Is there such thing as "luck" in chess?

the tales are different but the story is the same:  thinking errors in a game of logic.

Baseball is actually a great example of exactly the opposite point. Teams pour millions of dollars into their analytics departments looking to gain an edge in parsing skill from random variation. It's why e.g. a D2 college team beat the Phillies in a spring training game a few years back. Or why back when the draft was 50 rounds, there was tons of excess value to be had outside the high money picks (believe HOFer Piazza and future HOFer Pujols are prime examples); you can't just predict future success by looking at the stat leaders in college.

I'm reminded by the quote, I'm gonna butcher it and forgot who said it, but something like: "when you lose early in the season, they call it a bad start; when you lose in the middle of the season, they call it a cold-streak; when you lose late in the season, they call it choking". Point being that people constantly craft narratives around what's really just random variation. There's a reason the regular season is 162 games long. 

Same thing with hitters who are ostensibly struggling at the plate. "He's only hitting .270 this year, last year he was hitting .300!" Ignoring the fact that on a normal curve of performance, one would expect variation like that to happen something like 1/3 of the time, there was also an aha moment a few years back when teams started looking at league-wide BABIP and realizing that most people will regress toward league average on a large enough sample. So for example someone with an exceedingly low BABIP during a stretch is definitionally unlucky. Same thing for the converse

A series of poor swings is not some random chance, it's a lack of skill.  Loss of focus and concentration that causes someone to "miss" something is a lack of skill.  Skill is not just the ability to do something, but the ability to do something with increasing levels of consistency.

The "story" of each mistake is imprecise and inaccurate , ergo the need to formulate a narrative, to drive the agenda.  The second you put your foot on a slippery rock and your muscles attempt to balance you, it becomes a question of skill.  Luck represents random chance where skill cannot reasonably be applied and will not affect the outcome, like playing Roulette.

To add to the above, back when I played whenever pitchers threw a shutout the whole staff got to take batting practice (meaning pitchers would only ever even pick up a bat a handful of times each season). We'd always dream of getting an at bat in a game on the off chance there is ever a weird situation where it would take place. In one game, our pitcher got an at bat, and ended up getting a hit. He ended his career 1-1, after not being good enough to hit in high school, never mind college. In the same game, against the same pitcher, our programs all-time hits leader (and somebody who went on to play professionally) struck out. That's just the randomness of baseball at play, not skill.

It also points to what I've been saying about the number of trials needed to differentiate luck from skill. Against players of vastly different skill sets, the requisite sample size to always correctly identify the better hitter in baseball is pretty small. But among even semi-closely talented players, the requisite sample size to identify the better player is quite large. In any given sample of 10 at bats, it would not be weird to see Mike Trout have worse outcomes than someone who couldn't crack the starting lineup at an SEC school, for example.

The instance of the one hit was skill.  You can argue about the circumstances of his lack of attempts, but that is not part of the game of baseball.  When he had his chance, he took a swing and connected, and it was not random chance that he connected.  The moment you move 0.1% away from random chance, it's skill-based.

As a developer, it's more obvious.  There's only one way to achieve "luck" in programming...by using a random number generator and seeding it correctly.

Your incorrect interpretation is a text book logical fallacy that is very commonly exhibited in people without a background in stats https://www.logicallyfallacious.com/logicalfallacies/Texas-Sharpshooter-Fallacy

As to your prior point about increasing consistency, that is also incorrect. It's why people have differently shaped distributions; i.e. someone can be a better player and simultaneously have with wider variance in their play. In chess, a good example might be Ivanchuck. When he was at his best nobody could beat him. But he was never a world champ. In finance, it is why people look at risk-adjusted returns with Sharpe Ratios or drawdown metrics as opposed to exclusively average returns. This kind of thing permeates everything we do, and the larger the number of trials needed to determine one's relative competency with certitude, the more the activity falls to the luck-side of the spectrum. 

While I don't like the author, I suggest the book "Fooled by Randomness" to get more of a feel

If you accept that definition there is no luck in chess. But the very idea that things can be considered in isolation from the whole like that is ludicrous imo. 

