Contextual Cues
"As we said above, suspension lengths vary according to important contextual cues, which may not always be obvious externally — but the way we evaluate those suspensions is always consistent for all users."
If these are, indeed, consistently applied contextual clues, there's no reason why these clues shouldn't be publicly posted and / or visible.
Put another way, if these rules are to be totally transparent in their application, no part of them should be hidden. If they cannot be made public, then the word / phrase / implication of "transparency" should be struck from the Community Guidelines lexicon.
Sorry to be so blunt, but the idea of enforcement being "fair" without there being public and itemized factors is a pet peeve of mine. YouTube is notoriously guilty of it, as well.
The_Stebe
shared this idea
-
The_Stebe
commented
Good points.
While I realize that it is not feasible to have 1-on-1 support for every streamer, claims that every appeal is reviewed by humans and the appeal criteria is applied consistently from human to human simply does not ring true as a realistic assertion. Appealing to nuance is great, but sometimes broad strokes that allow for variance are a more realistic method.
Furthermore, I feel that Twitch needs to get better at "tossing out" TOS bombing. From "outside" it certainly seems there's a trend wherein Twitch automatically assigns more importance / urgency to streamers / streams that get a lot of TOS reports, which I think we can all agree is a system that is fraught with potential errors.
The core of this is the oft used phrase in the support world in general. Namely, "setting expectations" for those who are supported. The issue here is that the expectations are often at best translucent and at worst opaque. Twitch needs to do better.
-
MonkeysnUncle
commented
When expanding on this bit about "Contextual Cues", Twitch mentions that they look at the nuance of the behaviour and how it could be applied maliciously. You shouldn't be looking at how it could be used maliciously but rather at whether or not the behaviour which sparked the 'nuance' inspection was malicious or not! ANYTHING could be malicious when viewed in the proper context! The entirety of Twitch's behaviour to their user base and streamer base can be viewed as malicious in certain contexts and all you need to do is look online to see that this is becoming more and more prevalent as time goes on. Twitch needs to spend less time looking at nuance and more time actually communicating with their community in ways that do not include surveys with exclusively preformed responses or one-way letters in which Twitch 'politely' dictates to us "this change is happening. Be happy with it or put up with it cause it's happening with or without you."
To be more constructive in suggesting a solution: Better communication. In ALL aspects of Twitch. Allow two way communication that MATTERS. Give customer service, not just for casual input or billing or whatever, but also on disciplinary actions and let the communication be HUMANE. Understand that Twitch is not perfect, so don't poise like you are.