True or false? Chess will never be solved! why?

Sort:
TheGrobe
zborg wrote:
ponz111 wrote:

Why would we want to solve something when we already know the answer? 

Indeed, a perfectly played game, however construed, will end in a draw.  And that's the "personal prejudice" of most GMs as well.

That's an entirely plausible assertion.  Proofs (however construed), notwithstanding.

Yawn.

Yes, while many here take liberties with the defintion of "proof" or "solution" Ponz111 prefers to take liberties with the definition of "knowledge".

waffllemaster
Atlec wrote:
TheGrobe wrote:

So it's point #2 you don't grasp then.

confirmed idiot or troll

Just for fun, estimate the total computing power of all the computers on earth and generously assuming a billion positions per second for the equivalent of a personal computer what kind of time span would be required to solve chess?

A)  About the lifespan of a person
B)  Estimated lifetime of humanity (past up to now)
C)  Estimated lifetime of the earth (past up to now)
D)  Estimated lifetime of the galaxy (past up to now)
E)  Much longer than estimated age of the universe

The answer may surprise you!

zborg

No, Sir @Grobe.  Your universalism of maths (and physics) is the one taking liberties in that case.  

And it's still a YAWN.  Despite your heroic efforts trying to convince the hoi polloi of various elements of game theory, inter alia.

Knowledge is not a definition, nor can it be construed as some kind of binary choice between absolutes, unless your thinking is fin de siecle 19th century.  In that case, it's quite an easy assertion (as you demonstrate), made frequently in these nutty forum threads.  No surprise there.

Thanks for your continuing efforts in any case.  Much appreciated.  Smile

Irontiger
waffllemaster wrote:

[Atlec :] "The storage capacity of the universe is a lot lot larger than 1 bit per atom."

Indeed, it might be a fantastic 100 bits per atom.

Let's be generous, I will grant you one million bits per atom. A lot lot larger (and far far far far far ahead from every existing storage system, but let's forget about that).

 

Now, to store binary information about the 10^123 positions a strong solution needs (ie again drop the draws), you need instead of my enormous estimation of 10^123 atoms, the ridiculously smaller amount of 10^116 atoms. There is till a bit of way to go to fit into a reasonable size, though.

 

I don't even know why I write this, I know that math were never an issue for some people here.

TheGrobe
waffllemaster wrote:
Atlec wrote:
TheGrobe wrote:

So it's point #2 you don't grasp then.

confirmed idiot or troll

Just for fun, estimate the total computing power of all the computers on earth and generously assuming a billion positions per second for the equivalent of a personal computer what kind of time span would be required to solve chess?

A)  About the lifespan of a person
B)  Estimated lifetime of humanity (past up to now)
C)  Estimated lifetime of the earth (past up to now)
D)  Estimated lifetime of the galaxy (past up to now)
E)  Much longer than estimated age of the universe

The answer may surprise you!

They say there will be 2 billion computers in the world by 2016.

At 1 billion positions per second, Thats 2x10^18 positions per second.

For all estimated 10^43 positions, that's 5x10^24 seconds.

That's 1.58 x 10^17 years -- about ten million times the estimated age of the universe.

Nevermind the fact that it's not the 10^43 positions that need anlaysis (and storage), it's all of the legal transitions between those positions.

TheGrobe
Irontiger wrote:
waffllemaster wrote:

[Atlec :] "The storage capacity of the universe is a lot lot larger than 1 bit per atom."

Indeed, it might be a fantastic 100 bits per atom.

Let's be generous, I will grant you one million bits per atom. A lot lot larger (and far far far far far ahead from every existing storage system, but let's forget about that).

 

Now, to store binary information about the 10^123 positions a strong solution needs (ie again drop the draws), you need instead of my enormous estimation of 10^123 atoms, the ridiculously smaller amount of 10^116 atoms. There is till a bit of way to go to fit into a reasonable size, though.

 

I don't even know why I write this, I know that math were never an issue for some people here.

Assuming that there are 10^80 atoms in the universe (the upper bound of estimates), you would need 10^43 bits per atom.

Ziryab
Atlec wrote:
Ziryab wrote:
tmodel66 wrote:

Sorry to pick at a sore spot - but looking at your games archive, you last played live chess on this site in 2010, and lost your two games of record. Some people play chess, others like to talk (a lot) about how much they know about it.

If a decent chess rating were necessary to argue in these forums, they would be silent. TheGrobe's arguments are all that matter, not how well or how often he plays chess here. 

TheGrobe has stated well:

1) the need to define what is meant by solved

2) the magnitude of the problem

3) data storage challenges

I do not see credible rebuttals of his key points  

Please read the thread rather than just the last few posts then.

