life is just too tough for you poundball.
Will computers ever solve chess?
Sort:
Jun 7, 2017
0
#923
I have seen your insults, trolling in other threads. Stick to them chex as your wasting your time here.
Jun 7, 2017
0
#924
really ? my trolling ? you BS me and say I'm trolling you! cheeky little pup. will stay clear of your posts then in case you might start crying again.
s23bog wrote:
intermediatedinoz wrote:
s23bog wrote:
What do you mean by that?
By that I mean that if you go to any church, or you are religious, I will not consider you as my family. I hope I have made myself clear this time.
I have no problem if you, or anyone else, cares to consider me an enemy. That is your prerogative. I have many "enemies". I rarely face them. It seems they generally stay out of arm's reach.
Because you're in Chile?
bluepear wrote:
can you please define what you mean by "solve chess"
because i don't think chess is a math problem or a jigsaw puzzle.
do you believe that there really is a finite solution to it?
Read the thread. The definition and its source are probably posted a few dozen times.
s23bog wrote:
I would love a definition of it too, but I am looking for a psuedocode variation of it. If you were to write a program to "solve chess", what would be the point where you know that chess is solved?
Let me know if this helps define solving chess.
1) Is there any way that White can make a sequence of moves that guarantees a win? If yes, chess is solved.
2) Is there any way that Black can make a sequence of moves that guarantees a win? If yes, chess is solved.
3) If there is no way that either side can guarantee a win, and if EVERY game has been investigated, then chess is a draw. chess is solved.
If (1), (2), and (3) are unanswered, then chess is not solved.
As an example, the game diagram below shows a forced win for White. If this can be created for at least one complete game, then chess is solved.
Jun 7, 2017
0
#928
btickler wrote:
Read the thread. The definition and its source are probably posted a few dozen times.
Probably somewhere there somewhere.
In chess, each player usually has more than two choices for each move, so one section of the game tree might actually look like this:
Infinite chess is more complicated and infinite chess with huygens (they jump prime numbers of squares) is yet more complicated.
s23bog wrote:
I don't really think that it can be claimed that a game is "solved" unless every possibility has been exhausted. If the game starts 1. e4 e5, then you will always have to go back to that point and exhaust the possiblities for other black responses. ...
That's mostly true. But if you check all black responses, and none can avoid a loss, then you never need to check 1.a3, 1.b3, 1.c3, 1.d3,...,1.a4, 1.b4, 1.c4, 1.d4, 1.f4, ... and so forth. Each one of these can lead to a massive section of the game tree but there is no reason to investigate those, because 1.e4 led to a forced win. In this same way other major branches of the game tree can be "lopped" off during the analysis.
s23bog wrote:
...analysis is fresh every time and it is not stored anywhere. It just goes into the bit bucket.
That's another good point. If a programmer can define a method to put every chess game in a sequence (game 1, game 2, game 3, etc.) then there is no reason to store chess games that have already been analyzed. If games 1-10,000 are not a forced win for white, then you just store "n = 10,000". You don't need to store the actual games in memory. Saves a lot of computer memory.
btickler wrote:
bluepear wrote:
can you please define what you mean by "solve chess"
because i don't think chess is a math problem or a jigsaw puzzle.
do you believe that there really is a finite solution to it?
Read the thread. The definition and its source are probably posted a few dozen times.
Under the rules of Chess there are a finite number of possible games (admittedly it is a BIG finite number), therefore if there were a computer with sufficient power and memory to play and store all of them the answer would be known...
Jun 8, 2017
0
#934
bluepear wrote:
can you please define what you mean by "solve chess"
because i don't think chess is a math problem or a jigsaw puzzle.
do you believe that there really is a finite solution to it?
Can we prove definitively that chess is a win or draw with best play?
SmyslovFan wrote:
Can we prove definitively that chess is a win or draw with best play?
There's an algorithm to answer the question, but it would take a long time to run. As far as I know, no one has tried to start such an effort yet.
There are more than 1,000,000 people who have joined a distributed computing project to look for new prime numbers (GIMPS project = The Great Internet Mersenne Prime Search).
There's also more than 5,000,000 people who have joined computers to look for extra-terrestrial life (SETI@home = Search for Extra-Terrestrial Intelligence).
I wouldn't be surprised if a similar group is eventually formed to try to solve chess.
tesslovechess wrote:
plz watch my vid!
Good luck with your next account after this one is banned.
How about a different standard for deciding if computers have 'solved' the chess or not: computer playing black should always be able to draw(if not win) regardless of what white plays. I think that would also mean that computers have 'solved' chess.
I am just guessing but I think it would be easier for black to draw than white to win in computer vs computer play.
Jun 8, 2017
0
#939
Just about every serious chess player seeks to find a win for white, and seek to find ways to equalize as Black.
The strongest engines are closing in on demonstrating that chess is practically drawn. I believe that within the next 20-30 years, engines will be unbeatable, even against the strongest engines.
s23bog wrote:
Perhaps one of the most useful statistics in determining if chess has been solved is simply the win/loss/draw results. If it goes for months or years without a win for white, one can surmise that the computer's ability is approaching solving the game. Likewise, if it goes for months or years without white ever losing.
Not for sure...engines run using a set of valuations based off of human understanding of chess. Until engine developers bootstrap the engine's own results over years of play without using human opening books or position valuations to come up with a set of valuations built from the ground up without human "seeding", all that engines can prove is that they (on average) know how to play way better than humans and equal to themselves. That does not mean chess is then "solved".
Chex.. you missed the point entirely. What is difficult to grasp is the finite, that what we observe is a finite universe and everything in it.
Infinity is an abstraction which can not be applied to the real world.
What we see in front of our eyes is finite. We do not observe anything infinite. Observation of an infinite universe is not possible, if one thing were to be infinite, than everything must be so.
Either everything is infinite or everything is finite. Can't be some things are, some not. I choose to accept everything is finite, as to choosing the opposite. If everything were infinite, the laws of nature would not exist.