Sometimes it does literally feel like a coin toss going into every long game. Everyone is of course trying their absolute best, and of course everyone can easily make mistakes. The game result depends on any random thing that causes both players' mistakes, and you have no idea who will be lucky enough to win in the end.
Here is one simplified way of thinking about it: Imagine every game your opponent has an associated "play quality score" (PQS) between 0 and 100. If their PQS is 100, you will lose, if their PQS is 0, you will win. (Not necessarily just centipawn loss or accuracy, but some vague quantity that is hard to precisely measure in practice.)
So, there is some threshold q between 0 and 100 quality (likely depending on your overall skill level, mindset at the time of the game, and other technical things like opening choice etc.) where if your opponent plays at more than q quality you will certainly lose and if your opponent plays at less than q quality you can win or draw. Every player will have some q less than 100; even engines don't play absolutely perfect chess.
If you are playing a random opponent of the same rating, their PQS will be somewhere around q and you're essentially rolling the dice on whether it will be above or below the threshold. Of course, in principle, you can raise the threshold by getting better at the game, but then you will just be paired against people closer to the new threshold.
So, there is some threshold q between 0 and 100 quality (likely depending on your overall skill level, mindset at the time of the game, and other technical things like opening choice etc.) where if your opponent plays at more than q quality you will certainly lose and if your opponent plays at less than q quality you can win or draw. Every player will have some q less than 100; even engines don't play absolutely perfect chess.
If you are playing a random opponent of the same rating, their PQS will be somewhere around q and you're essentially rolling the dice on whether it will be above or below the threshold. Of course, in principle, you can raise the threshold by getting better at the game, but then you will just be paired against people closer to the new threshold.
@FunnyAnimatorJimTV said in #2:
> Everyone is of course trying their absolute best, and of course everyone can easily make mistakes.
Agreed. But I believe an important part of what differentiates strong(er) players from weak(er) ones is that in general, stronger players are also more consistent. It's not a strict rule and even on the same level, some players are more consistent than others. But big part of what makes our results random are factors that can be at least partially controlled: lack of sleep, feeling comfortable, avoiding distractions (both external and internal) etc. Learning to minimize these negative factors is an important part of chess improvement.
Of course, it can never be fully eliminated and there are also factors that are out of our control so that there will always be some luck involved. After all, it's even quantified in lichess insights, isn't it? :-)
> Everyone is of course trying their absolute best, and of course everyone can easily make mistakes.
Agreed. But I believe an important part of what differentiates strong(er) players from weak(er) ones is that in general, stronger players are also more consistent. It's not a strict rule and even on the same level, some players are more consistent than others. But big part of what makes our results random are factors that can be at least partially controlled: lack of sleep, feeling comfortable, avoiding distractions (both external and internal) etc. Learning to minimize these negative factors is an important part of chess improvement.
Of course, it can never be fully eliminated and there are also factors that are out of our control so that there will always be some luck involved. After all, it's even quantified in lichess insights, isn't it? :-)
chess as a game is purely deterministic, but considering that humans are in a world with dynamic and non deterministic nature adds more "luck" variables in the game, and it is widely known as "playing the opponent not the board", i personally consider this dirty play, and a specific group of GMs from the same nation play like that, it is simple, you see what irritates your opponent the most and use it against them, for example, people who are offered a draw are really afraid that if they don't accept it, they loose that game, this is one psychological trick, another trick is with face emotions, specially in classical games, looking confident after making a blunder makes your opponent think it is not a blunder (if the blunder isn't obvious), some players fake emotions too, in a drawish position, they sigh and give negative face expressions, indicating that they think they are lost, giving their opponent some relief, and some people stop thinking when relieved... in short, yes there is luck in chess if you take advantage of someone's human nature while playing the game, i.e. the psychological battle
@MasterOfSkillIssues said in #5:
> for example, people who are offered a draw are really afraid that if they don't accept it, they loose that game
I no longer feel like that. Instead, the natural question when being offered a draw should be: "Does he/she have a reason to be worried about something?"
Also, at some point I noticed that surprisingly often the move preceding a draw offer turned out to be a mistake or at least an inaccuracy. My theory is that when someone decides to offer a draw (sincerely, not e.g. as a form of trolling in a lost position) and still has to play a move first, they subconsciously expect the game to end soon so that they do not focus at the move as well as if they expected the game to go on. IMHO it's a similar effect as with the (in)famous "move 40 blunders".
> for example, people who are offered a draw are really afraid that if they don't accept it, they loose that game
I no longer feel like that. Instead, the natural question when being offered a draw should be: "Does he/she have a reason to be worried about something?"
Also, at some point I noticed that surprisingly often the move preceding a draw offer turned out to be a mistake or at least an inaccuracy. My theory is that when someone decides to offer a draw (sincerely, not e.g. as a form of trolling in a lost position) and still has to play a move first, they subconsciously expect the game to end soon so that they do not focus at the move as well as if they expected the game to go on. IMHO it's a similar effect as with the (in)famous "move 40 blunders".
The blog post points out various sorts of luck influencing results in order to give people more trust in their training plans and play in general. Jagged graphs are the norm.
I'd say in chess played by humans there is luck because of our limited horizon. We will eventually have to move without knowing if our move is good or bad. Maybe we can enumerate the respective reasons for our candidate moves and still not say which one is better, or probably better.
I played poker as a little kid with Grandma and Dad (for matches not for pennies). Would you poker players among us say that there is some luck in poker in the same way I just described it for chess- that it is impossible at times to make 'the right' decision, because the indicators within our limits look equally strong for several continuations and we will have to guess what's best? I mean independent of the luck that the random factor of shuffling brings in, i.e. independent of the incomplete information?
I'd say in chess played by humans there is luck because of our limited horizon. We will eventually have to move without knowing if our move is good or bad. Maybe we can enumerate the respective reasons for our candidate moves and still not say which one is better, or probably better.
I played poker as a little kid with Grandma and Dad (for matches not for pennies). Would you poker players among us say that there is some luck in poker in the same way I just described it for chess- that it is impossible at times to make 'the right' decision, because the indicators within our limits look equally strong for several continuations and we will have to guess what's best? I mean independent of the luck that the random factor of shuffling brings in, i.e. independent of the incomplete information?
Something to add is that Glicko-2, the rating system that Lichess uses, assumes that a player's rating is a random variable with normal distribution after they finish a rating period (that in Lichess consists of a single game). So a player's rating is a random variable from the very construction of the rating system.
As for me, I don't believe in Luck, but when someone wishes me luck or gives me luck, I feel lucky. :p
я тоже за этим "феноменом" наблюдаю давно и вот к каким выводам я пришел: мозг думает-функционирует циклично или еще что на мыслительную способность мозга могут влиять много разных факторов допустим настроение, обстоятельства и.т.д. но что бы ни было играть надо в шахматы всегда)