I've decided to undertake a bit of a data study project
It is my opinion that chat engagement, in terms of unique users and unique members seen per stream, is a more reliable measure of viewer engagement than CCV and likes. It should also be immune to the recent cull. We all know that CCV gets botted to hell, and nobody can ever agree on who is or isn't botted on any given day. Some might even argue that likes are being bought for some individuals. While some of you in response to my study might argue that chat engagement can be bought (I'm sure it can be), I would argue that we haven't seen evidence of this, and furthermore, unique member count is far, far less likely to be manipulated as it is far less likely to be noticed by anyone but autists in the first place.
My source for Holos, exHolos, and Nijis will be Hololyzer, and for Phase I will use vrabi. I'm definitely going to do Saba, Nimi, Clio, Biboo, and Mori, at the very least. If anyone has ideas for who else I should do, I'm open to suggestions. I'd like to have a NijiEN in there for fun but since I don't know anything about them I don't know who to choose. For now I'm limiting my scope to EN only and 2025; I have no plans to go beyond EN, but I would like to go further back in time to catch Fauna and Gura comparison data. I will not be including CCV or likes in my graphs; it is not my intent to attempt to prove or disprove that anyone in particular is being botted in these regards. I don't care if others use my data to that end, though, it's just not my focus. I also recognize that raids will influence/skew the unique user numbers for streamers who are frequently raided, I simply don't particularly see this as an issue; that's just the privilege of having a box to engage with, and I'd expect unique member count to be far less affected. Not to mention that bot raids won't be reflected in chat user counts at all.
