Notes on Ratings
Under the Elo ratings system, the winner of a game gains a number of ratings points and the losing player loses the same number. The number of points won or lost depends on the difference between the ratings of the two players; a player will gain more points by beating a higher-rated player than by beating a lower-rated player.
Where games have been completed, the chart above shows the actual number of ratings points passed from the loser to the winner. Where games are unfinished, two numbers are shown in the form A/B. A is the number of points to be transferred if the player in the row wins and B is the number of points if the player in the column wins.
For example
| Mildred Elo 5051 | If Joe wins he will gain 9.5 rating points from Mildred but if Mildred wins she will gain 14.5 rating points from Joe. Mildred stands to earn more than Joe from winning the game because Joe is the higher rated player and is expected to win more often. |
Joe Elo 5123 | E: 9.5 / 14.5 |
In the case of a draw, the lower rated player wins points from the higher rated player, but fewer points than if he or she had won the game.
The value of each game is based on the players' rating at the start of each round. At the end of the round, the points won and lost on each game are added up and the players' rating are adjusted and rounded to the nearest whole number for the next round.
The difference in two players' Elo ratings implies an expectation of how often each of them will win when playing each other. Under the system we use:
- a player with a rating of 400 more than her opponent is ten times as likely to win
- a player with a rating 120 more than his opponent is expected to win two times out of three
- a player with a rating 75 more than her opponent is expected to win three times out five.
Notes on Statistics
μ (mu) is the mean score, ν (nu) is the median score and Mo is the mode. Mean, median and mode are three different kinds of averages. The mean score is the total number of points divided by the number of games played. The median score is the middlemost score (or the mean of the two middlemost scores) when they are sorted from high to low. The mode is the most frequent score and is only helpful after a large number of games.
σ (sigma) is the standard deviation. This is a measure of how consistent the player is. The smaller the value of σ, the more consistent. 68% of the player's scores are within one standard deviation above or below the mean and 95% are within two standard deviations.
Sk is median skewness. Skewness is a measure of how asymmetrical the graph is. There are many measures of skewness; we've chosen median skewness because it's the easiest to calculate! A positive value means there are more scores above the middle of the graph than below, a negative value vice versa.