Will computers ever solve chess?

Yes, with perfect play Tic-tac-toe is a draw.

Also:

Connect Four is a win for the first player.

Chopsticks and Sim are a win for the second player.

Chess - Unknown.

😕

PatriotGames   We all make assumptions. My assumptions regarding chess are based on a whole lot of evidence.

You seem also to be making assumptions about chess.  One assumption is that a perfect game of chess cannot be played by someone who is not perfect.  But this, by logic, is not true as almost any chess player can play a perfect game just by chance alone.  And the stronger the chess player--the more likely he can play a perfect game and/or recognize a perfect game having  been played by some other chess player. And also, the shorter the game--the more likely it can be a perfect game.

Also, you seem to imply that there is only one or a small number of opening moves which will lead to a perfect game? And i believe from the evidence there are several first moves which will lead to a perfect game. 

You also seem to believe that assuming chess is a draw makes no sense "because it is an assumption"!! You could say that about every single assumption you ever make in your life!  We could not live without making assumptions.  We make the assumption that it is better to drive a car to work than to get up early in the morning to walk to work. We make the assumption that it is better to get up and out of bed than to stay in bed all day and all night. We make the assumption that it is better to get a flu shot than to go without a flu shot. We make an assumption that it is better to wear a seat belt than to chance  what will happen to us if we do not wear a seat belt.  When and if we marry--we make the assumption that it is a good thing to marry the person we are going to marry.  All or most of our assumptions are based on evidence.  My assumptions about chess are based on evidence.  

By the way, and this is an aside--I think the following game is one of thousands of perfect games which have already been played: 

If chess is a draw, then here is another perfect game:

1. Draw offered...Draw agreed.

Yes.

  Yes, ‘IF’, a big ‘if’.

Actually i don't think that chess is a draw is a big "if" at all!

Methinks it is about 99.99% likely!  

Regardless of what you think, it’s still a guess and not a fact.

I dont understand that at all. You start by assuming chess is a draw. Then a game is played where both players after a couple moves agree to a draw. Then you use that as evidence that chess is a draw. Vick gave a perfect example. If two players agree to a draw (whether they made any moves or not) do you consider that evidence chess is a draw? What about games where black wins? Isn't that evidence chess is a forced win for black?

Where you go wrong on this one is assuming that me showing a game that is considered a draw [by some] is proof that chess is a draw.  I am not using this as "proof" or "evidence" at all that chess is a draw.  I am just showing that a very  short game is more likely to be a draw than a longer game. [because the longer a game is--the more likely that one or both players will make a mistake]

As for evidence that chess is a draw with best play for both sides--i have previously given a ton of evidence that this is true.

Mathematicians want proof. Everything else is speculation.

i dont think it will solve

There is no 100% absolute truth. I am quite satisfied to state the evidence tells me that there is a 99.99% chance that chess is a draw when neither side makes an error.

We do not live our lifes by demanding 100% proof of most things.

You can demand 100% proof all you like but if you were to demand 100% proof that your actions are correct--you would be taking no actions. 

Also, "speculation" is the forming of a theory or conjecture without firm evidence.  I have firm evidence that chess is a draw when neither side makes an error.

The 99.99% is an expression of your strength of belief. It cannot be justified without relying on your confidence in extrapolation from experience of chess at a high but very far from perfect level (the latter being based on the strength of modern computers and the fact that even they are clearly still a long way from perfection). The reliability of this extrapolation is unclear: we don't have very clear arguments even for how strong perfection at chess is (we can guess, but guesses are not a solid basis for reasoning).

Imagine there is a deep, complex ending and this ending is played by expert chess players millions of times (please ignore the practicality!). They find that although they sometimes make mistakes and lose (leading to an average score of 55% for white) a lot of the time they draw.

So the players conclude the ending is probably drawn. What percentage confidence would be reasonable?

Suppose it turns out that very deep analysis (to 60 ply, say) finally shows that white has a forced win in the ending. Is this a huge surprise? Or just an interesting thing that is not such a surprise?

The question under consideration is very much the same, except the ending is replaced by the starting position, and the deep analysis would require more than 200 ply. This does not make conclusions easier or more certain.

It is true that it does not look like white has a win in the opening position in chess. But that was what all those experts thought about the initial position of that hypothetical deep complex ending.

So far your imagined situation has never happened.

Here i will give one piece of my evidence that chess is a draw when neither side makes an error.

This is my preliminary analysis of a position in the Ponziani Opening. The analysis starts with Black playing 5. ...Qe7 which is a known variation in the Ponziani. Black's move of 5. .... Qe7 has less practical chanches to draw then  some other moves. So by that criteria it is inferior.

Yes, White gets to the end game up a pawn but very probably cannot win.

This assumes a lot. It assumes that my guess as to the possibility of a non loss for either side translates also into options for a non  win.

It also assumes that both sides are playing best moves.

I will say one thing  99.99% likely is approximately 10,000 more times likely than 0.01%  [if my math is correct ]

of course it is--also a forced draw is a forced draw...also a pumpkin pie is a pumpkin pie... 

Yes!  You’re acknowledging it’s a guess!

Yes, it is my guess, But it is a guess based on a ton of evidence. I never said that i had solved chess.

Most everything we do is based on a guess. It is my guess that i should wear a seat belt. It is my guess that it would be better for me to get up out of bed rather than to stay in bed all day long. It is my guess that scientists are often correct regarding things of science. It is my guess  that I will live at least for one more day. It is my guess that i should not drive a car. 

Most of our guesses are routine and we make similar guesses each day.

Regarding the question: "is chess a draw if neither side makes a mistake?"

my guess it that chess is a draw and i am 99.99% sure based on a whole lot of evidence. 

Alphazero is a well develepoed program which teaches itself and stronger than stocfish But Could only beat 28 times and 72 matches were draw. Actually Alphazero is much more developed than stockfish because it plays itself and then memories much more variations and  moves  than stockfish. So this makes me think that there is a limit in chess playing because they draw 72 times. This shows that there is a limit in chess. if not ,it should have beaten stockfish in all matches wheter white or black. This shows that there is a limited quality moves after a while in match. So the chess is already solved by programs but just the time limit and the power of the computer will define the winner then...

  That’s it. I was gonna write about it but you did it already. 

  Despite that strong belief it may well turn out it’s not a draw. I wouldn’t be surprised in the least. But the person harboring this strong belief would certainly be shocked.

 Other than this difference, we’re back to square one: nobody really knows.