annoying plat seller tells

[Twistagain]

Yeah, I'm afraid, though, that they will wise up and figure out why /role and /anon work. I think I know the answer, but might as well let them do the figuring. Unfortunately for us, the "fix" shouldn't be difficult.

[Mukkul]

Sony not only fails to prevent this easily-identifiable plat seller spam, they also spam all the servers with ads for the Sony card game by abusing the systemwide message system via GMs.

Sony not only condones spam, they participate in it.

[Drawde]

What I think was being talked about, about the people running the bots, is that the owner of the company sets up rooms with a bunch of computers. They then hire one person to watch over each room of computers. That person is the poor one, and basically just sits there until the script alerts him to something happening. Like a GM asking the botted farming group a question, or the account being banned. He then follows whatever instructions were set up for that case. That's how many of those companies work. They're set up in some other country, where the labor is cheap, and virtually no laws from the US can hit them. The most SOE can do is ban any account found doing that, but the company just gets a new one then. And do you think a corporation (SOE) is going to fork out the money necessary to get rid of them? You'd pretty much have to hire several people just to watch for that stuff, and keep them on, doing pretty much nothing but watching for stuff that hardly happens anymore, to keep the farmers out. And you can't move the watchers elsewhere, because the farmers will keep scouting out for that to happen.

[melkorr]

Or just stop them from mass-spamming tells...

Add a counter. If X player sends over Y tells in Z amount of time, mute them until A hours have passed.

I really don't see why a normal person would be sending anywhere over 100 tells in a 10 minute block. That's 10 tells a minute, which is sending 1 tell every 6 seconds. The spambots just burn through some list of names, sending tells to everyone.

And if they can have a GM on to spam LoN crap, why can't that person deal with reports of spammers?

[Kcalehc]

Or just stop them from mass-spamming tells...

Add a counter. If X player sends over Y tells in Z amount of time, mute them until A hours have passed.

Er, no. Sometimes during a raid I know some people coordinating it are getting and sending a huge number of tells. Not fun ot have them muted halfway through; this would have to be set quite high to avoid that. And how would it discount /afk messages as tells?

[melkorr]

Your raid coordinators are individually sending out 1 tell every six seconds? For 10 minutes straight? You all must be *very* coordinated. And if you somehow managed to get 100 tells in 10 minutes while AFK, what harm would it do to be muted for 5 minutes? Assuming they can't just flag AFK messages not to up the counter.

[Minta Rose]

Some (mad) raid loot coordinators hold private auctions, where everyone bids in tells. Part of the bonus of this is supposed to be a quicker bid cycle, so lengthen the amount of time Z past the length of a bid cycle. However, the plat spammers seem to change their toon hourly, so a throttle wouldn't aid much in keeping the peace.

[melkorr]

But with people sending tells to the raid coordinator, it's *incoming* tells. I'm talking about if they send *out* more than Y tells.

[Minta Rose]

Well, from the point of incoming tell spam, you get two divergent strategies:

1. Bids by tells are meant to keep the bidding spam to just one person, not the raid. In this case, coordinator sets an AFK message with the current high bid(s) and doesn't send out tells.

2. Bids by tells are meant to keep all members bidding in a vacuum, uninfluenced by others. In this case, there are just as many outgoing tells as incoming ones (/rt Overbid!, /rt Tied bid, etc.) to keep the bidders informed of their status.

[willimena]

Now I am getting spam from what looks like a website. When I hit the reply key to respond, it tells me that this player does not exist. How is this possible??

Roleplay does not seem to work in blocking these.

There is now a person that is spamming on a regular basis going by the name of gmworker. It seems to hit my trader about once and hour or so. It comes in the form of e-mails too. They are looking to charge people to level their toons...Are we now paying people to play the game for us...

[Wanyen]

Well, from the point of incoming tell spam, you get two divergent strategies:

1. Bids by tells are meant to keep the bidding spam to just one person, not the raid. In this case, coordinator sets an AFK message with the current high bid(s) and doesn't send out tells.

2. Bids by tells are meant to keep all members bidding in a vacuum, uninfluenced by others. In this case, there are just as many outgoing tells as incoming ones (/rt Overbid!, /rt Tied bid, etc.) to keep the bidders informed of their status.

In this case there is what might be seen as a two-way conversation, or "dialogue". In the case of spam, it is unsolicited, "monologue," at least until you try and reply.

Direct dialogue should generally not count against a tell/say threshold. However, the threshold would either be much higher or non-existant in "private channels" -- channels requiring membership or a password (guild, raid, fellowship, group, etc). There really isn't or hasn't been a problem with private channels being infected with spam. The threshhold should apply most importantly to "public" or unprotected "shared" channels -- built in system-based general, level range, class-based, era/zone based, and character to character tells.

