Now, I'm not sure if this has ever been suggested, as I've never seen a thread on it, so here it goes: Auto-Silencing It's pretty much self explanatory, what it does is it detects Bypasses and spam (see spam detector), and it will still send info to devs so they cam review it to make sure it isn't a accidental silence. Pros • Mods will not have to deal with as much "There's a bypass/spammer in WC.", on their walls. • The bypasser/Spammer will be dealt with instantly. •Will reduce feedback about silences to devs. Cons: • There possibly might be a hiccup every once in awhile, where some people are silenced even though they didn't deserve it. _____________________________________ Spam Detector Again, another obvious one, it detects spam. How it works: 1) Player is sending spam, after 3 posts they will get a message saying "Do Not Flood Chat with messages" in their newsfeed. 2) If they try to continue, they will be warned one more time. 3) if they try to continue again, they will be silenced. (note: this is detecting the spam BEFORE it is posted.) The whole point of this system is to silenced real spammers (the ones that purposely spam WC every second) before they try to send more than 3 messages of the same subject. _____________________________________ Where I got this idea: I play a game called AQWorlds, the chat system has Spam detector and a Bypass filter (it blocks out any bypasses) And it has an auto-mute system which temporarily mutes you for the amount of Time it takes for you to go "Away From Keyboard". (which I think is like 2 minutes) _____________________________________ I'm sorry if I didn't explain it very well. I just thought this would be handy. ~Dragonite
Meh.. While this is a good idea in theory, it sort of takes away about half of a moderators job, which is moderating wc. Plus there would be loopholes in the system that, granted, could be fixed as they're seen, but I'm not sure about the whole idea. Not only are the mods moderating wc, they should be talking and answering questions in wc also, which could vanish if they don't need to be in wc anyways.
Well, I mean there's always the forums. And there's a possibility that they won't be detected by the system. In AQWorlds, there are still mods, even with this system. So they wouldn't be that useless.
All those things you said are true, but they already have a very good system in place to take care of bypasses and spam, to change things might be more than it's worth. But still, it's an idea I've never heard before, so kudos to you
Try to spam wc right now, if it's the same message it will warn you before the infraction, there was even a time when it would silence you. I am not sure if it still does. I like the idea but you also are right that there are several words that may trigger it incorrectly
Oh, and this would only apply to WC. (It may reduce ad spam) So if it's spam/bypass on someones wall, a mod will need to silence them.
I suggested this years ago shortly after I came up with the idea of the silence button for mods. The real challenge behind automated stopping of people who bypass is it requires a pretty well-developed database of bypasses. Like if the word MILK was a blocked word, the system would have to know that M1LK is a bypass (easy) but also that M.ILK and MI.LK and MIL.K and M1.LK and M..ILK and M.I.L.K and so on and so on... it is easy to implement the system and it's easy to have a really rudimentary database, but to effectively detect bypassing in automated way is just really difficult to do well. As far as detecting spam, it would be best designed to do a compare of your posts, if they are within the time limit, not looking only for identical posts but looking for posts that are like 90% similar. Otherwise just changing one character at the end of your post would sneak past the detector. Also, you're never gonna stop, in an automated way, the following: Hire my allies Hire me Hire all of my allies Allies for sale Ally sale going on now Like, post that within one minute, and you're clearly spamming but an automated detector wouldn't be able to tell that unless it was set super aggressively to the point where it would have all kind of false positives. It's tricky stuff, for sure. Which is probably why it wasn't implemented when I suggested it a couple years back.