1.  It has been defined, not only in this thread but there is a clear-cut definition of what a solved game is.  You being to lazy to read the thread or do your own research doesn't mean it isn't defined.

2. The magnitude is understood and not that daunting in the grand scheme of things. There are a lot of other mathematical and physics problems that eat up huge amounts of processing power.  The thing is, no one cares about chess (why should they?).  The government isn't giving out any grants to solve chess, solving chess isn't going to better our understanding of the universe or help us understand the complexities in knot theory and how proteins denature.  You know why chess isn't solved?  Because it's a game and no one cares.

3. See any of my previous posts.

Your previous posts are typical of the absence to which I referred.

Your #1 demonstrates that you understand neither TheGrobe's call for definition, nor my assent. It is not that definitions do not exist, but that they are not clear in the grandiose claims of the believers who litter this thread with answers in the affirmative.

Your #2 shows that you fail to comprehend the magnitude of the problem.

If you read the first four pages of posts, you will realize how utterly foolish you appear when you (a latecomer to the conversation) assert that I have only read the last few posts.
 

winerkleiner

True, chess will be solve when they determine two snowflakes are identical.

MorganGray

My guess is that given unlimited time and improved computer programs and capability-- chess will be "solved".  However, the solution will be basically meaningless... except perhaps to some future generation of sentient computer.

ponz111

Why so people on this forum keep debating if chess will be "solved" without giving their definition of "solved"?

Most already know the game of chess is a draw if played without error.

fburton
ponz111 wrote:

Why so people on this forum keep debating if chess will be "solved" without giving their definition of "solved"?

Most already know the game of chess is a draw if played without error.

It certainly appears to be a draw.

waffllemaster

Ok, so we'll say solve means we know the evaluation of evey position.  Either win, loss, or draw (the only three real evaluations).

With this you can generate many perfect games and always perfectly refute any bad move in any position.

In other words a 32 man tablebase.

Irontiger
ponz111 wrote:

Why so people on this forum keep debating if chess will be "solved" without giving their definition of "solved"?

Did you read the thread ?

zborg

A well played game will draw simply because there is no "forced win" for white.  You can be up a whole piece (3 points) but be unable to force mate in many cases.

Let's reformulate the question(s) -- 1) Is there a forced win for white?  NO.

2) Will we ever be able to prove that assertion?  That answer is (almost certainly) NO.

All the rest is idle speculation.  Neither Georg Cantor's work on transfinite numbers, nor Star-Trek-like quantum computing, will provide us with the key to the executive washroom.  So just live with it.

Put the onus of the debate on question #1.  Then have a few beers.  Laughing

fburton

Are any of the standard openings definitely losing (with best play) for white?

I'd hazard a guess that 1. Nf3 .. 2. Ng1 is, but would love to see it demostrated conclusively. It may be one of many as-yet-undiscovered counterintuitive results for chess. Oh, for a 32 man tablebase! Money Mouth And how many consecutive knight-return moves of are needed for a irrefutable, forced win for black?

qrayons
zborg wrote:

A well played game will draw simply because there is no "forced win" for white.  You can be up a whole piece (3 points) but be unable to force mate in many cases.

Let's reformulate the question(s) -- 1) Is there a forced win for white?  NO.

2) Will we ever be able to prove that assertion?  That answer is (almost certainly) NO.

All the rest is idle speculation.  Neither Georg Cantor's work on transfinite numbers, nor Star-Trek-like quantum computing, will provide us with the key to the executive washroom.  So just live with it.

Put the onus of the debate on question #1.  Then have a few beers.  

Your comment that there is no forced win for white is idle speculation. You might as well have written "Unicorns live inside the moon, we just can't prove it".

ponz111

The comment that there is no forced win for white is a lot more than idle speculation.  It happens to be true and If you ask a good player master or better--he will tell you chess is a draw with best play and there is no forced win for White.

There is very strong circumstantial evidence that there is no forced win for White. So strong that 98% of masters or higher will tell you this.

zborg

I agree completely with @Ponz111.

Unfortunately, too many people on this site habitually confuse the concepts of proof, knowledge, and warranted conclusion (compare classical rhetoric) with all sorts of idiot scientism ideas from the end of the 19th Century.

Chess players are a VERY eccentric lot.  Many are still stuck (intellectually) in "The Scientific Method" (a la Freshman year, class 101).  Give it a rest, PLEASE. 

Then (perhaps) you might decide to join the human conversation.

It's a very big tent.  Only pinheads exclude themselves.  Smile

Senator-Blutarsky

Chess is a draw ----> speculation

Chess is a win for white ----> speculation

Chess will be solved -------> speculation

Chess will never be solved ------> speculation

Saw it all with my dark spectaculations on.

zborg

Feel free to sit on those pencils.  Unless you'd rather speculate.