Dear
MOG
users,
We have updated the
rating calculation formulas
in our system. These are now based on the ones used by Codeforces, which can be found in this
post
. Additionally, there are some aspects we would like to share:
-
All users with at least one submission (which is not Compilation Error) will be rated. In the first rounds in
MOG
, it was done like this but in the last ones only the users with at least one
AC
were rated).
-
Non-rated
users will be considered to start with
1300
(base rating).
-
After rating each contest it holds that:
-
a) if the participant
j
had a worse rating than the participant
i
before the contest and finished the contest on a worse place, then the new rating of
j
can’t be greater than the new rating of
i
.
-
b) if
i
finished the contest better than
j
but
i
had worse rating before the contest then
i
should have equal or greater rating change than
j
.
-
c) a user whose rank in the contest is the last place cannot increase rating.
-
One of the steps for computing the new rating of a user is to calculate his/her
expected rank
. We created a
new page
for each rated contest, where you can see for each user: the
rank
in the competition, the
expected rank
, the
old rating
, the
new rating
and the
rating change
.
The Rating Divisions remain the same:
Green: [1, 1399]
Blue: [1400, 1599]
Violet: [1600, 1799]
Orange: [1800, 1999]
Red: [2000, -]
To keep the rating system consistent, we are going to
rerate all previous contests
according to the new formulas.
UPD1
: All contest have been rerated. If you spot anything that doesn't make sense, feel free to leave us a comment! You can check the rating changes of all the rated contests with the following links: