The rating system we use is a seed based rating system based on the ELO rating. It is a method of rating participants in the individual competitions. Introduced around the same time four years ago the ratings system is implemented in all our monthly contests (Long Challenge, Cook-Off, and Lunchtime).
To give you an insight on our ELO based rating system, we have described it below:
First for each pair of players (A,B) we calculate the probability of A defeating B and vice versa based on the ELO formula. Then for every player we calculate the sum of probability of defeating other players.
P[B, A(i)] = probability that B defeats A(i).
total[B] = summation of P[B][A(i)] over all A(i).
n = number_of_participants.
The expected rank of a player B in a contest,
E[B] = n – total[B].
We define regular factor for contestant B,
Regular_factor[B] = E[B] – Rank[B].
We have also added a bonus factor so that we can reward extra points to players who have been ranked higher in the contest.
bonus_factor[B] = n/rank[B].
Then for each participating contestant her rating is modified as follows:
New rating[B] = rating[B] + k * regular_factor[B] + m * bonus_factor[B].
Here k and m are constants to give appropriate weight to both the factors.
We have separate rating lists for all our monthly contests and you can check them all below:
We hope this will help you understand how you are being rated in our contests. However, if you still have any doubts/suggestions, do send them to us at: email@example.com and we will be happy to talk to you.