How does hacking work with the new feature of unrated registration? I have seen this question raised in the comments to the announcement several times, but there is no answer still and I think it's weird that Mike hasn't thought about it before announcing.
I was working recently on a div.2 Codeforces round (still might happen in the future) and I was told by my coordinator that pretests should be equal to systests in all the problems. From this I can conclude that it is a Codeforces policy and intentionally weak pretests do not happen. It doesn't mean that it is impossible to hack anybody, but it doesn't sound like a viable strategy. I find it weird that this information is not public, it's like I'm getting bonus information about all other Codeforces rounds by setting a round myself.
I have done exactly zero research on the matter, but my feeling is that 99% of hacks (I'm talking about during-the-round-hacking here, not open hacking of education rounds and such) happen in rare cases when authors didn't break some popular for some reason wrong solution, and every room has (15+-3) solutions with the same bug. So somebody in the room gets (1500+-300) points that have no correlation with their problem-solving skill, and even if somehow it is the top participant in the room, it still can affect the results in a random way due to rooms. It doesn't feel good to win a round like this. It feels even worse to lose a round in such a way.
I would be glad if during-the-round-hacking was removed.
