Chess will never be solved, here's why - Chess Forums - Page 164

TheChessIntellectReturns · 2022-01-16T22:50:19-08:00

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 coul

MARattigan

Jun 8, 2022

0

#3261

tygxc wrote:

#3318
That is a bad example.
1) It is an artificial construct, not a real game between humans or engines or ICCF.

So far as can be understood, you plan to use Stockfish to play from artificially constructed positions. How does that make it a bad example? It's Stockfish playing from an artificially constructed position.

The bad examples, so far as any of your suggestions about how you might go about solving "chess" are concerned are precisely the real examples between engines or ICCF which are played under different rules from the chess variant you apparently intend to solve. (The game I posted isn't even a game under TCEC rules - it ended on White's move 36 under the TCEC draw rule.)

At any rate you can now desist from posting "the 50 move rule is never invoked with more than 7 men" ad nauseam. You have a counterexample.

2) It is a clear draw; in a real game between humans or engines or ICCF they would agree on a draw and not play 50 useless moves.

Your statement has never previously included any caveats about how you would assess a position. Why do you start now?

You regard the starting position as a clear draw. Does that mean you regard ICCF games as all sequences of useless moves?

MARattigan

Jun 8, 2022

0

#3262

tygxc wrote:

...

2) Weakly solved means that for the initial position a strategy has been determined to achieve the game-theoretic value against any opposition. Hence black tries to achieve the draw and white tries to oppose against the draw.

...

There seems to be little correlation between the two sentences.

Even if you assume that the game-theoretic value is a draw and that the meaning in your first sentence is that a strategy is provided for both players, it's difficult to see why the strategies for the two players should have different aims. Shouldn't both aim for a draw in that case?

Moreover your description of Black's strategy appears to be in direct conflict with FIDE's

Art.1.4 The objective of each player is to place the opponent’s king ‘under attack’ in such a way that the opponent has no legal move.

and you assert that 1.e4 e5 2.Ba6 is a loss for White, which would obviously oppose the draw after 1.e4 e5, yet you also say you don't plan to consider that continuation.

Could you elucidate the connection between the sentences, please?

Optimissed

Jun 8, 2022

0

#3263

I think both players are aiming to win although white's expectations are somewhat higher. The nature of chess is that to win, an unbalanced position must be created. Some opening systems, like variations of the King's Indian, consist of both players co-operating with one another to create unbalanced positions with winning chances. It's perfectly feasible to aim to win. In 5/5 blitz I do it in the Sicilian by trying for a dynamic equality and by gradually improving black's chances compared with white's of creating a situation where white's position is overloaded. As white I do the same but more aggressively.

It's unrealistic to think that black is simply trying to draw. Both sides try to win.

playerafar

Jun 8, 2022

0

#3264

Of Course the " 10^17 " figure is 'Drivel' !!
He has selected it to correspond to 'five years' with 'nodes' computers.
Which is also Drivel.
One year would appear to push too hard.

Ten years also more difficult for 'selling'. More 'daunting'.
It is even more ridiculous than the 'Y2K' movement that predicted world disaster because of future change in the dating system.
A fringe lunacy !
But unlike flat earth craziness - this crazy 'five years' notion attempts to 'infiltrate' science rather than deny it.
Its a slight deviation from the usual pseudoscience and pseudomath.
Which even have 'mainstreams' of their own !

Phony computer projects !
There's probably a whole 'genus' of such species ?
There's another term ... phylum.
Its probably an easy google about phony computer projects !!

pcwildman

Jun 8, 2022

0

#3265

Which begs the next question, is Chess infinite or finite?

playerafar

Jun 8, 2022

0

#3266

pcwildman wrote:

Which begs the next question, is Chess infinite or finite?

I don't 'beg'.
This is a Statement ...
the number of possible chess position is Gigantic. Daunting.
But it is finite. The upper bounds on it are well known.
The first upper bound is 13 multiplied by itself 63 times.
13 to the 64th power.
Every square has 13 possible states. Six kinds of piece multiplied by two colors plus empty square. 13. Sixty-four squares.
So now - right away - you know the number of possible positions is Finite.
That number can be further cut down because of maximum two Kings and at least 32 squares must always be empty. And other maximums on pieces.

