Solving chess? With no BS. (moderated)

NASA has run experiments which yielded positive (non-zero) results... but still within the error limits of a null hypothesis. In other words, inconclusive but not discouraging.

White–Juday warp-field interferometer - Wikipedia

I'm all for this, but it would actually get more interest from me if they *weren't* positing a warp field right out of Star Trek.  This is less than ideal for the scientific method.  Of course the Star Trek warp field borrows from sci-fi which borrows from real science, but it's circular.  The same problem as quantum entanglement...some people immediately went right to "transporters will work, just as depicted"...

I teleported home last night

With Tom and Sid and Meg.

Tom stole Maggie's heart away

And I got Sidney's leg.

Wondering... 
Once Chess is solved -

Are our troubles over ?

Yes.  Once we solve Chess, our Creator comes down and shakes our hand..."I've been waiting for this answer a long time".  Then they snap their fingers and a new Big Bang forms a new Universe to solve the question "Can sour cream actually go bad?"...believe me, you'll be mighty happy not to be part of *that* Universe...

That's very interesting....but on a lighter side, you go first.

The EM-Drive is another one that's about half respectable scientific theory, half moon-shine. Again, it has just enough credibility that NASA has a project investigating it.

I remember reading about the possibility of Cesium Ion drives in World Book Encyclopedia when I was 6 or 7.  Still waiting on one .

e

Then we'll have to tablebase every possible move for all the possible Fischer Random opening positions.

So it'd be like solving chess all over again, 960 additional times . . .

I believe it will be possible to solve chess in a century or so. It seems far off but I believe that they will be able to do it. I, honestly, will be sad to see it being solved. Something about it just feels like it takes the fun out of chess.

I also believe it will not be "solved" that way, but by AI who play each other enough to know the utmost perfect style of play.

There's no reason to believe this point is coming.  In fact, the advent of machine learning engines proves that there's a ton of room for improvement in chess play far past what humans have been capable of understanding.  Engines are leapfrogging each other every several weeks and months.  It's very obvious that engines are not reaching a ceiling.  Even if engines did, that does not solve chess.  It merely means that the current "best players" have plateaued.

I don't believe chess will ever be solved.

Why would we want to solve chess? That would take the fun away.

Yea, thats true too. but its just interesting to think about.

Solving chess doesn't actually do anything to the game from a human perspective.  Saying it takes the fun away is like saying the existence of Formula One race cars takes the fun out of skipping and running.

The solution to the 10^46 positions is not going to be learnable or actionable by any human being, ever.  This is the same concept as human beings who used to be able to change out a single transistor that failed, but now there are millions of transistors in the same space, and you cannot even see individual transistors anymore, much less swap them out.  It doesn't stop anyone from enjoying devices with millions of transistors, nor does it even stop you from working with single transistors in some electronics kit .

I'm pulling in this post from an old thread that lays out how there are 10^33 possible positions in chess *before a single piece is captured or pawn promoted*...

https://www.chess.com/forum/view/general/how-many-different-chess-positions-are-there?page=8#comment-17620310

@johntromp has today published a set of work that drops the number of possible positions from 10^46.7 to 10^44.88...

https://github.com/tromp/ChessPositionRanking

https://www.chess.com/forum/view/general/on-the-number-of-chess-positions?page=1

This has yet to be poked or prodded much, so feel free to help him out by kicking the tires...there's a bounty of up to $256 for finding something.

Edit:  After further work, the number is is now 10^44.6.