Which rating can be reach without learning ?

kinda depends on your chess talent.

These ratings mentioned here...
Are these lichess ratings?

I ask because chess.com is about 400 lower than lichess. Lichess 1200 = 800 chess.com

Which is closer to the 'proper' ratings?

I think about Lichess ratings.

I didn't do any training at all until it seemed to plateau for me around 1700 Lichess blitz. Then tactics puzzles plus a few videos about the openings I had trouble against helped pop me up a couple hundred points.
Now I'm back on a plateau because I don't want to study real chess, just waste my time playing mindless bullet (the crack cocaine of chess). Hit the puzzles dude. You'll learn the tactics quickly and then start linking them together into a full game.

@hampy

Lichess is NOT strictly 400 points difference. I am personally between 150-200 lower. And a lot of my friends are about equal. Some have a higher chess.com rating than here. I think you're the first I have heard such a large gap. I wonder if it's low rated?

@MeWantCookieMobile

I'm sure I read somewhere that lichess sets the middle of the curve on the bell graph at 1500.

My Blitz rating of 1220 puts me better that 25% of the user base.
1500 rating puts you as better than exactly 50%

I think chess.com sets the 50% mark at 1200.

So the differences is 300 not 400 like I said. And it's also dependent on the ability of the player pool.

My 917 chess.com Blitz rating puts me at 43%...
So worse players there...

But I'm wondering which is closer to the proper rating you'd get playing real over the board chess.

@hampy you're looking at ratings wrong. A rating is how YOU as an individual does against the pool as an individual pool. You can't logically claim that because one average is consistent with your rating gap that it's going to be 400 difference across the board. And claiming there is worse players in a pool based on rating is a misnomer.

None is necessarily closer to over the board. If they seem to correlate it's because of coincidence and luck. Also it's bad juju to look at blitz vs uscf main. No matter what anyone tells you they don't correlate. You play differently blitz. Everyone does. So comparing blitz with uscf or fide or local Fed is bogus.

If you are interested though my USCF long play is about 2030 right now. I don't really play that much at all on chess.com but when I do I average around 2000. What does your stats say there?

@MeWantCookieMobile thanks for the feedback - I'm starting to understand it a little better.

Your ratings are around 2000 on lichess, and if you're around 2000 over the board that answers my question.

Anyway I'm abandoning chess.com, I just like seeing the country flags because I'm shallow.
lichess is much better even though there are a few bugs.

Anyway I've taken advice and am playing some slower (correspondence) games instead of blitz/bullet to try to improve and find out where I really stand!

@hampy that's great. But the best is to play 30-60 min with analysis. Preferably with humans. You tend to learn more.