Will computers ever solve chess?

Just wondering is there any game proven to NOT be a draw with Perfect play.

 Obviously, one can’t prove anything with perfect play without actually a perfect play in an actual game, which it means it could be anything, a draw or a win either way, despite several beliefs held here one way or another. 

Any single game of backgammon is always won by one side or the other; there are no draws, except by agreement because both players would rather cancel out the game than run the risk of losing (which is not a draw in terms of the game itself, just agreeing to abandon it and leave it unfinished).

On a more trivial note, the children's checkers/draughts game of Fox and Geese is a forced win for the geese.

For those who don't know the game:  Played on an 8x8 checkers/draughts board, using 32 squares of one color as in checkers.  The geese are four checkers that start on the back four squares, and can only move one move forward one square diagonally, like ordinary men in checkers.  The fox starts on a central square on the opposite end, and can move one square diagonally forward or backward, like a king in checkers.  No "jumping" or other captures are possible.  The fox moves first.  If the fox slips by the geese so that it cannot be prevented from reaching the geese's back row, the fox wins.  If the geese succeed in hemming the fox in so it has no moves, the geese win. With proper play the geese can always win.  

Obviously, it should = 3.That is far closer to it's now nonsensical 3.141592....and would lead to a fewer number of deaths. (on the road in thne air, in hospitals and so on).

One would think this is obvious, yet somehow, people can confuse themselves so easily that the can come to deny this in less that a paragraph. 

Apparently you are saying Kasparov was confused when he said that...However, he is the strongest player between you two. Which means that if you don’t have the intelligence to see the truth of that immediately, at least wait until you reach the same level of play and look again. You might see it differently.

Weaker players than you? Where is the evidence of your playing strength?

The comparison was with Kasparov, who made a truthful, honest observation, which can be instantly seen regardless of the playing strength. But if the overall intelligence is missing—as a human being—then the playing strength seems to be a factor that can help seeing it.

I didn't say it was true. Nobody knows. I said there has been a consensus among grandmasters for a long time that a perfectly played game will end in a draw. I also said I believe it and explained why - because I don't believe the first move advantage is decisive. Evidently you believe it is decisive. Yes, there was once a consensus that Earth is flat - despite the obvious evidence to the contrary (shadows at different locations, the bottoms of ships disappearing before the tops, etc.). Since you made the comparison, perhaps you will tell us what the obvious evidence is that a perfectly played game will end in a win for white.  

I am not saying we have access to the truth, but it is what good chess players are aiming for: i.e. they essentially never aim to play a move that is not theoretically correct. But as well as this, good chess players like to play moves that give their opponent more opportunity to go wrong. In all but bizarre circumstances, they will never compromise the former for the latter.

Your point is that this is not achievable, but it is very likely that the strongest human chess players play theoretically correct moves a large majority of the time - say 95% . Don't be fooled by minor evaluation differences when engines are used to analyse games: these are likely indicative of practical considerations, not theoretical results. And they can occasionally be dead wrong as well)

While all players (and computers!) have is belief In positions where exhaustive analysis is impractical, they not only aim for truth, they probably achieve it most moves. This point is a judgement call, but disagreement can only realistically be limited to how big the percentage is for the strongest players. It might be as low as 90%, but I doubt it.

https://www.slashgear.com/google-just-offered-its-magical-machine-learning-chips-to-all-12519182/

  But all that is a bunch of beliefs, at the moment. Whatever move humans or machines are playing we don’t have the final evaluation of it. What appears to be the best move may turn out to be a blunder, with best play on both sides. Since nobody plays the best moves on both sides at the moment, because all the variants have not been exhausted, and since computers have shown us most bizarre-looking moves that produce an advantage through what seems like forced sequences, we can only get the hint that we don’t know as much as we thought we did. That until all options are exhausted, all we have is guesses.

 Of course, it is what it is, and what else can we do? We play the game the only way we know how. But essentially we can’t take anything to be final, at the moment, and we will continue to keep guessing, ultimately speaking—since the whole picture has not been seen.

  Again, I gave an example earlier, of opening a relatively closed position with ...f5. Some will open it, some will not, depending on temperament, which is far from logic. A player like Karpov will probably not play it, while one like Kasparov will. Both will tell you they’re correct. But it’s a matter of belief in their style, not of logic. And in many games this point is being reached, where one plays strictly according to their belief in their style, not based on what position dictates. 

 And best players tend to have arrived at some system, which is ultimately a bunch of beliefs, like Kasparov with his belief in quality, time, material—all that MTK stuff—while Sveshnikov’s beliefs are sometimes ridiculous. After admitting that he is not 100% sure ( how could he or anyone else be?), he has the stupid audacity to criticize players like Botvinnik, who started the game with 1...e6 in response to 1. e4, in a big chunk of games in his long career, who according to his belief is inferior to 1...e5 or 1...c5, or Fischer responding with 2...d6, instead of ‘the best move’ 2...Nc6 in the Sicilian and so on. So he suggests to a bunch of kids who were favoring 2...d6 because Fischer ( and Kasparov) employed it, without ‘considering that they might have been wrong on their second move!’ ( his exclamation mark) 

 Of course, he forgets to mention to the kids that that was his belief, and that it is also possible they have not been wrong on their second move after all. 

