Chess will never be solved, here's why

The irony of posting "yap yap" repeatedly because you don't find other posters' content worthwhile on the thread seems to have escaped you.

You've already bought yourself 3 reports for spamming...how many would you like?

'The tablebases are lying'.
I like that.
Regarding whether that's right or wrong I have never seen Martin wrong yet.
In over two years of postings he appears to be consistently right.
Is it possible for two people to disagree with both being right?
Yes.
Why is Martin consistently right?
Same reason that Elroch and Dio and mpaetz and MEGA and llama are almost always right.
Because if they didn't have reason to be sure - they'd say so or not post whatever.

oh wait i didnt see how far back the history goes u right. although who's to say that the history before didnt also hit kc1 3 times.

let them call it yapping, it doesnt hurt us, and it seperates the thread from tygxc. no harm no foul.

Spamming is spamming, whether it's 'join my club" or "check out my AMA thread", or repeating the same phrase over and over or posting a page full of white space or gibberish.

@12343

"Mathematics is about abstract truth" ++ Were Fermat's Last Theorem or the four color theorem true or not before they were proven? Not true before date, true after date?
Are the Riemann hypothesis or Goldbach's conjecture true or not? It this digital true/false, or are there gradations of truth?
Were Ramanujan's theorems true or false when he formulated them before somebody else proved them?

"This truth is determined by deduction from axioms to theorems"
++ Truth is independent from the deduction.

"within mathematics, all the reasoning is deduction"
++ No, for example Labelle used a Monte Carlo method to establish a lower bound of 10^29241 for the number of chess games.
Also Tromp's method of counting chess positions, sampling 2,000,000, inspecting the fraction of legal ones, and then concluding on the number (4.82 +- 0.03) * 10^44 of legal positions is inductive.

"you can never prove it is true for observations to come"
++ The power of a theory is its ability to predict. All engineering and large parts of physics depend on the truth of theories for observations to come.

after so much time I'm here to bump this thread again.

bump

Da - joined chess.com 8 days ago.
And yeah it does look like 'the same thing'.

"Are the Riemann hypothesis or Goldbach's conjecture true or not? It this digital true/false, or are there gradations of truth?"

how about you look up the definitions of a hypothesis and conjecture. we dont know if they are true or not. IT'S IN THE NAME.

"unproven" is literally the the truth value desribed.

Check the ply count in the FEN of the starting position (setup parameter in pgn).

u win.

Note that the relevance of the history to affect optimal play from a tablebase position is only relevant in the odd situation where a tablebase had not been available to the player of interest for immediately preceding positions with exactly the same pieces on the board (which are certainly positions in the exact same tablebase).

So, barring the very odd circumstance of tablebases suddenly becoming available part way through play with a certain material balance on the board, it's only real relevance is the esoteric example where the players did not have access to a tablebase, played into positions in the tablebase, and then post-game analysis reveals that a player gave up the win by blocking it by a double repetition of a position after which he had needed to play differently to get the win, but now was doomed to run into a three-fold repetition.

Of course, the case of the 50-move rule (or similar) is similar. It can reveal when a player has faffed around to waste time to reach a position (all the time in the same tablebase - i.e. specific material), reaching a position which would be a win if he had enough moves but which is not without.

In this situation, the tablebase will tell us that the mate that would be available isn't because the 50 move rule scuppers it, but does it tell us when (and how) a player might still have a win by finding a different route, necessarily one that zeros the 50 move counter?

But does mean exactly that the tablebases are not strong solutions under competition rules, which is where we came in. @tygxc regularly asserts it (I think he thinks there's only one kind of tablebase) is.

The relevance is that he makes the same assumption that basic rules positions correspond univocally with competition rules game tree nodes when he applies SF quoted nodes/sec in his calculations to basic rules position counts and assumes transpositions of basic rules positions correspond with transpositions of competition rules game tree nodes.

As well as the question at the end of my edited post above, consider the odd case where one player is table-base assisted and has to take advantage of the other player having given up the win by faffing around and becoming blocked from the win as a result.

Do tablebases contain enough information to enable this assisted player to force the draw? There is only a question about play that takes place before the next irreversible move - after that point normal use of the tablebase becomes optimal again.

tbf how can you expect tygxc to follow that logic when he thinks that you can just ignore "unreasonable" moves in a game solution.

the tablebases COULD function as a "strong solution" (in practice) in the process of a weak solution to solving chess, as you would never reach a repeated position in such a game tree unless repetition would already be infinite.

Im well aware that tygxc doesnt have the knowledge to make such a distinction but id like to chime that in.

Except for castling rights the Syzygy tablebases are weak solutions of all ply count 0 positions with up to 7 men (and many others), but they don't fit the definition of a strong solution of almost any position (stalemate and checkmate positions being exceptions).

As you and @Elroch (and I) have pointed out, they don't need to.

[Advert removed; DS]