Chess will never be solved, here's why
Sort:
Wow so complicated topic
Oct 21, 2022
0
#4842
Donald Kingsbury's "Courtship Rite" is a stunning if little known novel. Some of the best world-building I've ever read. I highly recommend it if you have never read it.
TheChessIntellectReturns wrote:
Imagine a chess position of X paradigms.
Now, a chess computer rated 3000 solves that position. All well and good.
Could another computer rated a zillion solve that position better than Rybka?"
You are wrong, Chess computers have not solved such positions. A better computer can absolutely play that positon better than the older versions could? what are you even talking about?? You do know that we still have computer tournaments right?? ???
"No, because not even chess computer zillion could solve the Ruy Lopez better than a sad FIDE master could"
you're telling me a FIDE master can draw against stockfish 15 in the Ruy Lopez?? You do realize some of the best players in the world can't draw against stockfish even if it is forced to do the bongcloud right?? There is no way an FM is drawing in the Ruy Lopez against such a monster. You seem to be living in some alternate reality bro. The game is far from solved in a competitive sense, new ideas are constantly being discovered lol this is so dumb. You do know opening theory is still evolving right??
"the point is, there's chess positions with exact solutions. Either e4, or d4, or c4, etc.
nothing in the world can change that" Yes there are exact solutions, and no one on earth has ever even come close to grasping them knows neither computer, nor human.
"So if you are talking about chess as a competitive sport, then chess has already been solved by kasparov, heck, by capablanca"
This is perhaps one of the dumbest things ever said by anybody about chess, even uniformed people who wouldn't know capablanca would probably at least have enough common sense to know that players who were born more than 100 years ago could not play as good as the best players today can. This is just nonsense, you lack even rudimentary knowledge of chess history and lack even basic common sense if you believe that chess theory hasn't changed since capablanca . . .
@6433
"Chess can be solved. It would take a lot of money though."
++ Yes, 3 million $ to hire 3 grandmasters and rent 3 cloud engines during 5 years.
@6437
"Wow so complicated topic"
++ Yes, it is a complicated topic and it does not help some trolls derail it with mockery, false accusations, red herrings, and novel reviews.
@6444
++ I only post information true to the best of my knowledge.
By the way, the Hübner quote is from his Twenty-five Annotated Games, Berlin, 1996, pp. 7–8.
Oct 22, 2022
0
#4848
tygxc wrote:
@6411
"So, an appeal to authority" ++ You yourself asked for "an authoritative basis for your definition"
"It is important when you want to communicate with people to use terminology that is not at odds with normal usage." ++ That is why I use the common terms 'error' and 'blunder' and not your 'half point error' and 'full point error' that nobody uses, but means the same.
"No-one accepts your suggestion." ++ It is not mine, it is GM Dr. Hübner's and he at least accepts it. It is the only logical, consistent and objective meaning of error and blunder.
"you do not understand the established meaning of solving a game. ++ I do, you do not.
I quote from peer-reviewed literature:
- Next to brute-force methods it is often beneficial to incorporate knowledge-based
methods in game-solving programs. https://www.sciencedirect.com/science/article/pii/S0004370201001527
I understand what that means, but you clearly don't. "Knowledge-based methods" are precise and deductive. They are not crap like "Nh3 is obviously inferior to Nf3, so we can ignore it completely", - The checkers proof consisted of solving 19 three-move openings, leading to a determination
of the starting position’s value: a draw. Although there are roughly 300 three-move openings, more
than 100 are duplicates (move transpositions). The rest can be proven to be irrelevant by an
alpha-beta search.
Which is a deductive method, not taking any sloppy short-cuts. The pruning is not on the basis of evaluations, it is on the basis of transpositions to known positions dealt with in other openings.
https://www.cs.mcgill.ca/~dprecup/courses/AI/Materials/checkers_is_solved.pdf - A Shannon C-type strategy program, VICTOR, is written for Connect-Four,
based on nine strategic rules http://www.informatik.uni-trier.de/~fernau/DSL0607/Masterthesis-Viergewinnt.pdf
And the proof of solution is deductive, with nothing ignored in the way you repeatedly do.
