#315
Indeed all 7 men positions are exactly known. Work on 8-men positions is in progress.
You miscalculate your 50 million years: it is 5 years with 3 cloud engines.
I disagree.
But that's okay. Don't worry about it. 
"Work on 8 men positions is under way" ... long long way to '32 man positions'. Lol !
#298
The discussion about the 50-moves rule is irrelevant and confusing.
AlphaZero draws 88.2% of games @ 1 s / move and 97.9% @ 1 min / move, so it asymptotically approaches 100% draw with unlimited time i.e. perfect play.
With stalemate = win AlphaZero draws 86.06% of games @ 1 s / move and 97.1% @ 1 min / move, so it asymptotically approaches 100% draw with unlimited time i.e. perfect play just the same.
About stalemate = win:
"As such, it does not increase decisiveness of the game by much, as it seems to almost always be possible to defend without relying on stalemate as a drawing resource." - Kramnik
Thus a drastic change of the FIDE Laws of Chess, basic rules of play, 5.2 a. stalemate has no effect on the decisiveness.
Thus dropping or relaxing FIDE Laws of Chess, basic rules of play, 5.2 e. 50 moves has no effect on decisiveness either.
Millions of human and engine games show more games ending because of 5.2 a. stalemate and only rarely 5.2 e. 50 moves.
So for the purpose of solving chess we can disregard the 50 moves rule 5.2 e.
https://arxiv.org/pdf/2009.04374.pdf
https://www.fide.com/FIDE/handbook/LawsOfChess.pdf
#312
We know that ... although it is not worked out till checkmate.
We know things without a full proof.
That's your basic problem.
As @playerafar mentions in #315, many parts of basic endgame theory were fully proved prior to computer chess - down to checkmate.
Those parts held up when the EGTBs gave the real answers. An awful lot of the rest didn't (even though the results were "known"). And an awful lot of what didn't was far more fully analyzed than pretty much all opening theory,
Chess has become better not worse.
We're disagreeing on this.
And that's fine.
People hating the London System is fine too.
Computers have increased the popularity of chess rather than reduced it.
#312
We know that ... although it is not worked out till checkmate.
We know things without a full proof.
That's your basic problem.
Correct.
But its okay. Is anybody worried?
Some are.
Let them.
@tygxc
You've done the work of looking up those papers
and holding up the side of the debate that chess 'will be solved'.
That's often important in discussion - for there to be debate as well.
Especially if its not 'Da bait' ...
Earlier - I mentioned the idea that computers could use a 'solving method' whereby a position one 'ply' from a very solved position - could be quickly and easily solved by a computer based on its relationship to the other position.
And that's often how players approach positions. Related positions.
If sufficiently efficient 'algorithms' (kind of a buzz word but I'll use it) - using that 'relationship' principle can be developed - then maybe the game will be 'solved' someday.
In other words - any position at all entered in the supercomputer - and that computer 'has the complete story' on it.
Then will chess add a third dimension? A new piece?
Computers will then try to solve the game of 'Go' ?
If the sun burns out will the New York Mets and Manchester United have to play all their games at night? (yes I know they're different sports)
Lots of time could be invested in worry about same ...
#298
The discussion about the 50-moves rule is irrelevant and confusing.
Definitely not irrelevant. The solutions would almost certainly differ depending on whether or not the rule is in effect
You obviously find it confusing, but that's probably because on your own calculations it would make the difference between 5 years and 50 years on a cloud cuckoo computer and you don't want to admit it.
#298 is in fact perfectly accurate.
AlphaZero draws 88.2% of games @ 1 s / move and 97.9% @ 1 min / move, so it asymptotically approaches 100% draw with unlimited time i.e. perfect play.
With stalemate = win AlphaZero draws 86.06% of games @ 1 s / move and 97.1% @ 1 min / move, so it asymptotically approaches 100% draw with unlimited time i.e. perfect play just the same.
About stalemate = win:
"As such, it does not increase decisiveness of the game by much, as it seems to almost always be possible to defend without relying on stalemate as a drawing resource." - Kramnik
On the other hand AlphaZero and Kramnik are irrelevant, because there's no way to compare their playing level with perfect play and no way to connect any results derived from these sources with the facts.
Thus a drastic change of the FIDE Laws of Chess, basic rules of play, 5.2 a. stalemate has no effect on the decisiveness.
Thus dropping or relaxing FIDE Laws of Chess, basic rules of play, 5.2 e. 50 moves has no effect on decisiveness either.
Millions of human and engine games show more games ending because of 5.2 a. stalemate and only rarely 5.2 e. 50 moves.
More often agreed draws, which explains your comment. But as I pointed out, it's irrelevant anyway.
So for the purpose of solving chess we can disregard the 50 moves rule 5.2 e.
https://arxiv.org/pdf/2009.04374.pdf where does this paper say that?
https://www.fide.com/FIDE/handbook/LawsOfChess.pdf or this one?
This is article 5 (all of it).
The link you give points to the rules taking effect in 2009 and is well out of date. I did point this out before, but your communication circuit appears to contain an industrial grade rectifier.
Try this:
https://handbook.fide.com/chapter/E012018
If you want the 50 move rule it's art. 9.3.2 in the Competition Rules section.
