does luck exist in chess

Computers don't make blunders in calculation, but their vision is also limited. Let's say that computer A has calculated a tactical line 20 ply deep and figured that the line wins a pawn. However, it also leads to an endgame that is drawn. This was not foreseen by either computer, it just happens to be a drawn endgame. In this sense, computer B was lucky.

I have trouble believing that you honestly believe this.

Surely our ability to build complete five, six and now seven peice tablebases is compelling evidence to the contrary.

Is there something fundamentally different between the game with seven peices verus with eight that makes luck a factor in the latter and not the former?  Does this somehow mysteriously change when we acheive a complete eight peice tablebase?  If not, at what point between nine and 32 does it change?

The point is 32. If we reached a point where chess would be a solved game with 32 pieces, and if we had computers who could memorize tablebases of such volume, then luck would no longer be a factor. (Unless one of the computers crashed during the game.)

But isn't the point regarding computers that there is no luck: the only reason that chess has not been solved is that we have not yet built a powerful enough computer. The fact that e.g. Rybka vs Houdini produces wins for each engine is because of the compromises made to complete the game in a tractable amount of time. The errors arising from the time limitation are themselves likely to be deterministic.

Also, just because something appears random does not mean that it actually is.  An inability or failure to understand the underlying deterministic causality behind an effect does not mean that it does not exist and the effect is therefore non-deterministic.

Chess is a perfect information game, and as such, is devoid of luck.

Every OTB player I have known, after squeaking a ropey win, has said "I was lucky". Whether or not their experiences equate to some formal definition of luck is debatable but the fact that they felt lucky is not.

I still think it is luck if your opponent chooses an opening I have prepared very well for. and know a lot of variation.

- I have no mind control skills yet(still working on it lol). therefor I think this can go under the definition of the luck. because it is an event I cannot control.

When humans are involved luck certainly exists, in the sense of a player who you would reasonably expect to spot something missing it. You could argue that there is an underlying causality in the way humans make mistakes in chess but this could easily be a lifetime's worth of research, with no indisputable conclusions.

With computers, the errors they make in chess are a function of compromises in their algorithms and are deterministic.

Based on my  past severak years of playing online chess, I would say that luck happens when someone rings your doorbell and you run out of time going to answer it.  Likewise when the phone rings. I have lost many games this way.   In a tourniment these distractions do not exist so I say luck does not exist in 99.99% of the games played. 

There seems to be two issues here:

1. The role of external factors

2. The role of chance within the game itself

As for external factors, it is particularly obvious when playing on the Internet, eg unforeseen distractions, unpredictable disconnections/lag etc. However in a less visible way it exists in tournaments also, eg one of the players might be sick, tired, distracted by private problems etc. Even in WCH matches, Korchnoy's distractions with his family still being back in in the USSR and the troubles they had arguably affected the results. Karpov apparently wasn't involved in this, but arguably he was a bit lucky.

Like I said 2 pages ago, the game itself has no luck, but add humans to the mix and there is an element of chance.  Not just outside distractions but internal difficulties in the mind itself that are unpredictable.

This is also true for computers who are not perfect.  Playing with hash tables and thinking time you can observe engines on identical hardware prefer different moves.  Although rare, also aberrations where they seem to get "stuck" on one move.  Programmers of course work to eliminate these things.

The game itself in isolation has no chance, but such an observation has very little practical value.  For imperfect players, the complexity creates chance.  Unless you're arguing from a deterministic philosophical view, in which case you would have to at least agree that luck or chance is observed although no there in the truest sense.

In chess, the concept of luck exists... but not the one of happiness

It depends if you randomly prepared an opening and your opponent happened to play it or you looked up your opponent's games in a database and specifically prepared an opening you know he plays.

the first about random opening I would qualify as luck.

but preparing an opening that he plays is part of chess preparation the better player I meet the more the change is that I can read about their opening.

I was once reading everything I could find about e4 b6 against a players that always played b6 in this game for some unknown reason he played d6. :-) but I won anyway. 