"you don't respect what you are told."
++ Whenever I am told something that is wrong. Do you respect what I tell you?
Your proclamations are frequently definitively wrong. For example, asserting you can ignore 1. Nh3 when creating a strategy for black for chess. How much respect do crucial errors merit? As much as blunders in chess games, I suppose.
- In weakly solving Chess it is not forbidden to think and use knowledge
You are oblivious to the difference between deductive knowledge and inductive heuristics. The former is certain and may be used as part of the reasoning in a proof. The latter cannot - but it still has a valid role in the selection of strategies. It would be so good if you were capable of understanding this.
You are confused about the stark difference between the complete freedom to choose moves in a strategy (all that matters is that they work) and the complete lack of freedom to constrain the legal responses of the opponent (every single one of them needs to be dealt with). - When the 4 most optimal white moves cannot win, then the 16 least optimal cannot win either.
- 1 g4?, 1 e4 e5 2 Ba6?, 1 e4 e5 2 Nf3 Nc6 3 Ba6?, 1 e4 e5 2 Nf3 Nc6 3 Nd4?, 1 e4 e5 2 Nf3 Nc6 3 Nxe5?, 1 e4 e5 2 Nf3 Nc6 3 Ng5?, 1 e4 e5 2 Nf3 Nc6 3 Nh4? etc. lose for white and do not even try to win and thus are irrelevant in weakly solving Chess.
- 1 a4 cannot be more optimal that 1 d4 or 1 e4 and thus is irrelevant in weakly solving Chess
- 1 Nh3 cannot be more optimal than 1 Nf3 and thus is irrelevant in weakly solving Chess
It is most helpful of you to provide so many excellent examples of the things I have pointed out earlier. In a fantasy world where you were capable of improving your understanding, you would relate each of them to my points above.
@6446
"Nh3 is obviously inferior to Nf3, so we can ignore it completely"
++ That is not crap, that is knowledge.
"it is on the basis of transpositions to known positions dealt with in other openings."
++ No, 300 openings, 100 are duplicates (move transpositions), 19 are solved, 181 irrelevant.
"And the proof of solution is deductive" ++ Just like the solution of Connect Four,
the weak solution of chess can benefit from incorporating knowledge.
"asserting you can ignore 1. Nh3 when creating a strategy for black for chess."
++ It is unthinkable that 1 Nf3 draws and 1 Nh3 wins.
"difference between deductive knowledge and inductive heuristics"
++ Some heuristics are deductive, not inductive. For example the importance of the center is deductive. Nf3 is more active than Nh3 and controls 2 vs. 0 central squares.
"complete lack of freedom to constrain the legal responses of the opponent (every single one of them needs to be dealt with)." ++ I disagree. It is pointless to work out all possible variations of 1 e4 e5 2 Ba6? to checkmate in 82.
"things I have pointed out earlier" ++ But wrongly so. Solving a game does not forbid to think.
Oct 22, 2022
0
#4850
tygxc wrote:
@6446
"Nh3 is obviously inferior to Nf3, so we can ignore it completely"
++ That is not crap, that is knowledge.
It is inductive belief, a good bet based on imprecise, woefully incomplete evidence. It is fine for practical, imprecise chess play, and has absolutely no place in a rigorous strategy for black.
I accept 100% that you think it is knowledge, but everyone can see that is because you do not understand the difference between deduced facts and induced beliefs.
"it is on the basis of transpositions to known positions dealt with in other openings."
++ No, 300 openings, 100 are duplicates (move transpositions), 19 are solved, 181 irrelevant.
False. Alpha-beta pruning is used to deal with the inferior openings. It seems you do not understand what alpha-beta pruning is. Let me quote:
"Alpha–beta pruning is a search algorithm that seeks to decrease the number of nodes that are evaluated by the minimax algorithm in its search tree. It stops evaluating a move when at least one possibility has been found that proves the move to be no better than a previously examined move. Such moves need not be evaluated further. When applied to a standard minimax tree, it returns the same move as minimax would, but prunes away branches that cannot possibly influence the final decision."
