Elo rating system
Comments916this wiki
Redirected from ELO
The Elo rating system is a method for calculating the relative skill levels of players in two-player games such as chess. It is named after its creator Arpad Elo, a Hungarian-born American physics professor.[1] The Elo system was invented as an improved chess rating system, but today it has been adapted for use in many other games. Variations of it is also used as a rating system for multiplayer competition in a number of games and has been adapted to team sports including association football, American college football and basketball, and Major League Baseball.
In League of Legends the Elo rating of a player is used by the matchmaking in ranked games to find other players of a similar skill level to play with/against. Elo is not used for custom and Co-op vs. AI games. The Elo rating for ranked games is different for each queue types: 3v3 arranged, 5v5 solo and 5v5 arranged teams. The rating is only visible for ranked games after 10 games played in a certain queue type. A summoner's Normal game Elo remains hidden at all times and can only be guessed upon based off his or her win/loss ratio and the apparent skill of teammates and enemies, although not perfectly accurate as normal queues are less stringent as opposed to ranked queues. Players were awarded with medals in their summoner profile based on their Elo at the conclusion of Season One. These medals are given as follows:[2]
- Bronze: Between 1249 and 1399 (3v3: 1249-1409, pre-made 5v5: 1249-1409) (Top 25%)
- Silver: Between 1400 and 1519 (3v3: 1410-1489, pre-made 5v5: 1410-1499) (Top 10%)
- Gold: Between 1520 and 1899 (3v3: 1490-1699, pre-made 5v5: 1500-1749) (Top 3%)
- Platinum: 1900 and above (3v3: 1700+, pre-made 5v5: 1750+) (Top 0.1%)
Contents |
The math of Elo
Edit
The specific formulas used for Elo calculations in League of Legends are unknown. However, most Elo implementations share the same basics as that originally designed for chess. A brief summary is given below. For a more detailed discussion, see Wikipedia.
It is assumed that a person's performance varies from game to game in approximately a normal distribution and a person's Elo rating is the mean of that distribution. A person with a higher Elo will perform better on average than a player with a lower Elo. This score is determined entirely by win/loss statistics in relation to other players. For players A and B with respective Elo ratings of Ra and Rb the expected victorious outcome Ea of the game for player A is given by the following formula:
For every difference of 400 points, the team/player with the higher score is ten times as likely to win as the other team/player. This standard is for Chess and may be different in League of Legends. After a game the actual outcome is compared to the expected outcome and each team/players rating is adjusted to bring them closer to where they should actually be. As a result, if a team was expected to win and does their score changes less than if they where expected to lose and instead won. Successive games should eventually bring each player/team to a point where they are expected to win 50% of the time against opponents of equal score.
A player's change in rating is linear to the difference between the expected outcome and the actual outcome. It is given by the following formula where Sa is the result of the game and is presumably 1 for a win and 0 for a loss.
The magnitude of the score change is determined by the player's K value. In chess initially this K value is big (25 for their first 30 games) resulting in large changes in Elo. This is so a player can rapidly find his correct place in the ranking system. As their number of wins and losses becomes more even this K value is reduced to prevent dramatic changes in Elo against evenly matched opponents (K = 15 to 10). This also prevents inflation in ratings at high Elo play. It appears that League of Legends uses a similar system of changing K values: K appears to start around 60, eventually leveling out to about 25.[3]
All players start ranked play with an Elo of 1200 for their first 10 games at level 30. From there they are assigned a score and changes are made as normal. [4]
Elo decayEdit
Elo decays over time when you are above 1400 Elo:[5]
- Elo decays after 4 weeks of inactivity.
- For normal rating, inactivity is defined as no activity in any queue.
- For ranked rating, inactivity is defined as no activity in the specific queue (arranged 5x5, arranged 3x3, and solo/duo 5x5 are all tracked separately). Ranked decay also applies to people who are unranked and above 1400 rating.
- Elo decays at a rate of 25 every week after the 4th week of inactivity for both normal and ranked ratings.
- The decay timer completely starts over after a game is played in that specific queue
- Elo will not decay below a rating of 1400.
Elo hellEdit
Elo hell is a theoretical and controversial range of the ranked solo queue. It is claimed to be populated with griefers, players that intentionally provide a poor performance in the game to reduce the effectiveness of the players on their own team. For example, some griefers are known to intentionally feed, which refers to the practice of strengthening the enemy team by deliberately allowing enemy players to kill their champion without resistance. Griefers might also intentionally afk, disconnect, or refuse to cooperate with teammates within a game resulting in incomplete teams and consequently making the game more difficult to win for the player's remaining teammates. It is also common to find skilled players who have intentionally queue dodged in order to play against inferior opponents at lower rankings, thereby imbalancing teamplay and corrupting the intended function of the ranked match making system.
There is frequent debate about the existence and nature of Elo hell in the League of Legends community. Some players say that since they are constantly being matched with bad teammates, new players (since the starting Elo is 1200) and/or griefers, they cannot win enough games to reach their appropriate Elo since they cannot carry them.
Others claim it is an excuse of unskilled players for their poor rankings, since the enemy team is more likely to get poor players than the ally team (assuming that the player in question performs above their Elo rating). They say the Elo system is very self-correcting in this regard. Assuming the player in question plays above their Elo, the enemy team has 5 possible positions for a poor player, while the allied team only has 4. By playing enough games, the law of large numbers states that the player will eventually reach their actual Elo because they will win 5 games out of 9 until they reach their actual Elo, in which case both teams have 5 positions for the poor performing player and therefore the player in question can expect to win 5 games out of 10. Players who deserve a rating will rise or fall to it in due time. The odds of remaining at an undeservedly low rating are insignificant given a large enough number of games.
The real problem is volatility of the normal distribution. Elo was originally intended for chess where games are 1vs1 or team sports where the rating reflected the team's performance, rather than that of individual teammates. League of Legends, however, is a team game with a large number of dynamics that will affect an individual's performance. The More simply put, a player's Elo is comparatively imprecise, and in turn this results in one's Elo shifting radically between a certain level. For example in solo queue if a players true average Elo is 1300 his rating may fluctuate between 1200-1400 (from being rated mediocre to being a part of the top 10% of players). It's also more common to find unintended variables which can corrupt the system at lower rankings. The increased frequency of players intentionally feeding, griefing, rage quitting, or even skilled players who have queue dodged to lower rankings can all have substantial and unintended effects on the outcome of games. While some would argue a skilled player should be able to compensate for these events, or that over time these occurrences would theoretically occur as much in a player's favor as against, it's an unfactored variable and an obvious handicap not dealt with as often at higher rankings. It is this extremely high level of volatility and unpredictability that fosters a belief in the existance of Elo hell.
Anonymous User