It's not unknowable, but as humans don't know everything there is to know in a chess game, things can happen that were, to them, out of their control, and those things can be good or bad. Good is luck, bad is not. Like I said, when you add the human factor, it turns it into luck. If poker is not pure skill, then why should using intuition to determine the validity of a piece sacrifice they can't fully calculate be so? Like in poker, one can try to make good choices, but they can't know 100% if it works unless they saw the concrete 16 move variation through.

The "outside" factors are basically the potential ideas laid out in a position, and thinking of this like a poker game, we can loosely equate this potential of the opponent's unknown resources to his unseen hand. Even though the game (chess) technically is of complete information, since humans can't truly see all of that information, due to their limitations, some part of the outcome can be considered luck, as one may or may not be punished for not seeing something. Now obviously, as mentioned I think by Fezzik in his nice concise original post, a strong player is basically minimizing the effects of luck as he's seeing more pitfalls; obviously if you see the variation perfectly you don't have to pray it turns out well for you, yet at the same time the amateur can sometimes fortuitously make the same move and get the same result, even though he didn't see every challenge to the move. Strong players see more of that stuff so have less to be afraid of. Seeing face down cards in poker is of course not correlated to human ability, but it's still a similar effect.

Yeah if you recollect a recent thread, someone played a piece sac that looked attractive, but he couldn't actually calculate the winning line. (The winning line was there but it was sufficiently long and complicated that I doubt if any amateur could calculate it in a blitz game.) As it happened in the game, his opponent got scared and made a move that lost easily. It seems to me that he was lucky then, for if his opponent accepted the sac he wouldn't have been able to find the correct continuation.

Chess is a theoretically perfect information game. In practice, all decisions are made with less than perfect information because the people (or computers) making the decision can't use all of the information.

Not all, as in, if you see a forced mate in two and go for it, then you are indeed playing it as perfect information game. However, most of the moves in a game don't in practice involve such clear situations, and these situations are a relatively trivial part of the game in practice. 

Take tic-tac-toe as a much simplified example of another perfect information game in which the ability to use all of the information is well within our capabilities.

If two players who don't realize this (or understand how to do this) play each other, is luck a factor here or is does it simply come down to a difference in skill that dictates the result?

The vast majority of it is skill, but luck can be a small factor, for reasons mentioned above.

Ok, lets talk about tic-tac-toe a little bit. If one did not know the winning strategy, it's still possible for them to guess the right choice and perhaps draw against someone who does (a good outcome coming from poor skill; is that really a pure skill game then?). He is relying on things to work out regardless of the quality of thought behind his decisions, while if we take the guy who knows, he, as mentioned earlier, could be considered, effectively, minimizing the luck factor to the point of 100% reliability. All it is is just an extreme version of this minimalization process in chess.

Because there is so much information, the interpretation of that information is what determines how much you will miss, and is thus out of control, which will thus make you hopeful for a good outcome, which may or may not be true; in effect, you're taking some kind of chance.

A pure skill game, which practically is impossible to make, is one where you have to have the right intent behind each move (so not just get the move right) in order to be successful; if you don't you somehow get punished in the long run. But that's not how it always works - sometimes you find resources along the way that weren't planned.

So for example, in tic-tac-toe, for it to be pure skill it would have to be able to punish you for not knowing the best strategy every single time, but that's not how it always turns out! Sometimes you get rewarded for your ideas even if the idea came for the wrong reasons, which demonstrates flawed skill, yet if you draw a guy who knows the strategy the result of the game says you guys are equal. I call that bit lucky, don't you?

I really do think this uncertainty, though not exactly the same as the kind in cards, creates a chance element similar to a game like gin rummy, though to a much less significant extent.

I'm sorry to be fastidious, but I don't think chess can be considered absolutely 100% skill if it's possible to get away with playing with less skill sometimes. For example it's possible to make the same moves as someone, but have the wrong intention behind them, and sometimes you may get punished for it, sometimes you won't.

When you lose (chess), or don't have(cards), control, you create (or already have, in the case of card games) chance. In chess this creation is rather artificial, but nonetheless this element is created.