Chess will never be solved, here's why

Sort:
Avatar of Optimissed
Optimissed wrote:
tygxc wrote:

#1487

"knocking 10^44 down to 10^17"
++ Can we please agree that the 10^44 is too high? Tromp exactly counted the number of possible chess positions. Then he randomly sampled 10000 of these. He found 543 of these legal. Thus he arrived at 10^44 legal positions. However all of his 543 randomly sampled  positions found legal contain multiple excess underpromotions. Such positions can never occur in a reasonable game with reasonable moves and hence not in an ideal game with optimal moves. I gave a proof game for one of his randomly sampled positions. An ideal game with optimal moves would have an accuracy close to 100% and an average centipawn loss close to 0. The proof game for the randomly sampled position has an accuracy near 0% and an average centipawn loss close to 500. So none of the 543 positions play any role in weakly solving chess and the 10^44 is too high.

"Relying on Stockfish for perfect evaluations to bridge to the actually perfect evaluations of tablebases"
++ No, I do not rely on Stockfish for perfect evaluations.  I rely on the 7-men endgame table base for perfect evaluations draw / win / loss. I use Stockfish only to generate the candidate ideal game i.e. to guide the search towards the 7-men endgame table base. The proof that all black moves are optimal comes retroactively from the 7-men endgame table base giving a draw. The proof that all white moves are optimal comes from exhaustively investigating all reasonable white move alternatives.

"Removing all promotions"
++ The Gourion count 10^37 excludes all excess promotions, i.e. all promotions to a piece not yet captured or i.e. all promotions needing to borrow the promoted piece from another box of 32 chess men, not all promotions. I agree that some positions with e.g. 4 queens do occur in reasonable games with reasonable moves and thus can occur in an ideal game with optimal moves. Thus in that respect the Gourion count is too low. Thus the number of legal and sensible positions would lie between 10^44 (Tromp) and 10^37 (Gourion).

"Casting aside dozens of orders of magnitude for "nonsensical" positions (also an assumption) based on sampling a small set set of positions"
++ Tromp randomly sampled 10,000 positions without excess promotions. I presented 4 above. Inspection shows that these positions are not reasonable: cannot be reached by a reasonable game with reasonable moves and thus cannot occur in an ideal game with optimal moves. Tromp conjectured that only 1 in 10^6 legal positions without excess promotions is sensible: can occur in a reasonable game with reasonable moves.

"that limits your reduction to 6 orders of magnitude"
++ That is right: if Tromp's conjecture on 10^6 is right, then the figure of legal and sensible positions would not lie between 10^37 and 10^44, but between 10^31 and 10^38. We do not know if Tromp's conjecture is right. We do not have the number of legal and sensible positions with reasonable excess promotions. It should lie between 10^31 and 10^38, I assume 10^36 to stay at the safe side.

"Assuming Sveshnikov knew anything about solving chess vs. just analyzing openings when there's no demonstration that his statement is anything more than an offhand boast at a dinner party."
++ Sveshnikov knew something about chess. Unlike all here he was a grandmaster. Even if he had not become a grandmaster in 1977, he would have become a grandmaster in 2017 by becoming 65+ Senior World Champion. Sveshnikov was ridiculed for his statements that the Sveshnikov Variation (oh no, the hole at d5), the French Advance (oh no, white loses a tempo), the Alapin Variation (oh no, white gets an isolated queen's pawn) were correct. He was vindicated by Carlsen playing the Sveshnikov at his World Championship Match with Caruana. Maybe it was an offhand boast at a dinner party. Maybe it was a profound prophesy based on years of chess analysis without and with engines.
Sveshnikov claimed he could in 5 years with modern computers and good assistants analyse from the opening to the 7-men endgame table base. So his claim is about analysing openings. That is what he did for a job. Unlike all here he was a recognised world top expert at analysing openings.

"If Tygxc had the money and achieved his 5 year analysis goal, he would be able to produce an engine that plays exceptionally well, perhaps...but it would not be a solution for chess at all."
++ No, I would not produce any engine, I would use an engine that exists to analyse from the opening to the 7-men endgame table base. It would be a weak solution for chess as it would produce an ideal game with proof that all moves are optimal.

"You can weakly solve chess with brute force even without any further pruning" ++ I agree.

