Chess will never be solved, here's why

#3353

"Does your proposed method of solution involve playing any real games?"
++ Proposed method: take an ICCF drawn game. Look at 3 alternatives for the last white move. check they lead to draws as well. Look at 3 alternatives for the 2nd to last white move. Check they lead to draws as well. Then 3rd to last move, 4th to last move all the way down until another predecessor game is reached.

"If I kibbitz the position I posted it settles down to -0.66"
++ KNN vs. K is +6.00.
The provisional evaluations serve to identify the 4 candidate moves to be investigated.

"in winning positions that are out of the players' depth the players will also agree a draw"
++ No, ICCF grandmasters just play on if there is even the slightest possibility of the opponent going wrong even if it is objectively a draw. Remember 136 games, 127 draws, 9 decisive games.
If a player has no hope of saving a game, he resigns.
If neither side has hope of winning a game, they agree on a draw.

#3352
"the assumption of statistical independence between errors is not different from the example of the room"
++ It is different. I gave several examples from other sciences.
Assume something. Calculate based on that assumption. Verify the assumption. Assumption and calculation valid.
If you demand all calculations done with relativistic quantum mechanics, Navier-Stokes equations, Maxwell equations, then the world around you would not be as it is. We would live like in stone age.

In that post I explained why those kind of examples are different. Do you not agree? If so, why?

#3358
" You said: the classical model is validated by experiments" ++ i.e. induction
I say my method is validated by deduction.
I say the error distribution for 127 draws and 9 decisive games is 126 - 9 - 1.
If you were to say after investigation it is 125 - 9 - 2, then I am willing to accept that.
On one hand you say you can explain the data, but on the other hand you refuse to present an alternative error distribution saying you cannot tell.
So you can tell and you cannot tell at the same time.

If you say that dark matter is chocolate because chocolate is dark and sticky and I say I cannot tell, it does not mean that dark matter is chocolate. Data do not include the error rate. For the seventh time: what is the difference between your reasoning and the example of the room?

#3361
For the 7th time: try come up with a plausible alternative error distribution in lieu of 126 - 9 - 1.
I say 3 + 5 = 8. You say that is wrong because the continuum hypothesis is not proven. I then ask what is 3 + 5 in your opinion? You say you cannot tell.

You can tell that you cannot tell. This is perfectly clear to everyone else in the world.

#3363
On one hand he says he has an alternative explanation that does not need the assumption.
On the other hand he refuses to present an alternative plausible result.

How far is a car from its starting position after driving 400 km north and then 300 km east?
Me: assume Pythagoras, answer = 500 km. He: That is wrong. The Earth is not flat so spherical trigonometry is required. Euclides' axiom is not proven. We cannot tell.

Do you even read, or think, before you post?

People have not to come up with an alternative conclusion to challenge your conclusion. I explained why your assumption of statistical independence is not plausible, though. Do you have an objection to the last post about that?

If you really want to use such metaphor, you say x = 8 and I explained why (errors not statistically independent) x cannot be determined. Your attempt to prove otherwise is fallacious because it's circluar thinking. Do you want to prove otherwise? For the 8th time, what's the difference between your circular reasoning and the example of the room? Do you think that in that example the conclusion validates the premise?

I answered your questions.

Well, at least that answers @playerafar's conjecture that you're a flat-earther.

#3365
For the 8th time: assumption - calculation - validation are commonly used in many sciences.
It may appear circular, but it is valid.
Dark matter is unrelated to chocolate, and your room story is silly.
I answered your questions.
You did not answer my question about your alternative for 126-9-1.
If you cannot tell, then you cannot tell I am wrong either.

#3366
I am even heliocentrist.

False.

Carlsen was convinced he could get a desirable result against his opponent with this line. He had no confidence he could achieve the same result against Leelazero (say), so the game theoretic value of the position is not actually the issue. All that matters is practical chess is the statistical expectation of results.

#3369
Both Caruana and Carlsen ran LC0 on the cloud engines they rented for preparation.
If Carlsen was not sure of 1 e4 c5 and of 1 e4 e5 2 Nf3 Nc6, then he would have varied as black, but he did not. If Caruana and Nepo were not sure of 1 e4 e5 2 Nf3 Nf6, then they would have varied as black, but they did not.
If Carlsen had thought either 1 e4 or 1 d4 had a faint chance of winning, then he would have played that exclusively, but he did not: he varied 1 e4 and 1 d4 and 1 c4, and within that he varied variations as white.

This thread is surely the most nonsense thread of them all. Now to work ....

This thread must never end.  Here's why ...

I said I explained the data, and the draw rate is a datum; the error rate is not. I said that the statistical dependence of errors can be deduced without major assumptions.

First you compare your theory to Einstein's to explain why nobody thought of it before, then you compare it to elementary problems everyone can solve, to make objections appear overly complicated. It's convenient, but not much coherent.

Scientific reasoning is coherent with external evidence, too. Your reasoning is just coherent in itself. Systematic errors are due to the difference beween the "subjective" and the "absolute" evaluation, as you call them. The subjective evaluations of a move m by two strong engines are strongly correlated, and the ability to exploit an error is correlated with the evaluation too. So player₁'s systematic errors cannot be uncorrelated with player₂'s systematic errors. I explained that in more detail previously.

I did not mean your theory is just like "dark matter is chocolate". I meant that we are not forced to refute a theory by providing alternatives. If we don't do that, the theory is not necessarily good. We can prove it is not good in other ways. Nonetheless I apologize, because I stated that, in a way that could be easily misinterpreted. "The room story is silly" is just an "appeal to the stone", however.

You may be correct, but the numbers are unreliable if the errors are statistically dependent.

     It will end only when chess is solved.