there are many moderators who are moderators in many channels.
Through my own experience, I see again and again how pishing attempts/unauthorized advertising etc are posted by users and I would like to ban these then as a precaution in all channels in which I am a moderator.
I would now like to program a tool for friendly moderators, so that they can easily perform CrossBans.
To be able to develop this with reasonable authentication and against the Twitch API, I miss in the API the possibility to make the following request to the API:
GET USER MODCHANNELS
To get a list of all channels in which the user has moderator rights.
For the critical voices that say why publish this information?
The functionality is already feasible with 3rd party software, it would just be safer for an application if it didnât have to work against 2 datasets of which only one knows the absolute truth.
If I was blind and there is something like that in the API. I would be happy about a link to the documentation.
You would be better off reporting the account rather than cross banning.
This sort of accounts are âfire and forgetâ and generally in my observations by the time youâve banned them in channel 1 they already hit channel 2/3/4/5 or donât bother trying due to their fire and forget natrue.
So this sort of defence generally doesnât work in my opinion.
Not only that, but many of the common advertising bots are actually a vast pool of compromised accounts rather than dedicated bots. Which means by banning them youâre also banning legitimate users, which includes normal users, affiliates, occasionally partners, and even at one point a staff member was compromised.
Because the number of users is quite vast (for example, the âbigfollowsâ bots regularly use between 7,000 and 16,000 unique accounts per day), even if you encounter one of them the chance of that same one being the account that goes to the other channels you mod for is unlikely, chances are it will be one of the many other thousands of bots so pre-banning in this situation will have little or no impact other than using up your rate limit for no reason, while potentially harming the channels if they ban legitimate users.
It takes time from report until twitch reaction (if they react)
So this is not the one&only solution for you. Uâr right you should also report this users, but as first solution u want to have this user outside of all communitys u moderate.
And also for normal users itâs important to know âi have to be a good boy, otherwise I got banned on more channels - maybe a lot of channelsâ
As I can see, this helps at the moment very good.
To the example with the bot accounts.
Uâr right, but this is not what my tool should do.
There are a lots of puplic bot account databases available with more than 13 billion accounts. If you like to ban this (i guess this is not a war u can win), take this databases and ban all accounts u donât like
As I said, this is not part of my tool.
If there is no way possible and no chance to get this feature, then I have to use my other way to realize it.
I guess crossban discrod servers where user post staff like â!ban username reaseonâ is a bit oldscool, but I know people they use this kind of solutionâŚ
If the account is indeed a bot spamming advertising then Twitch usually actions the report within 2 minutes (this is based on my observations of when I was reporting hundreds of accounts per day and tracking their response time and handling of the bots).
Yeah, those supposed bot lists that people post on Reddit, or some devs use in their tools, are known to have significant numbers of false positives and the people who create those lists often do nothing to actually reduce those wrongly on the list. So I would discourage anyone reading this thread from using such lists.
For simply banning individual users across multiple channels, rather than bots, Iâd suggest having some sort of text input field where a user of your app can specify the channel theyâd like to use this app on, many broadcasters would not want such tools being used by their mods so it may be best to let the user configure what channels exactly they wish to ban on.
I canât help but feel somewhat awkward about this concensus of âItâs a fight you canât winâ.
Agreed, sharing accounts over Discord is cumbersome and requires administration in terms of false positives, but until thereâs an auto-report or some kind of report-api, itâs one of the only things streamers and moderators can do to keep their channels at least a little bit safer.
Itâs not so much that this is a fight that canât be won, but rather this isnât the way to go about doing it, and the real impact on combating this can only be done by Twitch.
Thatâs just it though, multi-channel bans often rarely keep channels safer at all and can even lead to a false sense of security by building systems that actually have very little impact when bot networks have so many accounts that a spammer in one channel is unlikely to be the same one that spams on one of the other channels you moderate for.
I donât believe there should be a need for an auto-report, or an API endpoint for reporting, because for safety reasons all reports should be manually handled by both the one reporting and the one reviewing the report.
Twitch used to provide a firehose to a select few 3rd party developers which did help in the tracking and reporting of the tens/hundreds of thousands of bots but unfortunately theyâve deprecated that now, and unofficial firehoses arenât as reliable, so the best option for now is to hope Twitch themselves handle the banning of these sort of accounts themselves, and as channel moderators/broadcasters we should make the best use of automod, blocked terms, and individually banning accounts on channels as and when thereâs an issue rather than trying to mass ban across multiple channels to often little effect.
