Anyone have any evidence doing puzzles helps your rating?

Yes, it improves pattern recognition and tactical awareness, and make your moves more precise.

At least in my experience, if i don't solve puzzles for some time, my game decreases in quality considerably.

But, is not only solve the puzzles that get you the benefit. You must understand what is happening there. If you go wrong on the puzzle, analyse it to see the correct responses, and even if you go right but have any doubts, put it to analysis too.

However, the rating gain, when it comes, will be some 20 to 50, at most 100 points, and may be gradual, not immediate. It can of course be more with time, but to expect too much gain can be unrealistic.

I think the "evidence" for puzzles improving chess skill is anecdotal, maybe coincedental at best. If you play long enough you'll do a bunch of puzzles, but you'll also learn some openings (which I think helps a lot in winning games by getting a better position early on in the game, particularly in the early stages of chess development) and endgame principles from reading a book.

Another huge factor people forget to take into account is your natural talent. Talent is the major dictator in how quick you pick something up and how good you're ultimately going to be. Puzzles won't help with that.

Something else to consider, when doing a puzzle you know that there's a winning move or tactic. It's just a matter of working out what the theme of the puzzle is. In a real game, you don't have that luxury. So when you succeed at a puzzle you've kind of 'won' a section of a game and that will give you a dopamine hit, and because you feel satisfied you'll trick yourself into thinking you've gotten better and that puzzles are the way to improve.

One reason I think puzzles are overrated is because of how much the idea that they help with development is pushed on by chess coaches/personalities. Ultimately, these people gain something (fame, money, whatever) if more people play chess and want to be good at it. Telling people "hey, here's an easy way to get really good at chess" is going to be more likely to attract and retain new starters (and thus increase potential financial gains) than saying "to be good at chess comes down to years of study and natural talent, and if you don't have it then forget it". It also makes for more appealing book and video titles via "clickbait" (which again, nets more profit for the people who push puzzles as the main way to improve).

The value in puzzles is in how you do them. It gets you used to searching for tactics and quickly assessing a position. If I get a puzzle wrong, I check it on the analysis board afterwards and figure out why my move didn't work.

That's how they help teach you things. And once you get good skills, you can get faster at spotting them in your game positions rather than the selection the website draws on from other people's games.

The trick is to treat every puzzle as a real situation in a long time-control game. First identify the tactics THEN move. Welding those patterns into memory.

Yes! It blew my mind when I realized the positions in puzzles are taken from games played on this website where the analysis engine marks a great or brilliant move.

Saying puzzles don't help you learn when you don't do them right is like saying books don't help you learn--after all, you ripped the pages out, shredded them, and used them as confetti. You can get through a dozen books a day at that rate! I've done so many books and I'm not getting better, must mean books are useless!

No, but they're fun.

Puzzles demand precise calculation and exact moves, improving your overall accuracy and reducing blunders in your games.