But regarding the number of possible games - without a repetitions rule the number of possible games would be infinite ...
and without a 50 move rule also ... Game length would be infinite too ...
unless you rule that the game ends when one players lifespan has ended.
The winner wins by 'default'

'Solving' using numbers of games - is therefore unrealistic.
Positions are used.
But the 'game' element ... sequences of moves ... is still in there.
Can't be avoided in 'solving'. Or not with today's computers.
Modern tablebase projects try to 'backtrack' from two Kings only.
But that task becomes Daunting with just 8 pieces on the board ...
So Daunting that even with just 7 pieces they couldn't factor in castling rights.

taychoe

Jun 8, 2022

0

#3267

This thread will never end. Here's why ...

JohnNapierSanDiego

Jun 8, 2022

0

#3268

I just think it's really dumb to talk about Chess as some sort of math problem... What do you mean "solved"? It's just such a silly thread

Daphne_Girlie

Jun 8, 2022

0

#3269

Even if we say chess (classical) can be solved, we still have fischer-random

playerafar

Jun 8, 2022

0

#3270

sachin884 wrote:

Instead of posting all this you could have played 100 chess games

Or - eaten 100 pints of Haagen Dazz french vanilla ice cream.

tygxc

Jun 8, 2022

0

#3271

#3341
"What do you mean "solved"? "
++ Ultra-weakly solved means that the game-theoretic value of the initial position has been determined,
weakly solved means that for the initial position a strategy has been determined to achieve the game-theoretic value against any opposition, and
strongly solved is being used for a game for which such a strategy has been determined for all legal positions
H.J. van den Herik et al. / Artificial Intelligence 134 (2002) 277–311

tygxc

Jun 8, 2022

0

#3272

#3330

"How does that make it a bad example?" ++ It is a clear draw. The only resonable thing to do is to agree on a draw in that position. That is why you will not find any instance of the 50 moves rule being invoked in a real game.

Let me restate: "the 50-moves rule is never invoked with more than 7 men in any real game"

2) It is a clear draw; in a real game between humans or engines or ICCF they would agree on a draw and not play 50 useless moves.

"how you would assess a position" ++ Base line: the modern computers calculate until they hit the table base or a prior 3-fold repetition. However, when a position is a clear draw, the 'good assistants' must step in and adjudicate it a draw, like in your position. It is pointless to calculate 50 moves deep and 4 moves wide in a position known to be a draw. That is why chess knowledge is beneficial in the brute force method, per van den Herik.

"You regard the starting position as a clear draw. Does that mean you regard ICCF games as all sequences of useless moves?" ++ I regard the starting position as a draw, but not a clear draw. I believe 1 e4 e5 is a draw, but about 1 e4 c5, 1 e4 e6, 1 e4 c6 I am less sure. Probably both 1 e4 e5 2 Nf3 Nc6 and 1 e4 e5 2 Nf3 Nf6 are draws, but maybe one of them fails to draw. There is a distinction between a draw and a clear draw. Your posted artificial position is a clear draw: human players or ICCF grandmasters agree on a draw in that position and do not play 50 useless moves. So ICCF grandmasters play on in drawn positions, but agree on a draw in clearly drawn positions. Likewise in weakly solving chess. The whole process is stepping from one drawn position to the other until a clearly drawn position is reached: either a table base draw, or a 3-fold repetition, or a position known to be a draw, like your posted artificial position.

tygxc

Jun 8, 2022

0

#3273

#3331

"Even if you assume that the game-theoretic value is a draw and that the meaning in your first sentence is that a strategy is provided for both players, it's difficult to see why the strategies for the two players should have different aims. Shouldn't both aim for a draw in that case?"
++ No.
Weakly solved means that for the initial position a strategy has been determined to achieve the game-theoretic value against any opposition.
There are 3 possibilities for the game-theoretic value, in order of decreasing likelihood
D) chess is a draw
W) white wins
B) black wins
All evidence points towards D) being true, though not yet formally proven
All evidence points against B) thus B) being false, though not yet formally disproven
If we rule out B) for the moment, then white should try to prove W) and black should try to prove D).
Hence black tries to draw and white tries to win.
If all reasonable white attempts to win fail then chess is weakly solved to be a draw.

tygxc

Jun 8, 2022

0

#3274

#3334
"is Chess infinite or finite?"
++ Chess is finite.
There are a finite number of legal positions: 10^44.
Because of the 3-fold repetition rule each of those 10^44 positions can be reached only twice. Thus the number of moves in a chess game is finite too.
For a finite number of moves and a finite number of positions only a finite number of games exists.

tygxc

Jun 9, 2022

0

#3275

#3328
"in a position P the best move A gives an evaluation in centipawns of 71 and the second best move B, of 70" ++ The move with the highest provisional evaluation is not necessarily a good move. It may be that the move with +0.70 evaluation wins and the move with +0.71 evaluation draws. That is why the 4 top ranked moves need consideration so as to make only 1 error in 10^20 positions, while there only are 10^17 legal, sensible, reachable, relevant positions.

