Luck in Chess

That is not luck. That is a good guess.

Looking at this thread, the discussion seems to be a straightforward one between determinists and non-determinists, where the environment of "specifically chess" is irrelevant and all we have is some people saying that, yes, probabilities, randomness and luck do exist, and others, who have been told by a higher authority that probabilities don't exist and Bohm's Hidden Variables interpretation of quantum mechanics is correct. If they would kindly reveal what this higher authority is, we may have a chance of arguing against it.

You can say there is luck in everything or you can say the universe follows deterministic laws making luck impossible. It doesnt matter. You are saying the same thing in both vases but simply chosing a different isomorphic model. Its semantics. Like most questions given in an ambiguous language like english this question is only semantics. We believe that chess is a game of perfect information as defined by game theory so the wording of the question is clearly misleading.

There does exist a perfect strategy. It is the inevitable human error which is in fact the reason anyone plays chess. Other things are irrelevant.

mmm

Deep Blue was lucky in its upset over Kasparov.

Is perfect strategy the Nimzo/Queen's Indian complex, Najdorf, or Petroff (depending on if 1.d4 or 1.e4 are best)?  The world may never know... or know with good enough machines.

"Who is going to be stupid enough to say that is not luck?"

That's not really an argument, and summarizing something in this way oversimplifies many of the things that the discussion depends on.

I could say that it's "extremely dumb and confused" to think that everything involving uncertainty simply must make a game a game of chance, or something else like that that makes the position sound good, but to just say that while offering no reasons isn't going to convince people.

I'm not sure why the issue of whether or not determinism is true needs to come into play. Determinism is pretty much stipulated as regards chess, even if not the world as a whole is deterministic. Otherwise the rules of the game would change every few seconds or something. We don't call something chess whenever a rule like how pawns can move changes, so we are talking about some set of rules that we created that indeed would operate deterministically. Again, otherwise, we are referring to some other thing that is not chess.

Ah, I guess determinism as a whole comes into play if we're trying to say that nothing is luck. Although it seems "naive" (see I can make my position automatically look good by inserting nice words like that huh? ) to think that there aren't ways to make sense of luck just because someone believes in determinism. All you really have to do is say that chance results when you have insufficient information to determine an answer. A 50/50 shot of a coin being heads is just telling you how good your prediction is; it's not saying that whether or not a coin is heads is in a constant state of flux.

Yeah I do suspect that the term "luck" for example here is being used in a number of slightly different circumstances yet in each case is being passed off as the same thing.

In chess, luck is an excuse made up by people who suck because they can't duck away from the inevitable yuck that is their chess skill of muck.

"If other people don't come to the same conclusion as me, it must be because they are wrong, and that I am right. I do not have the mental faculties to conceive of some other possibility."

Yeah I mean that's the sort of thing I've been saying, that when we are surprised and such it's because of the "probabilistic feelings" we decided to give ourselves. But in poker you would eventually run out of theoretical resources in predicting what card will be on the turn, river, etc., if you don't know the cards in the deck. You might be able to narrow your choice down, but there is a limit. In chess there isn't really that sort of barrier -- it's just a matter of how deep you look at the position etc, for which there will be human limitations.

So I guess what I have been trying to say is that the human reaction to something, e.g., thinking it might be this or that, is just that, a human reaction. It says something about humans and how they infer things. If you infer things in an imperfect way, yes, you will be right some of the time and wrong some of the time. That doesn't make your inference random.

Now true, in the poker example I am taking for granted that you were, say, not looking at the cards as they were being shuffled. In poker that would give you incomplete information, but in chess, the information is always available to you. It's not something you have to "catch" within a few seconds. But unlike seeing what card will come up next, the information you have in chess does not always make it clear what to do with it.

