Hey coders, I wanted to share my views on the huge increase in rating change in starter contests. In many past contests it can be seen foe Div2 that till C question are of ratin 1400 and suddenly next question comes of 2500 or 2200+. I just wanted to ask wouldn’t it be beneficial to add question of 2000 rating instead of this. Like if we have 6 questions in contest we may have rating like…
A-1400, B-1600, C-1800, D-2000, E-2300, F - 2800+
But what is happening in many contest is:
A-1200, B-1400, C-1600, D-2500, E-2800, F - 2800+
Due to this till C many coders just complete in less than 30 minutes and after that solving 2500 level is really difficult as it is much higher rating than any 3 or 4 star coder can do. Please consider this request and if it is not good please specify your point of view as well.
I made a post previously making the same point,
First, rating isn’t decided pre-contest, they are based on how many participants managed to solve the problem during the contest and hence decided once the contest is over, what you are trying to say is the difficulty is not distributed evenly.
The answer to this problem is not the difficulty distribution, it is cheating during the contests, some participants leak solutions, so, unless the problem is extremely hard and only the very best participants (who don’t indulge in cheating) can solve the problem, the solution gets leaked and the whole crowd catches up.
I hope this helps.
But due to some cheatings we should not make contests like this. It doesn’t make sense to create 3 easy problems then. Then if cheating is reason no problem should be easy, neither A,B,C.
And also while creating problem creator knows very well what level it would be as they are experienced. So can’t say it is decided after contest.
I have only this point that don’t distribute rating like this… We have platform like codeforces and atcoder it doesn’t happen there. All problems are evenly distributed.
A contest motive must be to improve participants i.e. if we upsolve problem we must get some important insights from that. But as of now upsolving 2500 problem for 3 or 4 star coder doesn’t lead to anywhere, it just creates confusion rather than any enhancement.
CodeChef should look into even distribution of problems like others do.