Let me get straight into point , I have played about more 100 blitz and bullet games . The rating i gain or lose after a game is kinda high . Even if I play against someone of same rating as mine I would sometime gain 8 rating points which is illegal.This means i can quickly gain points and lose in the same way ! Pls fix this ! For more information feel free to visit my profile .I am not advertising please...I would like to get it fixed it as soon as possible
@challenger363 said in #1:
> Even if I play against someone of same rating as mine I would sometime gain 8 rating points which is illegal.
What do you mean by illegal? What would you expect instead?
> For more information feel free to visit my profile
I visited it and found no glitches. Can you point out to a specific example?
> I would like to get it fixed it as soon as possible
Have you considered the option that you may not be understanding the rating system?
Please note that lichess uses Glicko 2, which is considered a more accurate system than for example the systems used by FIDE or chess.com
Find more info on lichess.org/page/rating-systems
> Even if I play against someone of same rating as mine I would sometime gain 8 rating points which is illegal.
What do you mean by illegal? What would you expect instead?
> For more information feel free to visit my profile
I visited it and found no glitches. Can you point out to a specific example?
> I would like to get it fixed it as soon as possible
Have you considered the option that you may not be understanding the rating system?
Please note that lichess uses Glicko 2, which is considered a more accurate system than for example the systems used by FIDE or chess.com
Find more info on lichess.org/page/rating-systems
The rating changes you are seeing looks correct to me, even if it were the ELO system, which would give you about +10 for beating an opponent of the same rating.
Lichess uses the Glicko-2 rating system instead of the ELO system.
How many points you gain or lose depends on two things:
1- The relative rating of your opponent. (Higher than you means you gain more and lose less, and lower than you is the opposite.)
2- How stable each of your ratings are. (Unstable ratings change a lot, stable ratings don't change hardly at all.)
The 1st one is just like ELO. The 2nd one is based on how many games you've played and on how frequently you are playing them. The longer you go without playing a game, the less stable it considers your rating to be. The more frequently you play, the more stable.
Hopefully this helps understand how the ratings on Lichess work.
Lichess uses the Glicko-2 rating system instead of the ELO system.
How many points you gain or lose depends on two things:
1- The relative rating of your opponent. (Higher than you means you gain more and lose less, and lower than you is the opposite.)
2- How stable each of your ratings are. (Unstable ratings change a lot, stable ratings don't change hardly at all.)
The 1st one is just like ELO. The 2nd one is based on how many games you've played and on how frequently you are playing them. The longer you go without playing a game, the less stable it considers your rating to be. The more frequently you play, the more stable.
Hopefully this helps understand how the ratings on Lichess work.
Glicko-2 is difficult to implement correctly but could be more stable than the current rating system.
@Toadofsky said in #4:
> Glicko-2 is difficult to implement correctly but could be more stable than the current rating system.
The current rating system being--?
> Glicko-2 is difficult to implement correctly but could be more stable than the current rating system.
The current rating system being--?
@Marlonc said in #5:
> The current rating system being--?
Good question; I'll give you page 1 of a Glicko-2 paper in case you want to check for yourself whether Lichess implements it (you don't even need to read "Step 1"):
http://www.glicko.net/glicko/glicko2.pdf
> The current rating system being--?
Good question; I'll give you page 1 of a Glicko-2 paper in case you want to check for yourself whether Lichess implements it (you don't even need to read "Step 1"):
http://www.glicko.net/glicko/glicko2.pdf
@Toadofsky said in #6:
> Good question; I'll give you page 1 of a Glicko-2 paper in case you want to check for yourself whether Lichess implements it (you don't even need to read "Step 1"):
> http://www.glicko.net/glicko/glicko2.pdf
I still don't understand your comparison of Glicko-2 to "the current rating system", though thanks to that paper I believe I can guess the etymology of 'Glicko'.
Let me rephrase my question. Given that we are talking about lichess, and further given that, as the others pointed out before me it says here lichess.org/page/rating-systems, "Lichess.org uses the Glicko 2 system" what do you mean by saying that "Glicko-2 is difficult to implement correctly but could be more stable than the current rating system"?
> Good question; I'll give you page 1 of a Glicko-2 paper in case you want to check for yourself whether Lichess implements it (you don't even need to read "Step 1"):
> http://www.glicko.net/glicko/glicko2.pdf
I still don't understand your comparison of Glicko-2 to "the current rating system", though thanks to that paper I believe I can guess the etymology of 'Glicko'.
Let me rephrase my question. Given that we are talking about lichess, and further given that, as the others pointed out before me it says here lichess.org/page/rating-systems, "Lichess.org uses the Glicko 2 system" what do you mean by saying that "Glicko-2 is difficult to implement correctly but could be more stable than the current rating system"?
