Settings and activity
6 results found
-
922 votes
We apply IP bans in Chat based on users' Moderation actions in certain cases. We don't currently plan to publicize the logic we use because we think it would make it easier for bad actors to bypass the restriction.
An error occurred while saving the comment An error occurred while saving the comment 
James
commented
If the Twitch devs won't implement this feature because banning IP addresses is hard, then I have a different solution. Instead of playing whack-a-mole by banning specific IP addresses, just ban ALL users that match a certain criteria that doesn't depend on IP address. See proposal here https://twitch.uservoice.com/forums/310201-chat-messaging/suggestions/10151265-give-channels-the-ability-to-ban-all-recently-crea
An error occurred while saving the comment James
commented
Uh, I forgot to add my reasoning to my last comment. I wanted to add that, although a temporary ban wouldn't stop a troll completely, it could be used to make it significantly harder for them to troll, at the cost of a slight(?) inconvenience to non-troll users at a certain IP address.
An error occurred while saving the comment James
commented
Rather than permanently banning an IP address from using chat, whispering people, and creating new accounts, how about using a *temporary* ban, such as a day? Anyone trying to sign up for a new account at the banned IP address would see a message displayed that goes something along the lines of
> This IP address is blocked from creating new accounts for the next X hours due to recent abuse.
Stack Exchange frequently uses rate-limiting of various activities from certain IP addresses and users,
1. http://meta.stackoverflow.com/a/259942/456814
2. http://meta.stackexchange.com/a/234518/163250and those are only *a few* of the posts their about it.
An error occurred while saving the comment James
commented
Looks like this is basically the same request as https://twitch.uservoice.com/forums/310201-chat-messaging/suggestions/7956255-ip-ban-from-channel.
-
57 votes
An error occurred while saving the comment James
commented
If you have 10 troll accounts that are 6 months old, then just ban 10 troll accounts. Done, won't have to be bothered by the same troll until X days/weeks/months later.
An error occurred while saving the comment James
commented
If this gets implemented, it would also be really helpful to streamers if statistics were provided on what their typical random viewer looks like. Like demographic data, but instead of being about human age, gender, etc, it would be about what the distribution of **account** ages are, and what the distribution of ages are for accounts that end of getting chat-banned.
Providing that "demographic" data would allow streamers to make more informed decisions about how old a new account has to be before it's allowed to chat in the streamer's channel.
An error occurred while saving the comment James
commented
If you want to give broadcasters even more fine-grained control, also give them the option of specifying how old an account has to be before being allowed to chat, like 1 week, 2 weeks, etc.
James
shared this idea
·
-
39 votes
An error occurred while saving the comment James
commented
Regarding Olle's comment,
"i am against this, not everyone wants to give out their phone number",
you already have to give Twitch your phone number if you want to use 2-factor authentication, which you should be using if you value keeping your account safe from hackers and social engineers.
An error occurred while saving the comment James
commented
If someone doesn't have a phone, then they don't make an account. How many people don't have access to a phone? I don't have the numbers, but it's probably not a very high percentage, and probably not high enough to justify *not* implementing this.
An error occurred while saving the comment James
commented
Also, rate-limit the number of times that a phone number can be used to verify a new account beyond just the X times per year.
For example, say that you can only verify new accounts 3 times per year using the same phone number. In addition to that, make it so that you can't verify a new account using the same number if that number was previously used in the last 24 hours.
So there would be 2 types of rate-limiting for account verification using a phone number:
1. The phone number cannot be used to verify a new account if it was previously used for verification in the last 24 hours.
2. The phone number can only be used to verify new accounts X times in total per year.
An error occurred while saving the comment James
commented
Also, if a troll get a chat ban from a broadcaster like Total Biscuit, for example, and then requests for his account to be deleted, and then makes a new account with the same phone number (or email) used to verify the previous account, then the **new** account should also inherit the same chat ban from the old account.
This is an anti-recidivism system that Stack Exchange/Overflow employs to great effect, which you can read about at
1. http://modnewsletter.stackexchange.com/2014/05/may-2014-newsletter/, and
2. http://meta.stackexchange.com/a/234518/163250An error occurred while saving the comment James
commented
This is another anti-troll feature to combat trolls who evade chat bans by just creating new accounts. Whenever new users sign up for an account, they have to verify their account with a phone number, much like how Gmail and Hotmail verify new accounts. Each phone number can only be used to verify an address X times per year (for example, 3 per year).
This will make it harder for trolls to constantly create new accounts to evade chat bans. You can even give broadcasters the option of preventing people who haven't verified their accounts with a phone number from chatting in their channel, and sending them whispers and messages.
James
supported this idea
·
-
20 votesJames
supported this idea
·
-
112 votesJames
shared this idea
·
-
68 votes
An error occurred while saving the comment James
commented
I can see this as only being useful to people who are frequently banned from many chat rooms. Most well-behaved users on Twitch don't receive chat bans, and in the rare instances that they do, they could probably easily figure out which 1 or 2 channels that they need to send a polite message to, requesting to be unbanned.
So basically, it's mostly only toxic users (even trolls) who would benefit from a feature like this.
So this hasn't been mentioned yet, but there are other reasons why IP bans are rather useless against dedicated trolls. People can mask their original IP address behind VPNs and proxy servers, or by using such as Tor to hide their original IP address.
IP bans are not a magic silver bullet. Sometimes they might stop a troll dead in his tracks, but it won't stop all trolls competely.