Can Chess Moves Be Predicted with Enough Computing Power?

[Treadstone 71]

Is chess a futile game?

 

[arildno]

Nope, the answer is, in fact, 42.

 

[matt grime]

Define futile for us.

 

[arildno]

It isn't tic-tac-toe, treadstone, if that's what you're asking.
(or, at least, we don't know if it is).

 

[Treadstone 71]

Futile game = game that always results in a draw if all players play properly. Kinda like tic-tac-toe. I was wondering if chess was the same.

Nope, the answer is, in fact, 42.

:rofl: I knew it.

 

[Tzar]

Is chess a futile game?

The question you ask is a good one. I have spent quite a bit of time thinking about it, and well got almost nowhere. I believe that the fact that white goes first does give white some advantage. Statistically speaking I think white wins slightly more often. So I believe that if white plays "perfectly" then black is bound to lose. I ofcourse have no proof of this, this is just an opinion. I think the only way to check that I'm right or wrong is by going through all possbile combinations, where white plays the "best" move and seeing what happens to differnt responses from black. Of course finding the "best" move isn't easy, so a few different combinations will have to be tried each time. However, This will WAY to long even for the fastest computers.

 

[Tide]

Futile game = game that always results in a draw if all players play properly. Kinda like tic-tac-toe. I was wondering if chess was the same.



:rofl: I knew it.

That depends on what you mean by "properly." I think you're being somewhat circular by implying that "properly" equates to "stalemate."

 

[Treadstone 71]

Properly does equate to stalemate in some games, such as tic-tac-toe.

 

[neurocomp2003]

draw out the strategy tree and find out...might take you several pieces of paper or you could code it. I doubt white has an advantage given the # of permutations a board could have at the start phase of a game.

 

[Useful nucleus]

The modern players of chess always think of their rating and afraid of loss, so a lot of games nowadays end with draw by agreement. But in the past, until 1950's, The players were more enthusiatic and they played to win only and draw by agreement was so rare .
the best move in chess is something relative. Some players make very good sacrifices in the game which , by human mind, are regarded as great moves. When analyzing these moves by computer engines, the engine regard these moves as bad moves.so the game depends mainly on the sense of players and my opinion there is no " PERFECT MOVE",i.e., any move coul be to some extent good and to another extent bad.

 

[quasar987]

The modern players of chess always think of their rating and afraid of loss, so a lot of games nowadays end with draw by agreement. But in the past, until 1950's, The players were more enthusiatic and they played to win only and draw by agreement was so rare.

For the anecdote:

Draws by agreement were indeed more rare, but they were quite common amongst the players of the soviet union. In fact, young Bobby Fisher once throwed a tantrum about it and left the qualification cycle for the place of challenger to the world champion (Botvinik at that time), accusing the ruskies of cheating.

I believe that the fact that white goes first does give white some advantage. Statistically speaking I think white wins slightly more often. So I believe that if white plays "perfectly" then black is bound to lose.

The fact that white plays first give them a slight initiative. This is recognized by the chess players as an advantage, and reading the analysis of Grand Masters (GM) about a game, you'll encounter at some point in the game, probably around the 12th move or so, the commetary "Black has equalized", meaning the slight advantage of white has disapeared. However, even if black never quite equalizes, it is believed by GMs that this slight advantage of white is insufficient to win.

 

[HallsofIvy]

Draws by agreement were indeed more rare, but they were quite common amongst the players of the soviet union. In fact, young Bobby Fisher once threw a tantrum about it and left the qualification cycle for the place of challenger to the world champion (Botvinik at that time), accusing the ruskies of cheating.

Indeed, the FIDE established a rule that players could NOT agree to a draw until they had played at least 30 moves each in order to placate Fisher. During the next championship round, Fisher himself offered a draw after only 10 moves! When he was told he couldn't do that, Fisher protested "that rule is for the Russians, not me!"

 

[vaishakh]

I have something to say here. An indian book for undergrad physics starts with such an explanation. The name of the book is concepts of physics by H C Verma. I will try to say a bit of what he mean.
He starts the book with the explanation of basic idea of physics. He says that the scientist's mind is like a good chess player. He tries to make more and more perfect guesses seeing or observing the surounding. But how much perfect is it depends on how much experienced he is. When we once start seeing chess games played between two players without knowing the rules, after a lot of games we slowly understand mmany rules by guessing at instants and then neglecting the once which are wrong when we see the opposite thing happenning. And thus soon we start getting more and more accustomed with the rules. So a perfect move in chess is very difficult to guess.

 

[JasonRox]

I have something to say here. An indian book for undergrad physics starts with such an explanation. The name of the book is concepts of physics by H C Verma. I will try to say a bit of what he mean.
He starts the book with the explanation of basic idea of physics. He says that the scientist's mind is like a good chess player. He tries to make more and more perfect guesses seeing or observing the surounding. But how much perfect is it depends on how much experienced he is. When we once start seeing chess games played between two players without knowing the rules, after a lot of games we slowly understand mmany rules by guessing at instants and then neglecting the once which are wrong when we see the opposite thing happenning. And thus soon we start getting more and more accustomed with the rules. So a perfect move in chess is very difficult to guess.

Now, this is just the opinion of the author and how he visualizes how physics is done. That doesn't imply that it's useless.

I don't think Albert Einstein had lots of experience when he made his "perfect move".