Presumably, the way this worked was that first all the decent openings were dealt with, then for the weaker ones it was demonstrated that a player could force a position that had already been dealt with and whose value was known for sure (this being a huge number by this stage).
"And the proof of solution is deductive" ++ Just like the solution of Connect Four,
the weak solution of chess can benefit from incorporating knowledge.
It can. But only in a valid way.
"asserting you can ignore 1. Nh3 when creating a strategy for black for chess."
++ It is unthinkable that 1 Nf3 draws and 1 Nh3 wins.
It is not at all difficult to think of it as a possibility. Regardless, if chess is a draw, solving chess requires a proof that black can draw against 1. Nh3, not a glib assumption.
To see this is so, pretend that the starting position in chess is the one after 1. Nh3. Anyone who claims this game can be proven to be drawn by saying "this game is obviously drawn" is very foolish.
"difference between deductive knowledge and inductive heuristics"
++ Some heuristics are deductive, not inductive. For example the importance of the center is deductive. Nf3 is more active than Nh3 and controls 2 vs. 0 central squares.
Thank you for confirming that you do not understand the difference.
"complete lack of freedom to constrain the legal responses of the opponent (every single one of them needs to be dealt with)." ++ I disagree. It is pointless to work out all possible variations of 1 e4 e5 2 Ba6? to checkmate in 82.
It is pointless to solve chess. So no need to start.
"things I have pointed out earlier" ++ But wrongly so. Solving a game does not forbid to think.
That does not make sense.
@6448
"Nh3 is obviously inferior to Nf3, so we can ignore it completely"
"It is inductive belief" ++ It is deductive knowledge.
Nf3 covers 8 squares, of which 2 central. Nh3 covers 4 squares, of which 0 central.
This peer-reviewed paper ranks in Figures 5 & 31 the 20 opening moves with no human input but the Laws of Chess, just performing boolean logic. https://arxiv.org/abs/2111.09259
d4 > e4 > Nf3 > c4 > e3 > g3 > Nc3 > c3 > b3 > a3 >
h3 > d3 > a4 > f4 > b4 > Nh3 > h4 > Na3 > f3 > g4
It is obvious that if the #3 move cannot win, then the #16 move cannot win either.
"Alpha–beta pruning is a search algorithm that seeks to decrease the number of nodes that are evaluated by the minimax algorithm in its search tree. It stops evaluating a move when at least one possibility has been found that proves the move to be no better than a previously examined move. Such moves need not be evaluated further. When applied to a standard minimax tree, it returns the same move as minimax would, but prunes away branches that cannot possibly influence the final decision."
++ So 1 e4 e5 2 Ba6? can be pruned.
"It can. But only in a valid way." ++ Glad you finally agree with van den Herik that it is beneficial to incorporate knowledge. Of course in a valid way only.
"solving chess requires a proof that black can draw against 1. Nh3"
++ Proof that black draws against 1 Nf3, and proof that 1 Nh3 cannot be superior to 1 Nf3,
together prove that black can draw against 1 Nh3.
"It is pointless to solve chess." ++ There may be better ways to spend the 3 million $ and the 5 years, but people may be interested to know how to draw against 1 e4, 1 d4, 1 c4, 1 Nf3.
None will be interested in how to checkmate in all variations after 1 e4 e5 2 Ba6?
I find the conversation interesting....on all sides. I may not understand all the math and other details, but I find tygxc's arguments compelling, even if I don't fully agree with his line of thinking. I recognize that my lack of full agreement could easily be my own deficiency.
I'm not yet convinced anyone is wrong here and that the disagreement isn't just a difference in how different people view "solvability". Doesn't matter....I'm not making any decisions or judging anything..... just a spectator. My $.02.
I don't enjoy the personal insults, however, no matter whom they are directed at.
Oct 22, 2022
0
#4853
I have to admit that tygxc is correct in his basic idea, much as I disagree with him. But my take on his theory probably doesn't correspond with his own view.