Lol.  Please lay out step by step how my position mirrors this fallacy, or is opposed to/different than your Ivanchuk example.  Don't make assumptions/extrapolations about what I said, and don't misspell Ivanchuk .  A distribution range does not imply luck, it can equally denote inconsistent application of skill.  Statistics is a piss poor way to judge luck, but a good way to ascertain probability.  Probability uncertainty is not the same as "luck".  Maybe that should be a fallacy...

Attributing ex-post subscribes meaning to something that doesn't necessarily have any. If you exclusively look at the outcome ("hey, he got a hit!" or "hey, he hit the bullseye!") without looking at what led to it ("oh, he just closed his eyes and prayed" or "oh, he just closed his eyes and pulled the trigger, the bullet had to go somewhere"), you are erring. On the flip side, maybe you are familiar with the phrase "do everything right and still go wrong"? Similar concept that again shows variance at play in everything humans do (which was my earlier point about nothing humans do being deterministic). 

I don't know why you keep bringing up sports, but I played baseball at a pretty high level and the game is filled with luck as my prior examples illustrate (BABIP stabilization, outlier analysis, hot-hand and freaky-streaks, etc.). Ask someone who is about to invest a million dollars in a player who they would rather spend it on ceteris paribus: someone who just went a season with a +1 SD batting average and +3 BABIP, or someone with a -1SD batting average but -3 SD BABIP. To say a skill-based game like chess or baseball has elements of luck in it is not diminishing it at all; you are acting like this isn't a huge field of study academically or a huge piece of several multi-billion dollar industry (insurance, finance, manufacturing, etc.). 

Chess, within the very game itself, by virtue of having to choose a move, has elements of luck built into it's fabric. I return to my prior example of two similarly ranked players who play a series of games. It is fallacious to think that if player A splits with player B 5/5, that in the 5 wins, player A wins exclusively because he is more skillful and loses because he is less skillful. That is like the definition of an attribution error. In reality, there are probably a handful of critical positions in each game where each player considers something like 5 candidate moves in the position. That means, ex ante, there is a 20% chance they select the winning move just by chance. Interpreting the results afterward through the lens of the conclusion is not sound reasoning.

The design of the game of chess would seem to be all skill, but its complexity keeps full mastery out of reach. Players rise to the top because they are more skilled. Even so, even the best find themselves in positions against weaker players where their skill is insufficient to control the outcome. What is the word for these failures of skill against the less skilled?

I never implied this. I said he is a player with higher variance than most, i.e. he can have a wide range of results. To attribute his poor performances exclusively to his lack of skill on that day and his great performances exclusively to sudden strokes of brilliance is poor reasoning. It is fitting a narrative with the benefit of hind-sight rather than acknowledging the natural variation of his play.  Worse players can have tighter distributions about their play, and still have their best days intersect with his worst days. It doesn't change the fact that they are worse players. 

Skill is not a constant.  It's a range.  Elo/Glicko operates on this premise.  Otherwise someone rated 50 points higher than you should win games against you all the time.  So it's entirely possible for a higher rated player to lose to a lower rated player due to lack of skill on the higher rated player's part.  When Karpov blundered and resigned in 12 moves, that was not luck.  It was a lack of skill, or if you prefer, a lapse in concentration/focus.  Karpov, as it turns out, is a pretty consistent player, and never repeated this type of blunder, but it was skill, nevertheless. Maintaining consistent concentration and not making simple, avoidable blunders is a skill, as any player trying to get from 1500 to 2500 could attest .

Without repeating the same thing, read my reply to coolout. Since I am not the most eloquent or short-winded: "The Texas Sharpshooter Fallacy is when outcomes are analyzed out of context, giving the illusion of causation rather than attributing the outcomes to chance. The Texas Sharpshooter Fallacy fails to take randomness into account when determining cause and effect, instead emphasizing how outcomes are similar rather than how they are different." 

When you said, "The instance of the one hit was skill...When he had his chance, he took a swing and connected, and it was not random chance that he connected" you are only looking at the outcome to determine the event's causation, rather than interpreting the event with any sort of prior context. 

