The history of Elo in chess and its modifications.
The origin of Elo
Elo emerged as an improvement on the numerical scoring system used by the United States Chess Federation (USCF) since 1939, devised by Kenneth Harkness. This system assigned each player an initial score of 1500 and modified it according to the result of each game, adding or subtracting a fixed amount of points. Harkness' system was simple, but it had some limitations, such as score inflation, a lack of statistical precision, and difficulty comparing players from different regions or eras.
In 1959, the USCF appointed Árpád Élő as chairman of the Scoring Committee and charged him with the task of designing a new, fairer, and more rigorous system. Élő, who was a physics professor and master-level chess player, relied on statistical data and game theory to create the system that bears his name. Élő's system is based on the idea that the probability of a player winning a game depends on the score difference between him and his opponent and that this difference can be expressed by a logistic function. Thus, Élő's system assigns each player an initial score and modifies it according to the result of each game, adding or subtracting a variable number of points that depends on the expected probability of winning and an adjustment factor called the development constant. .
Élő's system was adopted by the USCF in 1960 and by the International Chess Federation (FIDE) in 1970. The first FIDE ranking list was published in July 1971 and ranked American Bobby Fischer as number one. in the world, with a score of 2760. Since then, Élő's system has spread to other disciplines such as Go, Scrabble, team sports, and video games.
Elo characteristics
Élő's system has several features that make it suitable for measuring the strength of chess players. Some of them are:
It is dynamic; it is constantly updated according to the results of the games and reflects changes in the form and level of the players.
It is relative; it does not measure the absolute skill of the players but rather their performance compared to that of other players in the same system. Therefore, scores from different systems are not directly comparable and may vary depending on the size and quality of the sample.
It is predictive; it allows you to estimate the probability that a player will win, lose, or draw a game against another player and calculate the number of points that he will win or lose based on the result.
It is scalable; it can be applied to any number of players, from a few to millions, and at any level of play, from beginners to world champions.
It is universal; it can be used in any type of chess, whether classic, rapid, blitz, team, mail, internet, etc.
Elo modifications
Despite its success and popularity, Élő's system is not without its critics and problems. Some of them are:
- Score inflation occurs when the average score of players increases over time without implying a real increase in their level of play. This can be due to several factors, such as the addition of new players with low scores, the withdrawal of players with high scores, an increase in match frequency, matchmaking bias, etc.
- Player inactivity: this occurs when a player stops playing for a long period of time and his score remains frozen, without reflecting possible changes in his level of play. This can affect the accuracy and fairness of the system and lead to paradoxical situations, such as an inactive player being number one in the world.
- Player heterogeneity occurs when players have different characteristics that influence their performance, such as age, gender, playing style, preparation, motivation, etc. This can cause the system to fail to capture real differences between players and assign scores that do not correspond to their strengths.
To solve these and other problems, Élő's system has been modified and adapted by different chess federations and organizations, which have introduced changes in some of its parameters, such as the initial score, the development constant, the minimum number of games, the calculation period, the idle factor, the bonus factor, etc. Some examples of these modifications are:
- The Glicko system, developed by Professor Mark Glickman in 1995, is a variant of the Élő system that incorporates an uncertainty factor that measures the reliability of each player's score and is modified according to the number and quality of the games played. The greater the uncertainty factor, the greater the variation in the score after each game, and vice versa. The Glicko system is used on some online chess websites, such as Chess.com and Lichess.
- The dynamic Elo system, proposed by the Spanish grandmaster Miguel Illescas in 2009, is a variant of the Élo system that introduces a bonus factor that rewards players who achieve results above expectations and is calculated according to the difference between each player's performance and score. The dynamic Elo system is used in the FIDE Grand Prix circuit.
- The continuous Elo system, proposed by the Belgian mathematician Marc Van Herck in 2010, is a variant of the Élo system that eliminates the calculation period and updates the score of each player after each game, taking into account the results of all the players in the system. The continuous Elo system is used on the Belgian Chess Federation website.
Bibliography
Elo, A. (1978). *The rating of chessplayers, past and present*. Arco Pub.
Glickman, M. E. (1995). A comprehensive guide to chess ratings. *American Chess Journal*, 3, 59-102.
Illescas, M. (2009). Elo dinámico. *Peón de Rey*, 76, 6-9.
Van Herck, M. (2010). Continuous ratings: a new rating system for chess. *ICGA Journal*, 33(4), 221-228.
Wikipedia. (2021). Sistema de puntuación Elo. Recuperado de [Wikipedia](https://es.wikipedia.org/wiki/Sistema_de_puntuaci%C3%B3n_Elo)
0 Comments