The following is semantics, but it might be useful to talk about semantics here :) It's clear that there are different things that can cause uncertainty, and whatever "luck" this is can vary depending on that cause. I think it's pretty misleading to say that "luck" applies to both poker and chess -- at the very least you'd have to say that there is a different kind of luck in each, but honestly, if you're going to say that, you might as well use a different word to describe each, because you're talking about two different things. If you want to say that you don't know whether your move will work or not in chess, ok, but that doesn't suddenly change the rules of chess into poker's rules.

To just say "I got lucky" in chess doesn't tell you so much about what actually made the result come about. What really happened is that your way of thinking came up with better moves than the opponent. You can consider that fortunate or not, but again, if you want a description of how the result came about you'll have to go deeper than that; for example, were the moves that you played the result of incomplete information and just seeing that new info be revealed, or was it the result of taking information and seeing how your interpretation of it fits with reality. In the former case you are describing entities you have no connection to -- the logically unpredictable; in the latter case you are connected to the entities (no info about them was hidden from you); you merely didn't understand what it all meant -- the uncertainty stems from your subjectivity alone.

So, yeah, we can call it all luck or not, or different kinds of luck. I mean, sure, we can relate to any risk-taker's uncertainty even in a deterministic game as chess, no one would disagree with that. I'm just not sure why everyone is so concerned about attaching the same banal, ambiguous, "luck" descriptor to everything, when we can probably have less confusion by just explaining what we mean rather than attaching a generic label in a simplistic, inaccurate attempt to summarize whatever we're talking about.

"If I have 2 candidate movements, and I dont have the skill to determine that one is a blunder and the other is a good movement, and I make my decision flipping a coin, I can be lucky an choose the good one."

Yeah I mean this is tricky to respond to but it seems a bit suspicious. So if you are 99% sure of something we would agree there is a lot of skill leading to that, but suddenly when it's 50-50 we just "become" probabilistic or something? Yet the sort of stuff that led you to your 50-50 belief is not any less deterministic than what led you to your 99% belief.

And the 50-50 shot is not, in this case, based on two literally equal possibilities like in a coin flip, but is just your subjective representation of the complete info available to you. For all you know one of the possibilities you think might be good might not make any sense. Again this is different from a coin flip scenario because all of the info available to you can't suggest one possibility or the other; that's not subjectivity, that's just logic. (Again yes I am assuming that our minds are not so quick as to read a person's mind as to how they will flip a coin -- I'm sure people get what I mean here. If you don't know the physics of the coin flip there is absolutely no way to tell how heads or tails are even different, let alone which choice makes more sense)

I am assuming that when you said 2 candidate moves, you meant two moves you were willing to consider. If it's literally just two possible moves, and you are able to randomize the choice with a coin flip, ok, yes, I would agree, that is indeed luck. But that's not how chess is played, frankly. People think of moves; they don't, and simply can't, flip coins in their heads.

Ancares wrote:

MuhammadAreez10 escribió:

I guess it is resolved! Chess has no luck.

We might argue if luck exists...

Consider the following example:

I program a computer to play chess, with random movement.

If this computer plays against a computer that is programed the same way, wouldn´t it be pure luck the result of the game?

My 4 year old son plays chess this way, almost ramdonmly. So we can say there is a lot of luck involved in his games against other 4 year olds that play the same way.

When we learn to play, we stop making all random movements, but sometimes he have to chose between 2 or 3 options and we evaluate those options to be equal (at least this happens to me). So sometimes I choose 1 option almost ramdonly, and here you can be lucky (if for example you choose the best option without knowing).

I´m not a grand master in chess, but given a position there is a chance that I make the best movement even without knowing what I´m doing. If I choose the best movement for the wrong reasons, isn´t that lucky?

Areez writes:- There's a problem in your explanation. You seem to apply the term luck to a variety of situations where luck doesn't have a chance to be there.

When two computers play randomly, the result of the game is not decided by luck. I don't think I can explain this well as I'm just 11 yr old. Okay, the computers play the moves with no thinking, without any plan and regardless of the position. But it will only play 'legal' moves. Thus, from all the moves that are available in that position, only one will be played. This can also make a perfect game. But the computer does nothing from its part. It's just probability. In a complex position, the probability can be 0.00000000000000000000001%. This makes the moves pointless. And a stupidish game.

