Can chess be perfect?

I saw a movie the other day called 'Match Point'. It starts with a tennis match: the ball hits the net and it can either bounce backward or forward, means you win the game or you lose it. Luck. Then the movie really starts and we meet a poor irish guy that moves to London. There he meets 2 women, one he got married with and the other one he is crazy about her - played by Scarlett Johanssen ^^- . He even got a baby from her, while still married with the other woman. Scarlett forces him to tell his wife that he got an affaire and wants to divorce her, but he can't. Instead he kills her ( yeah i know).  He throws the only piece left of evidence in the river but it hits a bar, if it falls forward into the river he's free, if not he is charged with murder. It falls into the river.

After i saw this boring film, i thought chess must be all about luck also, like in every other game. In non-coordinate games it's clear, if you got the luck, you're gonna win, like Monopoly or something. But in coordinate games you have the most chance with using your brains, though you can't play it perfectly right? I mean after the opening, in the middlegame everyone needs some luck, even Kasparov or good old Bobby can't play perfect chess, at least that's what i think..What do you think? Can chess be perfect, or do you always need some luck?

Not an expert, but from what I've read of game theory, Chess is theoretically as luck-based as tic-tac-toe. In other words, the game is "drawn" right from move one.

Now a very famous chess player called Steinitz asked pretty much the same question you did and postulated a very famous theory about chess, that also relates to most other games of strategy, or non-luck as you put it.

Steinitz' Laws

1. At the beginning of the game the forces stand in equilibrium.
2. Correct play on both sides maintains this equilibrium and leads to a drawn game.
3. Therefore a player can win only as a consequence of an error made by the opponent. (There is no such thing as a winning move.)
4. As long as the equilibrium is maintained, an attack, however skilful, cannot succeed against correct defence. Such a defence will eventually necessitate the withdrawal and regrouping of the attacking pieces and te attacker will then inevitably suffer disadvantage.
5. Therefore a player should not attack until he already has an advantage, caused by the opponent's error, that justifies the decision to attack.
6. At the beginning of the game a player should not at once seek to attack. Instead, a player should seek to disturb the equilibrium in his favour by inducing the opponent to make an error - a preliminary before attacking.
7. When a sufficient advantage has been obtained, a player must attack or the advantage will be dissipated."

So what is he saying in a nutshell? That, really good players will take a position and keep it for what it is worth till the end. So Kasparov or Anand will take a good position (i.e. better than just drawn/equal) and RARELY make it worse.  Give them a bad position and they will still probably just keep it bad, not make it deteriorate further.

However .... if I gave you or I a position that was "good", we are not GOOD ENOUGH to uphold the postulates of Dr.Steinitz 100% of the time on each move... we may be a pawn up in the middle game, but as our opposition gets stronger and stronger  (as you have noticed) it is MUCH harder to win from "good" positions as well ... one mistake (or weak/error move as Steinitz puts it) and our super-strong opponents will jump back in.

Luck really has very little to do with things, the stronger you get! Though the element of surprise....well, that's a totally different story :)

Hope this gives you something to think about!

S.

Nice read.  Thanks.

Yes, and no.

I look at chess as a combination of science, and art.

From the science perspective, yes, it can be "perfect." Fritz, if you leave it thinking long enough, will come up with the "best" move.

But the reason that chess will never be "solved" or "dead" is that there is an artistic aspect to it. Adjusting on the fly, making brilliant sacrifices, the subtle dance of pieces, can never be "perfect" as it is largely a matter of taste. You may think that a sharp, tactical bloodbath with many sacrifices is "perfect" whereas another may prefer the slow, subtle positional battle down to the endgame.

That is what makes chess such a great game. There are as many styles as there are players.

Didn't you just said yourself that it can't be played perfectly, chess is a combination of science and art, a computer like Fritz is only science. If one super intellegent guy has the science and art, has his unique style and thinks more moves ahead than the computer, isn't it possible to beat him? Or can Fritz think every possible game throgh from move 1? Is he THE best chess computer ever made? I don't know Fritz, but i'm gonna look it up now.. Is he ever got beaten by computer or human?

To Shivsky, the Steinitz Law is very interesting, but i read somewhere that white always has an advantage from the beginning, even if it is 0.1%..but that's probably not enough for a win..

My take:

1. "Luck" can save you from your imperfect play, but only in the form of less perfect play from your opponent (which is arguably skill and not luck).
2. Perfect play takes "luck" out of the equation altogether.
3. Perfect play only exists theoretically -- no human or computer can currently achieve it and no human ever will.

So is there luck in chess?  In theory, no -- in practice, maybe, but only if you equate luck with skill.  I do not.

