Question about FIDE ratings

I'm a little curious about rapid ratings...
So I know this could be quite a pointless discussion, but I can't help but be curious about the difference between FIDE ratings, chess.com ratings and lichess ratings... (as in for rapid)
I have my own perspective on this, feel free to disagree.

I think for chess.com, a 1500-1600s rating would be a 1100s rating in FIDE, but it could differ from person to person.

Anything below that, like a 1400-1500s rating would probably either be a rating below 1000 (unrated) or a rating between 1000 and 1100.

In my opinion, chess.com ratings are inflated by about 400-500 points for lower rated players, and probably 200-300 for higher rated players..

Now for lichess, here's what I think.

Anything below 1600 is pretty much unrated (<1000)
1700s are probably unrated too...same goes for 1800s, but it could be 1000-1100.
1900s is probably somewhere in 1000-1300.
2000s probably 1100-1300, 2100s could potentially be a 1300-1500s, and so on. 
In my opinion, lichess ratings are inflated by like 700-800 points for lower rated players, and 500-600 points for higher rated players. 

But this is simply based on my knowledge, if anyone else has anymore real life examples or more knowledge, I'd love for them to share!

There isn't any conversion formula between FIDE ratings, chess.com ratings and lichess ratings.

How could there be? Chess ratings are a distilled summary of your results against a SPECIFIC pool of chess players. FIDE, chess.com and lichess have three DIFFERENT pools of chess players. 

We can make very general comments, such as "Chess.com ratings are higher (and worth less) than FIDE ratings" or "lichess ratings are even higher (and worth even less)"... both of which I agree with... but that's about as far as we can go in comparing them.

I completely agree with you!
But I also do have to say, lichess and chess.com ratings are definitely overinflated. I'd say it if we were to literally compare them, it would look something like this...
Lichess ratings > Chess.com ratings > FIDE ratings
For a person with all three ratings, playing frequently at their normal level.

See my edit above.

@1

"In my opinion, chess.com ratings are inflated by about 400-500 points for lower rated players, and probably 200-300 for higher rated players."
++ This makes no sense.
The probability of player A beating player B should be the same.
The probability of player A beating player B is given by the rating difference.
Thus
Chess.com rating A - Chess.com rating B = FIDE rating A - FIDE rating B
Thus
Chess.com rating A - FIDE rating A = Chess.com rating B - FIDE rating B

Sure, if we compare the difference from the same platform, the subtraction result will be same, and same also for the probability.

If we calculate the difference between the same person's FIDE, chess.com, and Lichess ratings, we can find the amount of inflation for each platform.

Then, calculating more person will give more accurate result.

Yeah, I get what you're saying, it really doesn't make sense if you look at it... But that's cuz after consideration, I honestly believe a 2700s chess.com player would be like a 2500s FIDE, instead of 2200s. 
But for lower rated players the gap just seems bigger...Yeah I'm not making much sense, but I do hope you understand xD

@8

Maybe you base your difference a lower levels on not yet established rating.
I estimate chess.com rating - FIDE rating = 100

I can't help but disagree with that... 
A 1200s player on chess.com is nowhere near 1100s FIDE. They could get super lucky (for instance, play in a competition where they face two 2000s FIDE and one 1000s FIDE, and they draw against the 1000s. This would get them a good rating even though they only scored 0.5/3, that half point originating from a low rated player). But then they would instantly loose it all in their next tournaments, assuming they play at their level and nothing strange happens (like some 2000s suddenly forgot about time and lost on time)
3 years ago when I was 9, I played many tournaments. I was 1300s in rapid on chess.com. Yet I was nowhere near a 1200s rating. I did draw a 1200s, and beat a few 1100s, but in the end, I only managed a rating of 1013. Even then I was super happy, but proceeded to loose all of it in my following tournament, making my rating drop back to 0. 
So I don't think it's only 100. I know a few people for reference.
I've got a friend who's rating is around 1200s. Their chess.com is 1600s.
Another friend of mine has a rating of 1100s. Their chess.com is 1400s.
Last reference has a rating of 1300, but their chess.com is 1800s.
I really don't think it is only a 100 gap. I can't help but completely disagree on that. But either way, have a good day!
Peace out.

According to your example,
(FIDE) --> (chess.com)
1100 --> 1400
1200 --> 1600
1300 --> 1800
(+100) --> (+200)
Quite interesting though

Yeah, I don't think there is an exact value of difference when converting the ratings, but there is one thing that I am absolutely certain about. Chess.com ratings are for sure inflated, by quite a lot.

If I had to give an exact value, I would say by 400 points. Lichess is probably by 600-700...
If you have a rating of 1400s in chess.com, it's likely that you'd stay unrated. (But a chance that you get a small amount of rating as well). 1500s is decent-ish, and probably gives you like a 1100s or a 1200s. Yeah. My first friend almost has 1500s.

This makes no sense.  You are trying to compare numbers that are completely meaningless to compare to draw a conclusion that 2 are "inflated" in comparison to the other.  Suppose instead of numbers, chess.com used fruits, lichess used vegetables, and FIDE used meats as their ratings.  You might be a banana, spinach, and a NY strip.  How would you compare them?  You wouldn't, because they are not meant to be compared.  Rating measures your performance against the pool using the settings of the initial algorithm.  The only thing you can take between them is the probability of winning given 2 ratings from the same pool (since they are all using 100 as the standard deviation).  Thus, a 1700 has the same probability of winning against a 1600 regardless of what pool you are looking at, but you have no idea what a 1700 on chess.com would do against a 1600 FIDE.

I agree with you if we look in that perspective, but I think they were trying to say that lichess and chess.com ratings are just higher than fide. I completely agree. I was once a 1300s player on chess.com, but at that time I was nowhere even near 1300s fide. 
My final conclusion:
Chess.com & lichess ratings tend to be higher than the player's fide rating, but there is no exact way to convert them (only perhaps estimations), and they are 3 seperate things.

But the only reason people are thinking they are "higher" is because they are given a "value" that looks like a number.  That number only means something within the system it exists.  Comparing it to numbers in a different system means nothing.  It is like comparing your bank account to my 100m dash times.

To break into a separate example:  years ago, I used to play a DotA clone (HoN).  They used an MMR system that was basically a Glicko algorithm with the mean set to 1500 and a standard deviation of 50.  Comparing the rating of 1800 in HoN to a Chess Rating of 1800 would obviously be completely meaningless - not just because they are completely separate games, but because they are completely separate player pools with different configurations for the algorithms.

Chess.com used to use the Glicko2 system with the mean set to 1200 and a standard deviation of 100.  LiChess uses the Glicko1 system with the mean set to 1500 and a standard deviation of 100.  FIDE uses the ELO system with an expected mean of 1200 (IIRC) and a standard deviation of 100.  As you can see, each rating pool uses a slightly different system (the Glicko systems are based on the Elo system, but they try to accelerate to the appropriate rating faster by adjusting the k-factor, called the rating deviation).  With each system expected to be a normal curve, you would expect LiChess to be "higher" than the other 2 because its mean is higher (meaning the curve shifts ~300 points to the right).  Additionally, since the Glicko systems try to accelerate to the appropriate rating, they will move faster than the Elo system (which expects players to make slow changes).

Mhm this makes more sense now. Thanks for clarifying!

I'm on average 1550 FIDE at the club and 1700 on this site in rapid. I know a 1250 who is 1100 here. So the 300 point difference OTB rises to 600 here. Otherwise, check chessgoals for more precise comparisons, you'll notice USCF and this site correlate well.

I am 1675 here and 1607 FIDE.

sure i dont l know about for FIDE

xxxx