I personally believe weighted rating is more fair than the standard IOI rating, because solving a difficult task should earn you more points than solving an easier one, even if the number of test cases for both tasks is the same.
That's why I decided to see what the scores would have been like for the past USACO Contests if that system was the one used.
It's not the exact same system USACO used to employ, because I didn't limit a problem weighting to 400 (ie: btree weighting in the November contest is 519).
Check it out and let me know what you think.
"You will receive an email if and when the owner of this item approves your request."????
shall we wait??
Same problem here
Sorry! Didn't know I had to share the document.
Now it's public.
hmm .. for me I prefer equal weighted problems system.
in case of this system all contestants will try to solve the easier problems first that's why it's fair
that my opinion
Hhhmmmm, it's just a matter of opinions, but I disagree.
Maybe a contestant will solve the easier problem fast and move on to the harder ones. He'll spend a couple of hours devising a good algorithm for the hardest problems, and earn 100% of its points, but because of rushing the solution for the easiest problems, a silly error in his code will earn him 10% of the points for the easiest problem. And he will score the same as the contestant who only solved the easiest problem.
It looks interesting.Could you make weighted results of January Contest?
Yeah, I'm gonna do it tomorrow. I'll post it here.
I updated the document. Now it includes January contest too.
Document updated to include February Contest results.