"which is what I mean with "evaluation"" ++ Fair enough. I call that 'provisional evaluation' or 'subjective evaluation' in contrast with 'objective evaluation' or 'absolute evaluation': draw / win / loss derived from either the 7-men endgame table base, or a 3-fold repetition, or a clear draw. In my lingo the initial position would have an objective evaluation of a draw and a provisional evaluation of +0.33.

"If Player₂ does not recognize the error, how does he know how to exploit it?"
++ He does not 'know' but it eventually becomes clear after the game is over. Say the Tal engine plays against the Petrosian engine. The Petrosian engine sees a provisional evaluation of +0.30 and allows a sacrifice. The equally strong Tal engine sees a provisional evaluation of +0.50 as by its settings it values some non-material factors higher and it sacrifices. Now they continue play and after say 20 moves the game ends.
Possibility 1. The Petrosian engine wins. The Tal engine in retrospect made a mistake sacrificing and the Petrosian engine exploited it by defending well, though not 'knowing' the sacrifice to be incorrect.
Possibility 2. The Tal engine wins. The Petrosian engine in retrospect made a mistake allowing the sacrifice and the Tal engine exploited it by sacrificing, though not 'knowing' the sacrifice to be correct.

playerafar

Jun 9, 2022

0

#3276

https://scambusters.org/fakeinvestment.html

Its maybe time to see where fake computer projects fit in the 'phylum' of pseudoscience scams worldwide.
And from there - where a fake project to solve chess would fit.

haiaku

Jun 9, 2022

0

#3277

tygxc wrote:

"in a position P the best move A gives an evaluation in centipawns of 71 and the second best move B, of 70" ++ The move with the highest provisional evaluation is not necessarily a good move.

I did not state otherwise.

That is why I say the 4 top ranked moves need consideration so as to make only 1 error in 10^20 positions, while there only are 10^17 legal, sensible, reachable, relevant positions.

Non sequitur.

"which is what I mean with "evaluation"" ++ Fair enough. I call that 'provisional evaluation' in contrast with 'objective evaluation': draw / win / loss derived from either the 7-men endgame table base, or a 3-fold repetition, or a clear draw.

Subjective.

"If Player₂ does not recognize the error, how does he know how to exploit it?"
++ He does not 'know' but it eventually becomes clear after the game is over. Say the Tal engine plays against the Petrosian engine. The Petrosian engine sees a provisional evaluation of +0.30 and allows a sacrifice. The equally strong Tal engine sees a provisional evaluation of +0.50 as by its settings it values some non-material factors higher and it sacrifices. Now they continue play and after say 20 moves the game ends.
Possibility 1. The Petrosian engine wins. The Tal engine made a mistake sacrificing and the Petrosian engine exploited it by defending well, though not 'knowing' the sacrifice to be incorrect.

How do you know that "Tal" made a mistake!? I said: "now let's suppose that v is 71, but in fact the game-theoretic value of the move is a loss"; that does not imply that we know that the move is an error. Even supposing that the game value is a draw, "Petrosian" may have won after other correlated errors, made in any of the 20 moves played after the sacrifice.

Possibility 2. The Tal engine wins. The Petrosian engine made a mistake allowing the sacrifice and the Tal engine exploited it by sacrificing, though not 'knowing' the sacrifice to be correct.

Same as above.

Plus, you deliberately ignore the case that would challenge at best your assumption of statistical independence of errors:
Possibility 3. The game ends in a draw, because the move is an error but both players (and you too, even in post analysys) fail to recognize it and exploit it. If either player knew how to sistematically exploit the error to a win, the evaluation would not be 0.30 and 0.50 respectively.

For the fifth time, your circular reasoning to justify the assumption of statistical independence between errors is not different from the example of the room I made in a previous post. What's the difference to you?

MARattigan

Jun 9, 2022

0

#3278

tygxc wrote:

#3330

"How does that make it a bad example?" ++ It is a clear draw. The only resonable thing to do is to agree on a draw in that position. That is why you will not find any instance of the 50 moves rule being invoked in a real game.

