Will computers ever solve chess?

You didn’t explain anything.

How does an engine beat a person without the person making a mistake?

If you have no good answer, please indicate so by deflecting with something like, “I already answered that.” 

From the theoretical point of view, this is of no importance. All that matters is that there are a finite number of legal positions. When each of these has a precise value (as in a tablebase) the much greater number of possible legal games can be perfectly analysed, and every mistake precisely identified.

  He keeps missing it, because he considers a game entering into Ruy Lopez ( after 3. Bb5 ) to be perfect because two idiots agreed on a draw.

  What he doesn’t get—his 60 years of experience apparently didn’t affect his day-to-day logic—is that nobody can tell for sure that 3, Bb5 doesn’t lose by force and that his experience is just a tad above zero, when we consider the sheer number of games that can be played. He wouldn’t let go of his meaningless experience ( relative to knowing what a perfect game is ), a mere trick of the ego, attachment and pride.

  But why stop after three moves? You might as well consider a ‘perfect’ game a draw agreement between two idiots after 1. e4 e5, or the Symetrical English, 1. c4 c5.

 And we still don’t know—his ‘60 years of meaningless experience ( in regards to this discussion ) not withstanding—whether 1...c5, 1...e5, 1. c4 and 1. e4 lose by force or not.

 It is simply a fact that we don’t know. When attachment  to prior experience gets in the way, that fact is obscured.

The engine does not beat a person without the person making a mistake.

Another person does not beat a person without the other person making a mistake.

To lose a game it is required that you make a mistake.

And none of this has relevance regarding the question "Will computers ever solve chess?"

"Wrong, I'm afraid. A game of chess could be infinitely long. Not really sure why you think it couldn't be. "

Neither one of you is wrong or right...yet.  You need to properly define "game of chess" in this context.  The number of moves in a chess game can only be infinite under the basic rules that ship with your first chess board .  Under tournament rules, 3 fold repetition and the 50/75 move draw rules, the maximum number of moves in single "official" chess game is under 10,000.  I had calculated it out in another thread, but I can't find it.  The basic calculation is not that hard:  since a pawn must move or a piece be captured every 75 moves for a game to not be drawn, you take the maximum number of pawn moves * 75, plus the number of pieces x 75 (note that for this calculation a pawn promoting to a queen or other piece is still just one piece).  This gives a soft upper bound of 10,800 moves.

Certain complications have to be subtracted from this upper bound, in that pawns cannot promote without the opposing pawn being captured or making a capture itself to get out of the way.

I would like to say the answer was between 9-10 thousands moves for a game.  This would require both players to forego claiming a draw under the 50 move rule, and forcing the arbiter to wait for the 75 move rule, then ducking under it by 1 move each time.

Note that you will also have to define "move" for this debate.  Are you talking about an official chess "move" comprised of a play by white and a response by black, or are you talking about each ply?

What you are missing in all of the above is that my 60 years of experience is only one piece of a mass of evidence regarding the question "Is chess a draw when neither side makes an error which would affect the theoretical outcome of the game?"

With circumstantial evidence you often have several different pieces of evidence when taken all together can add up to a very sure result.

However if you take one piece of evidence and examine it by itself--then that would not be sufficient proof of anything.

The police found a man in the same house that a murder took place--this in itself is not enough evidence that the man did the murder.

add The man was found in the same room that the murder took place.

add The man had a gun on him which was the murder weapon.

add The man had gun shot residue on his body,

 add The man had blood from the victim on his body

add The man was observed by another person shooting the victim

add The man had made many threats to the victim in the past.

When you add all of these circumstantial pieces of evidence you would reasonably conclude the man was the murderer. Or you might even need more evidence?  It could have been a case of manslaughter. So then the police would try and gather more evidence.

The point being that there is a mass of evidence that the game of chess is a draw and any one piece of evidence will not prove the game of chess is a draw.

The other point is there is a whole forum regarding this subject but this forum asks the question "Will computers ever solve chess?"

Great. Humans lose to the most powerful engine (with optimal setting and hardware) everytime, right?

I do not know if a human will lose to the most optimal setting and hardware  EVERYTIME ? I know that humans will lose ALMOST   every time. Maybe someday a human such as the world champion may get a draw in such a situation?

