Chess will never be solved, here's why

I wonder how does the algorithm filter the moves? Just because 2.Ba6 is a piece sacrifice? There are lots of piece sacrifices that are unclear or even win the game.

No I'm not missing any point. I'm probably the only one here who isn't missing anything. Probability isn't involved, except in the minds of those people who don't understand that this is a cut and dried situation. In the context of the larger conversation with tygxc, it's this kind of mistake that makes these people extremely ineffective in their discussions with tygxc, because they don't realise that ty is completely entitled to assume that Ba6 loses. They tried to use an innocent point that ty was making against him.

If this were a philosophy debate, which it isn't, their ideas would be dismissed out of hand, since the entirety of science rests upon inductive evaluations and deduced proofs only proceed from there. So the people who demand deductive proof in all circumstances can never really find it. At best they'd be laughed at, sympathetically.

If you can't tell that 2. Ba6 is definitely a loss then you really shouldn't be commenting here. Read my posts. Probably best to ignore theirs.

The position is either a win, draw or a loss. In this case, it's a win for black. Probability isn't involved. How can it be, except for someone who believes that the win is down to chance?

It's only inappropriate in your mind.

If Magnus was here and he said there's no need for proof, would you tell him he's wrong?

If you, as a chess player, can't see that particular position is a win for black, you aren't a very good chess player are you. One or two people have incorrectly and rather foolishly brought in the completely unrelated analogy that if there were gazillions of positions then I would get some wrong. That kind of argument is out of desperation. There's one position and it's a win for black. Are you telling me that you'd have this argument with Magnus and make yourself look like a complete fool?

I don't think you would. At the very most, it's down to a personality difference but I'm the one with the philosophy degree and epistemology is my specialisation. I'm aware it comes into statistics a bit but only in a simplistic form.

So you're still going to tell Fischer he's wrong? My impression is that certain people here want to control how others think. However, the only probability is the uncertainty in your own mind.



I think your main mistake is the assumption that probability is involved and that the optimum or theoretical result of that position has some relationship with chance. Everything else flows rather awkwardly from there.

The only possible origination of this confused idea is that doubt exists in your minds, concerning the result. Therefore all your speculation about the necessity for proof is subjective only. If it was Q + K vs K, would you have the same difficulty in assessing the outcome?

I think opening the door of an aeroplane at 20,000 feet and jumping out would be a good test. After all, there's no proof that you'll die. Go for it .... you could learn from it.

Yes, the position is either a win, a draw, or a loss. But until every line has been calculated, there is no way to prove that it is a win for black. I believe it's a win for black, I'm almost certain of it. But there are many positions where a player is down a bishop but the position is a win or a draw. It's possible, no matter how unlikely, that white has a line that gets to one of those win/draw positions by force, even after 2. Ba6. And it will remain possible until someone, or something goes through every possible variation of 2. Ba6.

That seems pretty obvious. There are sacrifices in chess that often lead to winning. So it's also possible there is a very deep sacrifice (that no computer has even come close to discovering) somewhere in the opening or middle game. Which leads to a forced win, from the opening position. For either black or white. 

In that case, the kind of proof demanded is incorrect, if it cannot possibly be given. It becomes necessary to think in different ways. But in reality there's no need to look at every line because we can know it's a win for black.

If this were a philosophy debate, which it isn't, their ideas would be dismissed out of hand, since the entirety of science rests upon inductive evaluations and deduced proofs only proceed from there. So the people who demand deductive proof in all circumstances can never really find it. At best they'd be laughed at, sympathetically.

For someone who pretends she believes that chess is a forced win for white, everything is obvious if it's contrary.

D.B. Cooper jumped from 10,000 feet. He's sort of a local legend. So from 20,000 feet you'd have the advantage of about twice as long to decide your next move. 

I was just agreeing that it's obvious it's a win, loss, or draw. And also obvious that it's possible that what currently is believed to be a loss could in fact be a draw or win. Over time the ability to assess chess positions gets better, so opinions change. 

Did he live?

What I can't quite understand is that these people think probability is involved in some way.

