Chess rating seems to be exponential

I've played chess online during two separate periods, a few years ago on ChessCube and currently been playing on here for a few months. Both times I've gotten to the mid 1600s. When I quit playing on ChessCube I was fluctuating between 1550 and 1650 in 10 minute games. I'm starting to notice the same pattern developing here.

It just feels like the skill level of a 1700 player is dramatically better than 1500. Whereas, I can barely tell the difference between 1300 and 1500. It makes me think that a chess rating is somewhat exponential, kind of like the Richter Scale. As a hypothetical example, one rating point at 2000 might be worth 10x as much as a point at 1000 level. I think the ratings graph seems to support this.

Has anyone else noticed this dramatic increase in skill level at a certain point, whether its around 1700 like me or at some other level? Or are there natural plateaus where all of a sudden you just find a different class of player?

Now that I'm actually studying the game, which I had never done before, I figure my rating will go up, so I'm just curious what's around the next corner.

I think this is the biggest reason you feel like a 100-200 point difference is dramatically better.  Your strengths are propping you up vs the players you face now, but when you play someone a bit better the role the holes  in your game play dominates everything else.

Mathematically speaking, a 1300 has just as much chance to win vs a 1500 as a 1600 does vs a 1800 etc.

But there is something to what you're saying, and I think it's true for almost any skill in life.  In the beginning it's very easy to make progress, and as you go further the more you have to work to see improvement.  To go up a class (200 points) you do have to know many more patterns than it took you to advance from your previous class section.  There was some study that did show GMs have 10x the patterns IMs do or maybe it was masters, I don't remember exactly.

The bell shaped cruve you show is only loosely rated to the fact that it gets harder to improve as you get better, it's not showing it directly.

Right on. Thanks for your take on it. I'd pretty much agree with everything you said. It makes sense that just being a good game player in general will only take you so far and then you'll hit a wall, and have to put in some work to get through it.

You hit it perfectly when you talk about my strengths getting me by because less skilled opponents are not so keen on exploiting weaknesses. Essentially I probably have gotten away with too much sloppy play and formed bad habits because of it. Time to start plugging some holes I think.

Like I said in a different thread. Getting from 1200 to 1500 is not that hard, but getting from 1500 to 1800 is terribly hard. I thought I hit my limits after I got 1500 long long time ago. I was going in circles, not being able to pass over 1600 for longer than one game (I also stopped playing on chess.com for some time to change environement).

After some time I thought that maybe my problem is with the basics. I started watching ChessNetwork on youtube, read several books about the openings and did more tactics than before (although I'm still really poor when I'm pressed by time). I also started playing in real life, opposed to playing only on the Internet.

After I returned to playing on chess.com I've stumbled a couple of times but then I started rapidly advancing and now I'm a stable 1800 player. I even have to purposefully play badly against my real life friend, so that he won't just give up on playing with me. xD

Statistically, a 200 pts difference translates into a 75% points expectancy. So you could say that a 1700 player, that would score on average 7,5/10 against a 1500 player, is three times stronger (as the 1500 player would score 2,5...). Here's where the exponential feeling may come from

I think we called that geometric progression when I was at school.

Arithmetic series is plus x

Geometric is times x

Exponential is to the power x

I think you're right about playing over the board more often. I hadn't done it in years. Now I'm helping teach at a middle school chess club, and I struggle to see the whole board as I do in a 2d computer board. 

We had a 3 minute ladder tournament and the guy who runs it put me at the top of the ladder. The onslaught of 11 year olds was pretty intense but I managed to win every game. I never considered how important hand-eye coordination is in speed chess, so that you're not knocking half the pieces over in the process.

Other than the kids I don't really have anyone to play against. I was thinking it would be a good idea to play on a board while playing a 10 minute online game every once in a while so it doesn't feel so foreign.

Ooh. Data. I like data. Thank you.

I don't think comparatively the rating system is exponential. However, I do very much believe, as most people probably do, that it gets exponentially more difficult to improve your rating the higher up you go (in a manner of speaking).

"There was some study that did show GMs have 10x the patterns IMs do or maybe it was masters, I don't remember exactly."

Wow. Very striking indeed!

You're right

OTB, this is still a crushing feeling