Will computers ever solve chess?

Computers can analyze over 8 million positions/second .

(we can only analyze up to 5 positions/second)

It searches for all the elements that make a position winning (material, seeing moves ahead, position, etc.) in every single position, then it determines the best possible way to get to the best possible position.

They are facts.

You can solve each individual opening by name under the ECO codes.

Did it... did it stop?

Almost...

I think that human error is just error which is what loses a game. White has to attack more and black has to defend more. 

What if you could codify chess as a set of higher level rules. Sequences of moves that work together.

For example,

• Check, checkmate.
• Take piece not sufficiently defended.
• Fork, Pin, Skewer, Discovery.
• X-ray.
• Trap (no good places to move to).
• Standard checkmate algorithms
• ...

So maybe all opportunities for advantage could be encoded according to these rules. This is kind of what people do.

So maybe a complete set of these higher level rules could be considered a solution. In the sense that you no longer consider low level moves, only higher level moves, and always by fitting all the rules to the board, never by mini-max multiple ply branching.

Chess is vast!!

Although I believe computers CAN solve it, it is highly improbable.

It requires extremely high human work , effort, time and so on...

Not that everyone is reading 8k in, chess is a perfect information game- it is capable of being, and will be solved.

If you want to know where we are on the timeline, Connect 4 was solved in 88, Checkers was solved in 07... We have two decades on front of us, maybe. That said human play is going to continue, even Anand takes time to look at positions now, humans can't memorize every line.

But one good thing for humans is that it is not going to affect our human chess, we will not be able to memorize even 0.1th part of the 'winning' line.

So, I feel there is going to be no harm to us

I'd think about it this way:

Even a simple game like tic-tac-toe can be fun.  Even if it is solved and you know exactly how to never lose, and have several ways you know of to try to win, in case your opponent doesn't know how to stop them.  I think eventually chess will be like that.  There will always be a way your opponent can draw your strategy with perfect play, but you can try different setups different games to see if you can find a line of attack that they don't know the drawing strategy against.  As in, since no one can ever memorize every single possible position, you'll still be able to play games just like today where you try to find a way to beat them that they don't know how to stop.

Also I don't think there is a 100% win strategy.  There is probably a way to draw any of them, and so the best computers will ever be able to do is to have a list of strategies that can win if your opponent doesn't know how to defend, but draw in worst case, and have the defense to each such attack in order to avoid losing.  It's really just like tic-tac-toe but with closer to 10^100 positions instead of like hundreds or whatever.  Although most of those positions would never be reached in a real game.

Chess would be "solved" to a draw, since a tempo is not sufficient to win. And it might be a win for black, if he plays the Sicilian. 

I was thinking about the same, because you can't win, or convert to a winning position with a 0.50(roughly) advantage, which is that tempo advantage, according to an engine

At the very start of a game a tempo is worth roughly about a third of a pawn...

 In 1996, Deep Blue—which is nothing compared to the present engines—was analyzing 200 million positions per second.

yes quantum computer 20bits, will easily be able to handle all the variations

I have often discussed this possibility but, as far as I am aware, no-one knows how to program a quantum computer to do this yet. Programming a quantum computer is not like using a general purpose computer language (which can effectively do anything a Turing machine can).