Chess will never be solved, here's why

Sort:
Avatar of Optimissed
mpaetz wrote:
Optimissed wrote:


The entire project of solving chess depends on rejecting bad moves and using good ones. A decent algorithm needs to be written, because I'm afraid that tygxc's "three grandmasters on ice skates" just doesn't cut it. Real AI needs first to be developed in order to write the algorithm.

Nice try.

     No, solving chess involves finding out whether or not there are any lines that will win by force from the initial position for either side no matter what the opponent might try. Whatever algorithm might be developed to sidestep this task cannot yield a complete proof.

No it doesn't. In their extremely confused terminology they call that the ultra-weak solution but it isn't relevant. It isn't even really possible to achieve. Best to read what's being written and try to catch up. I'm sure btickler would guide you through. Otherwise, I can't help you.

Avatar of Optimissed
btickler wrote:
mikekalish wrote:

Elroch, tygxc, mpaetz, and MARattigan are having a fascinating discussion of the topic. Much of it is over my head, but I'm following it best I can and appreciate the effort they're making to provide valid, topical information. 
I would respectfully ask that others who are using this thread to wage a personal feud back off and let the rest of us focus on the real discussion without having to sort out long posts that are nothing but personal insults that no one else cares about. 

I would respectfully ask that if you do not like seeing confrontation on the forums that you report people anyone that calls others "idiots", "imbeciles", "dullards", "morons", etc.  Not just once, but every single time.  Report actionable behavior.  When the forums are largely clear of such trolls, there will be a distinct drop in confrontations overall.  If the mods don't follow up on "verbal abuse" reports with visible results over the long term, consider escalating to staff.


He means you, btickler. You're the troll.

Avatar of MARattigan
tygxc  wrote:

...

"You are hawking a second-rate alternative."
++ No, I prefer a smart way that works over a stupid way that does not work.
...-

So does everyone, but so far all you've managed is a stupid way that doesn't work.

You can easily discount your own calculations by applying them to the games I posted here.

You claim they will tell you the result of the starting position and the number of errors in each game. Why don't you do that and we can check your conclusions with Syzygy?

How long does it take?

Avatar of DiogenesDue
Optimissed wrote:

He means you, btickler. You're the troll.

You can keep pretending he isn't talking to both of us, but it makes you look pretty oblivious.  The difference between us is you confront all kinds of posters.  I confront people who confront other people wink.png...and I do it in a measured and non-abusive manner.

Avatar of mpaetz
Optimissed wrote:
mpaetz wrote:
Optimissed wrote:


The entire project of solving chess depends on rejecting bad moves and using good ones. A decent algorithm needs to be written, because I'm afraid that tygxc's "three grandmasters on ice skates" just doesn't cut it. Real AI needs first to be developed in order to write the algorithm.

Nice try.

     No, solving chess involves finding out whether or not there are any lines that will win by force from the initial position for either side no matter what the opponent might try. Whatever algorithm might be developed to sidestep this task cannot yield a complete proof.

No it doesn't. In their extremely confused terminology they call that the ultra-weak solution but it isn't relevant. It isn't even really possible to achieve. Best to read what's being written and try to catch up. I'm sure btickler would guide you through. Otherwise, I can't help you.

    It seems that you are the one who is confused. Perhaps you neglected to read most of the posts here or just can't remember what others have written. I have repeatedly said that the only real solution is a calculation to checkmate or a draw in all possible lines. The kinds of half-a**ed solutions some others may suggest I don't consider entirely valid.

Avatar of Optimissed

Ah so you were the fount of wisdom here all along and didn't let on. I really had no idea you're so multi-talented!

Avatar of Optimissed

The trouble is that even if you could perform a calculation of all lines, it wouldn't achieve anything because you would have to sort and retrieve it all. That can't be done. You would have no access to the calculations, after they were finished in several million years' time.

Avatar of Elroch
MARattigan wrote:

[snip

Of course not "any legal moves".

This is a frustrated win under competition rules.

White to play

White can win against any legal moves but he can't win against any legal moves and draw claims.

You will find I have already provided the answer to this.

When solving games you can either include draw claims as legal moves that end the game or make the draws automatic. It doesn't matter to solving the game which one you do (a strategy that is trying to win that repeats moves can either be improved by avoiding a pointless return to a position already visited.  It's like if you win a game over the board after repeating positions once, you could have won more quickly by a more direct route.

Avatar of Optimissed

It's definitely a problem regarding solving chess .... when to call the draw? After a long time or a very long time?

Perhaps there is no point even attempting it without an algorithm, written by AI, that turns chess into discrete mathematical units, to be analysed further.

Avatar of Optimissed
btickler wrote:
Optimissed wrote:

He means you, btickler. You're the troll.

You can keep pretending he isn't talking to both of us, but it makes you look pretty oblivious.  The difference between us is you confront all kinds of posters.  I confront people who confront other people ...and I do it in a measured and non-abusive manner.

Mike was probably being polite. Even your concession that he was talking to both of us is completely astounding, however.

Avatar of mpaetz
Optimissed wrote:

Ah so you were the fount of wisdom here all along and didn't let on. I really had no idea you're so multi-talented!

    No, that would be your constant claim. I have proposed no method to make the solution easier. I have said that a solution awaits the development of much better engines or a breakthrough in methodology.

Avatar of MARattigan
Elroch wrote:
MARattigan wrote:

[snip

Of course not "any legal moves".

This is a frustrated win under competition rules.

White to play

White can win against any legal moves but he can't win against any legal moves and draw claims.

You will find I have already provided the answer to this.

