A series of unfortunate events has been going on for a few days since I moved to Innopolis, Russia. First, a nuclear explosion upon my arrival. Then the a small fire in the dorm I am staying at which lead to alarms firing and me running around in my pajamas while carrying only my laptop (Yes, I had to save one thing). And then I login to codeforces to find my contribution has decreased. Well now that's bad.
So as continuation for last year's blog here is some data from the handles we have of IOI participants.
Total Score Distribution
Day1 Distribution
Day2 Distribution
Bonus
Try to find tmwilliamlin168 in those pictures since Benq is obvious.
Why there are no participants in the graph between 100 and 200 in day 2?
Ye give me a minute.
Yup, now it's fixed. How many users are in the graph? Could you also colour the background in the total score image according to medals?
tmwilliamlin168
Your guess for day 2 is wrong.
I can easily find myself lol.
That's because your rating is 3000+and the were only two participaters in IOI2019 whose ratings are above 3000. orzzzz~
You also have relatively the same position in all 3 graph. Kinda interesting
It would be amazing to have a graph like this comparing users' previous rating in Codeforces with their actual score/rank in some recent contests (in Codeforces), thus, we could use it as a "blank experiment" and compare the correlations.
If the correlations are similar, it would mean that Codeforces ratings are as good predictor of performance in IOI as in Codeforces, and if the correlation is much better for Codeforces competitions, it would mean that we can't really trust a lot in Codeforces ratings for IOI competitions (which is what I conjecture).