"just not within our lifetimes by any foreseeable technology."
++ I respectfully disagree. It depends on the speed of the engine and on the number of positions needing to visit.
Cloud engines have reached 10^9 nodes / second, that is 1000 times faster than a desktop.
The number of legal and sensible positions lies as said above between 10^31 and 10^38 I assume 10^36. That would be the number of positions needed to visit for strongly solving chess i.e. a 32-men table base. The time and the storage are prohibitive.
Weakly solving chess requires to visit less positions than strongly solving chess. Each pawn move and each capture make huge numbers of positions unreachable and thus irrelevant. How many positions are relevant? We do not know until it is done. We may have a better estimate after one opening say C67 is analysed, that would take about 3 months of cloud engine time.
Checkers was solved using the square root of the number of legal and sensible positions. Checkers is no chess, but is a draw as well as chess is conjectured to be.
Losing Chess was solved using the 4th root of the number of legal and sensible positions.  Losing Chess is no chess, but it is close in its rules.
It is thus plausible to assume that for chess it is about the square root too. That leaves 10^18 relevant positions i.e. 10^9 seconds for the whole of chess i.e. all 500 ECO codes.

To weakly solve chess it is not necessary to look at all ECO codes.
If 1 e4 e5 draws, then it is not necessary to establish if 1 e4 c5 draws as well or not.
Instead of 200 ECO codes B00 to C99 only 19 of these are enough.
That is another reduction by a factor 10 i.e. 10^17 positions or 10^8 seconds for the relevant ECO codes.

Concentrating on multiple, excess underpromotions is a flea-bite in the percentage of riduculous positions that are legal. Possibly only approximately zero positions out of every 10,000 are both legal and non-ridiculous, meaning that they are legal and relevant; not containing random moves and gross blunders. Approx. zero out of 10,000 still would lead to a very high number of relevant positions overall. So the first part of your argument is correct.

However, you do fall into error very soon. <<<++ No, I do not rely on Stockfish for perfect evaluations.  I rely on the 7-men endgame table base for perfect evaluations draw / win / loss. I use Stockfish only to generate the candidate ideal game i.e. to guide the search towards the 7-men endgame table base.>>>

SF or another engine can only "guide the search" if its evaluations are perfect, so you are relying on it for perfect evaluations. There's no escape from that. There's no argument you can effectively make, to the contary, since an algorithm is the only thing between a guided solution and a so-called strong solution. A strong solution is meaningless in competitive chess terms and so each game has to be assessed and SF is not up to the task. There's no question about that. It is not debateable and you should accept it. Otherwise you cannot guarantee an accurate solution, which this is all about.

I wish some of the others were capable of arguing coherently and decisively and yet any argument is only as good as the readers' perceptions of it. No-one can be forced into understanding, if indeed they are capable.

<<To weakly solve chess it is not necessary to look at all ECO codes. If 1 e4 e5 draws, then it is not necessary to establish if 1 e4 c5 draws as well or not.>>

Yes, that's obviously correct if the only form of the idea of a "weak solution" is to determine if black can actually draw. But as Elroch correctly asked, what if 1. e4 ...a5 wins for black? I think the idea is ridiculous ,so do you and so does Elroch but he is correct to enquire. What if it does? Or, differently, what if the object of a weak solution is to determine whether white can win against any possible reply by black to 1. e4? That doesn't mean "against all black replies". It means "against any single one of them or more". Then, all black first moves become relevant. The point is that artificially constraining the "weak" solution to one drawing reply by black isn't sufficient. It's exactly equivalent to solving a chess puzzle and claiming that chess is solved.

Isn't it?

 

Avatar of tygxc

#1526
"SF or another engine can only "guide the search" if its evaluations are perfect, so you are relying on it for perfect evaluations."
++ No, I do not rely on the Stockfish evaluation. I could even do without any evaluation function at all and just use Stockfish to generate moves and positions that result from a given positions and select between the moves at random. It would take more time but it would reach the same result. There are 2 peer reviewed papers where AlphaZero does it at random.

"Otherwise you cannot guarantee an accurate solution, which this us all about."
++ Yes, I can. If the result is a table base draw then all black moves are retroactively validated as optimal. That all white moves are optimal follows from exhaustively exploring all resonable white alternative moves.

"what if 1. e4 ...a5 wins for black?"
++ It goes contrary to experience from centuries and millions of games that A) black would win  and B) it would be by 1 e4 a5. If there is such a thing as a strategy in chess then the center, not the wings, develop pieces, make no weaknesses is part of it. 1 e4 a5 goes against every knowledge humans and engines have gathered independently over centuries.

"what if the object of a weak solution is to determine whether white can win against any possible reply by black to 1. e4?"
++ Well it is. White tries to win, black tries to draw. If the ideal game with optimal moves leads to a draw then black succeeds to draw and white fails to win.

