Hello,
I have a concern about the rating. In my last contest (Round #657, only Div 2), I got about 1200th rank out of about 8600 contestants. That's roughly top 15%. This boosted my rating by +11 points. I am guessing that indicates the performance of a 1900 rated contestant. However, the contest before (Round #652, only Div 2), I got ranked again 1200th, but this time about 14500 contestants participated. That put me on top 10 percent. However, that time I actually lost -4 points, which this time is about the performance of a 1840 one.
My question is, even though I did quite better on my 2nd previous contest, which did I get 15 points less? Is it because of rating inflation? Or maybe because the last contest was an outlier (much less participation than the COVID era amount of users)?
Auto comment: topic has been updated by _nicktz1408_ (previous revision, new revision, compare).
I believe it depends on your expected rank as well.
If you get a better rank in a contest but all the people behind you have much less rating than yours, then you won't be benefitted that much. You can't really predict rating change just based on your rank and total participants.
Also maybe the timing played a role on that? Maybe a percentage of users, that can potentially participate in a normal contest, could not participate on this one due to the unusual starting time? That could lead to a partially different set of participants (in a macroscale perspective), so maybe the structure of the finalized ranks got altered a bit? Just speculating!
i think it depends on the no. of participants above you and have current rating more than your current rating so it will not help to be in top 10 percent if the avg rating of participants is less. also i think the overall rating i.e the sum f ratings of all participants must remain constant