my glicko RD is 52.
Glicko RD
Sort:
Apr 19, 2016
0
#23
I got an RD of 24. So that means my most recent rating stats for live blitz - which is around 1840s in the mean time - are relevantly accurate. I think average opponent is another good indicator to figure a player's strength too. But I wonder how accurate the chess.com's general blitz rating system is compare to, say FIDE ... ?
so the lower the rd rating the more accurate your chess rating is..my rd rating is 36, and in blitz it's 1325. not bad for just over 1 year into it, and working full time plus school.
This is how the rating system works:
for example: Won 1047; Lost 1093; Draw: 1111
With Glicko RD as : 16;
You will won possibly against player with rating: 1047 +/- 2* 16; the possibility of your winning is 98%
You will lose to player with rating range at: 1093+/- 2* 16, with 98% possibility.
3* Glicko RD is your 100% chance.
The winning or losing or draw possibility sometimes overlapping, that would be your grey area which only your effort in the game controls the outcome.
There is more: Your winning or losing potential can also be calculated by; highest best win rate +/- 2* Glicko RD, if you experienced that you can win someone is out of your league, that extra rating is your potential, or outcast.
[COMMENT DELETED]
I HAVE A GLICKO RD OF 31 IN RAPID
Jul 14, 2017
0
#28
As I understand it then, your rating is say, 1000. And your Glicko is say, 100. Then, your real rating is anywhere from about 900-1100.
does that sound about right?
Joe1izzo wrote:
As I understand it then, your rating is say, 1000. And your Glicko is say, 100. Then, your real rating is anywhere from about 900-1100.
does that sound about right?
That's not how standarddeviation works sadly. Look up the 68-95-99.7 rule for an easy way to see the implication of standarddeviation in practice. I'll do it here:
If your rating is 1000 and your standarddeviation(or glicko) is 100 then:
You can be 68% confident(in around 68 out of 100 cases with players who'd have this) does your real rating fall within the range 900-1100.
You can be 95% confident that your real rating falls within the range 800-1200
You can be 99.7% sure that your real rating falls within the range 700-1300.
The explanation behind this is pretty complicated but if you really want to know more, it has to do with the density function(which gives probabilities of a variable) of your rating is supposed to be a normal guass curve, and that's one of the easiest probability curves they have all the probabilities exactly calculated of. Also there is a law that all density functions resemble the normal curve as we are talking about bigger amounts of test cases.
But nevermind that, it's complicated and I gave you the numbers you wanted to know, so I hope that helps
Good day how do i know im playing a rated game on chess.com or is everygame rated?
batgirl wrote:
interesting. How low the glicko Rd can go?
RD partly differentiates Glicko fro Elo.
Glicko is a rating sysem. RD= rating deviation. As a person plays more games the RD lowers under the assumption that the more games played, the more accurate the rating. if a person doesn't play much over a period of time, the RD will increase slowly. So you can look at a person's RD to determine how accurate his rating is. But also, you can look at his RD to determine how many points you'll win/lose by playing him. If he has a high RD (somewhat inaccurate) lose to him will affect you less to compensate for the inaccuracy. If you have a high RD you'll have a lot of rating change after each game, if you have a low RD you'll have slighter changes. Your RD doesn't just lower the more you play, but rather lowers in relationship to the RD of your current opponent. So, everything in sort of interconnected and relatively confusing.
Starchip, I know this was four years ago, but I don’t really care. You said,
"mean" is not the same as "average". This is what happens when non-mathematicians start using mathematical terms.
"mean" has a certain meaning and "average" has a pretty average one.
Do not confuse them.
but this is completely wrong. Average is by definition the mean. Mean=average. You also said that you drink while you play sometimes. Maybe you had one too many before writing that down. Everything else besides correcting that guy’s spelling was an opinion so I couldn’t care less about it. Within the last four years I hope you’ve learned a thing or two about what an average ... means.
Does winning or losing a lot affect your Glicko RD?
Aug 28, 2019
0
#35
vaakash wrote:
Does winning or losing a lot affect your Glicko RD?
Only the number of games played (regardless of the results) and the passage of time affect the RD.
ok
Nov 12, 2020
0
#37
batgirl wrote:
RD partly differentiates Glicko fro Elo.
Glicko is a rating sysem. RD= rating deviation. As a person plays more games the RD lowers under the assumption that the more games played, the more accurate the rating. if a person doesn't play much over a period of time, the RD will increase slowly. So you can look at a person's RD to determine how accurate his rating is. But also, you can look at his RD to determine how many points you'll win/lose by playing him. If he has a high RD (somewhat inaccurate) lose to him will affect you less to compensate for the inaccuracy. If you have a high RD you'll have a lot of rating change after each game, if you have a low RD you'll have slighter changes. Your RD doesn't just lower the more you play, but rather lowers in relationship to the RD of your current opponent. So, everything in sort of interconnected and relatively confusing.
Thank you batgirl, this was the best explanation I could have hoped for. A complicated thing clearly explained in a short paragraph. Well done!
Nov 17, 2020
0
#38
integratedcrazy wrote:
Basically, it's an indication of rating acceleration after a win or loss?
Higher RD = higher acceleration.
lower RD = low acceleration.
That's a good rule of thumb, at least concerning your own RD and rating. More specifically, it measures how uncertain Chess.com is about your rating. If there's a lot of uncertainty about your rating, then your rating will change more each game. When there's less uncertainty about your rating, the adjustments are smaller. This way your rating is calibrated quickly.
However, if your opponent has a high RD, that doesn't mean your rating change against them will necessarily be large. If they have a very high RD, that means Chess.com is very unsure of their rating (they probably haven't played many rated games recently). I think this can lead to situations where your rating changes by less against an opponent with a high RD because there's a good chance their rating is inaccurate.
caimimi wrote:
[COMMENT DELETED]
i dont think so
From the freechess.org page linked above:
"Each player can be characterized as having a true (but unknown) rating that may be thought of as the player's average ability."
The only problem with this assumption is that it assumes that each time a player plays a game, that put as much effort and thought into that game as they do for every other game. I know for myself this is entirely untrue. Sometimes I play for fun, sometimes I try new things, and mostly I could care less what my rating is. Sometimes I will beat people a couple hundred points higher than myself, and sometimes the reverse. So, for myself, the whole rd itself is probably highly innacurate, because I really do not play each game in the same way.
Actually, it says the player's average ability, so it accounts for all of that.