>>91972974There are 42000 streams in the dataset, a good deal of them being streams played in groups, these outliers are WAY off the mark.
For instance, searching the dataset here is another stream by Lisa
https://vrabi.net/video/i1v3nZwKctgIn a group all the same, similar avg, HALF the stream length and four times the number of chat comments.
>>91973300>a graph point where chat is like 5 pm when the neighboring points are 150 or whateverYou might be confusing the anons you are talking about, every post made about this subject by me is backed not by the instant in time but the average chat speed of the whole stream. Silly picking a point in the stream would be silly because lulls in a stream happens but 1k messages in total for a five hours stream that never dropped below a 4k average is definitely an outlier
>>91973309>You can easily find completely normal non-botted streamsI am so far calling these streams "outliers" because slow chat doesn't necessarily mean botted streams.
See for instance
>>91968832, it's a detected outlier but there is a simple explanation
>slow mode, subs onlyboth of which can be checked.
For that Lisa one it is not that.