I know very well I think differently from many or maybe from most people. I try to adopt the mental attitude that has the best chance of making the World conform to my wishes. If I want a football team to score, I visualise it, complete with bulging net and rapturous joy of the scorer. Probably, most people here believe that can't possibly work and that I must exist in a Word of delusion. But it's just about different ways of thinking. Some people think their way of thinking is the only rational or reasonable way. I like certainty because it empowers the mind. They prefer self-doubt because it causes them to feel that they're wise. So we're all winners really, but some much more than others.

If you didn't like that example, the fact is that White cannot necessarily force mate in KRK even from White to play positions (which rules out stalemates and hanging rooks); in fact from the majority of White to play positions.

Here is another you might like to try. See if you can force mate from the final position (shown) against the computer.

But it can be given (whether hypothetically or when the technology is developed in the future), because there are a finite number of possible positions after 2. Ba6. I'm not sure where you stand in the original "Will chess ever be solved?" question, or even if you've been involved in that discussion at all, but the very question of whether or not it is possible to analyze every position (from the starting position or after Ba6) is the entire point of this thread.

I'll tell you what my opinions are then.

Chess is undoubtedly a draw. Nothing has occurred to change that assessment. Chess will never be solved. I asked my son, who's a mathematician, if chess can be represented mathematically. He said not. Therefore at least at the moment it isn't possible to develop algorithms that are perfect, which would be needed if the long route of analysing all possible chess games is to be avoided. Many here think that "only" positions need be analysed. that's incorrect, since there IS no algorithmic analysis tool that's accurate, so the long route must be taken. I'm only assuming my son is more qualified than others here to judge. I may be wrong.

So in my opinion, chess will never be solved, if that means all possible meaningful games. Even with meaningful games, the number of move permutations is effectively infinite, in that there may not be enough time in this galaxy to do it, given present technology.

The question of Ba6 came about because tygxc asserted that irrelevant lines can be recognised and rejected. That is, lines where an obvious mistake has been made need not be analysed. He gave the Ba6 line as an example. Their objections are correct in principle, because if Ba6 is just an example, there will be many other lines, less clear.

Unfortunately, we have a lot of people who are pretty poor at debating. They tended to take the Ba6 line as a concrete example in that they claimed it isn't possible to know the result of that 100%. Although their point is good in principle, they made it in the wrong way, since it is definitely possible to know absolutely that Ba6 loses for white. It's what comes of trying to amuse ourselves by debating things with people who miss nuances and get the emphases on points wrong. It happens, it's life, but when you combine that with dogmatism, they always have to be right. They haven't answered me when I asked them if they would correct Carlsen equally vehemently if he claimed it is a loss. Dogmatists congregate on threads like this one. "Normal" people tend to talk elsewhere.

Interesting, I appreciate you explaining your perspective, so thank you for that

I suppose it really boils down to how far technology develops. There was a time where the current level of engines was believed to be impossible. 

These people, talking about "proof", miss the point because they're only interested in deductive proof and if the premises don't exist, then a syllogism is impossible. So, instead of becoming nihilists, which is what they're doing, they need to calm down a bit and, in particular, they should desist from telling other people how to think.

@4222
"whether hypothetically or when the technology is developed in the future"
++ Present technology can weakly solve chess in 5 years now

"there are a finite number of possible positions after 2. Ba6." ++I showed it is a forced checkmate in 82. There is a finite number of possible positions in the initial position as well.

"Will chess ever be solved?" ++ Chess can be weakly solved in 5 years, but if it will depends on money to hire the assistants and rent the computers.

"whether or not it is possible to analyze every position" ++ The point is that not every position needs analysis, some positions like 1 e4 e5 2 Ba6 can be dismissed immediately

@4208

"There are lots of piece sacrifices that are unclear or even win the game."
++ That is right, but 1 e4 e5 2 Ba6 is no such case.
In game theory unclear does not exist: it is either a draw, a win, or a loss.
Per Capablanca any material advantage no matter how small is enough to win the game,
when all other factors are equal.
In this case black is up a bishop, a considerable material advantage.
All other factors are equal.
So that position is a loss for white.
I even proved above it is a forced checkmate in 82.
So weakly solving chess does not need to burn engine time on 1 e4 e5 2 Ba6.