I'll repeat simply what Glickman says... when a player has exceptionally good runs or bad runs, without doing the nonsense which is the US Chess Rating System it is valuable to allow a player's rating to vary, and when a player has predictable performance their rating shouldn't be volatile:
Every player in the Glicko-2 system has a rating, r, a rating deviation, RD, and a rating volatility σ. The volatility measure indicates the degree of expected fluctuation in a player’s rating. The volatility measure is high when a player has erratic performances (e.g., when the player has had exceptionally strong results after a period of stability), and the volatility measure is low when the player performs at a consistent level. As with the original Glicko system, it is usually informative to summarize a player’s strength in the form of an interval (rather than merely report a rating). One way to do this is to report a 95% confidence interval. The lowest value in the interval is the player’s rating minus twice the RD, and the highest value is the player’s rating plus twice the RD. So, for example, if a player’s rating is 1850 and the RD is 50, the interval would go from 1750 to 1950. We would then say that we’re 95% confident that the player’s actual strength is between 1750 and 1950. When a player has a low RD, the interval would be narrow, so that we would be 95% confident about a player’s strength being in a small interval of values. The volatility measure does not appear in the calculation of this interval.
Every player in the Glicko-2 system has a rating, r, a rating deviation, RD, and a rating volatility σ. The volatility measure indicates the degree of expected fluctuation in a player’s rating. The volatility measure is high when a player has erratic performances (e.g., when the player has had exceptionally strong results after a period of stability), and the volatility measure is low when the player performs at a consistent level. As with the original Glicko system, it is usually informative to summarize a player’s strength in the form of an interval (rather than merely report a rating). One way to do this is to report a 95% confidence interval. The lowest value in the interval is the player’s rating minus twice the RD, and the highest value is the player’s rating plus twice the RD. So, for example, if a player’s rating is 1850 and the RD is 50, the interval would go from 1750 to 1950. We would then say that we’re 95% confident that the player’s actual strength is between 1750 and 1950. When a player has a low RD, the interval would be narrow, so that we would be 95% confident about a player’s strength being in a small interval of values. The volatility measure does not appear in the calculation of this interval.
@Toadofsky said in #8:
> I'll repeat simply what Glickman says... when a player has exceptionally good runs or bad runs, without doing the nonsense which is the US Chess Rating System it is valuable to allow a player's rating to vary, and when a player has predictable performance their rating shouldn't be volatile:
>
> Every player in the Glicko-2 system has a rating, r, a rating deviation, RD, and a rating volatility σ. The volatility measure indicates the degree of expected fluctuation in a player’s rating. The volatility measure is high when a player has erratic performances (e.g., when the player has had exceptionally strong results after a period of stability), and the volatility measure is low when the player performs at a consistent level. As with the original Glicko system, it is usually informative to summarize a player’s strength in the form of an interval (rather than merely report a rating). One way to do this is to report a 95% confidence interval. The lowest value in the interval is the player’s rating minus twice the RD, and the highest value is the player’s rating plus twice the RD. So, for example, if a player’s rating is 1850 and the RD is 50, the interval would go from 1750 to 1950. We would then say that we’re 95% confident that the player’s actual strength is between 1750 and 1950. When a player has a low RD, the interval would be narrow, so that we would be 95% confident about a player’s strength being in a small interval of values. The volatility measure does not appear in the calculation of this interval.
Then your terms may be defined as follows:
#4 > Glicko-2 is difficult to implement correctly but could be more stable than the current rating system.
where
#4 > the current rating system
is equal to
#8 > the US Chess Rating System
Y/N?
> I'll repeat simply what Glickman says... when a player has exceptionally good runs or bad runs, without doing the nonsense which is the US Chess Rating System it is valuable to allow a player's rating to vary, and when a player has predictable performance their rating shouldn't be volatile:
>
> Every player in the Glicko-2 system has a rating, r, a rating deviation, RD, and a rating volatility σ. The volatility measure indicates the degree of expected fluctuation in a player’s rating. The volatility measure is high when a player has erratic performances (e.g., when the player has had exceptionally strong results after a period of stability), and the volatility measure is low when the player performs at a consistent level. As with the original Glicko system, it is usually informative to summarize a player’s strength in the form of an interval (rather than merely report a rating). One way to do this is to report a 95% confidence interval. The lowest value in the interval is the player’s rating minus twice the RD, and the highest value is the player’s rating plus twice the RD. So, for example, if a player’s rating is 1850 and the RD is 50, the interval would go from 1750 to 1950. We would then say that we’re 95% confident that the player’s actual strength is between 1750 and 1950. When a player has a low RD, the interval would be narrow, so that we would be 95% confident about a player’s strength being in a small interval of values. The volatility measure does not appear in the calculation of this interval.
Then your terms may be defined as follows:
#4 > Glicko-2 is difficult to implement correctly but could be more stable than the current rating system.
where
#4 > the current rating system
is equal to
#8 > the US Chess Rating System
Y/N?
Terms "Glicko-2" and "US Chess Rating System" are Prof. Glickman's intellectual property. Here is his PDF for the latter:
new.uschess.org/sites/default/files/media/documents/the-us-chess-rating-system-revised-september-2020.pdf
By "the current rating system" I refer to the nameless system used on Lichess today.
new.uschess.org/sites/default/files/media/documents/the-us-chess-rating-system-revised-september-2020.pdf
By "the current rating system" I refer to the nameless system used on Lichess today.
This topic has been archived and can no longer be replied to.