When solving games you can either include draw claims as legal moves that end the game or make the draws automatic. It doesn't matter to solving the game which one you do

My comment was about why van den Herik would use the phrase "all opposition" rather than "all legal moves".

In chess the term "legal move" is defined in art.3 of the FIDE handbook and is distinct from "legitimate move". Not all people would understand "legal move" to mean a transition between consecutive legitimate game states. Not all people would class a draw claim as a move.

As you say, " you can either include draw claims as legal moves that end the game or make the draws automatic" in your abstract rules, but it's false to say that it doesn't matter which.

With an automatic draw claim a solution would be a strategy for one or both players that did not include a recommended draw claim at any point.

If the draw claim is not automatic and the game theoretic result is a draw, the solution for one or both of the players could include recommendations to claim a draw at some points.

The game trees would be different and the solutions in the two cases could be different.

(a strategy that is trying to win that repeats moves can either be improved by avoiding a pointless return to a position already visited.  It's like if you win a game over the board after repeating positions once, you could have won more quickly by a more direct route.

Agreed, but I don't see how it relates to my comment. I posted a ply count 0 position.

 

Avatar of DiogenesDue
Optimissed wrote:

Mike was probably being polite. Even your concession that he was talking to both of is is completely astounding, however.

Lol, only to you.  Part of the delusion.  I am routinely more aware of things.  Even now, you feel like MikeKalish was *mostly* talking about me...

Let me give you a little insight.  When people do try to talk to me about you, they say things like "you should stop picking on him" or "why do you lower yourself to his level?".  The answer is that I treat you as an equal by default.  Always have.  You are still, in spite of all evidence to the contrary, capable of change.  You could decide to stop insulting people around you.  You could decide to stop pontificating on topics you have no expertise in and then trashing those who do have expertise.  You could decide to stop pretending there's a cabal of posters in cahoots to take you down.  In a general sense, I expect better of people because I believe they are capable.

You are quite the opposite...nobody is your equal, and you expect everyone to acknowledge that or you immediately dislike them.  How many times just in recent weeks have you said something like "I thought you were [some attribute], but now I realize [something derogatory about the target]" to somebody after they have come out in direct disagreement with you?  Nervesofbutter is one example, there are more.  It's petty and small-minded, ultimately. 

You'll notice over time if you were to pay attention that whenever you say something, it's usually about the person, when I say something, it's usually about the behavior or mindset.  That's the difference between making a pointed observation in an attempt to get someone to see something about themselves without rancor (the latter), vs. being bitter and vengeful (the former).

Avatar of vikric
Chess will be solved eventually. Perhaps not quickly but given enough time chess is already solvable.
Avatar of MARattigan

@btickler @Optimissed

Could you boys not do that stuff via messages?

It's taking me ages to find the posts I'm looking for.

Avatar of MARattigan
mpaetz wrote:
Optimissed wrote:


The entire project of solving chess depends on rejecting bad moves and using good ones. A decent algorithm needs to be written, because I'm afraid that tygxc's "three grandmasters on ice skates" just doesn't cut it. Real AI needs first to be developed in order to write the algorithm.

Nice try.

     No, solving chess involves finding out whether or not there are any lines that will win by force from the initial position for either side no matter what the opponent might try.  ...

That is an ultra-weak solution as @Omnipissed rather surprisingly points out. He must have run out of wine.

A weak solution is finding the lines themselves (or a strategy for producing them which amounts to the same thing) or finding lines for both sides that prevent a win by the opponent from the initial position no matter what the opponent might try.

It would be easy to post such a solution or even a strong solution.

The critical factor, and what is being discussed, is time, which is referred to only in the definition I posted  somewhere on the thread (now lost in the mists of @Optimissed's IQ). It should also have mentioned space, but at least it was more relevant.

Avatar of DiogenesDue

Let me spell it out.  These confrontations can and should end, just as soon as the mods/staff do something about it.  If they don't care about the forums, then the spirit of the ToS will keep getting violated in perpetuity.  If they do care, then they will start to enforce the ToS, including namecalling.  When that happens, I (I can only speak for myself here) will no longer have need of confronting such behavior.  Until then, you're going to have to suffer through it. 

As I mentioned previously to MikeKalish, you can make this go faster by reporting namecalling and verbal abuse.  You can report Optimissed if you don't like the namecalling and verbal abuse, or you can report both of us for spamming/arguing too much.  Either way.

Avatar of Mike_Kalish
btickler wrote:

\

As I mentioned previously to MikeKalish, you can make this go faster by reporting namecalling and verbal abuse.  You can report Optimissed if you don't like the namecalling and verbal abuse, or you can report both of us for spamming/arguing too much.  Either way.

As a staunch believer in free speech, I would always prefer to see problems involving speech solved voluntarily rather than by enforcement. If it doesn't rise to the level of vulgarity or serious threats, I likely would choose not to report it. 

Avatar of Optimissed
NervesofButter wrote:
mpaetz wrote:
Optimissed wrote:

Ah so you were the fount of wisdom here all along and didn't let on. I really had no idea you're so multi-talented!

    No, that would be your constant claim. I have proposed no method to make the solution easier. I have said that a solution awaits the development of much better engines or a breakthrough in methodology.

Finally a post i can relate to.

He seems to have taken those words straight out of my mouth, since I have always said that, in those words. I know I like people to learn but that seems a bit much, all the same.

Avatar of tygxc

@5495
'Who gets to judge what is/is not "stupid"?'
++ The 3 (ICCF) (grand)masters. That is why they are necessary.
Examples of what is stupid:
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?
We know all of these lose by force.
It is a lot of irrelevant work to calculate all of these until checkmate in all variations.
It is waste of engine time.

This forum topic has been locked