I have no doubt that his proposal to put five years of supercomputer time and the efforts of a team of GMs into the study of the most popular opening moves, using the accumulated experience of years of high-level correspondence games as a starting point, would add a lot to our knowledge of which lines in which openings offer white the best prospects and which strategies by black are most likely to successfully thwart those attempts. It is likely that some selected lines may be explored all the way to known ending tablebases.
This will b no means answer the question of whether there is ANY line in which chess is an irrefutable win for either side.
I find the discussion interesting, whatever that says about me. And that is a simple fact. It makes no difference whether it's interesting to anyone else....
It's like watching a tennis match. The ball gets hit back and forth until someone misses. Some people find that interesting and some don't. No one gets to say, "tennis is not interesting". To some it is, and to some it isn't.
If someone wants to attack or refute tygxc's logic or his conclusion, I'm interested. If someone wants to attack his motives, his character, or his general state of mind, my enjoyment level diminishes. Again, that's just a fact, as it's simply about what's going on in my head. And I believe that everyone's logic and conclusions should stand on their own. If they are foolish or invalid, then the better arguments will prevail. If there are people like me observing...who can't follow all the logic, then it seems unlikely that we will be swayed by attacks on motives and such. I certainly won't.
That said, my dominant priority is my belief in freedom of speech. So whether I like insults or not, I would fight hard to protect the freedom to speak them.
Oct 22, 2022
0
#4855
Optimissed wrote:
Just wondering what his basic idea is, which is correct. Surely it can't be the five year plan? For me, he seems to agree with my position, that we have to take a pragmatic viewpoint and we can assume chess is a draw. However, I believe we must accept that as knowledge, because there's no source that can replace the pragmatic viewpoint, since chess will never be solved. So far as I can see, I'm not alone in believing it will never be solved.
His correct idea (in my view) is that taking only the opening moves his committee of GMs think are best, accepting the ICCF games he chooses as "perfectly played", discarding all variations the committee of GMs don't like, and spending the amount of computer resources he proposes it is possible to link the chosen opening lines to the proven ending tablebases.
His incorrect idea (for me) is that this is a definitive solution of the game.
Optimissed wrote:
If I might be allowed to make a suggestion, perhaps you could think about motives more, since when there are irreconcilable difficulties, it seems reasonable, to me at least, to accept the possibility that the differences are artificial and caused by someone's attitude or perhaps by a clash of attitudes, from either side.
I'm not comfortable trying to judge or assess his motives. I have no evidence to base any judgement on, nor is it a concern to me. When I read tygxc's comments, I see no evidence that he is not operating in good faith. I realize that you, Elroch, and MARattigan all do, and I fully respect your opinions and your right to express them, even if I'm not comfortable with those expressions.
I don't think my lack of conformity here should be a concern to anyone. If it weren't for tygxc, this discussion would have probably been over long ago, and many of the questions I have pondered would have never entered my mind....so right or wrong, I'm grateful he's participating.
Very compelling arguments, a bit of a read but seems worth reading 
@6456
"I realize that you, Elroch, and MARattigan all do"
++ Some do not want Chess being solved and thus sabotage the thread,
e.g. by posting pictures of wives and discussing novels instead of the topic.
"I fully respect your opinions and your right to express them, even if I'm not comfortable with those expressions." ++ Some see discussion as a gunfight of insults.
Those insulting do so for lack of good arguments.
@6454
"discarding all variations the committee of GMs don't like"
++ No, I mean: discarding all variations that are clearly worse e.g. 1 e4 e5 2 Ba6? and occasionally adjudicating positions that are clear draws e.g. many opposite colored bishop endings. It is not a matter of liking or disliking, but of being 100% sure.
"His incorrect idea (for me) is that this is a definitive solution of the game."
++ Why incorrect? Then a strategy is determined to achieve the game-theoretic value against any opposition. That satisfies the definition of weakly solved.
@6451
"some selected lines may be explored all the way to known ending tablebases."
++ Not some, but all calculated until the tablebase, or a prior 3-fold repetition,
or a clear draw e.g. some opposite colored bishop endgame.
"This will by no means answer the question of whether there is ANY line in which chess is an irrefutable win for either side."
++ Yes it will. If the good lines cannot win, then the bad lines cannot win either.