/vt/ - Virtual Youtubers

>>52467021
I agree with 30 days being way too short, and one year may be a touch too long. I prefer the longer one since, as you mentioned before, with short graphs you have things like Sana's graduation buff skewing results, which wouldn't be an example of actual incline. With a longer graph, while it might be harder to see the induvial impact each event has, it still gives a good idea of general treads.
>>52467007
>>52466655
>>52466406
Original graph provider here. Trendline anon is not exactly accurate. There's a clear shift in the trend starting August 22. As a clearer example of EN's graph, this is AdmiralBahroo's graph. As you can very clearly see, his general trend on an incline until a very clear shift where he reclined and remains stagnant. Because of this shift, it would be more accurate to use two separate lines. EN has a sharper decline on August 22, but the cause is unclear especially on the 30 day graph.
>Because your shit graph is literally made to flatten most of the context around said peaks and reclines
Which makes it less shit than the 30 day graph. In the short graph, it appears EN inclined to the highest they've been pre-covid. The reality is that Sana was graduating, an incline that obviously wouldn't last and doesn't speak on EN as a whole, only that something big happened. Longer graphs aren't week to freak outliers like a graduation.