Live Streaming Moderation Best Practices for Event Organizers

Key insights

6 takeaways

01
The document states that a moderation plan should be in place before an event starts, because once it begins it is often too late to find labor, set up tools, or put policies in place.See it on page 1
02
The paper says the absence of a Code of Conduct often signals to people that a space may not be safe for them, and recommends making the code public and visible in online spaces such as chat.See it on page 1
03
Citing Twitch's 2017 research, the document says a chat rules tool requiring an "I agree" click has shown a "statistically significant reduction in timeouts and bans".See it on page 4
04
The document recommends setting AutoMod to at least level 2, which removes sexually explicit language, hate speech, and abusive language, or to level 3 for a broader filter, while warning that AutoMod cannot be an organizer's only anti-harassment tool.See it on page 5
05
The paper argues that third-party bots are more responsive than AutoMod because any moderator can change their filters, whereas AutoMod's blacklist can only be changed by one account-holder.See it on page 6
06
The document recommends that organizers consider compensating moderators, especially for larger productions from companies generating revenue, and notes that head moderators bear an especially high workload.See it on page 4

Summary

AnyKey published this 2018 best-practices white paper for event organizers on moderating live stream chat, primarily focused on Twitch. The guide covers three planning components: codes of conduct, staffing, and tools, and includes AnyKey's "Keystone Code" as an appendix. It argues that chat is an important, potentially valuable part of live streaming, and that a moderation plan should be in place before an event begins. On staffing, it recommends a moderator liaison, two-factor authentication for moderators, and considering compensation for moderators, particularly for larger productions. On tools, it describes Twitch's built-in options such as chat rules, AutoMod, slow mode, follower-only chat, and timeouts and bans. It also covers third-party bots, with moobot and Ohbot named as common examples. The document states that AutoMod cannot be an organizer's only anti-harassment tool, because specific or targeted messages can slip past it and users can evade it through letter substitution and similar methods. It also notes that chat rules are a simple first step, citing Twitch's 2017 research on a "statistically significant" reduction in timeouts and bans. The paper ends by stressing that chat is a core part of the broadcast that organizers are responsible for.