If a 4 year old who does not understand anything about chess plays random-like moves from his intuition, he will use what he knows about the game and his understanding - no matter how absurd - will get him to play a move he want. Something you control entirely isn't luck, not even chance or probability.

Now you say that when we learn the game, we stop making random-like moves and apply the principles and understandings to come up with candidate moves. Yes. But from where do those moves come from? Your thinking process. So, the move you choose to play, even if it's the best move without thinking properly, has been played by you. You are responsible for making that move. You cannot blame luck for it being bad or good. You played it, then why not acknowledge it?

Thus, the conclusion is, There is NO Luck in Chess.

Elubas wrote:

"If I have 2 candidate movements, and I dont have the skill to determine that one is a blunder and the other is a good movement, and I make my decision flipping a coin, I can be lucky an choose the good one."

Yeah I mean this is tricky to respond to but it seems a bit suspicious. So if you are 99% sure of something we would agree there is a lot of skill leading to that, but suddenly when it's 50-50 we just "become" probabilistic or something? Yet the sort of stuff that led you to your 50-50 belief is not any less deterministic than what led you to your 99% belief.

And the 50-50 shot is not, in this case, based on two literally equal possibilities like in a coin flip, but is just your subjective representation of the complete info available to you. For all you know one of the possibilities you think might be good might not make any sense. Again this is different from a coin flip scenario because all of the info available to you can't suggest one possibility or the other; that's not subjectivity, that's just logic. (Again yes I am assuming that our minds are not so quick as to read a person's mind as to how they will flip a coin -- I'm sure people get what I mean here. If you don't know the physics of the coin flip there is absolutely no way to tell how heads or tails are even different, let alone which choice makes more sense)

I am assuming that when you said 2 candidate moves, you meant two moves you were willing to consider. If it's literally just two possible moves, and you are able to randomize the choice with a coin flip, ok, yes, I would agree, that is indeed luck. But that's not how chess is played, frankly. People think of moves; they don't, and simply can't, flip coins in their heads.

Excellently written Elubas! +1

Even within the context of deterministic forces (in this case playing and study experience that influences our decision making process) luck can still come into play.  What if someone recalls a strategy Kasparov successfully employed but one small detail in the position means it's a complete dud if he tries implementing it?  Sure it isn't completely random but your opponent's carelessness in that case means you were lucky he didn't play better.

I think anyone who has played chess long knows that there is some luck in chess .  

Do you know how computers play chess? They evaluate movements and assign a numeric score to each movement. Then they pick the movement with the best score. And what happens when two o more movements are tied with the same score? Yes, they pick one ramdomly (exactly the same as flipping a coin). So there is absolutely no doubt that there is luck when a computer is involved.

If you ask the computer to analyse the movements with more time to compute, they will be able to assing a different score to the movements and select better.

With humans it´s exactly the same. I don´t know about other people, but I can assure you that many times I come to the conclusion that several movements are equally good and choose one ramdomly (and if you are short of time this is even more common). Of course those movements are not equally good, and sometimes I end up choosing a blunder and some other times I choose a good movement.

Your skill only determines the quality of your candidate movements (a grand master will never consider a blunder as candidate movement and most of them will be excelent movements, while a lower level player will have a lot of blunders among his selected candidate movements).

This is exactly the reason why I used the example at the crossroad.

When flipping a coin you have no control. But when turning right or left in a crossroad, that decision is yours.

If you don´t know the way (exactly as you don´t know which movement is the right one in chess) and just choose to turn right for no reason at all (exactly as you can choose one movement out of the possible ones), what exactly determines that you choose the right way or the wrong way?

If among two or more possibilities you (or something else as a coin or a dice) choose one for no reason, but randomly, thats exactly when luck enters into play.

The fact that the random decision is made by a coin, a dice, a computer or a human being is irrelevant. Always that a decision is made randomly we say the output is determined by luck.