"artificially constraning the weak solution to one drawing reply by black isn't sufficient."
++ If 1 e4 e5 draws, then the same method with more time and money could be used to see if 1 e4 c5, 1 e4 e6, 1 e4 c6,... 1 e4 a5, 1 e4 b5, 1 e4 f5 draw as well or not.

"It's exactly equivalent to solving a chess puzzle and claiming that chess is solved."
++ No, it is not.  Chess is weakly solved when for every reasonable white move there is at least one black move that leads to a table base draw.

Avatar of Optimissed
tygxc wrote:

#1526
"SF or another engine can only "guide the search" if its evaluations are perfect, so you are relying on it for perfect evaluations." ++ No, I do not rely on the Stockfish evaluation. I could even do without any evaluation function at all and just use Stockfish to generate moves and positions that result from a given positions and select between the moves at random. It would take more time but it would reach the same result.

It wouldn't reach any result, because you include no assessment, so therefore no means to guide the search. You wouldn't be aware if any one line produces a forced checkmate with best moves, since there would be billions of lines ending with checkmate and the machine would have no means of choosing between them, to discover that one of them is forced with best play.

There are peer reviewed papers where AlphaZero does it at random.>>>

You actually trust peer revue, in a difficult research subject, where no route to a solution has been demonstrated? I'm afraid that's nonsense. What does the ratification of a few peers with no greater abiluity than the author even matter?

"Otherwise you cannot guarantee an accurate solution, which this us all about."
++ Yes I can. If the result is a table base draw then all black moves are retroactively validated as optimal. That all white moves are optimal follows from exhaustively exploring all reasonable white alternative moves.

No they aren't, because the path to the draw may not be optimal. It isn't a guided search. You said yourself it's random.

I wish some of the others were capable of arguing coherently and decisively and yet any argument is only as good as the readers' perceptions of it. No-one can be forced into understanding, if indeed they are capable.

<<To weakly solve chess it is not necessary to look at all ECO codes. If 1 e4 e5 draws, then it is not necessary to establish if 1 e4 c5 draws as well or not.>>

Yes, that's obviously correct if the only form of the idea of a "weak solution" is to determine if black can actually draw. But as Elroch correctly asked, what if 1. e4 ...a5 wins for black? I think the idea is ridiculous and so do you but what if it does? Or, differently, what if the object of a weak solution is to determine whether white can win against any possible reply by black to 1. e4? Then, all black first moves are relevant. The point is that artificially constraning the weak solution to one drawing reply by black isn't sufficient. It's exactly equivalent to solving a chess puzzle and claiming that chess is solved.

Isn't it?

That remained uncontested.

 

Avatar of tygxc

#1527
I am sorry, I accidentily posted prematurely by a combination of keys, so I had to edit after you posted.

Avatar of Optimissed

Oh I beg your pardon. I get frustrated when that happens to me, too.

Avatar of champion-19

Lol, the initial comment was deleted already_

Avatar of Optimissed

<<<++ No, it is not.  Chess is weakly solved when for every reasonable white move there is at least one black move that leads to a table base draw.>>>

That's only one of 1000 interpretations of "weakly solved". Anyway, according to that set of definitions, which in any case I think is nonsense, what you gave there is an ultra-weak solution and also, it ignores the possibility of a black win. What about white being able to draw for each pair of opening moves?

Unfortunately, the definitions you're using aren't much use. They don't correspond to anything like what is necessary, even to know whether chess is a "forced draw with best play for both sides". The fact of "an optimal continuation for each possible black move on move one" or whatever, ignores the necessity that continuations for both sides are optimal, in order to mean anything.

There's no doubt that the people who put together these definitions are not intellectually trustworthy. A group of reasonably intelligent adults cannot even agree on what they mean and that's down to the authors of this claptrap.

Avatar of haiaku

But if you use a non perfect engine to analyze the most promising nodes first (according to its evaluation function) you indeed prune the tree, because if it finds one way to e.g. win a won position, you have found an optimal line (the others being equivalent at best, who cares). But this can cut the tree maybe by a factor of 2, 3, 5, don't think more than that. If one proves, through such an engine, that Black can draw against all possible White moves (not only the reasonable ones) from the very beginning, and if (likely) White can lose a tempo to replicate this drawing (White cannot win, because a drawing strategy for Black has been found) "strategy" (in technical terms), then the author can claim that chess is solved (@Optimissed, indulge me) weakly. But even if White is in zugzwang from the beginning (i.e. we cannot be sure there is a drawing strategy for White), I think the author could claim to have solved chess, because peers would not be very "choosy" about definitions ("Ok, good, it is finally proven that Black can at least draw").