OK. This is relatively simple.

Chess is a finite game (thanks to the 50 move rule), therefore it is possible to have perfect play by working out all possible moves (though not possible with current processing speeds). For example, checkers, once considered to be on a par with chess concerning intellectual pursuits, has been proven to be a draw with best play with this method. Something similar could easily happen to chess.

Anyone that says chess will never be solved "because there's an artistic element to it" is plaing wrong. Chess is solvable (just beyond our current capacities), and it is humanity's unique interpretation of chess strategy that makes it artistic for us, not anything intrinsic about the game, and it makes chess no less solvable. Consider this: there's artistry in a witty sacrifice leading to a clever checkmate, but that is perfectly solvable, a computer will work that out in a microsecond flat.

Because of the strategy-stealing argument (as in, any strategy for black is even better for white with his extra move), either white has an advantage (not necessarily a winning one) or every single possible first move is a mistake (as in, white, on the first move, is in zugzwang). I find it far more likely that white has the advantage, but I doubt think it's enough for a win - especially considering that there's a strong correlation between player skill (given two evenly matched opponents) and chance of a draw (just think of how many draws there are in world championship games). If it is, it's probably a complicated way to mess with black's pawns, exchange off enough pieces and convert to a winning endgame by promoting pawns.

Some interesting discussion on the topic (solvability, not luck) can be found here:

http://www.chess.com/forum/view/general/the-future-of-computer-chess?quote_id=2308578

Nobody knows what luck is.  The best I could offer as a definition would be: unpredictable consequences.  Beneficial unpredictable consequences = good luck.  Since, as has been noted, no human can play perfectly, and will at some points play unpredictably, we can therefore all experience good or bad luck on the chessboard.

I still say that what you're calling good and bad luck at the chess board I call bad or good play by my opponent and as a result it's the product of skill and not luck.

But whether they play well or poorly is outside of my control - and therefore it is MY good or bad luck to be the recipient of it.

If you could bring one chess player back from the dead....who would it be? Today's winner would be Wilhelm Steinitz. Thanks for posting those laws Shivsky! Interesting.

That's an awfully subjective definition of luck.  The way I see it, in a contest of skills, such as chess, my opponent's good or bad skill is just that -- skill -- despite the fact that it's outside of my control. 

Luck to me is something that is outside of the control of all participants and not skill to one but luck to another.  The roll of the dice or the deal of the cards are rooted in luck.  In chess, all the relevant informtion is on the board -- it's skill that dictates how well you and your opponent process that information and decide how to act.

As you said, of course, it really does come down to how one defines luck.  Fundamentally, it sounds as though you and I have different definitions.

To expand on my previous post, I'd counter with this definition of luck:

Outcomes that cannot be affected by anyone.

Seems reasonable.  But still, there are times when a player will make a poor move that is below their skill level - for any of a number of reasons beyond their control.  Could be poor health, distraction, momentary lack of concentration, fatigue.  Whether any of these factors come into play at any given moment in a chess game is beyond the control of both players - and therefore is that not luck, or chance?

Even in the highest levels of chess - including computers - chess is still confusing enough that a move that looks perfectly good may prove to give a tiny unforseen advantage to the opponent, and that advantage may be enough to win. This kind of situation has nothing to do with skill. It is genuinely good/bad luck (until chess can be understood better). As much as it is a logical truth that chess can be solved eventually, currently it is not well enough understood to be that deterministic as to say that everything is skill and nothing is luck.

A simple proof that chess is solvable: the 50 move rule means that if 50 moves have passed since the last pawn move or capture, the game ends in a draw. There are 6 moves for each pawn possible (but not all of them can do all these moves) and 15 captures possible. That gives an absolute maximum length of a game before a result at 49 initial moves before the first pawn move or capture, plus 6x8x2 + 30 = 126 captures and pawn moves resetting the 50 move counter, so an absolute maximum game length of 6349 moves (actually, the number is much much much less, because not all pawn moves for both sides and captures are possible). Work out all possible move sequences for these 6349 moves, and you have solved chess. It's insanely difficult but not impossible.

The number's actually around 500 moves less than that:

http://blog.chess.com/kurtgodden/the-longest-possible-chess-game-revisited

and it assumes that a draw would be claimed any time 50 moves is reached, although that's not actually how the rule reads.

Chess games need not be finite because of the way those rules are written, but any infinite positions are inherently drawn and detectably so.

I think we just said the same thing.

Thanks for the link, TheGrobe :) fair enough about the having to claim, but in that case, whichever side wants to claim has a forced draw, so that's something.