ex0du5, many thanks for the pointers. That has given me some reading, which I look forward to. (The full text of the citations appear to be accessible through my institution.)
Will computers ever solve chess?
Sort:
Feb 24, 2016
0
#223
s23bog wrote:
What is this talk of 8 queens? Surely 9 are possible. I think it would also be possible to achieve a position with 10 of any other piece plus one other piece. That being for each side.
Well, this would require 8 pawn captures and there would be 10 pieces that could play the role so, yes, K + 10N v K + 10B is legal.
I feel inadequate attention has been given to this important theoretical ending.
The nQ's problem is an interesting solve. Place 8 Queens on an open board such that no Queen is in a position to capture another Queen.
Feb 24, 2016
0
#226
s23bog wrote:
Important? My sarcasm detector seems to be going off.
It needs recalibrating: the tone was gentle humour.
I don't know about important, but it could be interesting.
I did feel genuinely interesting in how such a matchup would go.
I bet some extremely complicated positions could be created if you had two players willing to work together to create them.
Or against each other. This may be so: I really don't know.
CP6033 wrote:
+
in the next 30-40 years no, 100-200 who cares? i mean none of us will be alive then.
Feb 25, 2016
0
#228
Go home. Eat and have a nice sleep. Why do u even care if chess will be solved. Don't use your tiny brain too much.
Well, there is a good filosophical aspect to this question.
First of all, an opening can have three different states: Win, loss or draw. However, this is something that a computer probably never will be able to calculate due to the unimaginable large numbers of different moves that can be played. So a better question would be, does this particular opening result in a theoretical win for white?
What you are saying is that white should be able to use his advantage for a win. However, there can not be proven that this is enough for a win. Your question should sound something like "Is white's advanmtage enough to produce a forced win? And in which openings?"
Or an opponent losing on purpose for whatever reason. Computers are forced to play chess, while people choose to, via free will.
True. There are oodles of ways a game can be terminated by other means. The many variables of real life can affect chess.
Storfiskarn52 wrote:
Well, there is a good filosophical aspect to this question.
First of all, an opening can have three different states: Win, loss or draw. However, this is something that a computer probably never will be able to calculate due to the unimaginable large numbers of different moves that can be played. So a better question would be, does this particular opening result in a theoretical win for white?
What you are saying is that white should be able to use his advantage for a win. However, there can not be proven that this is enough for a win. Your question should sound something like "Is white's advanmtage enough to produce a forced win? And in which openings?"
The argument fails, because a 'won' opening becomes lost, if it leads to a loss with best play on both sides. We can only say something about the opening--won or lost--in relation to the final outcome. As a fact, not a belief. So until the computer can compute all the best moves for both sides--which means calculating all the possibilities, each time, from start to finish--nothing is certain.
Artificial intelligence is no match for natural stupidity.
Perfect moves can be deemed 'perfect' only when all the possibilities will be taken into account. Otherwise what seems perfect today will be proved a blunder tomorrow, and this is what we call progress in chess. It's like a blind being able to see more and more, but until he can see the big picture he remains blind.
As for computers, in the opening, where the possibilities are so huge, are clueless. Humans are doing better only because they have beliefs, such as gambit this pawn in this line at move 2, 3, or 4, or start a closed game as opposed to an open one, and then others copyycat that and an opening is born.
But computers don't hold any beliefs whatsoever, so they don't know how to proceeed, left to their own devices. There is no reason why to start this way and not that way, as far as they're concerned. But humans already altered their 'thinking' and made them think like them, in terms of judging opening moves according to man's own beliefs. They're like little humans with much better calculating abilities.
A real computer, independent of human prejudice ( when it comes to chess), is yet to be born.
Can an entire game be played perfectly? How many perfect games are possible? There is no "perfect opening". So how could computers ever "solve" chess when there isn't a perfect first move?
Apr 28, 2016
0
#237
I don't think it matters too much since the vast majority of chess players are amateurs and probably won't be effected by some computer "solving" chess.
Chess wasn't meant to be played against computers. Human brain vs human brain is the real challenge.
Apr 28, 2016
0
#239
I'm pretty sure that perfect play from both sides will always be a draw—and there are hundreds (if not thousands) of paths to reaching those draws.
So the only outcome I foresee happening with engines is: they confirm that, yes, with perfect play in any sound opening/defense, black will always draw against white.
Not much of a surprise there! :D
Chess can only be solved by a revolution in Game Theory, which would lead to an abstract analysis of chess.
A concrete analysis by calculating all possibilities will not happen, even with the fastet computers unless someone finds an ingenious way to build a quantum computer for chess analysis which could solve chess completely.
You can try to find a quantum algorithm for solving chess.
I ruined my Fidelity chess unit back in the 1980's by programming all the pieces to be kings. 🎎