chess.com ratings are deflated against USCF

Yes, chess.com has a good "toughness" to their live players.  I'd say overall not as strong a feel as ICC, but compared to just about any other site, I think chess.com has a stronger average live chess pool.

e.g. their 1300s are stronger than other's 1300s, their 2000s are stronger than other's 2000s, etc.

But also just the quality of play... hard to explain.  Regardless of rating there is a certain quality of openings, tactics, and technique.

I think there are a lot of people in Bangladesh and India and Argentina, etc, who see no reason to spend money at ICC. I played some guy in Syria the other day, apparently they still have internet in the middle of civil war.

This thread, top of page 4, basically says it all, "What was the highest rated player you beat?" Some guy responded that he beat a 2200 USCF, and a 2000 at chess.com. In both cases, the opponent was a national master. The ratings are basically equivalent.

yah, that's why i focused upon people who were actively doing both. lots of blitz players have come here laughing at the idea that they could 'gain' 2-300 points otb. well, that's not really exactly what i said, which is why i keep chiding people about reading comprehension. slower time controls require more endurance and discipline, and if you don't have that...

my interest was in the difference between players who maintained active ratings at both time controls, what is the average difference? this information isn't terribly helpful to a blitz player at chess.com who wants to estimate his USCF rating. its more geared at USCF players who want to estimate their probable blitz rating, as it takes much less effort to get up to speed on blitz, than it does to transition into slower chess.

Kinda noticing a pattern:

1) Adam Rinkleff expresses his point.

2) Someone attacks him.

3) He attacks back.

4) The same person attacks back or someone else joins in.

5) Eventually it stops when a mod comes in.

6) A new person to this thread joins in with questions/accuses Adam of something.

7) Adam attacks with his point, tells him/her to read the past pages, and discourages further arguing.

8) People are too lazy to (It's 21 pages) and sort of agree with Adam and express their supporting points.

9) A non insulting information and point exchange is made.

10) The cycle starts again.

LOL.

First of all I do play in uscf tournaments and just played in a g/90 30second increment uscf/fide dual rated tournament just this weekend.  It makes no sense for you to tell me that I should be a national master just because I have a 2150+ blitz when i'm constantly low 1900s uscf.  And then you say I need to maintain the endurance during longer time controls.  What does that even mean lol?  I almost always use all my time or nearly all my time in long games if you are suggesting I am not.  The fact is blitz and otb long games are different.  Some people are just bad at blitz.

oh yeah I just looked at that game.  A 1700-1900 definitely plays qe6 in the scandanavian and makes all those horrible blunders.  You are quite right. 

LOL

I don't no why it would surprise you.  The high use of engines in online blitz, would by itself lower the average online rating.  Coupled with the fact that chess.com rating is structured so that the average rating is 1200.  A USCF tournament player with a rating of 1200 or less is generally not going to be attracted to blitz play.

I will point out how chess.com's aggressive attack on cheaters contribute to a deflation of rating points in the chess.com's rating system; and I mean deflation in its actual meaning and not the misused meaning often used in this thread.  The cheater enters the rating pool and starts taking rating points from the system until he obtains a high enough rating that he shows up on the staff radar and his account is  closed.  When his account is closed, he is no longer in the pool, but the points he stole are not restored to the rating points pool and are lost.  He then opens another account and the process of the system losing rating points is repeated perpetually.

I rarely encounter cheaters online, I don't recall ever playing someone I thought was using a computer on chess.com.

Starting at 1200 means very little because in the beginning you affect your opponent's rating little while yours can swing hundreds of points catching you up to where you should be.

Your next point about banned cheaters makes sense though.

Bovine excrement! Us rather lower rated players, who tend to have poor short term memory and limited concentration span prefer playing blitz.

Besides, the average rating is not 1200. That is a starting point, and the median ends up a bit lower, for the obvious reason that it is possible to be more than twice average rating, say 2600, but losing the same number of points puts you in negative territory.

The interesting thing about this whole thread is that it has ignored the available data in favour of a fixation on the idea of USCF - 250. Chess.com shows the average USA member rating at about 1275. Running a quick weighted average over the top 20 countries by membership (covering 80% of members according to chess.com)  gave me an average rating figure of 1304. On the plus side, that says that the average American is a bit below the world average, but so is the rest of the  English speaking world.

When you look at the graph supplied by Chess.com, it claims an average well under that, which is intuitively likely, for the reason mentioned above. In fact it was so low that only 4 small nations have a lower average rating. But that doesn't make sense

astronomer999 wrote:

"Besides, the average rating is not 1200. That is a starting point, and the median ends up a bit lower, for the obvious reason that it is possible to be more than twice average rating, say 2600, but losing the same number of points puts you in negative territory"

This statement makes it clear that you have no idea what you are talking about.     A pure ELO rating system is a closed system Which means that points can neither be added or subtracted from the system.  Neither chess.com nor USCF is a pure ELO system but the exceptions to the zero sum gain for the system don't apply until you have a rating above 2200.  Since points cannot be added or subtracted from the sytem it is clear that average will equal the starting point.  It is true that if you take two players starting at 1200 points and they only play each other, it is impossible for one to obtain a rating of 2600 without one obtaining a negative score.  However if you have 3 players in the pool it is possible for one of the players to obtain 2600 points by getting rating points from the other 2 players (3x1200=3600 which is greater 2600 points.)  The math doesn't go beyond the grade school level.  

Theoretically the rating pool includes all account ever created including accounts that have been closed or are no longer active.  The charts and statistics you refer to, only include currently active accounts.  Also when I use the word average I am refering to the mean not the median.

Would that this were true.  Unfortunately there is software widely advertised, that  cuts out the extra keystrokes needed to cheat and allows the cheater to move just as fast as a non-cheater (perhaps faster since the cheater doesn't have to think at all, but merely press the enter button).

"If you are going to venture your opinions in areas you have no understanding, an  ad hominem attack does not disguise your ignorance, but rather makes it stick out like a sore thumb."

Meeeiow. 

"A pure ELO rating system is a closed system Which means that points can neither be added or subtracted from the system."

Yes, but chess.com does not use ELO.  We have things like RD which can cause the average rating to deviate from the starting point.  If I play a brand new player and lose, the new player's rating will increase by a large amount and my rating will decrease by a somewhat smaller amount.

"Neither chess.com nor USCF is a pure ELO system but the exceptions to the zero sum gain for the system don't apply until you have a rating above 2200. "

Untrue.  RD affects all ratings ranges.  

Also, unless they've recently changed it, the starting point of a USCF rating is not any fixed point, but is based on your opponents.  If you only ever play one USCF game, it's against a 1743 opponent, and you draw, your rating will be 1743.

If you have only played one rated game in the USCF, you are still unrated.  You have to play at least 8 rated games before you receive a provisional rating.  

no.

Actually you need 20 something...I'm a USCF member myself, and i played over 20 postal games to get mine and played in many OTB games to earn my rating

You currently need 25 games (it used to be 20 games) to get an established rating, from 8-24 games you have a provisional rating OTB.  USCF postal ratings are a different system entirely.

That is correct Blake :)