| |
| Based on the rating system developed by Prof. Árpád Élő |
| |
| The rating is
an integer, the minimum is 0. The starting rate of an unknown player is 0.
|
| |
| Ratings are amended
each time you kill someone by adding small amount from your
previous rating value. On the other hand, a small amount is subtracted if you've been killed.
The amount added or subtracted depends on the ratings of the players.
|
|
| It is an absolute
rating system, i.e. the rating can be used to determine the winning odds of a specific
player against another specific player. The system has been worked out according to
statistical and probability theory.
|
| |
| Rating difference | Probability of win |
| 0 | 0.50 |
| 100 | 0.54 |
| 200 | 0.59 |
| 400 | 0.68 |
| 800 | 0.82 |
| 1200 | 0.91 |
| 1800 | 0.97 |
| 2400 | 0.99 |
|
|
|
|
|
The following formula is used:
|
Rn = Ro + K(W-We)
|
Rn is
the new rating.
Ro the old (pre-event) rating.
K a constant (4 for 0-2099, 3 for 2100-2399, 2 for 2400 and above).
W the outcome: you kill=1; you're killed=0.
We the expected score (Win Expectancy), from the following formula:
We = 1/ (exp10(-dr/1200) + 1)
dr equals the difference in ratings.
|
| |
| This formula has the property that the sum of the
rating changes is zero. It turns out that if the rating difference is more then
719 points, then if the strong player kills the weak, there is no change in
either rating.
|
Overall ratings are decreased each day by an amount depending on the current rating.
The following formula is used:
Rn = ( Ro * Ro ) / 125000
Rn is the new rating.
Ro the old (previous day) rating.
| Rating |
Amount subtracted |
| below 500 |
1 |
| 600 |
2 |
| 800 |
5 |
| 1000 |
8 |
| 1500 |
18 |
| 2000 |
32 |
So players with higher rating will see their rating decrease quicker.
A 500 rating will reach zero in approximately 1.5 years without playing.
|
Series of games |
| |
Playing a series of games against the same opponent within a given
time interval will adjust the calculation as follows:
The expected score is calculated for the series of games from the pre-match rating-difference of the two players,
and the ratings are amended based on the expected score and the actual score of the series of games.
These are basically the extensions of the above calculation for more games played.
However, as more matches will reflect the skill difference more precisely, the amendment is raised based on the number of games played.
Number of games in
a series of games |
Maximum
rating change |
| 1 |
4 |
| 2 |
10 |
| 3 |
15 |
| 4 |
20 |
| 5 |
24 |
| 6 |
28 |
| 7 |
32 |
| 8 |
35 |
| 9 |
38 |
| 10 |
41 |
| 15 |
50 |
| 20 |
56 |
| infinite |
64 |
In other words the constant K in the above equations is multiplied by a factor of 8*{2-[(9/10)^(n-7)]}, where n is the number of games played.
Note that if you play a 3rd player in between playing the same opponent, it will still be treated as a series of games.
|
|
| Rating
categories:
|
 |
General of the army |
 |
General |
Above 700 |
 |
Lieutenant General |
600 - 699 |
 |
Major General |
550 - 599 |
 |
Brigadier General |
500 - 549 |
 |
Colonel |
450 - 499 |
 |
Lieutenant Colonel |
400 - 449 |
 |
Major |
350 - 399 |
 |
Captain |
300 - 349 |
 |
First Lieutenant |
250 - 299 |
 |
Second Lieutenant |
200 - 249 |
 |
Staff Sergeant |
150 - 199 |
 |
Sergeant |
100 - 149 |
 |
Corporal |
50 - 99 |
 |
Private 1st class |
0 - 49 |
|
|
