How does Glicko-2 rating works? • page 1/1 • General Chess Discussion • lichess.org

since I saw numbers like -496, -4, +-0, etc...

Fundamentally, Glicko ratings use "confidence intervals" when calculating and representing your rating. When you first start using the site, your rating starts at 1500 ± 1000. The 1500 represents your rating, and the 1000 represents the confidence interval.

Basically, the system is 95% sure that your rating is somewhere between 500 and 2500. It is incredibly uncertain. Because of this, when a player is just starting out, their rating will change very dramatically, potentially several hundred points at a time. But after some games against established players the confidence interval will narrow, and the amount of points gained/lost after each game will decrease.

Another point to note is that, as time passes, the confidence interval will increase. This allows you to gain/lose points more rapidly to match any changes in your skill level over time .

Ratings are calculated using the Glicko-2 rating method developed by Mark Glickman. This is a very popular rating method, and is used by a significant number of chess organisations (FIDE being a notable counter-example, as they still use the dated Elo rating system). Fundamentally, Glicko ratings use "confidence intervals" when calculating and representing your rating. When you first start using the site, your rating starts at 1500 ± 1000. The 1500 represents your rating, and the 1000 represents the confidence interval. Basically, the system is 95% sure that your rating is somewhere between 500 and 2500. It is incredibly uncertain. Because of this, when a player is just starting out, their rating will change very dramatically, potentially several hundred points at a time. But after some games against established players the confidence interval will narrow, and the amount of points gained/lost after each game will decrease. Another point to note is that, as time passes, the confidence interval will increase. This allows you to gain/lose points more rapidly to match any changes in your skill level over time .

TalMagicalDragon

You have accounts on several chess servers, and your ratings don't match from a server to another. Why is that?

This is a very legitimate and frequently asked question, which also has a simple answer:

It's because chess federations and servers use different rating systems.

There are 3 main systems:

Elo
The Elo rating system is relatively ancient and mathematically simple. It is still in use by many sports and federations, including FIDE.

The Elo system does not recommend any initial rating.

Glicko 1
The glicko rating system has a better prediction accuracy than Elo, but is more complex mathematically. It is a popular choice for new games and sports.

The Glicko system paper specifies that ratings start at 1500.

Glicko 2
An improvement over Glicko 1, adding the concepts of volatility, for even greater accuracy.

Ratings still start at 1500.

Which federation/server uses which system
By chronological order of creation:

FIDE
The International Chess Federation uses the Elo system. Back in 1971, it was the only choice.

New players get a rating equal to the performance in their first rated game results.

FICS
The Free Internet Chess Server uses the Glicko 1 System.

New players get a rating equal to the performance in their first rated game results.

ICC
The Internet Chess Club uses the Elo System.

New players get an initial rating of 1400 or 1600 depending on the pool.

Chess.com
Chess.com uses the Glicko 1 system.

New players used to get an initial rating of 1200. Nowadays, new players get to choose their initial rating themselves.

Lichess.org
Lichess.org uses the Glicko 2 system.

New players get an initial rating of 1500, as recommended by the Glicko definition.

Chess24.com
Chess24.com uses the Elo system.

New players get an initial rating of 1500.

Comparing ratings between federations or servers
As tempting as it sounds, it doesn't make much sense.

They use different rating systems, and also differ in the way they implement them.

FAQ
Are Lichess ratings inflated?
Lichess ratings start at 1500, as is recommended by the Glicko system definition. The median player rating of Lichess is close to 1500, and there is no significant deviation of this value over time.

The Lichess ratings can often be higher than ratings from other systems that start at 1200, such as FIDE or chess.com. This only shows that ratings from different systems cannot be compared.

Are players from server X stronger/weaker, because their ratings are higher/lower?
No. Ratings can't be compared from a server to another. They only serve to compare the skill of players within the same rating pool.

Which rating system is best?
The purpose of rating systems is to predict the outcome of games. Therefore, they can be objectively better or worse, according to their ability to make such predictions. Glicko 1 makes better predictions than Elo, and Glicko 2 makes better predictions than Glicko 1 (source).

Why don't they all use the same rating system?
Because the first rating system historically used, Elo, is pretty bad. Glicko 1, then Glicko 2, make considerable improvements, and offer greater accuracy. Chess servers didn't want to pay the cost of legacy forever, so they moved on to superior systems.

We hope it clears it up

You have accounts on several chess servers, and your ratings don't match from a server to another. Why is that? This is a very legitimate and frequently asked question, which also has a simple answer: It's because chess federations and servers use different rating systems. There are 3 main systems: Elo The Elo rating system is relatively ancient and mathematically simple. It is still in use by many sports and federations, including FIDE. The Elo system does not recommend any initial rating. Glicko 1 The glicko rating system has a better prediction accuracy than Elo, but is more complex mathematically. It is a popular choice for new games and sports. The Glicko system paper specifies that ratings start at 1500. Glicko 2 An improvement over Glicko 1, adding the concepts of volatility, for even greater accuracy. Ratings still start at 1500. Which federation/server uses which system By chronological order of creation: FIDE The International Chess Federation uses the Elo system. Back in 1971, it was the only choice. New players get a rating equal to the performance in their first rated game results. FICS The Free Internet Chess Server uses the Glicko 1 System. New players get a rating equal to the performance in their first rated game results. ICC The Internet Chess Club uses the Elo System. New players get an initial rating of 1400 or 1600 depending on the pool. Chess.com Chess.com uses the Glicko 1 system. New players used to get an initial rating of 1200. Nowadays, new players get to choose their initial rating themselves. Lichess.org Lichess.org uses the Glicko 2 system. New players get an initial rating of 1500, as recommended by the Glicko definition. Chess24.com Chess24.com uses the Elo system. New players get an initial rating of 1500. Comparing ratings between federations or servers As tempting as it sounds, it doesn't make much sense. They use different rating systems, and also differ in the way they implement them. FAQ Are Lichess ratings inflated? Lichess ratings start at 1500, as is recommended by the Glicko system definition. The median player rating of Lichess is close to 1500, and there is no significant deviation of this value over time. The Lichess ratings can often be higher than ratings from other systems that start at 1200, such as FIDE or chess.com. This only shows that ratings from different systems cannot be compared. Are players from server X stronger/weaker, because their ratings are higher/lower? No. Ratings can't be compared from a server to another. They only serve to compare the skill of players within the same rating pool. Which rating system is best? The purpose of rating systems is to predict the outcome of games. Therefore, they can be objectively better or worse, according to their ability to make such predictions. Glicko 1 makes better predictions than Elo, and Glicko 2 makes better predictions than Glicko 1 (source). Why don't they all use the same rating system? Because the first rating system historically used, Elo, is pretty bad. Glicko 1, then Glicko 2, make considerable improvements, and offer greater accuracy. Chess servers didn't want to pay the cost of legacy forever, so they moved on to superior systems. We hope it clears it up

StingerPuzzles

This explains the basic concepts of rating volatility and rating deviation:

http://www.glicko.net/glicko/glicko2.pdf

But LiChess also includes special code for evaluating your first game based on what color you had; likely you're more likely to win with white or black, and at the same time the result will have a large impact on your rating.

More info here: http://www.glicko.net/glicko/glicko-boost.pdf

This explains the basic concepts of rating volatility and rating deviation: http://www.glicko.net/glicko/glicko2.pdf But LiChess also includes special code for evaluating your first game based on what color you had; likely you're more likely to win with white or black, and at the same time the result will have a large impact on your rating. More info here: http://www.glicko.net/glicko/glicko-boost.pdf