Quoted By:
>Here we can see botting targeting HoloEN, Towa, and Phase Connect. It looks like either the person running the bots isn't trying to inflate views or that YouTube's live bot detection has improved.
>the way this sub diagnoses bots is and has always been really dumb. Suspiciously dumb, even.
>This stream by NijiJP's Rena is also present on the graph you posted, only its line isn't jittery, so I guess it looks fine to you despite having a half of Towa's chatters on a stream more than 5 times as long and more than 3 times the views. Kaelix's stream is botted by the same provider as well. Convenient to choose a metric that leaves these two off.
>The community has somehow decided to diagnose bots based off of graph shape and subjective perception about what a stream "ought" to have. Both of these are controlled by the botting service.
>The wider streaming community has always used, metrics like viewer engagement and whether ccv fluctuates organically with what is happening on stream instead. If you've been following any twitch discussion about recent changes targeting viewbots, you might get some whiplash contrasting how differently they approach this topic.