As I mentioned, I don't really like the author, but he does a good job here: https://www.amazon.com/Fooled-Randomness-Hidden-Markets-Incerto/dp/0812975219

No, it does not mean that at all (that I am only looking at the outcome).  How exactly does what I said imply that I am only looking at the outcome?  I am looking at the components of a swing of a baseball bat.  Did he attempt to swing the bat at a certain elevation (literally any elevation)?  Then he applied skill, and his attempt was successful in this case.  It's not luck.  If he even *chooses to face in the right direction*, that is also an application of skill (a very poor one, but an application of skill nonetheless) , not luck.  If he takes a breath and holds it for the pitch, that is an application of skill.  If he lowers the brow of his cap to avoid getting the sun in his eyes, that is an application of skill.  Basically anything he has learned that he decides to do that is ergo not random, is an attempted application of skill, which will be more or less successful, but regardless each such attempt is skill-based, not luck-based.

That's why I said don't make assumptions.

There was no assumptions. It is definitionally what you said. There are only two cases to consider:

1) you reject any prior (i.e. you always say the outcome, whether hit or miss, "was not random"), in which case you are vacuously demonstrative of the fallacy or

2) your prior is the information I provided, which only points to the negative outcome, and so your conclusion, that "it was not random chance that he connected" rests exclusively on the information that he actually connected. Whereas if you actually did have information that lent credence to his ability, like seeing the "components of a swing of [his] bat" or something similar (like maybe he played behind two major leaguers in high school, and was a great hitter in middle school), then your conclusion might have been built on the prior information (and you might have been able to more correctly identify the skillful batter who missed from a non-skillful batter who connected, for example). 

The fact that you need multiple trials to parse skill/competency from luck/randomness, whether it be in baseball, law, finance, or chess, means that luck is inherent in the activity. Nothing humans do is fully deterministic.

Ending with "period" doesn't make something true and will only ever serve to detract from a point you are making.

"consistency determines pros from the amateurs"- I just gave you a tangible example of where that is not the case. An amateur can be very consistent, and a pro can be very inconsistent. That is the deviation half of the mean/variance makeup of a player. Skill is the mean half (or if you'd rather, skill is the peak).

"People rank where they rank based on their skill not luck."- over enough trials, this is certainly the case. It also has nothing to do with what I've been talking about, which is that, by virtue of having to select a move in a position, there is luck involved (after all, there is say 1 in 50 probability of selecting the computer move). As the number of moves/games increases, the influence of luck decreases. This is true of any human endeavor, and how we determine where an activity sits on the luck to skill spectrum is the number of trials needed to reliably classify relative competencies

     I once saw Willie Mays duck out of the way of a wild fastball coming at his head. The ball missed him but hit the bat and bloopped down the third-base line for a double. According to the definition of "luck" some posters here insist upon using, that must have been skill. Willie stood on second base laughing uproariously,. Teammates and oppponents joined on,and postgame comments all talked about how the best player in the game didn't really need such a stroke of LUCK.

     Any activity in which something as inconsistent and unpredictable as a human being is involved cannot fail to have some element of chance involved.

"When you claim that lower skilled players only beat better players by luck"- I never said this and don't think it. Of course, a lower rated player can be more skillful in a single game and win as a result of that skill. This is true of any major sport; it's why we have series to determine champions. You are clinging to the false belief that someone only can beat someone 100% based on luck or 100% based on skill, and that attributing luck to something is derogatory. A poker pro is dealt random cards (luck) and yet over sufficient trials, proves his worth (skill).  

"ou mean information like seeing him actually swing the bat and not some other device doing it besides his human person?" No I mean visually, if you are a trained scout for example, and watch Mike Trout swing a bat and a middle schooler swing a bat (prior information), without knowing the outcome of the pitch (post information), you will be able to form a hypothesis about the relative skill level of each player that is not guided by the outcome of the pitch. 

"Why do you ignore the definitions of luck verse skill?"- You are the one mischaracterizing them. Human skill necessitates luck. We are not able to ascertain ability without comparing it to chance outcomes.

Have you considered that it's not the academics who've spent their life studying entropy, or the businesses pouring billions of dollars into researching causality who are wrong, but you?

    What a load of bull. You think your ideas MUST be correct so you will twist anything to show you are right. Willie made NO effort to hit the ball. Willie admitted it was sheer luck I think I'll choose his view on baseball over yours.