#327
That is all besides the question.
The stalemate rule is applied much much more than the 50 moves rule in real human or engine games.
In the longest world championship game ever they never came close to the 50 moves draw.
https://www.chessgames.com/perl/chessgame?gid=2127373
Alpha Zero approaches 100% draws with more time per move, even with stalemate = win.
Thus we can disregard the 50 moves rule for the purpose of solving chess.
Breathtaking logic indeed! Congratulations, sir!
(Mind you I didn't quite understand how the game you mentioned was closer to stalemate. It did come very close to an agreed draw.)
#329
Nepo missed a forced perpetual check, i.e. 3-fold repetition.
#327
That is all besides the question.
The stalemate rule is applied much much more than the 50 moves rule in real human or engine games.
In the longest world championship game ever they never came close to the 50 moves draw.
https://www.chessgames.com/perl/chessgame?gid=2127373
Alpha Zero approaches 100% draws with more time per move, even with stalemate = win.
Thus we can disregard the 50 moves rule for the purpose of solving chess.
I think its reasonable to disregard the 50 moves thing.
A position's a position.
And anytime an en passant looks possible - it can just become two positions without bothering with how it got there.
With castling - it could get tougher?
In any position with any one of the four kinds of castling theoretically possible - then that's two positions - not one.
With two kinds of the four - that's four positions.
With three kinds - that's eight positions ...
with all four kinds of castling theoretically available in a position - that's short and long theoretically possible for either side - then I believe its sixteen. Its sixteen positions - not one.
1 + 1 + 4+4 + 3! ... that's '3 factorial' which is 6.
Notable that the total doubles each time. Which makes sense - the addition of each yes/no possiblity multiplies by two.
Note that tygxc indicated 7 men positions and lesser have been 'solved' I believe. In many of such positions - castling would be theoretically possible. Not in percentage - but in absolute numbers yes.
You could even have each side with their Kings and rooks correctly placed and another piece on the board somewhere.
Such a position's possiblities would have to be multiplied by 16 !!
This is a tiny fraction of the 7 piece (or any) tablebase with very special piece locations. For every position with king on e1 and rook on h1, there is a large number (>1000) with the same material and the two pieces not both in those places.
Hence it scarcely affects the numbers.
#334
Yes, all positions of 7 men or less are known exactly.
However, castling is not allowed.
In the next position after 1 O-O-O O-O it is a draw, but before O-O-O and O-O the position is not in the endgame table base.
However, solving chess does not need a 32-men table base.
Checkers was solved without a 24-men table base.
Checkers was just calculated from the opening towards a 10-men table base and thus proved to be a draw.
#327
That is all besides the question.
The stalemate rule is applied much much more than the 50 moves rule in real human or engine games.
In the longest world championship game ever they never came close to the 50 moves draw.
https://www.chessgames.com/perl/chessgame?gid=2127373
Alpha Zero approaches 100% draws with more time per move, even with stalemate = win.
Thus we can disregard the 50 moves rule for the purpose of solving chess.
I think its reasonable to disregard the 50 moves thing.
A position's a position.
...
Well, no it's not.
See e.g. posts #131 and #138 here https://www.chess.com/forum/view/general/chess-will-never-be-solved-heres-why?page=7.
#335
There are 3 ways to win a chess game: by checkmate or by resignation or by time-out.
Time-out has no theoretical relevance. Resignation happens when checkmate is inevitable. So it is just down to checkmate.
There are 5 ways to draw a chess game: stalemate, dead position, agreement, 3-fold repetition, 50 moves rule. Agreement happens when one of the 4 other conditions is inevitable. As per the AlphaZero/Kramnik paper stalemate has no impact and thus could be done without as other means of defending i.e. one of the 3 other conditions can be used to defend.
As the 50 moves rule is invoked almost never between strong players and has even less impact than stalemate, we can do without that too.
That leaves 2 fundamental drawing conditions: dead position and 3-fold repetition.
Perpetual check is 3-fold repetition.
Nepo could have forced perpetual check i.e. 3 fold repetition.
...
However, solving chess does not need a 32-men table base.
Checkers was solved without a 24-men table base.
Checkers was just calculated from the opening towards a 10-men table base and thus proved to be a draw.
I think most posters would agree with that. It's your estimate of how long it would take that is contentious.
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?
No, because not even chess computer zillion could solve the Ruy Lopez better than a sad FIDE master could.
the point is, there's chess positions with exact solutions. Either e4, or d4, or c4, etc.
nothing in the world can change that.
So if you are talking about chess as a competitive sport, then chess has already been solved by kasparov, heck, by capablanca.
If you are talking chess as a meaningless sequence of algorithms, where solving chess equates not to logical solutions of positional and tactical prowess, but as 'how many chess positions could ensure from this one?'' type of solutions, then, the solutions are infinite.
So can chess be solved? If it is as a competitive sport where one side must, win, then it has already been solved. Every possible BEST move in chess has been deduced long ago.
If chess is a meaningless set of moves, with no goal in sight, then sure, chess will never be solved.
I dont think you understand what the notion of chess being 'solved' is.
#315
Indeed all 7 men positions are exactly known. Work on 8-men positions is in progress.
You miscalculate your 50 million years: it is 5 years with 3 cloud engines.