/vt/ - Virtual Youtubers

>>77268759
I did try to go with a regression to check what influences VOD views, checking to see what makes sense to compare in these graphs.

With Advent
R Square (average CCV): 0.362855403979952
R Square (max CCV): 0.353389100894621
R Square (likes): 0.777997283586724
R Square (duration): 0.000106762468149002

Before Advent
R Square (average CCV): 0.450962059298423
R Square (max CCV): 0.450058097682371
R Square (likes): 0.757685044948277
R Square (duration): 0.0007035611095

The stream length didn't really seem to be a factor, so I threw it in there just for fun as bubbles. After pruning the data and removing karaoke and ASMR, the R square was pretty high, so I thought it would be interesting to see how it's different for individual streamers (and to keep the botting in check).

A lot of the noise in the data was indeed unarchived stuff and ASMR. If we're looking at normal streams, CCV starts to account for a bit more of the overall VOD views this way. There's still about 40% of the growth of a typical video that's not explained.

The tendencies remain the same though, dudes in Japan watch streams more and VODs less in comparison to EN, basically, so a video needs a lot more CCV to perform well.