I think we should forget entirely about ECO codes, because:

1) If a line is so bad that it leads to a quick loss by the user, than it will not impact so much on the resources

2) Otherwise, it is not so bad, it requires analysis to prove it bad, it might even win.

But if we intend Sveshnikov's "from the opening" as "from every one of the final positions of the most promising opening lines", then it would simplify things, of course.

As tygxc himself recognizes, though, the main problem is that number of sensible positions, which is quite uncertain, and we actually don't know if we can use its square root, because chess, with its 6 types of piece and their possible moving options, is far more complex than a game with only 2 types of men on the board and limited moving options. In losing chess, on the other hand, captures are compulsory, right? That reduces the options very much.

Avatar of Elroch

A more precise and stronger definition of "losing tempi" would be that you have a strategy for black to draw and a (hopefully easier) strategy for white which aims merely to force reaching a position in the strategy for black with positions reversed.

Given that it is an obvious practical efficiency to always deal with positions with "colour to move" and "colour not to move" rather than "white" and "black" and boards with "rank from side to move" rather than "rank from white side" (so you can use them in both strategies), this is  a natural feature of the joint strategy.

It can be considered as a set of positions with moves with a subset comprising the black strategy and an additional set of positions with moves that are added for white so that she can reach the black strategy (the latter hopefully being quite a lot smaller because there is an overlap).

Avatar of tygxc

#1532
"But this can cut the tree maybe by a factor of 2, 3, 5, don't think more than that."
++ No, it cuts much more. Not caring about black moves as long as they produce a draw is a major cutdown. For the white moves I extrapolated that 4 alternatives per move at 60 h / move suffice for 1 error per 10^20 positions. There are only 10^17 positions to visit, so that should suffice. I would start with one candidate ideal game with Stockfish autoplay at 60 h / move and then verify that all white moves are optimal by peeling back and retracting all white moves in 3 verification passes.

"1) If a line is so bad that it leads to a quick loss by the user, than it will not impact so much on the resources"
++ No, that is not true. We know that 1 e4 e5 2 Ba6? is bad and loses for white, but to calculate all variations to checkmate does impact the resources.

"2) Otherwise, it is not so bad, it requires analysis to prove it bad, it might even win."
++ Yes, that is correct. There are interesting AlphaZero - Stockfish games with unexpected sacrifices that the Stockfish evaluation function misjudged. It is another reason why we cannot rely on the evaluation function, however sophisticated. Only brute force calculation towards the table base is sure.

"But if we intend Sveshnikov "from the opening" as "from every one of the final positions of the most promising opening lines", then it would simplify things, of course."
++ Yes, there probably are several tabiya per ECO code. It is also a way to chop the large task of weakly solving chess into 500 more manageable subtasks. Even at 10^17 positions it still is a formidable task.

"the main problem is that number of sensible positions, which is quite uncertain, and we actually don't know if we can use its square root"
++ That is right: we do not know, the square root is plausible, but it may be more or it may be less. We only know after it is done e.g. for 1 ECO code e.g. C67.

"chess, with its 6 types of piece and their possible moving options, is far more complex than a game with only 2 types of men on the board"
++ Yes, but losing chess has exactly the same types of men as chess and only needs the 4th root. So this argument is invalid.

"In losing chess, then, captures is compulsory, right?"
++ That is right, also in checkers capture is compulsory.

"That reduces the options very much."
++ That is right, but not only the captures but also the pawn moves are irreversible in chess. Moreover the pieces compete for the same good central squares and that kind of compels captures. If white plays a rook on an open file, then black is more or less compelled to oppose with a rook of himself to contest the file. This then leads to a trade of rooks i.e. 2 captures. If white installs a knight on an outpost say Ne5, then that hinders black so much, that he will be compelled to play say Nd7 and Nxe5. So for the rules of chess capture is discretionary, but for the game itself it is often compelling.

Avatar of Optimissed

This learned conversation is so funny! Round and round .... OK bye bye.

Avatar of playerafar


From @tygxc #1522
'Cloud engines have reached 10^9 nodes / second, that is 1000 times faster than a desktop.'
So what?  That '1000 times faster' isn't even a pinprick.
Cutdowns in time needed need to be in the many billions of times faster - not just a thousand times faster.

