I'm on the phone right now and have no laptop, so sorry for the low quality post. But I wanted to say that my solution for problem C was to set a random food for each boy and then randomly change a boy's food until the second restriction was met. Surprisingly, this passed the system tests! Can anyone relate? Did anyone else have a wrong solution that passed? Weak testcases, or what happened?
That seems believable, depending on your implementation.
One solution I saw was to simply force 2i+1 and 2i+2 (for i = 0 to n-1) to be different resulting in a bipartite graph. If you implemented it in a certain way I can see your solution passing because it was accidentally like this. Otherwise, maybe the test cases are weak :)
Either way, your honesty is valuable to CF community.
This is definitely not "VERY wrong". For me, it was highly likely that such solution will pass (I thought about exactly the same), but came up with a real solution, but if I hadn't have right idea I would have coded that local search.
And why exactly is this solution (very) wrong?
maybe the "very" part is a little click-bait :)