Let me restate: "the 50-moves rule is never invoked with more than 7 men in any real game"

It's not a restatement, but you can add that caveat (for the time being).

Does your proposed method of solution involve playing any real games? Pseudocode would be nice.

2) It is a clear draw; in a real game between humans or engines or ICCF they would agree on a draw and not play 50 useless moves.

"how you would assess a position" ++ Base line: the modern computers calculate until they hit the table base or a prior 3-fold repetition. However, when a position is a clear draw, the 'good assistants' must step in and adjudicate it a draw, like in your position. It is pointless to calculate 50 moves deep and 4 moves wide in a position known to be a draw. That is why chess knowledge is beneficial in the brute force method, per van den Herik.

Some of my original appears to have gone misssing. I reproduce it:

Your statement has never previously included any caveats about how you would assess a position. Why do you start now?

The question was why you introduce a caveat about the assessment of a position now, when you have never done so previously. That is not related to how you propose to solve chess.

"You regard the starting position as a clear draw. Does that mean you regard ICCF games as all sequences of useless moves?" ++ I regard the starting position as a draw, but not a clear draw.

I believe 1 e4 e5 is a draw, but about 1 e4 c5, 1 e4 e6, 1 e4 c6 I am less sure. Probably both 1 e4 e5 2 Nf3 Nc6 and 1 e4 e5 2 Nf3 Nf6 are draws, but maybe one of them fails to draw. There is a distinction between a draw and a clear draw. Your posted artificial position is a clear draw:

Funny, I could have sworn you said you knew the starting position was a draw. But I'm getting old; maybe my memory is playing tricks.

SF15 doesn't appear to agree with your distinction in this case. If I kibbitz the starting position the evaluation settles down to +0.34. If I kibbitz the position I posted it settles down to -0.66.

But then it doesn't have your direct line to God. You should write a Perl script to implement it. The SF developers would surely be interested.

human players or ICCF grandmasters agree on a draw in that position and do not play 50 useless moves. So ICCF grandmasters play on in drawn positions, but agree on a draw in clearly drawn positions.

I agree that is one of the reasons recorded games that terminate under the 50 move rule are rare. A more significant reason is probably that in winning positions that are out of the players' depth the players will also agree a draw after playing a few moves without getting anywhere.

Likewise in weakly solving chess.

The whole process is stepping from one drawn position to the other until a clearly drawn position is reached: either a table base draw, or a 3-fold repetition, or a position known to be a draw, like your posted artificial position.

See this post.

Elroch

Jun 9, 2022

0

#3279

One amusing failure of reasoning struck me there.

@tygxc - "Probably both 1 e4 e5 2 Nf3 Nc6 and 1 e4 e5 2 Nf3 Nf6 are draws, but maybe one of them fails to draw."

This is a classic example of assuming the conclusion and inferring what is true in order to make what is already believed true!

Let me elaborate. @tygxc states unambiguously that it is possible that 1 e4 e5 2 Nf3 Nc6 is a win for white. He also states unambiguously that it is possible that 1 e4 e5 2 Nf3 Nf6 is a win for white. So presuming only that he assumes that one of the given second moves for black is among the best moves (if not, the argument needs to be extended to other black second moves, starting with 2. ...d6, presumably) he also believes two truly magical things:

if it happens that 1 e4 e5 2 Nf3 Nc6 is a win for white, this makes it certain that 1 e4 e5 2 Nf3 Nf6 is a draw!
if it happens that 1 e4 e5 2 Nf3 Nf6 is a win for white, this makes it certain that 1 e4 e5 2 Nf3 Nc6 is a draw!

As I said, the power (ahem) of assuming the conclusion is demonstrated very nicely here.

tygxc

Jun 9, 2022

0

#3280

#3354
Carlsen and his team of grandmasters and cloud engines apparently was convinced 1 e4 e5 2 Nf3 Nc6 and 1 e4 c5 draw as he played it consistently in his last two world championship matches. Nepo and Caruana and their teams of grandmasters and cloud engines apparently concurred as they avoided the main lines.
Nepo and Caruana and their teams of grandmasters and cloud engines apparently were convinced that 1 e4 e5 2 Nf3 Nf6 draw, as they played it consistently in their last world championship matches. Carlsen and his team of grandmasters and cloud engines apparently concurred as he avoided the main line and tried different sidelines.
1 e4 e5 2 Nf3 d6 presumably loses by force, that is why nobody plays it at top level.
My calculated rule of 4 candidate moves is even excessive in this case: 2 candidate moves suffice.