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

Correct.  On a universal scale, chess could be solved completely in fractions of a second.  Harnessing such huge power I'm sure we'll have found a way around storing the data as it's calculated too.

But... I hate to state the obvious, but humanity wont be at this level in 100 years... I dare say in 1 000 or 10 000 either.  100 000?  1 000 000?  I think perhaps the more relevant question for this time scale is if humanity will still be around.

Yeah, no.  If you could solve the game tree of chess in one Planck interval you'd still never solve the game tree of 19x19 go before the heat death of the universe.  Even evaulating 10^43ish positions, to have a 32-piece tablebase for chess, is intractable.

Empirically, chess seems to be a draw and Go (before komi) seems to be a win for black by about 7 points.

Perceive Chess like a bell curve, and also a reduction algorithm. In the bell curve the opening moves, and endgame moves are finite (usually under 100 possibilities per move of which just a few are reasonable.) The middle game which peaks out the bell curve, are ~ 20 to 50 moves, each of which has only a few reasonable moves. So by this reduction technique one can surmise that chess may be solved.

• Like tic-tac-toe the objective to draw but never to lose. Hone away the imbalances per position. We have seen this in the career of Karpov during his championship reign in the 70-80s. So the question is whether we find an even more complex game like Go, or chess with more variables. However, the computer will be able to sort things out better than the human competitor by brute force of memory combinations on any future timeline of game play. 

What we come to is the question of human endeavor. Is there any original music yet to be composed, any more math problems to be solved, any more scientific achievements to be made? If we took all the brilliant minds off the chess board, and put them into solving universal challenges, what will the computer as a tool do to help us to work it out?!

To clarify.... atoms aren't what computers use to store information, they utilize electrons which are in far greater numbers than atoms. However even with the amount of electrons (in the solar system, lets not go nuts here) you still wouldn't be able to calculate every chess position there is. Of course... variables within chess like legal positions have to be taken into account. You can't have a pawn on the far ranks and have it be a legal position... so knowing that, you can take out a very large amount of positions
Even still, the numbers represented would be too much to be contained, and even if they were, the expense of creating something that could retain all of the information in those legal chess positions is not something we humans are yet capable of. Maybe once qbits get figured out we can begin to reach a solution to chess as... that has infinite potential so far as we know.

Neither chess nor go (without the appropriate komi) provides equal chances of winning.  In both cases the first player has an advantage.  The rest of what you wrote is not even wrong.

The number of legal positions is estimated only 2*10^52 for classical chess.

Yours 10^123 is number of possible games, i.e. possible sequences of legal positions made by legal moves from initial position.

*urgh*

1- electrons are not in "far greater" numbers that atoms. It's more like between 1~100 electrons per atom (but here we are talking about gigantic numbers like 10^80, so your 10^2 factor is nothing).

2- computers do "use electrons" to store information, because they use electricity, but that's a weird presentation. You can as well say I use electrons to walk because elctromagnetic forces are what prevents me to go through the ground.

Computers do store bits on atoms on hard drives.

What the poster you quoted mentioned is (simplified) that assuming we have 10^23 storage atoms of one bit each (simplification of something that has reasonable dimensions), the highest number we can store is not 10^23 but 2^(10^23) which is significantly monstruously higher.

However, the conclusion (that storage is not an issue) is wrong because assuming we want the database saying for each legal position whether it is a win, a loss or a draw, the amount of data to store is still equivalent to more than one bit per position, so that's also 2^(N) instead of N (where N~10^52 according to Polar_Bear's correction).

Of course, "1 atom = 1 bit of data" is a huge simplification, but not far in order of magnitude of the truth.

The computer received a handicap. I wonder how tic-tac-toe ends if one player starts with a handicap. 

In Andre Philidor's day, chess players were rated by the degree of handicap they needed to play even with a "first class" player.

You're only up-to-date if you read the articles that you post.


I stand by my claim. 

You absolutely need the former in order to determine the latter.  Solving chess is a proof.  If you are discarding entire subtrees because they are "clearly" suboptimal you must first prove that they are suboptimal -- with very, very few exceptions this is, out of necessity, a brute force exercise that requires you to examine the entire tree just in case.... meaning you haven't discarded it at all.

The definition is objective and well established in game theory.  I provided it in post #82.  Here it is again:

Also, the argument that atoms are not the smallest components of the universe, so maybe there is room to store it all after all completely misses the point.  The comparison is meant to be illustrative of the sheer impracticality of tackling the problem.

In theory, theory and practice are the same.  In practice, they are not.

I don't know whether chess can be solved, but I believe strongly that it won't be solved. 

The problem is so gigantic that whatever workable plan might eventually come along, the execution is bound to take more time and resources than anyone will think is worth the effort. This is likely to be the case up until the time that humans and chess no longer exist. 

And imagine how things would progress if we spent any and all resources we had at our disposal on futile pursuits?

Your hyperbole misses the mark anyway -- I don't think anyone here said nothing can be accomplished, just not this.

Um, no, not the same thing at all.

Feel free to pursue futile endeavours, I'd prefer to spend my energy on things that are achievable and meaningful.

true

Playing chess is not the same thing as solving chess..., though you seem to equate the two.

Playing, even occasionally winning, a game of chess is acheivable and meaningful to me.

Trying to solve the game of chess is not acheivable at all, and for the resources that would be required to even attempt it certainly not meaningful when held up against other endeavours that could make far better use of those resources.

By all means, dispense with history because it is in "old books". That way you can foist your comic strip version of the paleolitic upon those who are equally gullible.

The point about Philidor was clearly too obscure.  

I'm a very good tic tac toe player - I only lose about 1 in every 10 games. That's about the same as the percentage of chess games that Anand loses.

Tic-tac-toe and chess if played with no errors both end in a draw.

I'v lost one game of tic-tac-toe since 1970. I'm embarrassed to admit how recently.