Will computers ever solve chess?

 You’re right, the group did not say “proved.”  This article was written by Gary Kasparov.  HE said it has not been proved.

“Or perhaps it would prove that a game of chess played in the best possible way always ends in a draw.) Perhaps a real version of HAL 9000 would simply announce move 1.e4, with checkmate in, say, 38,484 moves. These gloomy predictions have not come true” -  Gary Kasparov 

 Yes, silly us for saying the only way we can declare the outcome of a given line in chess, is to examine those lines.  Silly us for disbelieving you don’t have magic powers to know the outcome of lines you haven’t even looked at. 

 Just because we think your so-called evidence is laughably absurd doesn’t mean we haven’t looked at it. It means we have. 

--------> Thank you but religion should be discussed on Open Discussion, not here.

I was sarcastic..if GK said chess is a matter of believes (if it were true i should eat a live cat ) i invoke divine grace in order not to suck at this game.

Well since most of what you typed is not true, I would say you are sounding more and more like that Lyudmil guy. The one who says he's the best chess player in the world, grandmasters are weak, and his chess rating/ability is 3500. He believes it's true, therefore, what anyone else says about it must be lying.

You are right about one thing though, I have my mind made up. I believe it's possible for computers to solve chess. I also believe the outcome will be a forced win from the beginning for white or a forced draw from the beginning. But it's not true that I wont change my mind. The reason is because I have not seen all the evidence. There might come a time where it's proven where the game of chess is a forced win for black from the beginning. I dont know which is why I have to say I believe what the outcome will be. I can't say I know what the outcome will be.

Posting again for posterity, with a larger chunk of the article.  This should be useful later...although, as I have pointed out in the past, GMs and other titled players are no more experts in solving chess with computers than a chef making scrambled eggs is an expert on avian biology...

Kasparov wrote:

Another group postulated that the game would be solved, i.e., a mathematically conclusive way for a computer to win from the start would be found. (Or perhaps it would prove that a game of chess played in the best possible way always ends in a draw.) Perhaps a real version of HAL 9000 would simply announce move 1.e4, with checkmate in, say, 38,484 moves. These gloomy predictions have not come true, nor will they ever come to pass. Chess is far too complex to be definitively solved with any technology we can conceive of today. However, our looked-down-upon cousin, checkers, or draughts, suffered this fate quite recently thanks to the work of Jonathan Schaeffer at the University of Alberta and his unbeatable program Chinook.

The number of legal chess positions is 10⁴⁰, the number of different possible games, 10¹²⁰. Authors have attempted various ways to convey this immensity, usually based on one of the few fields to regularly employ such exponents, astronomy. In his book “Chess Metaphors”, Diego Rasskin-Gutman points out that a player looking eight moves ahead is already presented with as many possible games as there are stars in the galaxy. Another staple, a variation of which is also used by Rasskin-Gutman, is to say there are more possible chess games than the number of atoms in the universe. All of these comparisons impress upon the casual observer why brute-force computer calculation can’t solve this ancient board game. They are also handy, and I am not above doing this myself, for impressing people with how complicated chess is, if only in a largely irrelevant mathematical way.

This astronomical scale is not at all irrelevant to chess programmers. They’ve known from the beginning that solving the game—creating a provably unbeatable program—was not possible with the computer power available, and that effective shortcuts would have to be found.

it doesn't hurt--computing is used for more things than chess?

I'm pretty sure quantum computing et al will advance just as quickly without the input of anyone on this particular thread ...

He said perhaps it would prove--that article was written some time ago.

 It indicates that maybe [perhaps] in the future it would prove that chess played in the best possible way always ends in a draw.

If you ask him now, i am sure he would say that chess when played in the best posible way will always end in a draw.

However your opinion is worth nothing, nada, zip as you have already stated that you have a closed mind by stating "find the defendant guilty without checking all the evidence"

Also you have made statements which were obviously untrue. Also you have several times used Ad hominem attacks.

You are clearly not worthy of debate on this subject. 

You contradict yourself-- you stated "without checking all of the evidence"

Ok, if you have looked at my evidence then give the evidence [without using strawman] that i have given...........................................

But until then i will take you at your word that you pronounced against me "without checking all of the evidence"

The mathematical truth is that you cannot determine if chess is a draw without solving it (no need for quotation marks: this term is made precise by definitions). You can believe it is and even be certain that it is (these are states of mind) but you cannot actually justify absolute certainty.

It's much like you can't determine if a number is prime without mathematically proving it.

And like in that case, it is possible to have good reason to believe such a proposition is true without a proof by statistical evidence. The problem is that you can't really be sure how relevant the statistical evidence you have is to the proposition, so any such belief remains just that - a belief which may possibly be false (however unlikely you think that is).

I challenge you to quantify your sureness about the conclusion, or justify it with an argument that does not include any guesses.

ponz111 said:  He said perhaps it would prove--that article was written some time ago.
It indicates that maybe [perhaps] in the future it would prove that chess played in the best possible way always ends in a draw.

 Which means he does NOT believe it has been proven that when solved, chess is a draw with best play on both sides. 

If you ask him now, i am sure he would say that chess when played in the best posible way will always end in a draw.

Riiiight. So let’s recap. I have what Gary Kasparov actually said, and you have an unsubstantiated wild freaking guess what he might say if he were asked again. I’ll just let others decide for themselves which of the two holds more weight. 

However your opinion is worth nothing, nada, zip

 I get that you are very impressed with yourself,  but I assure you I hold your opinion to be of equal value. As a matter fact, I’m not just going to take someone’s opinion on this.  I’m going to weigh actual evidence, irrespective of who it comes from.  All your nonsense about “based on my 60 years of experience“ amount to nothing more than “because I said so.“ 

   I don’t think you know what “contradict yourself” means. Which two things have I said that contradict each other?

a- People who think White wins are in charge to find the forced mate.

b- People who think chess is draw gets various arguments

c- People who think you can't say still claims you can't say until hal 9000 solve this.

Since i don't know people researching mate from the first move i think b it's a comfortable statement.

P.s.: please notice i know my english sucks

"without checking all the evidence"  "it means we have" [looked at the evidence"]

 Great. In what post number did I use this sentence fragment?

Myself, I used to be near certain chess was a draw. But the slightly increased advantage for white in top computer games (foreshadowed by some, but not all, world championship human play in the past) has weakened my belief. I don't know how much further this is going to go. I still think it is most likely the correct result is a draw, but I would need to do some work to quantify my belief properly (as a probability, of course, as a good Bayesian).

It may be that what we currently see as a useful advantage but not a winning one is more commonly a winning advantage with supremely precise technique.

Let humans (working with engines) prepare the openings as black and then see how well engines score with white.

For example IIRC nearly half of alpha zero's published wins were out of the Queen's Indian. Ok, but if black is playing something like the Lasker QGD how does white force a win? There's just no way... IMO.

That’s the really important part. 

From now on I’m just going to name the hypothetical computer that solved chess “Hal” - for simplicity.

I also strongly think (not know) it’s a draw but I don’t think that’s a particularly strong argument. Alpha Zero, Stockfish, all of what we have today haven’t even scratched the surface of solving chess. We just can’t base outcomes on today’s engines. Stockfish playing Hal would be like human who just learned the game 5 mintutes ago playing Stockfish.

For me, a white win would mean black is already in a mating net before the first move is played. The number of possible game variations is about (1 with 120 zeros after it). I find it very unlikely there isn’t a single line where black can’t force a draw.

Whoa, Nelly.

You're going to get charbroiled on this one...