Is Chess on the verge of being solved?

Yeah they are pretty big. The size of the problem depends if we want to be exhaustive in our solving or if we are content in throwing all of the 9 queens vs a king and 9 knights vs a king kind of rubbish positions into a trash bin. I'm all for using a sieve of some kind, an algorithm, to get into the information that is really relevant (so it's really a question of do we want to solve chess or do we just want to find the best/perfect play?). So, all interesting positions, I stick with my 30 year prediction, all positions, I have no idea, could take a long while but I think we will get there.

Explain. I can't be more off than Kasparov was.

Chuck Norris has solved chess, but we are not Chuck Norris. Theoretically, chess is already solved. It's awhile before computers and us can make the perfect move.

Chess is FAR from being solved!!!

Pretty crazy.  Comparing the search tree complexity to number of possible positions, is comparing apples to oranges.  It is clear that the number of positions in checkers is smaller than 5^32 (two types men x 2 colors and empty square; and 32 squares)  This search tree for checkers is much smaller (than the search tree for chess) because of forced moves, and more limited moves.  With the use of hash tables which could be shared with different computers working on the problem, you would be crazy   to say that checkers couldn't be solved. 

Still, the 100-square checkers are not solved.

The number of possible chess games vastly outnumbers the number of atoms in the known universe by many magnitudes, a tablebase of all possible chessgames would require every atom in this universe plus another 1000000000000000000000000000000000000000000000000000000000000000000 universes equal in size to store,

chess will never be solved

We don't really care about the number of possible games when it comes to solving a game, just the number of legally reachable positions.

Even if chess is solved by computers, it doesnt ruin the game.  Humans dont have the capability to solve the game.

What computers must work out to solve chess, is simply an accurate evaluation for each of those 10^43 positions. Inasmuch as each can be reached by some other in the database, the computer need only connect them all to one another.

Certainly, many are rubbish, and so the evaluation is simple. Many others--usually several in the course of every GM tournament--are incorrectly assessed by our best current computers.

It is likely that the universe itself contains enough atoms to store this data. However, we're not gonna be able to access the whole hard drive any time in the next 30 years. At least that's my tentative prediction.

Yeah, if you could break them into sets and then solve each set that would work.  You may never have a 32 man tablebase but it doesn't seem like it would be required to have it all stored at once.

The only way to avoid brute force calculation of all possible chess games is to reduce calculation depth (from game length) and use positional evaluation at the end of the variations. That's the way chess computers work. But a chess engine can't come up with its own positional evaluation algorithm, because it requires reasoning which is beyond its capabilities. It can only be given to it by a human by means of programming. That's why different chess engines play chess differently - because they evaluate positions differently, according to how they have been programmed, in turn according to their creators' chess strategy understanding implementation.

This is not objective reasoning though. If you had watched the whole video, you could have found another part where Kasparov says how he can guess which engine he's playing against after playing 15-20 games - exactly because of being able to recognize those subjective features.

I believe Kasparov is well aware of that fact, as any other player who takes chess seriously. That's why he's an authority - because he's spent enough time to consider what he's speaking, because he has had great success, and because almost all (if not completely all) the rest of the chess world has come to the same conclusions. Along with programmers and mathematicians. "The raw numbers" and "concepts" have been known for long and are available to anyone who's interested in them, be it to support or disregard them, but after getting acquainted with them.

You should come up with a better refutation of their opinions than just dislike, or search for a way to deal with your "hate".

GIex said it well!

BTW, thanks for the link to the video.

Well, he has been known to focus it on women, so there's that.

a game cannot "be solved" if it involves the ability to make error.

if you mean is it possible for a computer to play every possible game in the world and then win each possible game then, yes, yes, it can be "solved"...

but no human every playing the game can do this and all humans have more or less ability and either make errors from ego, physical tiredness, age, stupidity, bad judgement, lack of knowledge, etc.

no, no, chess is not about to be solved !

The game will be solved and  from there more and more sophisticated

human training methods and medical chess chip implants will eventually become

available.

Any game will eventually be solved  so why bother to fight it- just play it and love it!

a "medical chip implant" is just to computize the skull...so my point remains the same....the game is solved if you solve it with computers but it is not solved if you use real people.