Your time in between games doesn't affect the rating change.
Chess.com rating fluctuation.
Sort:
It's the Glicko system, which gives your rating a reliability factor measurement based on how many recent games you've played -- there's a built in assumption that if you don't have a lot of recent data points then your rating is less reliable than if you do, so a new game after a window of inactivity will have a larger impact -- up or down -- than if you'd been playing games throughout that period.
RyanMK wrote:
Your time in between games doesn't affect the rating change.
Actually, it does. The longer you wait between games, the more your rating will fluctuate. For more about how the ratings work here, see this page: http://www.chess.com/article/view/chess-ratings---how-they-work
TheGrobe wrote:
It's the Glicko system, which gives your rating a reliability factor measurement based on how many recent games you've played -- there's a built in assumption that if you don't have a lot of recent data points then your rating is less reliable than if you do, so a new game after a window of inactivity will have a larger impact -- up or down -- than if you'd been playing games throughout that period.
That's fine, but for a week of inactivity to effect the rating fluctuation to that extent is insane.
I stand corrected. My apologies for the false information. I was under impression that only the total # of games you have played affected the Glicko, not the density of those games as well.
You really think 22 rating points is a severe shift? That's less than a 2% shift.
ichabod801 wrote:
You really think 22 rating points is a severe shift? That's less than a 2% shift.
That is a somewhat still a large drop....take away 2% from a rated player of 2500 and see if he likes the drop
ichabod801 wrote:
You really think 22 rating points is a severe shift? That's less than a 2% shift.
I lost 2% of my entire rating against someone almost 100 points higher than myself.
Yes I think it is a severe shift.
omgCHECKMATE wrote:
ichabod801 wrote:
You really think 22 rating points is a severe shift? That's less than a 2% shift.
That is a somewhat still a large drop....take away 2% from a rated player of 2500 and see if he likes the drop
Agreed. That would end up with the person having a rating of 2450... a 50 point drop!
I think you're making a mountain out of a mole hill. If you couldn't get those points back just as easily, you might have an issue. Also note that your rating deviation depends on your opponent's RD as well as yours. So if he has a high RD, that 100 point difference means less.
I have since regained all of my rating points.
However I could have just as easilly played a few bad games in a row (against players ABOVE my level) and lost up to 70 points (only three games), that's a huge rating drop.
Well bear in mind that as you played those games you would have anchored your rating with a smaller RD value, so you wouldn't have lost nearly as many points on your third game as you did on the first.
I think the major beef here is not rating point drop but the disconnect penalty. I don't want to lose anything when it's beyond my control. Chess is one human mind against another. If I lose to an opponent, so be it. If I lose to a bum server, I'm upset.
My rating dropped substantially (like several hundred points) awhile back due to the fact that I had a long series of games where whenever I achieved a winning position, I was disconnected, while I didn't seem to get disconnected almost ever when I had a losing position (that's the way it goes sometimes...). My solution? I decided to quit playing live chess until Live Chess 2.0 (the more stable version that is in the works) comes out.
They say it's coming, guys:)
a 22-point rating drop would be huge in OTB chess, but is really comparatively small on this website, especially on live chess, where people of similar strengths often have huge rating differences due to things like disconnects and the aforementioned frequency of play.
The only thing that can reasonably be done about that, though, is to implement a adjournment and adjudication system. Not penalizing the disconnecting player will open the door for rampant abuse whereby players would disconnect from losing positions rather than playing them out or resigning in order to preserve their ratings.
TheGrobe wrote:
Well bear in mind that as you played those games you would have anchored your rating with a smaller RD value, so you wouldn't have lost nearly as many points on your third game as you did on the first.
Ok so it would have gone like this.
Myself against a 1550, loss -22
Myself against a 1550, loss - 19
Myself against a 1550, loss - 15
Total rating loss of 56 for losing three games, all against opponents 100+ points higher than myself.
Now my rating is going to be HARDER to increase because the fluctuating levels will set normally, and had I not stopped playing for a week (and that's not much time) I would have only lost about 20 rating for those three games.
This is just a hypothetical, but I really think it's crazy how it goes so askew so fast.
Earlier today I went to play some rated live chess games after not doing so for a week.
My rating was 1460+ at the time and my opponent was rated 1550, I managed to achieve a winning position and I was going to very easily win the game before I disconnected, oh well.
But when I managed to reconnect I noticed to my horror that I had lost 22 rating points, why do the ratings fluctuate so severely after only a week of not playing?