I do know that chess engines, even the best chess engines are not perfect as i myself have quickly solved a couple of puzzles that the very top chess engines could not solve.

Why would you ask the question regarding the relative skills of the best chess engines as compared to humans? Almost everyone knows the answer.

IM Kenneth Regan, a statistician as well as a great chess player, has argued that any player rated 2800 would be able to draw at least one game out of 50 against even the best engine. Therefore, the highest rating an engine could get would be 3600. 

That makes excellent sense to me. In fact, I bet the highest possible rating will be slightly lower than that.

For the Fischer fans, that means that Fischer could draw "God" at least occasionally. 

Unless "God" would play unfair and mess with Fischer's mind?!

You’re kidding right? Human get pimp slapped. Even when given multiple major handicaps the worlds top rarely manage a draw.

With NO handicap, the top engine running full force on a supercomputer? Humans, no chance, EVER. They would go 0 - 10,000 and beyond. And today’s engines are nowhere NEAR solving chess.

Now here’s the point:

You claim THOUSANDS of perfect games have been played. That means humans are capable of playing PERFECT games. And since computers can lose to each other, and since humans can occasionally play PERFECT, that means humans can occasionally BEAT the most monstrous chess engine we have.

  Same difference: put those 60 years of experience, add ALL the knowledge and all the games ever played, and you still have pretty much nothing. The amount of knowledge we don’t have is colossal.  We      don’t       know       anything.

 ‘Is chess a draw when neither side makes an error?’ We   don’t     know. We don’t know what happens when neither side makes an error, because nobody had ever played a game without errors, in the sense that we can be positive they weren’t errors. That includes the present engines. We can only be positive when all the variants have been exhausted. Until then we can assume or pretend we know, but the fact remains that we don’t know.  

  As long as we haven’t exhausted all the variants we    don’t      know.

 Until all the variants have been exhausted we don’t know whether 1...e5 ( in response to 1. e4) or 1...c5 ( in response to the same move ) are errors or not. We don’t even know whether 1. e4 is an error and leads to a forced loss, or not.

 But it takes humility to see the fact that we don’t know.

One problem with your thesis is that the very best humans do not play matches against chess engines. Sure humans who are rated below 2800 lose to chess engines. But if say, Carlsen played a 50 game match vs the strongest chess engine--probably Carlsen would get at least one draw.

Chess ratings is  fairly good math. 3247 may be the highest chess engine rating and Carlsen about 2843. With that difference in rating the lower rated person would at times draw with the chess engine.

Chess engines, even the very best, are not perfect.

Another problem with your thesis [your thesis that humans cannot play a perfect game]

is that the weaker your opponent is--the better chance you have to play a perfect game.

So while humans rarely play a perfect game [quite rarely play] against a strong chess engine--they may play a perfect game against another human.

I have played games where i made no errors several times. 

The universe is so big that there are trillions of things we do not know about the universe.

However even though there are trillions of things we do not  know about the universe--there are some things we DO KNOW about the universe.

Stick to chess, as that is overwhelming enough for your brain. Leave out the universe.

   Notbody knows anything about what a perfect move is,  not just one slow-poke who says ‘ I know.’ No—bo —dy knows.

   The evidence that we have is not colossal , because the number of moves and games that have not been played is colossal. It follows logically, that if you accept this as you have, you cannot also say that we know is is also colossal. It’s like saying what we don’t know is 99.99%, but what we do know is also 99.99%. There isn’t enough to go around...

 if what we don’t know is 99.99% then what we know is the remaining 0.01%, any slow-poke in kindergarten can see that.

  Unless all the variants have been exhausted, nobody knows what a perfect move is. Even a knucklehead can see that.

 This has nothing to do with your ability to play chess, it has to do with simple logic. Apparently, good players can lack the most elementary logic outside those 64 squares. Like the brain forgets to cach up in other areas of development. 

  Apparently this particular player’s 60 years of exoerience have blocked their vision/logic.

  But despite the knuchlehead’s absurd and illogical claims, nobody knows what a perfect move is, be that Kasparov, Karpov or any supercomputer of the present day. It’s so obvious why...

  That’s the point this particular brain keeps missing.