "chess consists of different sets of rules" ideas probably because he is stuck with the belief that solving chess consists of solving positions rather than games. Naturally, he then has no means to determine if his myriads of positions are legal, which means he has to refer to rules of chess to determine that. Meanwhile, Elroch insists that you cannot determine that 1. e4 ... e5 2. Ba6 loses for white. If that's the case, then we have absolutely no means to assess positions and therefore chess is insoluble, because without assessing positions, it's impossible to distinguish between winning, drawing and losing lines.

I don't believe chess is soluble, until certain parts of chess become mathematically analysable, in such a way that the analytics themselves may be solved. I don't necessarily believe that this will ever happen, either. However there seems to have been no one here who picks up on the unclear thinking of MAR, Elroch and others and challenges it sufficiently. Consequently the two of them impose a kind of smokescreen or blanket of indecipherability on the entire conversation. In their different ways, they are both just as confused as tygxc. So it's good to see It happening.