You can statically evaluate a position and that doesn’t imply looking ahead. But that alone would be incomplete. A required dynamic evaluation is required, which implies ‘moving the pieces’ and ‘looking ahead’.
Will computers ever solve chess?
Sort:
I was talking mainly about a human player. This is how the brain works during a game of chess. An engine is similar because no matter what codes you want to implement, that ‘read’ a position so thoroughly that it includes all the threats as well, it is still based on past games, which are incomplete and imperfect. Who can guarantee that there is no serious threat after 50 moves with perfect play on both sides? So the engine must still look ahead to see if it can detect a new threat down the line. Unless it has exactly the same horizon limitations, in which case it’s pointless to look ahead, so it just relies on some pre-established codes that immediately evaluate the position.
But if it is a new, stronger and faster engine, it is worth trying to look ahead, maybe it can spot something previous engines couldn’t.
Apart from the first sentence, the decent-sized last post is all about engines, not humans. But some humans with butterfly attention can’t read more than a sentence in a larger post. For them you have to break it up in smaller portions, which I’m unwilling to do. Treating ADD is more important than understanding a larger post.
Chess is converging to draw!
90% of top level play ends in draw. It will reach 100% in future
90% of top level play ends in draw. It will reach 100% in future
Top level chess and perfect chess (best moves on both sides) are two different things.
Dec 26, 2018
0
#6547
This thread has been closed due to inactivity.
############################################################################################################################################################
Yes.
This thread has been ridiculous at times, but with all the technological so called 'progress' from quantum computing to AI, it seems untimely to close it now. Myopic even.
Dec 28, 2018
0
#6553
Whether chess will be solved or not only matters for science but not for us mere mortals.
You never know.
"The future is unknowable, but not unimaginable." Or maybe it is unimaginable if the singularity is near. My first computer, a Tandy laptop, had no harddrive just two 3.5 low density floppy drives, less than 1 MB of RAM, maybe 8 Mhz. I played Psion Chess with it. I remember when it was an honest debate whether a sitting world champion would lose to a computer program under match conditions (Kasparov). My phone with its capabilities today was science fantasy back to many in 1988. Alphazero just crushed Stockfish. What will tomorrow bring? This forum is sometimes a running commentary as we move through time, experiencing new developments, towards answering the OP question...
Ponz has pointed out that even if chess is solved a human has no way of checking the solution...He will just have to trust the machine, but he will never know for sure.
s23bog wrote:
I'd say that a forced win might very well be unique, if it exists. There might be about a dozen saved draws. People will probably find a way to put an asterisk by that name, in a book somewhere. They'll find a way to nix it from the great book of chess games ....
s23bog: Is this a passage/quote from Life and Games of Mikhail Tal?
Couldn't someone write a sufficiently complicated eval that takes into account every nuance of the position and not even need a look-ahead.
You cannot evaluate a position without assessing the threats, which implies looking ahead.