Skill Conformity Hypothesis

Why does one's amount of experience not to correlate with his exact level of skill?
Why are some less experienced players much more skilled than seasoned players?
Why do different geographical regions seem to have different average levels of skill?

Considering these questions and other observations, it seems to me that one's skill level is strictly determined by the skill of the opponents he plays.
This may be obvious to many, but despite that, the typical responses to "How do I get good at chess?" are puzzles, books, and studying master games.
My hypothesis is that, while some theoretical knowledge can kick-start learning, the only way to really improve at chess is to exploit this correlation between a player's skill level and his average opponent's skill level; in other words, by almost exclusively playing opponents that are much stronger than oneself.
There is most likely a sweet-spot in terms of how much stronger one's opponents should be. For example, if someone with no experience played a Grandmaster thousands of times, he would likely become better extremely slowly, if at all. However, if a 1500-rated player played thousands of games with 2000-rated players, it seems very likely that he would eventually reach the skill level of his 2000-rated opponents (though he would not significantly surpass it).

Despite study, clubs (consisting of players around or below my level), and a few years of monthly OTB tournaments competing in reserve sections, my skill level plateaued a few years ago, and I eventually became complacent with my skill level and lost interest in getting good at this game. But now, using my obnoxiously titled hypothesis, I'm going to give it another shot.
During this experiment or whatever you want to call it, I will almost exclusively play much stronger opponents. I won't do any external studying--e.g., books, coaching, etc.--aside from reviewing my games and some tactics puzzles for fun.
I don't expect to win any games for months, but I think if I stick it out I'll finally make some real progress at this game. I may update every few months, mostly just for personal record or for the off chance that anyone's interested.
Thanks in advance to any strong players who are willing to play a noob like me.

Current Lichess rapid rating: 1600ish
3/2/2021

Hey Top. Interesting post and theory here. I actually agree with your hypothesis, and I would even say it helps explain why many long-time players plateau from 1700-2000 OTB, say. Fact is, the better you get, the more you don't wanna get crushed anymore! So many players who once sought out the toughest opponents they could find, eventually avoid the tougher players. These days when I'm feeling like that, I just take a break. I'm a net-losing player by a significant margin as you can see by my win/loss history, and I really think this is a decent brute-force attempt to improve, yeah. just play the toughest players you can

Awfully wordy way of saying something that seems fairly obvious.

Hey Dan, I fully agree. We become obsessed with winning and scared of losing, which stunts our progress pretty badly. Playing tougher opponents and losing most of your games like you do is definitely the way to go.

I agree that I'm explaining something that may be obvious to a master, but for me a few years ago and for people around my skill level, the actual importance of playing stronger opponents is underestimated, while puzzles, books, and playing those of equal strength are seen as the only way to get better. To many it seems pointless to lose over and over again, and others think they learn more by winning against equal players rather than losing against stronger players.

It is a commonplace that one should play stronger players in order to improve. That was as true when I was a beginner as it is now.

The answer is almost purely Neurological science rather than anything like "afraid of losing" unfortunately. If it were that easy we would all never plateau.

Plenty has been written on this subject in much more reputable sources so, as dismissive as it sounds, I do just have to recommend a google search to get to the real science behind it.

With that being said it basically comes down to 2 types plateaus:

1. Temporary plateaus. These are just natural and will end eventually as long as you are training correctly (which most players don't do unfortunately). This is where the playing stronger players will help as it is one of the (several) methods for improving your chess and may power you through a plateau.

2. Permanent peak. Can simply be determined by a lot of factors that can't be changed (natural intelligence, spatial reasoning ability etc., how early/late you started). I'm not about to come out and say intelligence isn't fluid but because chess is not just concrete knowledge based it is more similar to learning a language than it is getting a mathematics degree or something.

An important point is that your knowledge of the game can definitely increase with experience (and it almost always does) but that doesn't always translate into practical play (where 70% of non-master games are just decided by tactical blunders which isn't helped by knowledge).

I know a lot of older players from my area who are lower rated than I am but they have been playing chess consistently for a long time and have played 1000+ OTB tournaments and read dozens of books, so even though I am stronger than they are some of them have fantastic knowledge of the game and I still learn lessons from them today. My point is actual results are not purely tied to (though they are helped) by chess knowledge, which is exactly why it isn't just tied to experience.

"You may learn much more from a game you lose than from a game you win. You will have to lose hundreds of games before becoming a good player." - Capablanca