Chess will never be solved, here's why

@11540

"five candidate moves earlier (a couple of weeks since) but I also mentioned that to do the job properly, maybe nine candidate moves"
++ Chess is full of transpositions.
There are 10^38 legal positions without underpromotions to pieces not previously captured.
An average ICCF WC draw lasts 39 moves.
10^38 = 3^80
There are no more chess positions than there are chess positions.
So in 40 moves with 3 choices per ply that do not transpose we have the whole of Chess.
In some positions there are more possible legal moves: 20 to start with.
In some positions there is only 1 legally forced move, e.g. Qxd8+ requires Kxd8.
The number of 3 choices is that low, because there are so many transpositions.

I quote Schaeffer: 'The perfect Alpha-Beta search will halve the exponent'
That is where the square root comes from.

Saying blitz chess is completely useless for years with NOT GOOD evidence just like this sitution

Meanwhile switched to blitz chess...

@11544

"the number of errors is still unknown" ++ But it is not plausible there would be 70 games with 0 error, 37 games with 2 errors and 0 game with 1 error. All errors would have to magically come in pairs. There are some arguments for some errors to come in pairs, but not all errors.

"most of the time stockfish likes drawing moves" ++ ICCF correspondence is between humans that use engines. ICCF players do not play whatever move the engine evaluates best,
that is a recipe for losing, as the opponent uses an engine too and can find out weaknesses.
Please read the interview with reigning ICCF World Champion Edwards I linked to.
It shows how he plays: checking ideas with his engines.
He prepared his tournament analysing all games of his opponents.

How about 3 errors and ends up a draw (even though the other side didn't make any )

You would not be able to tell unless stockfish or humans got even stronger at chess claiming you can is the falacy

Or 2 errors on both sides and each where winning and some point like you can make no mistakes in chess but still miss winning moves even if according to stockfish (not trustworthy) that here is no "errors "

@11551

"How about 3 errors and ends up a draw (even though the other side didn't make any )"
++ That is impossible. Definition (per GM Hübner):
error (?) = move that changes a draw to a loss, or a win to a draw
blunder or double error (??) = move that changes a win to a loss

"You would not be able to tell" ++ I cannot tell where the errors are.
In a decisive game the errors is usually the last move of the losing side.

"Or 2 errors on both sides"
++ Yes, a game with 4 errors can be a draw. However, an error distribution of say 60 - 0 - 30 - 0 - 17 is not plausible: games with 0 errors and games with 2 errors, but none with 1 error.
Games with 2 errors and games with 4 errors, but none with 3.
All errors would have to magically come in pairs.
There are arguments for some errors to come in pairs, but not all errors.

"according to stockfish" ++ I do not go according to Stockfish, I go according to the final result.

according to stockfish" ++ I do not go according to Stockfish, I go according to the final result.

Same thing I was talking about the moves in the game where the actual errors happen not the result the result is meaningless if both sides miss wins and dont know that (which would look like no errors ) so technically it is according to stockfish (or whatever human blugged in the move ) since I'm talking about the actual moves of the game 

@11553

If chess is a draw, then each drawn game must contain an even number of errors,
and each decisive game must contain an odd number of errors.
If chess is no draw, then each drawn game must contain an odd number of errors,
and each decisive game must contain an even number of errors.

4 blunders 4 blunders

White makes the 4th blunder last right after black blundered a 4th time

Black wins because it spots the last blunder

Now read my post.

@11554

"A genuine weak solution would provide you with enough information to draw against ANY opposition"
++ Yes. I say they are almost there. A strategy to draw against any opposition is to follow an ICCF WC Finals draw for as long as possible and then use an engine at 5 days/move.

No. It needs to specify the moves to make in any situation that is reached. No-one was foolish enough to claim that Chinook had solved checkers because it was so strong.

Your confidence in the current Stockfish is especially ludicrous given its known lack of reliability in tablebases positions: it is highly likely it will lose some games against future engines, like all its predecessors.

Almost is not a weak solution almost is still almost

Your arguments are all based on "well we are close "

@11156

"White makes the 4th blunder last right after black blundered a 4th time"
Let us start with the initial position being a draw.
White makes an error (?): black has a winning position.
Black makes a blunder (??): white has a winning position.
White makes an error (?): the position is a draw.
White makes another error (?): black has a winning position.
Total: 5 errors: 3 simple errors (?) and 1 blunder (??)

@11560

"Almost is not a weak solution almost is still almost"
++ The 107 ICCF WC Finals draws form at least part of a weak solution of Chess.
Not yet complete, but redundant: more than 1 way to draw.

Exactly, @BigChessplayer665.

It would have been possible to say 'we are close' when Karpov and Kasparov had a long drawing run. It is still possible to say 'we are close'. That's empirical chess.

@tygxc is not consistent, and is merely woefully deluded when he thinks a tiny sample of 106 games is a solution of chess - and then 'almost a solution' - any more than a couple of dozen Karpov-Kasparov games. At least they had two different players!)

@11564

"when Karpov and Kasparov had a long drawing run"
++ The strongest humans in over the board play at 3 minutes/move are at average 1 error/game.
The strongest engines at 3 minutes/move are lower.
The strongest engines at 5 days/move are even lower.
ICCF WC Finalists with engines at 5 days/move came closer and closer to 0 error/game.
ICCF WC Finalists with engines at 5 days/move are now at 0 error/game.

@11564

"they had two different players"
++ The ICCF WC Finals has 17 different players.

Is your new plan to fill the forum with irrelevant uncertainty and count that as a proof, @tygxc?

Hint: go learn what a proof is.