 Secondly, using the same logic, he omits to criticize himself for playing in a big chunk of his career games 2. c3 or 2. f4 in the Sicilian, for the same reasons he criticized the other two mentioned above: for practical reasons, because he wanted to make a living out of chess, (and those two moves avoided a lot of theory)and so according to his own system of beliefs that he’s promoting, he did not play what he criticized others for not playing all the time: the so-called matemathically correct first replies into the game, which are nothing but more beliefs on his part.

 All great players achieve some system or other of beliefs, which then they apply faithfully in their games. But the unexperienced players have not accumulated enough experience, with enough study given to their games, to arrive at such a system of beliefs, which doesn’t mean they are better players, on the contrary. The better players always arrive at certain patterns which tends to give good results for them. They are nothing but beliefs. 

 Anyone who investigates deep enough into this will see it. And that’s the way it is: let’s stop pretend beliefs are not there when it’s obvious that they are.

My point is that the beliefs (to be more specific, beliefs that moves are good) of very strong chess players are very probably mostly true, and aimed primarily at achieving that truth.

Do you disagree? If you feel inclined to do so bear in mind that every position and every move has a precise value 0, 1/2 or 1 (the value of a position is the max of values of the moves if the player has white or the min if he has black, of course), and that the empirical evidence is that most (not all, of course) positions have multiple optimal moves. I say this because the success rate of players in picking moves is helped by the fact that usually there are multiple correct answers.

If computers did prove (mathematically) chess to be a draw with perfect play would it lessen the game for you in any way? 

There's the pile game that can be a win for the first or second player depending on the the number of marbles. Chopsticks is a win for the second player (but the rules can be revised so it's a win for the first player). Not sure what else. That would be interesting to have a list of all games that aren't a draw with perfect play.

The pile game is described in this video:

The ending gets strange because it shows some positions in infinite chess that are a win for one player, but it requires ridiculously long games.

Not me. This would only change my belief set by replacing "chess is in all likelihood a draw" by"chess is a draw".

I know you didn't say it was true. You said "I don't think they're guessing." If they aren't guessing, what are they doing? And I dont even know if there was a consensus the earth was flat. It's just something people still talk about. I think most people didn't know or care. And it's just my guess that a perfectly played game means white always wins. I dont even know what a perfectly played game is. I've never seen one, it's never been done. But since it seems that people that play white win more often than people who play black that maybe that tiny advantage, coupled with "perfect play" would lead to a win every time. Afterall, nothing can beat perfect play.

  Right there: ‘very probably...’—that’s a belief still. When you say ‘optimal moves’ it’s the same: ‘optimal’ for now. But for now we don’t have the complete picture. Until then the so-called best move remains just a guess, a good guess, but just that. Maybe some day we will have the complete picture—although it seems unlikely, we don’t know so we can only speculate one way or another—but in the meantime we don’t know what the best move really is, even when it wins a game, because we can’t be sure that the defense played the best moves.

 Apart from this, like I said above, strong players are prone to play according to a system of personal beliefs. Sveshnikov doesn’t deny he has a system of beliefs. Kasparov either. Besides consciously forming a system of beliefs, if a winning forced sequence is not found in a position then the best move becomes a matter of belief. Which is why there is disagreement in evaluating positions: one player thinks he has the advantage and starts to recklessly attack, while the other may think just the same when in fact only one or neither has the upper hand.

Logic seems to be playing the only part in selecting a move in the same way a politician uses logic to fix various problems society is confronted with. And in the process of solving one specific problem they are creating a dozen others, which will need future fixing by other politicians. Their solution seems logical, but it doesn’t address the whole picture, therefore it is pseudo-logic. True logic only comes from seeing the whole picture. And it’s the same in chess. 

 But it is what it is: in life we can see the whole picture; in chess we can’t, so we do what we can. No other option available at the moment. We are stuck in partial logic. No big deal.

  The point is not to falsely say that life is chess, or ‘chess imitates life’ or vice versa (!) and proceed in life with partial logic, because in life ( which seems more complex) it IS possible to see the whole picture, so there’s no need to be stuck in partial logic.

  Which means one must admit that chess is but a game and nothing more than that. Which requires humility and a lack of identification with it.

As I've already said, guess means "to arrive at or commit oneself to an opinion about something without having sufficient evidence to support the opinion fully." Proof is not necessary to support the opinion fully. If there is proof, it's not even an opinion; it's a fact.

Nobody knows if a perfectly played game will end in a draw or a win for white. But knowing and guessing aren't the only options. We are sure of many things in life that we cannot absolutely know. There are scientific theories that we have great confidence in because they've been tested many times. We still don't have proof that they're correct. But they're certainly not the product of guesswork.

Most grandmasters believe a perfectly played game will end in a draw. That's not a guess based on nothing. It's an opinion based on knowledge and experience - an understanding of white's first move advantage after having played countless games on both sides.