How close are we to solving chess?

Does God play chess? And if he does does he always win?

It is said that God knows the end from the beginning; would chess bore Him?

who would win in game of chess - God or Ditka?  trick question . . . Ditka is God.

-daaa bears

mhooner> Does God play chess? And if he does does he always win?

God never loses at chess... that's why he plays dice!

I personally disagree with many people who say that there is more to chess than just numbers, and that there is some sort of "mystical" quality to it, but I suppose this is for the same reasons that I am somewhat 'atheistic' or agnostic. In my view the purported "spiritual dimension" of the world merely arises from a void in human understanding, much in the same way as people may percieve a timeless or mystical aspect in chess simply because they are unable (of course) to analyse it to its full conclusion.

I think the only way we can use the word 'solve' in the context of chess is if we have a computer program that knows every permutation of every line in any position that leads inexorably to checkmate in the fewest moves. If any two such programs were to play, it would lead to a draw (unless of course white is proven to have a natural advantage over black, which I would imagine is the case as I fail to see how the starting position is a form of zugzwang!)

I hope I haven't been too inflammatory in my remarks. 

pvmike> likesforests, fermat's last thereom was proposed in the 1600's

That is true.

pvmike> not quite a thousand years.

Pardon, I took some liberties. Fermat's Last Theorem is that a^n + b^n = c^n has no positive integer solutions for n>2. Around 975 AD, Al-Khujandi asserted and attempted to prove that a^3 + b^3 = c^3 has no positive integer solutions--you see the relation of course--but another fellow named Al-Khazin discovered holes in his poof.

 This is the central issue here. If it is necessary, chess will not be solved in the foreseeable future with foreseeable technology.

likesforests > Step #1: Assume such a solution exists--or else it's even more complex. :)

Yup. Suppose that, with perfect play, there are no forcing lines which result in mate in under 150 moves. Even if we only look at logical lines and ignore all possible silly permutations of chess games (e.g. both players move knights out then back then out then back, then move a pawn, move knights back and forth again, move another pawn, etc. until one of the draw rules kick in), there are still a heck of a lot of permutations to look at.

What if there is no unique solution? Consider the example some alluded to where they suppose 1.e4 leads to a forced mate for white with perfect play, regardless of black's response. What if, instead, for any opening move for white there are multiple possible 'correct' solutions for black that lead to undecided positions 150 moves down the road with perfect play? What if, with perfect play, most positions have multiple possible 'correct' solutions which only lead to more positions with more multiple 'correct' possibilities?

which lead to a draw.  ie. perfect play on both sides always ends in a draw.

 I believe this to be true. The thing is, if we are talking about "solving" chess then we are talking about a rigorous proof of this statement (are we not?). The question then arises, how do we do that? The "simplest" way would be to enumerate all possible chess games and use this information to show that given white's first move, black has at least one move such that any response by white can be met by a move that leads to a draw (and show that this condition can also be satisfied on black's second move, and so forth, and show this also holds for white). The problem with this approach is that the number of possible chess games is so huge, it can't be done with our current technology and likely won't be done any time soon.

I think that what JG27Pyth was suggesting is that you can cut down massively on the number of necessary calculations if you could show that, for example, in any situation there is only one move which doesn't lose (given correct play from the opponent). We don't really have a way of achieving this, but he asks us to suppose we could find one. Then, solving chess becomes realistic.

My point was that it is possible that the solution to chess is not unique. In many situations there may be many possible correct moves. In that case, even if we could streamline the process as efficiently as he suggests, 'solving' chess could still require an astronomical number of calculations well beyond our current (and foreseeable) technological capabilities.

johnny263 wrote:

which lead to a draw.  ie. perfect play on both sides always ends in a draw.

Chess has NOT been  "ultra-weakly" solved.  Many people believe that if chess is solved it will be as a draw, but there exists, as yet, no compelling argument for why this must be the case. 

CatoTheElder> If Google wanted to, they could (they have enough space to store the results).

The estimated diskspace required to store a minimal bitbase of all chess positions is:

11,000,000,000,000,000,000,000,000,000,000,000,000 TBs  

Google's server technology is expected to scale to:

100,000,000 TBs

There's a bit of a discrepency there. ;)

He all! I read carefully all posts in the topic. It's 5:30 AM now, I suffer from a little insomnia, and I decided to write something from my point of view and my level of expertize (MD in philosophy).

