Alternative rating.

http://ratings.fide.com/calculator_rtd.phtml - I think this is a sufficiently authoritative option :smile20:
http://georgejohn.bcentralhost.com/TCA/perfrate.html

http://ru.wikipedia.org/wiki/%D0%A0%D0%B5%D0%B9%D1%82%D0%B8%D0%BD%D0%B3_%D0%AD% D0%BB%D0%BE

In the Elo rating system, it is accepted that the transition from one class of play to the next occurs approximately every 200 rating points. If the difference between two players is 200 points, the stronger player wins with a probability of 76%; if the difference is 400 points, the probability will be 91%. A difference of 600 points means that the stronger player wins almost always (97%). If the ratings of both players are equal, the probability of victory for one of them is 50%. These probabilities, of course, do not take into account a player's sporting form at any specific moment. At the lower class level, the Elo rating more often gives incorrect predictions of the result, as players of this class make unpredictable mistakes more frequently.

The less often a chess player loses, the more accurately their rating can be estimated. The most accurate rating can be obtained based on tournaments in which players of approximately equal strength compete. The Elo rating system is based on the assumption that the strength of each chess player can be represented as a random variable following a normal distribution. The calculation of a specific player's rating based on the results of any tournament is based on comparing the number of points they scored with the expected number of points predicted based on their rating. If, by the end of the tournament, the number of points scored is greater than the predicted value, then the player's rating increases. If, by the end of the tournament, the number of points scored is less than the predicted value, then the player's rating decreases.

http://research.microsoft.com/en-us/...lculators.aspx

http://www.namibianchessfederation.c...calculator.php