Chess will never be solved, here's why

@MaetsNori
"Case in point: as long as engines continue to advance in strength, the chess that we once thought of as "flawless" will eventually be shown to be inaccurate ... "
That's right.
But tygxc never gets that.
And he'll be stuck on this one about the 116 draws for a long long time.
Maybe even another two years or more.
But have you noticed his attempt to get 'wiggle room' by claiming that todays engines would draw future better ones 'by getting five days to move'?

Your posts don't deserve answers. You would be crazy if you really thought they did!

True.

Though, I imagine the players in 2013 were using fives days per move, or more, with the analysis tools they had back then, as well ...

What's impressive about that (to me) is it took SF 16.1 only about 5 seconds to see that 20. Nd4 was a missed opportunity ...

Engines continue to get stronger and stronger.

A few days ago you made a big thing over correcting me whan I typoed. So one good turn deserves another and I gave you some help with your punctuation.

So when you do something, that's a legitimate attempt to troll and it's to be applauded. When I do the same, you say I'm over-reaching. Is that like table manners, when you reach too far for the butter or for the water jug, instead of asking someone to kindly pass it?

You say that I have a fragile ego. Aren't we really talking about you? All the time, everything really about you? And you really believe that every step you take isn't a step further about telling us what you really are. I'm sure you do.

Yes, you've made a good case that engine accuracy is improving constantly and will continue to improve. I've made the case here that a new kind of algorithm is necessary, which approaches chess "from the middle" in order to identify what I have called "watershed positions", which are positions that are most susceptible to non-reversible and sudden alterations in the game balance. There's a lot of work to be done because a way has to be found to identify them systematically, before they can be analysed to identify common factors.

I would guess that work has started at least somewhere on this project since it it's apparent to me, so there will be others who have seen its desirablity.

That's an interesting idea. I'm sure there are arguments that can be made against it, but I do like the concept ...

instead obsesses over a 'cartel' against him

hes not alone. he has me.

...in your parlance refers to only two, Lola and...

[i wish hed feel antisocial now & again...] next time ?...TRY to leave me outta ur thoughts ?...plz ?

Thank yes, it was a few years ago in one of these solving chess threads. When the conclusion was reached that it's impossible to solve chess via a brute force method, due to volume of analysis and related storage, retrieval and analysis difficulties, as well as the fact that it would take too long a time, I started thinking about the possibility of turning chess into a set of equations. So I asked my son, who is a mathematician, although I think he's a kind of glorified computer programmer nowadays (he tells me he's a data scientist) and he told me that it is definitely impossible to turn chess into a set of equations. No ifs or buts ... it's so far beyond the ability of any mathematician nowadays that he thinks it can be regarded as impossible.

So that's when I started thinking about algorithms and how to improve them. I arrived at the conclusion that it was first necessary to design algorithms to identify positions that can be thought of as "watersheds" and then, algorithms to analyse them, before trying to use the results to produce game playing algorithms.

I outlined the idea here but apparently nobody could understand what I was talking about.

@14794

"We can't call these "certain" draws, because we can't claim certainty over the accuracy of them"
++ The 114 games in the ongoing ICCF WC Finals all end in certain draws: either a 3-fold repetition, or a 7-men endgame table base draw claim, or by agreement in a position where neither player has any hope to win.

"I've pulled up a game from the ICCF WC from 2013 (one decade ago)."
++ In 2013 there were 20 decisive games and thus errors.
Now they have reached zero decisive games and thus zero errors.
Humans got better, data bases got better, engines got better, hardware got better.

"the chess that we once thought of as flawless"
++ In 2013 it was not thought of as flawless, as there were 20 decisive games.

"This applies to the current ICCF games, as well." ++ No, because all 114 games are draws.

"engines from the year 2035 and beyond will almost certainly find mistakes"
++ They cannot drop below 0 error/game. They will be able to do so in less than 5 days/move.

"We won't be able to declare absolute certainty over the accuracy of our chess until a hypothetical 32-man tablebase arrives" ++ That is not true. We are able to declare absolute certainty over the accuracy of Checkers, though it is not strongly solved to a 24-men table base.