They could track monologues (and in fact already do in a way through /report), but should prohibit/throttle monologues that exceed either a depth or bredth. A monologue could be simplified or generalized as a a tell not responded to -- understandably, some people don't respond to the "right" tell when participating in a conversation, but the net effect would still be the same, wouldn't it? Sort of a chat "gesture." They could track quantity of public monologues per minute destined to single characters and to any/all characters. Obviously the public monologue threshold destined to a single character should be somewhat higher than what the average monologue per minute for all characters (perhaps you are just giving instructions or guidance to someone). Furthermore, if the public monologue threshold is exceeded, look at the history in /report. Compare the similarity of the messages, and if the messages are 60-80 percent the same or greater (in terms of relevant words or character strings -- accounting for leet speak and substitution of symbols that could be read as something similar and excluding what could otherwise be simple outliers in words like 'this' 'that' 'us' 'you' 'them' 'it' that would not really change the fundamental message), then it is pretty safe to say it's spam and add more weight to the flagging of the monologues.

Tracking and throttling commands isn't something completely unknown or undevelopeed -- other slash commands like /who and searches in general (lfg searches, item searches, tradeskill combines) are subject to limits, /say /<number>, /tell, <default: when directed to any of the predecessor channels> should not be that much harder to throttle.

We don't need delays built into reuse of tells or says or other chat, as that would not result in the desired net effect. We just need to maintain or better utilize a short history list of monlogues, and if the history list fills up the teller is blocked -- probably temporarily, but I think the next step would be obvious -- get blocked through a filled history and you get a counter. After so many unknown/non-publicized counters, you are flagged. Fill in the rest...

EDIT:

One more note: Station Chat/Station Launcher: Chat. Has anyone considered that a contributory culprit?

[Jinaabu3169]

Any attempts to squelch the spam will ultimately only interupt real players. The spam is generated once, they delete the character, make a new one, and continue spamming. You can't send them replies mainly because they probably disconnect the moment after they sent the spam.

I wouldn't be surprised if they hadn't figured out how to hack the chat server directly.

And another thing, it really bothers me when people claim that 'Sony doesn't do a thing'. How exactly would you propose they prevent it? Short of taking a mercenary squad to China (or wherever) and rooting out the people, I don't see how they can do a dang thing.

What I find amazing is that they have a market that nets them enough money that the effort to continue to spam is worth it.

The only way to stop the spam is to totally eliminate the market for which they are advertising.

[Wanyen]

I am not sure that it would be as complicated or as futile as it sounds...

If average, real characters, when outliers are removed, typically send 5 tells at most a minute (once every 12 seconds -- which is pretty often, don't you think), performing a followup action if the velocity of tells exceed 10 per minute shouldn't be that hard. In some cases it is probably even more egragaious -- more than 3 tells in 3 seconds? That should get 'heavy weight' when determing who to pay attention to and take action against.

If in a minute and at or above 10 tells per minute, are the tells being responded to in large part by the people they are sent to? If only a small percentage of tells are responded to in that high tells per minute area; temporarily suspend the ability to send tells. Let them know there is far too much talking and not enough killing -- Minsc was it? At the least, "Forsooth! Methinks you are no ordinary talking chicken!" by the protagonist...that quote I could find.

If the number of tell suspensions within a given period of in-game activity (being or going offline would not help you) exceeds a secondary threshold, suspend the account.

If a threshold ratio of accounts within a subnet become subject to suspension, suspend access to the subnet, and issue pro-rated refunds for service lost to non suspended accounts.

Next, if not already in place, character creation rate per IP and per subnet should be better measured and considered. If creating more than 2 characters an hour on an established account, there might be a problem. If creating more than 3 characters for a new account per hour, there might be a problem.

Finally, Station Chat should adhere to in-game rules. It should only be supplemental in terms of chat capability; not allow circumvention.

EDIT:

The idea behind server-side tracking and blocking based on activity rates, instead of historical activity is that characters might reform. Character names might be recycled, and thus would need a new 'reputation' within the tracking/filtering system if the names by themselves were simply blacklisted. By not using client side filtering you won't unintentionally make that character unable to socially integrate.

By using traditional 'key word' filters, you will just cause *unconventional* spellings to percolate. How many ways can you convey 'platinum', block those variations, yet allow the rest of the population to continue to use that terminology in its intended context without penalty -- as long as I am not spamming?

Spamming is a 'gesture.' Where this differs from email for instance, challenging the sender is acceptable, since the sender, the medium/mode, and recipient are on a closed system, there is little reason to not 'moderate' the use of those resources and block the sender, rather than to try and attempt and filter/file at the recipient end.

[Caddarick]

I am with you!