'The number of legal and sensible positions lies as said above between 10^31 and 10^38 I assume 10^36. That would be the number of positions needed to visit for strongly solving chess i.e. a 32-men table base. The time and the storage are prohibitive.'

'Prohibitive' is the key word there.  
Again underlining the key point that nobody in the forum thinks the pure task is feasible with current technology.  

There's been much discussion of how the task might be 'pruned'.
That's been going on anyway.  Over the centuries.  'Pruning'.  happy.png
First - by players (ever since the game developed)
and then - by players and computers.   Again 'ongoing'.
(Mikhail Bottvinnik one of the pioneers of that I believe)
And now - also by supercomputer research projects added in.
Its ongoing.
But only on a 'pruned' basis.
Even the 7-piece tablebases are 'pruned'.  They had to skip castling/en passant.
And the 8-piece is still 'prohibited' for now.
Meaning - they haven't been able to do it yet.  happy.png

So - the discussion seems to be - how would supercomputers 'prune' the openings of chess in such a way that people would 'trust' it?
Well its already 'trusted' that supercomputers are already doing a better job of that than the best players can.
Apparently they're now 500 rating points stronger than the top GM's.

With the gap widening.
But its not considered 'solving'.
Are the semantics of 'solved' whatever whoever wants them to be?
Yes.   

Avatar of mpaetz
Optimissed wrote:

I mean, you are a complete barm cake. A nut job. Just look at that post. Stream of consciousness but unhealthily paranoid too. You'd do better NOT to post your thoughts about others, if they're negative. Could get you into trouble.

     Are you familiar with the old maxim "Practice what you preach"?

Avatar of playerafar


Earlier - in post #1507 I posted a chess puzzle -
to illustrate how invalid dismissal happens in chess. 
Including invalid 'solving'.
The chess puzzles greatly underline the principle
'things aren't always what they seem to be'.

So now the Solution move !  Time to post it !  
Which I'll do by blue text at the bottom. 
There's an analysis button in the diagram for those who want to see confirmations and lines.

The winning move is Not Kc3.  That's a blunder.
Its Kb1 !!    

Avatar of Optimissed
mpaetz wrote:
Optimissed wrote:

I mean, you are a complete barm cake. A nut job. Just look at that post. Stream of consciousness but unhealthily paranoid too. You'd do better NOT to post your thoughts about others, if they're negative. Could get you into trouble.

     Are you familiar with the old maxim "Practice what you preach"?

Yes I am. Practise what you preach. I really wish you'd stop trolling, because you ought to be able to see the situation. Much of the time, you're reasonably intelligent. Such a maxim is not an inviolable rule and occasionally it's good to tell the truth.

Avatar of mpaetz

     Yet engaging in the very behavior your post claims to be condemning just makes you look the fool. Anyone reading through this forum and seeing this may well develop a very low opinion of your reliability, tending towards dismissal of your arguments in general.

     And how is it "trolling" to advise against such personal attacks? They contribute nothing to the forum.

Avatar of Optimissed

I no longer really care & I'm sure you'll all have great fun here. Nobody seems even to understand my arguments and they're all far too concerned about continuing talking rubbish and perpetuating their myth to care about individuals. Meanwhile, a mentally ill person is trying to dominate everyone else and form cabals. I'm told it's happened also in clubs, where the atmosphere has been made quite toxic. I just want to leave you all to it. Thanks & if you have anything further to say to me, my message box is at your disposal.

Avatar of playerafar

'he' (without referring to anyone in particular) gets attention by doing the exact opposite of what he recommends.  Even in the same sentence.
That is the 'cheddar'.  If he can't get away with that and double standards and numerous other provocations and 'power plays'  then there is no 'cheese'.  Defeating 'his' purpose of being here.

Avatar of playerafar
StumpyBlitzer wrote:

Let's stop the name calling again please 😁, let's be the adults here and discuss the topic of solving chess. Opinions are there for everyone whether you agree or not. Let's keep it on topic again. Thanks all 👍😁

@StumpyBlitzer is a chess.com moderator.  A member of the staff.
If somebody calls you a 'barm cake' or 'crackpots' or 'troll' or other name - or constantly spams such terms at the forum and at the people in it - you can report that to the moderators by 'PM's'.  (personal/private) messages.

Avatar of Optimissed

Yes. I invited him to look over the thread, see what you were doing and to make a comment. I had already told him I'd told you what I thought of you, told him what you're doing (endless trolling) and what I think is wrong. I didn't wish to report you officially. I specifically asked that there should be no muting or banning. I'm really trying not to comment here any more because no-one understands my arguments regarding the topic, so what's the point?