Sometimes I wonder...

Sort:
orangehonda

To prove a non-trivial move is forced though... in the solving chess sense, you would have to calculate more than what's physically possible.  Even with tremendous computing power, you're not going to be able to store the previous positions in your analysis as the tree branches further and further because as already said there simply isn't enough space even if one bit of information took up a space as small as an atom.

Sure computers are strong now, but the different between perfect play and strong play is larger than the difference between my play and Rybka 4's play.

For example any engine today would lose 100 out of 100 games against a 32 piece EGTB... at least in my opinion :)

jesterville

Even if the time comes (...and the comet misses earth...) when chess is "solved" by this future silicon monster, it really would not be relevant to professional play, since no human would be able to memorise all the best moves in all the possible lines anyway (currently even the strongest masters forget  their home prep every now and then...), plus no one really wants to see computers play each other.

...on the issue of if it is possible...anything is...we just need to discover the correct system.

orangehonda

It's not imaginable to me because the numbers are too big.  If you can even dream of a system that could solve chess then good for you, you're quite a thinker :)

pattrik

Another question- even IF there was a computer that can solve chess, is it possible to implant a computer chip into one's brain to contain all this data? But then chess wouldn't be fun, so I'd rather not...

Vaaelenko
[COMMENT DELETED]
AndTheLittleOneSaid

Now computers are so strong they can beat any human player... What more do we need?

SimonSeirup
Deranged wrote:
SimonSeirup wrote:

I have thought about that to. Here is my thoughts:

If a computer calculate all moves in chess / solve chess then...

...White winning is impossible, because black can play defensive for a draw.

...Draw is impossible, because there will be zubswang.

...Black will win, because of a forced zugswang from move 1.

My conclusion is that black will ALWAYS WIN. But i dont think that such a computer will ever be made. Of course im not sure about this, and im not saying this is fact, this is just what i think.


But white can play defensive too. I'm guessing it would be a draw.


Yes he can play defensive, but at some point he will become zugzwang.

Am i the only one thinking chess is a win for black?

Archaic71
ChristianSoldier007 wrote:
Sofademon wrote:
Coopah wrote:

There are already tablebases for games with 6 or less pieces.

Should only be a couple of years until they can do it with 32 pieces


 A couple of years. . .

Ok, chess has a state state-space complexity of about 10^47 power. (meaning the number of legal positions).  It has a game tree complexity (meaning the possible number or legal games of chess) of around 10^127 power.  To give you an idea of how FREAKING GIGANITC a number that is, many physicis estimate that if you take all of the matter in the entire observable universe and lumped it all together, you would have around 10^80th power atoms.  Chess has 100,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000 times more possible legal games that the number of atoms in all the matter in every planet, star, solar system, and galaxy, everywhere, in all of existance.

We are more than a couple of years away from having a table base for a brute force solve of chess.


I think that the saying that there is more chess positions than atoms should be taaken figurtivaly, not literally. It is not true. Either way, it is fun to look at the complexity of God's creation.


Um, actually - it IS true, its called math and it trumps faith pretty consistantly in blitz games.

I'm pretty sure that NM Weaver Adams already proved that a win is forced with the Frankenstein-Dracula variation of the Vienna (I do like me a F-D Vienna)

We can build a robot that can play golf better than any human, it has not hurt golfs popularity much (though it pretty well killed off correspondence golf)

My poor computer whines like a gutted lawnmower running a 5 year old version of Rybka - I suspect cranking through a 16+16 tablebase would cause my CPU irreperable grief

blake78613

I agree that we are a long way off from computers solving chess by brute force;  But one thing you are forgetting about in your calculations that there are 10^127th power games , is that many of these games are transpositions.   Computers use what are called has hash tables to avoid analyzing the same position over and over again.  You can have 1,000 computers in a network, each checking different lines and working off shared hash tables.   Such a system, while it probably would not solve chess in the mathematical sense, might come close for practical purposes.

pompom

Hey, that's what I was thinking, too.

Although there would be trillions of positions to caculate.

I sometimes think that white has a forced win from the starting position.

Well, at least you could probably just "not lose" from the starting position, if you play each move PERFECTLY.

pattrik

Another theory is the forced draw, which seems so much more realisitic...

pattrik

...since both sides start with the same amount of pieces. What do you people think?

oy83

I personally think it's a draw if both sides are played equally.

pattrik

Same.

pompom

Okay, maybe it is a draw.  Who knows?

 

I've figured out a few years ago that in tic-tac-toe, if each player plays correctly, it is always a draw.  We can already figure out all possible variations in tic-tac-toe.  Can't we use this same logic in chess?  Obviously, there are much more variations in chess than in tic-tac-toe.  But I think that in chess, is one player plays each move perfectly, it will never lose.

pattrik

Yeah- but chess is like a quintillion times more complicated than tic-tac-toe...

fburton
ChristianSoldier007 wrote:

If every side makes the absolute best move eevery move, it will always end in a draw.


Is there evidence or a logical argument to support this claim, or is it "just" a belief?