adeam's blog

I Am Curious... How Are Problems Getting Rated From 800–3500?
By adeam, 5 months ago, In English

I noticed something: many problems are rated low even though they feel hard, and some are rated high even though they seem easier than lower-rated ones. Can anyone explain why this happens and how the rating system works? For example, sometimes an 1100-rated problem feels harder than a 1300–1500 problem,and so on.

  • Vote: I like it
  • +2
  • Vote: I do not like it

»
5 months ago, hide # |
 
Vote: I like it +12 Vote: I do not like it

Not everyone has the same weaknesses, I believe the Problems are based on the ratio/amount of participants that complete the problem, if you struggling on a problem you consider low rated, it probably means your weak in the a subject, and should target it when you practice next time

→ Reply
»
5 months ago, hide # |
 
Vote: I like it +9 Vote: I do not like it

There are some harder (2000+) problems that I can solve fairly quickly, and there are some easier problems I still can't solve (there's a 1000 rated problem that I've probably spent over an hour on and I still can't solve it). The rating is based off of how most people solved it, so your personal results might deviate from the actual rating (and sometimes, there are easy problems that you simply can't solve, like I mentioned in my first sentence).

→ Reply
»
5 months ago, hide # |
 
Vote: I like it +37 Vote: I do not like it

theres a very complicated algorithm. It can be reduced to something like this

int getRating(int problemID){
    return ((rand()%28)+8)*100;
}

hope that helps!

→ Reply
»
5 months ago, hide # |
← Rev. 2  
Vote: I like it +8 Vote: I do not like it

The theoretical rating (the rating of a participant that has a 50% chance of solving) is calculatable from the rankings. For example, https://clist.by calculates them. The rating on CF is usually quite close to this rating, but maybe a bit lower (sometimes the [theoretical] rating can be inflated).

→ Reply