First of all, about terminology, and basic metaphysical concept (yes, theword "metaphysic" is forbiden in Analytic Philosophy - a boring and primitive pseudopsilopy prevailign today in philosophical departments - but I am in "european continental" tradition: any try to discart metaphysics lead to be a servant of some kind of bad, por metaphysical concept - oh well, forgot it, if u r not a "profesional philosopher"! ) behind every-day terms. You all said: "computer play chess"; polosportply started it (post #6), with a question:

How close are we to making computers that are 99% unbeatable

However, it's not a computer who play; it's a program.

 For human beings, it's a subject of philosophycal discusions and religious belives is it a basic diference between body and soul (or we need more then two categories - as "soul", "spirit", mind"...). For computers, it's clear: "hardware" is only a vessel for a "software", as St. Paul said that  the body is only a vessel for the spirit.

And, as it become visible in some posts, a presumpsion that "computer" is some kind of person is in back (as HAL 9000 - all know this film, I hope! ); this is why say "one day computers will rule a world".  

In the post #18, shaxmat64 demonstrate a confusion of ideas:

Those who think chess as a science claim that the era has come for computers because it is very hard to beat a program nowadays.

And then he add:

that would be pretty much the end of human race once machines start to think themselves...

However, machines didn't think in an essential sense. Programs are still wroten by people - nothing changed from the times of Enigma (WW II - btw, if u know russian, I just read at chesspro.ru an excelent article about the game Fine-Alexander in New York 1944. - Alexander come to USA for they both worked with computers for Defence! ).

I remember a report about the match Kramnik - Fritz 10 (they also wrote about the "man vs machine match". Before the match, Spiegel Online wrote, and it's translated into englis in chessbase.com:

If the world champion should lose this match against Deep Fritz, and lose it badly, one would have to admit that our electronic slaves have overtaken their human masters in yet another area of intellectual activity.

What a confusion of ideas! Projestion of human relatins where there are not aplicable! However, I can't find a source at the moment, but I remember that a team of programers said that Deep Fritz didn't play a world championship for progmas that yeear, for they prepared it especially for a match agaist human. They changed it's algoritms for possition estimation. They (humans!) changed a program! Vladimir Kramnik didn't play agaist "computer", but agaist a team of pepole using a computer as a tool ("vesel" for their invenitons).

And that' the point: we phylosophers say that computers (as like as animals) is not "self", i.e. "self-concieous". It's conencted with a diference in German between "Verstand" and "Vernunft", it could be translated into English as "intelect" and "mind" - sorry, english is not a good langage for a (continental) philosophy! (In Croatian: razum and um, it is nice said: "raz-um" is conented with analysis, "um" with syntesis - ti is famous "creativity" what shaxmat64 said computers havn't - but u need a littly philosophy to understend what they realy have not!).

And in cybernetics, the point is: a human beingb (as a "self", as a "person"), is capable to write his/her own programes - to programing him/herself!  (We have "free will".) And computers, from the time of Enigma to Fritz 11, show no more capabilly for self-programing then a hammer - the simpliest tool. Hardware and software - there are only a tool ("slave").

Do you know the film "War Games"? A computer Joshua at the end concluded for himself that The Nuclear War is a pointless game, and that chess is more interesting. I agree, but was it was writen by a human.

So, I use computer and a lot of programs as a tool, at the moment to write and also to play chess, and in turn-based chess I use on-line databases of games as chessok.com, as like books, as my "external memory" (CPU is mine alone! ). In the corespodent chess, acording to ICC rules, it's also alowed to use any kind of programes! They can come to the end of research and "solve" chess, but they will be only a tool we use! Don't wory - we will find a zilion of other interesting activities to express our creativity!

 OK, it's 7 AM now, a begining of a normal day (cold and rain, unfortunatelly..). It's more  I like to write, maybe nxt time.  About God and chess.

 I am a simple man -- a VERY simple man -- therefore, the only thing I understood in your post was this:

 lol - CTC processors - when they dont win a game they go back in time and start at move 1. so in *some* universe of the multiverse they will solve and hence have solved chess - but in that universe they dont know it :) in other universes of course they cannot get info from that universe [by definition of universe]

 Lydia blonde, i found that interesting and i agree with almost everything you said. i may be English but my ontological philosophy is Kantian and Sartrean [my political philosophy is a combination of anarchism and J.-P. Sartre's version of existentialism but that is a different thing]