 

[Leonardo Sidis]

My opinion is that chess is not a futile game because a "perfect" game cannot be played. As someone posted above, there is no perfect move, because the strength of a move is largely dependent on so many factors, especially in OTB games. Thus, there cannot be a perfect game. Some openings such as Alekhines Defense laugh at whites pawn center after e4, Nf6, e5, Nd5, c4, etc. in hopes that whites over-extended pawn structure becomes weak. Others focus on the pawn structure in the center and consider that the most important thing to establish in the opening. Even world-renowned grandmasters may disagree about which openings are better because they have had different experiences and have different opinions.

I'm curious, does anybody have a USCF rating in here? Mine's 1429.

 

[Hurkyl]

My opinion is that chess is not a futile game because a "perfect" game cannot be played. As someone posted above, there is no perfect move

Chess is a deterministic, turn-based game of complete information. Thus, it can (in principle) be completely analyzed with game theory, and it could be classified into one of three cases:

(1) White has a strategy with which he will always win.
(2) Black has a strategy with which he will always win.
(3) Both players have a strategy with which they will never lose.

The only problem is that we don't yet have enough theoretical and computational power to figure out which case is correct.

 

[Treadstone 71]

That's what I wanted to know. Thanks.

 

[kahless2005]

Define perfect...

Chess is on a 10x10 board with the moves designed so that there is a possible counter. So technically, if both players play their best and are able to counter each other's moves, then yes it will end in a draw. Yet at the highest levels of play, draws are rare.

 

[Hurkyl]

Chess is on a 10x10 board with the moves designed so that there is a possible counter. So technically, if both players play their best and are able to counter each other's moves, then yes it will end in a draw.

8x8.

If you have a proof that perfectly played chess will end in a draw, then please share. Otherwise, let me refer you to my previous post. :grumpy:

 

[George Jones]

The only problem is that we don't yet have enough theoretical and computational power to figure out which case is correct.

And we'll never have enough.

I seem to remember reading in Scientific American about 10 or 15 years ago (around the time of either Deep Thought or Deep Blue) that the number of possible chess games is about the same as the number of particles in the universe.

Regards,
George

 

[kahless2005]

8x8.

If you have a proof that perfectly played chess will end in a draw, then please share. Otherwise, let me refer you to my previous post. :grumpy:

Im sorry, I've been playing too much Martian chess...


I was saying that theoretically, if the two players knew every single move and counter to the moves, then it would end in a draw

 

[neurocomp2003]

are we trying to proof NPvP by chess?

 

[siddharth]

I was saying that theoretically, if the two players knew every single move and counter to the moves, then it would end in a draw

That's not right. As Hurkyl pointed out, there are 3 possible cases, and the game need not end in a draw.

Let me give you an example. If black has a position such that, even if black knew every possible legal move, white has a sequence of moves which assures him victory (such examples are easy to construct. For example, a forced mate). There is no "counter" in this example.

There are only a finite(but very large) number of legal moves possible in chess. So, is it possible for white to find a strategy that assures him of victory, no matter what black plays (Like in the previous example)? The answer to that question, is not known as of now. But it is known that the answer will be one of the 3 possible cases which Hurkyl pointed out.

 

[Numbnut247]

doesn't this mean that if we get the best computer for chess possible and make a copy and get the two computers to play each other they will never win (only draws)

 

[arildno]

Not necessarily, no.
Again:
We do not know which of the three cases hold:
1. That there exist a first move for white, so that irrespective of black's moves, there will always exist at least one move for white so that victory for white is a possibility.

2. that irrespective white's moves, there will always exist at least one move for black so that victory for black remains a possibility.

3. That irrespective of white's moves, there will always exist at least one move for black so that draw is a possibility.

 

[matt grime]

I was saying that theoretically, if the two players knew every single move and counter to the moves, then it would end in a draw

Right, let's play a game. Like noughts and crosses. I get to go first and place two symbols, you get to go second and play one symbol at a time. Three in a line wins. Every possibly move is knowable, and analysable. By your reasoning it will always end in a draw, so let's play for $1,000 per game.

Unless of course by 'counter move' you mean one that prevents someone winning, but then how do you know that one of these always exists? Sure, if it did then the game'd be a draw like ordinary noughts and crosses.

 

[vaishakh]

Grime, you are blasting.
Sidharth - you cannot directly start from a position and say that from here there can be forced mate for one of the colours or there is a trivial draw. Forced sacrifices or or draws can be predicted once a stage of chess called opening is over. The thing that you told mostly happens during the endings of the middle game when the one who makes the best possible combination at the earlies wins. I certianly agree with Leonardo because many people like different things. I would blindly play an e4 followed with Nf3. Why I don't know? Many of my friends insist me on d4. the thing is becoz the minds run fast when we are somewhere accustomed to. In fact one of my friend has a habit of playing f4 and this infuriates me a lot but he wins and hence I cannot criticise. Well, the Evergreen game and all are something very interesting when he is down a queen and more than a couple of pieces as well as rook but still wins. They go totally against the theory but brings a victory to make us keep quiet.
So the main problem to discuss on this issue is the first move. Then comes the second move. And gradually after atleast 20 moves comes the situation Siddharth is talking about( If I assume that no one makes blunders).

 

[Hurkyl]

Sidharth - you cannot directly start from a position and say that from here there can be forced mate for one of the colours or there is a trivial draw. Forced sacrifices or or draws can be predicted once a stage of chess called opening is over.

Yes you can. Given enough computing power, it is theoretically possible to determine if the opening position is "white to mate", "black to mate", or "drawn".


I'm going to lock this. No offense intended, but we have several people who don't seem to know anything about game theory, yet are insisting that they are correct.