Well, I wouldn't say I'm obsessed by it, Lola. I have other stuff to keep me busy. But thanks, it isn't lost either on me or on a lot of people that you're intelligent and also, more importantly perhaps, you are a good person.

That there is a cabal of trolls here is beyond question and only a fool would deny it .... or one of the cartel. They are not "against me" in particular. For some of them it's a game ... maybe an exciting or fun one. Some are alts, of course. The general idea is to cry with one voice "Unbeliever!" when anyone doubts whatever message they are attempting to instill (brainwash) into others. The fact that they continually deny to be "one voice" and pretend they're acting independently is obvious nonsense, since if that were true, why do they continually agree with one-another? It certainly isn't because they are right or correct.

So there is a cabal of trolls in these forums and that's beyond question.

Just to provide a useful translation of Optimissedlish to English.

'cabal of trolls' -> 'large number of people who disagree with Optimissed'

Four or five trolls = a large number?

Does this mean that you're one of the cabal, Elroch? I hadn't implicated you. This is an eye-opener.

OK I know I just asked for it. You're now going to pollute this thread with a host of trolls, their nephews and uncles and also their alts and they will all anonymously (because of course, they aren't connected with each other in any way and always act independently) cry with one voice that the cabal will triumph and Optimissed will perish.

The game I showed was not a decisive game - it was a draw.

By your suggestion, an ICCF WC draw means "zero errors". I've shown this to untrue.

This game was a draw - and it had errors in it.

Therefore: ICCF WC draws can still contain errors.

114 draws in current ICCF WC games simply means that there are 114 draws in which errors may (perhaps "will") eventually be found ...

Unfortunately, this point of mine can't be proven or disproven until the year 2034 arrives (or until some other revelation arrives in the meantime). Until then, time will have to tell ...

No, rhetorically speaking, it is merely a result of a belief in a certainty that chess, from its starting position, is drawn with best play.

You really are not keeping your eye on the ball. Agreeing with the cartel may have upset your accuracy.

Hi I wasn't following the games but by "error" do you mean it in the sense that it's being used here .... that the errors in the games were such as to change the game evaluation (from a draw to a loss and back again, perhaps?

I use the idea or phrase "good move" to refer to any hypothetical move which doesn't alter the game state (the game evaluation). Anything else is rather subjective, since for instance Emmanuel Lasker used to make dodgy moves on purpose.

Well, yes I'm merely using tygxc's term for ease of conversation ...

In this case, SF 16.1 found a sequence of moves that pushed the evaluation to +1.5 in White's favor, after a few moves ... in terms of engine play, I'd consider any move or sequence that allows a +1 or -1 change in eval to be a "mistake" (or "error", or whatever term we prefer ...).

But yes, in terms of human chess, I agree that we can sometimes call dubious moves "brilliant" for their impact on the game - especially if they give the opponent a hard time, despite not being the best.

If you're using it in same way, then 114 games all drawn does mean that the odds against the errors all occurring in pairs is astronomical. I disagreed with Elroch regarding the expected error distribution. Since chess hasn't been solved, we do not know the distribution of errors. I believed that he was taking advantage of the cloudiness of the situation to portray the odds as being much lower than they really are. I think this is a case of interpreting a statistical device favourably to a belief, rather than adjusting the belief to reality, whatever that may turn out to be (meaning that as yet, we don't know).

It means that tygxc certainly isn't the only one using assumptive thinking and the attempt is being made, wrongly in my view, to portray him as using assumptive thinking whilst the trolls are not, which is incorrect. I'm using "troll" to refer to those who have bought into the idea that chess can be solved by deduction, since the people acting trollishly ALSO believe that. They try to do anything to win their argument and are not concerned with truth or realities. Just with their egos, really.

I've explained why it can't be solved deductively and my explanation has not been (and cannot be) refuted. This is slightly off the point but Elroch uses assumptive thinking just as much as tygxc. The rest of the people here apparently aren't capable of making an accurate judgement regarding the matter.