If you’ve ever played a competitive online game, you’re probably familiar with the concept of MMR, or Matchmaking Rating. MMR is a system used to determine the skill level of players, and it’s used to match players against opponents of a similar skill level. One popular MMR system used in games like CS:GO is Glicko.
Glicko is a rating system developed by Mark Glickman in 1995. It’s an improvement on the Elo rating system, which was developed in the 1960s for chess. Like Elo, Glicko is a rating system that assigns a rating to each player based on their performance in matches. However, Glicko takes into account more factors, such as the number of matches played and the variability of a player’s performance over time.
In Glicko, each player starts with a rating of 1500. As they play matches, their rating goes up or down depending on the outcome of the match and the rating of their opponent. If a player wins a match against an opponent with a higher rating, their rating will go up more than if they had won against a lower-rated opponent. Conversely, if a player loses a match against a lower-rated opponent, their rating will go down more than if they had lost against a higher-rated opponent.
One of the unique features of the Glicko system is the concept of a “rating deviation.” The rating deviation is a measure of the variability of a player’s performance. If a player has a high rating deviation, it means that their performance is inconsistent, and their rating will be less accurate. On the other hand, if a player has a low rating deviation, it means that their performance is consistent, and their rating will be more accurate.
Glicko has been used in several competitive games, including CS:GO, and Chess. The Glicko system is used to determine the matchmaking rating of players. Players are matched against opponents of a similar skill level, based on their Glicko rating. This ensures that matches are competitive and fair, and that players are not matched against opponents that are significantly better or worse than them.
Overall, the Glicko system is a popular MMR system that offers a more accurate and fair rating system for competitive games. It takes into account more factors than traditional rating systems, such as the number of matches played and the variability of a player’s performance. This makes it a more reliable system for determining a player’s skill level and ensuring that matches are competitive and fair.
| Formula | Description |
|---|---|
Δ = (v² * Σg² * E)⁻¹ * Σg * (S - G(μ, Σg², R)) | The amount a player’s rating should change after a match |
v = (Σg² * E)⁻¹ | The reliability of a player’s rating |
E = 1 / (1 + exp(-sqrt(π² + 3 * Σd²) / ρ)) | The expected outcome of a match for a player |
μ' = μ + τ * Δ | The player’s new rating mean |
Σ²' = (Σg² * E)⁻¹ | The player’s new rating deviation |
In these formulas, Δ represents the change in a player’s rating, v represents the reliability of a player’s rating, E represents the expected outcome of a match for a player, μ represents the player’s rating mean, Σ² represents the player’s rating deviation, and τ represents the system’s “tau” parameter, which controls the degree to which a player’s rating can change from match to match. these formulas allow the Glicko system to take into account a player’s performance in matches, the number of matches played, and the variability of a player’s performance over time, in order to assign a more accurate and fair rating to each player.
Examples of Glico based on matches in CS:GO
Let’s say you play a match in CS:GO and you win against an opponent who has a higher ranking than you. According to the Glicko system, you will receive more rating points than if you had won against an opponent with a lower ranking than you. On the other hand, if you lose against an opponent with a lower ranking than you, you will lose more rating points than if you had lost against an opponent with a higher ranking than you.
The amount of rating points gained or lost after a match depends on the difference in ranking between you and your opponent. If there is a large difference in ranking between you and your opponent, you will gain or lose more rating points than if there is only a small difference.
The Glicko system also takes into account the number of matches you have played and the variability of your performance over time. For example, if you haven’t played a match in a while and then come back and win, you will gain more rating points than if you had been playing consistently and won the same match.
The system also assigns a “reliability” value to your rating, which indicates how confident the system is in your rating. This value is based on the number of matches you have played and how consistent your performance has been over time.
The Glicko system aims to provide a fair and accurate ranking system for players in CS:GO and other games, taking into account factors like performance, consistency, and the ranking of opponents.
Glicko in Dota 2
Recently it was reported that in the April Dota 2 patch will be introduced the same system Glicko
