Chess rating

Hi fellas! I was wondering about how much stronger do you think is a 2500 elo rated player than a 1000 elo rated one.

Please don't tell me 2,5 times :). I mean, the difference between 2200 and 2000 rated players is way higher than the difference between 1500 and 1000 rated players? Is it kind of exponential?

Oh and what's an estimate ratio between chess.com rating and elo one ?

Yeah, it's exponential. It's the delta rating that determines the winning expectancy. A 2500 vs. a 2000 would have the same winning expectancy as a 1500 vs. a 1000.

And a 1000 rated would expect to win about 1 in every 5,624 games from a 2500 rated.

Wow ! Thanks. So... it's completely useless to play against a 2000-2500 rated player to learn if I'm rated around 1000...?

It's not useless if you can get him to do a post-mortem analysis afterward, so you can find out what your biggest mistakes were.

Ok, but against a computer and trying to analyse it by myself?

The more you get away from a stonger human analyzing with you, the more in the dark you're going to be. An engine can point out the tactical mistakes, but it can't explain anything else. You'd have to figure the other stuff out by yourself by reading chess books, watching chess videos, etc.