Curse

Hengis · Jul 11, 2006, 05:45 PM // 17:45

I had an idea a while ago about how chat could become self-regulating, but I totally forgot about it until I read Mr Interrupt’s thread on trying to encourage spammers to spam less etc, and as I did not want to hijack his thread I decided to make a new one here.

The idea is as follows:

In addition to the regular ignore list you have a small maybe five slot temporary ignore list.

If someone spams the Trade channel or All channel with WTS etc, or if they are rude or abusive, or generally behaving like a jerk, then you can add then to your temporary ignore list. (You would still have the option of adding them to your permanent ignore list if you so wish).

The next time that person sends a message to the All or Trade channel they would receive back a message in a distinctive colour saying “1 person is temporarily ignoring you” and their chat input box would be disabled for 1 second.

For each person who is temporarily ignoring him or her, their chat input box would be disabled for an additional second. E.g. If they are being temporarily ignored by 30 people, they would receive a message saying “30 people are temporarily ignoring you” and their chat input would be disabled for 30 seconds.

The temporary ignore would automatically cancel itself after 10 minutes.

This would give every member of the Guild Wars community some degree of control over those whose aim is to spoil the game for the rest of us. Not only would it help prevent excessive trade channel chat, but it could help with things like general abuse, racial hatred, sexist/obscene comments etc.

The person being temporarily ignored would not know who is ignoring them so would not be able to take any form of retaliatory action.

The Ignore messages would only be issued and the chat box lockout would only apply in towns and not in mission, explorable areas or in PVP.

The downside to this idea is that someone could add YOU to their temporary ignore list for no reason what so ever – However the 1 second chat lock out that this would cause would be almost insignificant. I suppose to that it could be possible for a group of people to pick on one person for no reason too, but I don’t feel it is that likely and if you are for example just standing around watching the chat, or chatting in Guild chat or whispering to other people, or if you are just not in a town, then it would have no effect on you what so ever.

This is only of course the framework for the idea and things such as timings etc. would have to be worked out properly. I also of course have no idea how difficult it would be to implement or how much extra network traffic it might cause.

I am sure that there will be both positive and negative responses to this idea, but I wanted to post it anyway just to see what kind of reaction it gets.

Thanks for your time in reading this rather long post.

Lets Get to Healing · Jul 11, 2006, 05:46 PM // 17:46

Would be so abused lol . . .

lyra_song · Jul 11, 2006, 06:53 PM // 18:53

This is a very interesting concept.

I kinda like it, however....once you add someone to the temp ignore list, they are removed from it after a certain time limit.

The potential for abuse is there...but in the end its just local or trade chat right?

It doesnt affect team/guild/alliance.

Hmn...i really think its an interesting idea and should be developed more.

Carth` · Jul 12, 2006, 12:31 PM // 12:31

I like it

inscribed · Jul 12, 2006, 04:00 PM // 16:00

I like it. Whenever I see someone selling the same stuff I am, I could just temp block him and keep him from spamming as regularly. Or when two groups are trying to find monks in a mission area, I could just block the leader of the group I'm not on, keep him from spamming 'GLF Monks" as regularly.

bamm bamm bamm · Jul 12, 2006, 06:40 PM // 18:40

I've been thinking about something similar for a while now. It's kind of cross between bayesian spam filtering and meta-moderation, like some websites use. So basically, you can label any chat line in the public channel as spam by clicking on a small icon next to it. So if you see "\/\/TS req 8 zodiac axe blah blah blah" you can click on a little button and declare it as spam. That's the end of it on you part.

Now, every so often the gw servers generate a report for an Anet staff member to look at. This report is the result of 100,000 clicks of the spam button or a week passing since the last report, whichever comes sooner. Although that kind of gives spammers a week to use a new combination of symbols, so maybe once a day? So you have something like this:

\/\/ts
\/\/tb
whisp offer
wt:s
w t s
lfg

So this is just a crappy short example. For each entry the staff member has 3 options: "spam", "undecided", "abuse". The reason for this is that context is important. So anyway, he would label the obvious ones as spam, they would get added to the spam filter and from now on if these are used in a message, the message is automatically moved to the trade channel. The filter acts as the learning dataset. The bottom one, "lfg", is abuse, possibly by a lot of people in kaineng having a laugh at 3am or an alliance looking for something to do. So you label that as abuse, and each person who clicked has their abuse integer increased by 1. If it hits, say 5, they are no longer allowed to use the spam system.

So I don't know if it would work. It probably has some huge flaw that I haven't seen yet. The biggest problem that I can see is that for the bayesian filtering to work it has to monitor ALL messages, not just spam messaging, which might be a serious load. It's not completely automated as I think you would end up having the same problems as we get occasionally with email filtering: false positives. Also, when in first use, they would get a torrent of reports, but after a while it should end up reduced to once a week (or day). I think eventually spammers would just give up. If an auction house is ever introduced, I think spam will still exist, even if the trade channel is removed, so maybe a system like this will always be needed.

So, could it work?