How can we determine the playing strength of a player?
If we look at a single tournament, the result is heavily superimposed by the form of the day. Have a look at the rating table to see how the performance of players fluctuates. Then again, we know that a strong player will win more often than a weak player in the long run. Therefore we need to consider more than one tournament to get a reasonable estimate of the playing strength.
However, the playing strength can change with the time. Averaging over a period with different playing strengths, the result loses its timely resolution. The reference is given by the development speed of the rating. It is said that it takes about ten years to obtain the mastery of field, let it be arts, science, or craftsmanship. The same applies for the opposite direction. Once being a virtuoso does not change immediately. Arpad Elo (The Rating of Chess Players Past and Present) plotted the rating of chess players over the time. It showed that the rating of adult players changed by no more than about 10 points a year, independent of in-activities and internments. This does not apply for juniors: At the age of 14, Robert Fischer gained more than 100 Elo points a year. Though the history of the Stratego rating is young, the data already shows the same behaviour like the chess rating.
There is a trade-off between an effective suppression of fluctuations and up-to-date measurements. An aperture of about 2½ years seems to be a good choice.
Therefore this system uses a GAUSSian function with a time constant of about 2½ years. The tournaments start with a significance of 1.0 which fades away with the time.
The loss of significance starts slowly. During the first 12 months, a game loses about 14 % of its value.
Over the years the significance lessens faster and faster. After 7 years, about 0.06 % is left.
Best viewed with any standard conformant XHTML-browser.
Your browser is outdated and does not accept XHTML.