Chess will never be solved, here's why

Quoting for posterity, Mr. Travolta...and yes, that is narcissism (at least when you add it to your supposedly being the best at everything you talk about). Disco? Really?

@12882

"The ICCF does this. If a position is reached where Syzygy tablebase shows a win for White and White claims it, the game ends right then and there 1-0 and vice versa for Black."
++ Yes, but table base win claims do not happen in ICCF World Championship finals,
only table base draw claims.
Even win claims that exceed the 50 moves drawing rule are acceptable, but never happen.

The 112 draws out of 112 ICCF WC finals games thus provide an ultra-weak solution to Chess: the game-theoretic value of the initial position is a draw.

That is in line with the initiative being a white advantage of +1 tempo = +0.33 pawn, not enough to win and each further move diluting the advantage.

The 112 draws constitute at least part of a weak solution of Chess and redundantly so.

The games represent 10^17 positions considered and 90*17*2*2 = 6120 CPU years.
Schaeffer weakly solved Checkers considering 10^14 positions and 50*2=100 CPU years.

an ultra weak solution is a rigorous mathematical proof, by definition the games cannot be an ultra-weak solution.

why do you keep sidestepping mathematical rigor tygxc?

O - that's Optimissed - finally got something right.
He began a post correctly - but again I didn't read the rest of that post.
He began 'It isn't a case of authority'
Yes. I straightened him out on that.
He has no authority.
Correct.
So for a few seconds he was relieved of his continuing delusion that he does.
Of course Dio is not a psychopath and O is slandering Dio in saying so.
O continuing with his trolling.
-------------------------------------
But - there's instead the forum subject.

tygxc apparently wants to hide behind the semantics of 'game-theoretic value'.
A fancy term that could be applied in any invalid way to many things.
tygxc has been told over and over again that today's chess engines drawing each other proves Nothing because as engines progress they would beat older engines.
Why does he not concede that point?
The answer: because if he does then he's Got nothing.
If a salesman wants to sell paper encyclopaedia books he does not concede to customers that its easier and more updated to look things up on the internet.
That's if such sales even still exist.
Will tygxc ever have 'customers'?
None so far.

Let's be quite clear about this: the part of ICCF games after the game enters a table base is reliable and could be part of a weak solution. The part before the table base is unreliable and can't.

An excellent reason this is so is that Stockfish blunders in 6 and 7 piece table base positions without the table base, so why would anyone be sure it won't blunder in 8, 9,10 piece table base positions without a suitable table base existing?

There is no established 'perfect game' reaching the table base positons.

I've done no such thing.
BC is projecting. 
Plus nobody gets 'exemption from criticism' because he/she is not tygxc or Optimissed.
Plus tygxc does infinitely better than Optimissed does.
But then - everybody does.

@12900

"the part of ICCF games after the game enters a table base is reliable"
++ There is no such part: they claim the draw and that is it.

"The part before the table base is unreliable and can't."
++ It is because the positions considered 10^17 and the effort 6120 CPU years is more than how Schaeffer weakly solved Checkers, and also because it is redundant and thus fail safe.

"Stockfish blunders in 6 and 7 piece table base positions without the table base"
++ Won positions are irrelevant to weakly solving Chess.
Stockfish blitz games are irrelevant, ICCF is average 5 days/move.
ICCF WC Finals is much stronger than Stockfish, that is why the 17 finalists made it through Preliminaries, Semifinals, Candidates, and finally the Finals.

Russian ICCF players use worse hardware because of sanctions,
but nevertheless 4 of the 17 finalists are Russian. How can that be?
The late ICCF GM Dronov, 3 times ICCF World Champion answered: general chess culture.

And for ten years.
Because he's got problems. Problems he denies. And falsely accuses others of.
He's been muted by chess.com on multiple occasions including two very recent mutes. And blocked by good posters.
The real conversations of forum topics have to go on around him not through him.
And - the real friendly conversations too.

@12781

"ICCF snoozefest"
++ That is disrespectful. ICCF games are quite sharp and often end with spectacular sacrifices to end in a perpetual check to secure the draw by 3-fold repetition.
Here is one recently finished game: black accepts 2 poisoned pawns,
is later forced to give up the exchange and saves the draw by 3-fold repetition.
https://www.iccf.com/game?id=1360192

tygxc it would serve you to actually look up the difference between a mathematical proof and a casual proof.

@12579

"do we have the 1st 20-ply completely described yet ?"
++ Here are 14 ply:
1 e4 (also tried 1 d4: 54 draws, 1 Nf3: 18 draws)
1...e5 (backup 1...c5: 17 draws, 1...e6: 2 draws)
2 Nf3 (also tried 2 Nc3: 1 draw)
2...Nc6 (backup 2...Nf6: 6 draws)
3 Bb5 (also tried 3 Bc4: 7 draws, not tried 3 Nc3, 3 d4)
3...a6 (backup 3...Nf6: 4 draws)
4 Ba4 (not tried 4 Bxc6, other tries dismissed by logic)
4...Nf6 (no backup)
5 O-O (no other tries)
5...Be7 (backup 5...Nxe4: 1 draw)
6 Re1 (not tried 6 Bxc6)
6...b5 (no backup)
7 Bb3 (no other tries possible)
7...O-O (backup 7...d6: 1 draw) 1 draw

^^^ (fyi this is not a complete description as it literally does not address any of the positions)

It is great that @tygxc is such an advocate for the work ethic, arguing that if someone consumed more computing time than was used to solve checkers, they should be assumed to have solved chess. But this is NOT how things are proved.

Note that the 116 draws are in no sense a single attempt to solve anything: they are independent games without any access to each others analysis. Solving checkers required a single proof tree which was able to verify that every legal opponent move gave a position already in the proof tree or added a new one. It is blind to think 116 independent games is similar to this, still less actually this.

You need a proof tree to solve chess. Chess being 10^24 times more complex than checkers, the computing time is not practical.

In a game where a program analyses a given number of positions, more positions needed for a proof tree are left unanalysed past a zero depth evaluation. This is just not good enough.

(Note that it is efficient to use a single node in a proof tree for the same position as black and white (a symmetry of the game) and for left right symmetries. When there are no pawns up-down symmetries can be added. This reduces the number of truly distinct positions by a factor between 4 and 8).

its hilarious watching tygxc downvote factual rebuttals because he cant address them

it's so goofy that tygxc seriously tried to claim that the air tight, mathematically rigorous algorithm used in the connect 4 solution gives him the excuse to disregard entire branches based off of "game knowledge".

'oh they both used game knowledge so every other aspect of their logic must be the same'

Yes, he thinks in an unmathematical way.

The test of his ideas is if he tries to publish a paper trying to convince people that chess has been solved. The contrast with anything worthy of publication is gargantuan.

I disagree that it's a good test, since inevitably it would be merely "further evidence". I do think that the much vaunted GM team has gone about it in the wrong way because the new evidence doesn't take into account the obvious truth that if a forced win were to actually exist (which I disbelieve in but then this is an hypothetical) it would exist in a place as yet unsearched. That is, 2000 to 3000 move games with a preponderance of single pawn moves and a tendency against early simplification.