INflated ratings?!

Is it real? Are ratings in a surge of inflation right now? I mean, I was in the 1300's a week ago, and now I'm over 1900 in bullet. I don't think I've improved that much. Others have risen up with me as well. And my opponents don't feel as good as their ratings. My question is did all the higher players have some sort of mass exodus? Why does 1900 feel like 1500 a week ago?

Are you in college? Students are experiencing inflation far beyond the average.

I don't get that joke at all, sorry. but yes I'm serious about the bullet ratings inflating

hmmm well i've recently gone from 1700 -1900 blitz .... so, maybe there has been a tad of inflation?

College tuition has increased far beyond the average (monetary) inflation rate.

I suppose it could also be double entendre if you really wanted to get crazy with it.

For me is the same, there is for sure a big inflations in the ratings!

Could somebody of the staff try to explain this?

The ratings were deflated so we made a small adjustement to them.

I am the only one who doesn't see that :(

What was the adjustment?

Looks like it is a live chess issue. The original poster mentions bullet and there is this topic in the Live Chess forum.

http://www.chess.com/forum/view/livechess/rating-system4

I've noticed on several occasions that a 2700+ player can gain rating points by repeatedly beating a 2200 player, while the 2200 player's rating does not change.

Is this the "tweak" that the staff is referring to?

I think the rating system is not entirely accurate, Carlsen had the highest rating in the world and came in 3rd place in a tournement. Your rating does not reflect your true strength.

A 2500 player could get that rating multiple ways.

Fight it out with other people around 2500 and roughly win 1/2 the matches. This is the  hard way.

Fight it out with people lower rated like 2000-2300 and dominate them most of the time. This is the easy way, and the person never has to beat a person as strong as them.

Take the 2500 that did it by beating many people under him, and pit him against the 2500 who focused on people as strong as himself, and odds are the guy who was fighting people of even strength will have the advantage even though they are rated the same.

To be more accurate with Chess ratings I think players need not ONE but TWO ratings.

Your rating + Your average opponents Rating.

A 2500 with an average opponent of 2500 is better than...

A 2500 with an average opponenet of 2250.

And the ratings should be 8 digits not 4 so it reads 2500/2500 and 2500/2250

I could become the best boxer in the world if I fought only 9 year old kids, in boxing it is called a "paper champion" if you are only fighting weak opposition then your reputation will be inflated higher than it should be, and when you get in the ring with someone who can throw a punch you get your teeth punched down your throat.

In my personal opinion 4 little numbers is not enough to gauge a persons strenth at chess accurately enough and it should be more complicated.

 I agree. My rating is 1274 and my opponent's average is 1291, so I am about equal. There should be some sort of formula so that a player like me can't be matched (either in a tournament or TM) with a player who's rating is 1274 and whose OA is 1600 or something.

So according to that logic the #1 rated player should never lose a game.  Anand is currently rated 2803 to carlsen's 2801 so he should win every game because his rating is higher?  Chess ratings are based off of very large amount of data and confidence intervals to estimate ones strength relative to another.  You can't come to any kind of conclusion about a player's "true strength" based on one tournament.

Thanks kohai, I know you answered my on live chess, but I made this topic before that. So I really was listening to you

I have noticed as well that the Live Ches ratings have gone up recently (mine did also, a little) and it's been time because they were deflated for quite a while.

I hope that we can avoid inflation now.

Ratings reflect performance. They are not intended to predict winners and losers with certainty. If Carlsen finishes third in a tournament, it only means his performance in that tournament was not equal to all of his performances measured to that point. Or possibly that the other players played better than their ratings would indicate. A few hundred rating points won't intimidate a grandmaster.

If you are interested in how ratings work, wikipedia has a pretty good explanation. http://en.wikipedia.org/wiki/Elo_rating_system

There is inflation because

1. Computers in live are too easy

2. More players have come to give away theit points from 1200