The Silence Penalty

[Dumbledoremd]

Greetings!

Thanks for all the feedback and ideas on how to better restrict and prevent spam messages in FINAL FANTASY XIV.



We are pleased to share that the dev. team is currently looking into a new system to address RMT advertisement spam, and we're planning to implement this in the future. Rest assured, we're making sure that this system won't inconvenience normal players.

Please continue to share your thoughts, and we will keep you posted with any future updates.

This is great news thanks for the reply rmt is still a big problem on low/med pop servers

[TaranTatsuuchi]

Or just remove the "Send Tell" option from Player Search > Right Click. Their job becomes much more difficult if they have to type names or use other methods. The reason they're so "efficient" is because they use tools to search/spam tells, making it harder to automate makes it harder for them to work.

Part of me hopes we can use the upcoming context menu report from the player search...

All those delicious rmt bots just sitting there, all displayed in a nice list for reporting...

[AutoWhit]

....or, we could just have gms sitting at every MB to mark and ban all the toons with "no names" constantly logging in and out....Limsa is ridiculous.

[MilesSaintboroguh]

....or, we could just have gms sitting at every MB to mark and ban all the toons with "no names" constantly logging in and out....Limsa is ridiculous.

And there's enough GMs to do that? On every server?

[Niwashi]

Part of me hopes we can use the upcoming context menu report from the player search...

All those delicious rmt bots just sitting there, all displayed in a nice list for reporting...

But what is it that you could report from there? That a character exists? I don't think even spam bots put spam in their search info message, and unless they do, there's nothing you can see from player search that would be reportable.

No, the chat box is the best place for it. For the spam bots, it's the messages they're sending that are reportable, so it's easiest to report those where they're at. And for farming bots, well, those really need to be spotted by SE themselves who have access to the server information.

You can report exploits if you see someone doing something that would clearly be impossible without hacking the software, like fighting or gathering from below the ground or something. But in that case, you have to write up an explanation of what they're doing that would be reportable. That explanation means it's not as well suited to an automatic right-click report feature. (Being able to right-click on a character or their overhead name to initiate a report filling in fields like their name as well as the server, time, location, etc, but leaving the description field for you to fill out yourself would still be helpful, but that's a different feature from just reporting them directly from the right click menu.)

[Amnesia]

Or just remove the "Send Tell" option from Player Search > Right Click. Their job becomes much more difficult if they have to type names or use other methods.

A program can parse the text on the screen very easily... Your solution would just make things more difficult for real players, not bots...

[Khanscott]

I would say have an option to blacklist accounts that the RMT bot s use. Unless the botters are using some Jedi voodoo I don't know about, I would assume that it takes a lot more effort making/taking an account than it takes to make a new toon.

[FoxyAreku]

Dota does silence penalties and it works pretty awesome, I'd be for it.

Limit people with the penalty to only using auto translate.

[LineageRazor]

They only keep churning out spam bots because the bots are profitable to them. As soon as they're not, they'll quit replacing them. The current system is that a spam bot can spend weeks sending out thousands upon thousands of ads before they're finally banned. That's long enough to bring in some sales, so replacing the spam bot is just the cost of advertising.

Weeks? Hardly. Your average spam-bot lasts only a few hours before getting banned, and pretty much never lives longer than a day. You're correct, though, in that this system would reduce this lifetime from hours to just minutes. I still believe that RMT have the resources to continue to proliferate their spam-bots at this enhanced pace while still remaining profitable. So, a placebo it remains.

Gather-bots and level-bots tend to live quite a bit longer. They do not directly harass players, and so players don't care as much about them, see no need to blacklist them, and under this new system will probably be much more likely to neglect reporting them. Which is a shame, as these bots are how RMT actually makes the gil that they sell. Getting these guys regularly banned would be much more harmful to RMT than scooping buckets of spam-bots out of the spam-bot ocean.

Of course, the BEST solution would be a punishment system directed against gil-BUYERS that is draconian enough that only an idiot would buy gil. RMT only advertise, because they have customers to sell to, after all... I don't see this happening, though, as it can get ugly when a company that gets a reputation for mass-bans of its own "legitimate" playerbase. :P

Or just remove the "Send Tell" option from Player Search > Right Click. Their job becomes much more difficult if they have to type names or use other methods. The reason they're so "efficient" is because they use tools to search/spam tells, making it harder to automate makes it harder for them to work.

As Amnesia noted, this would not slow spam-bots down in the least. There aren't any human players behind these bots, right-clicking names and choosing "Send Tell". That would be way too inefficient. These are fully-automated programs. They most likely use the Search menu to bring up a list of the names of everyone on the server and methodically send tells to each name, one by one. The only folks likely to use the right-click method are players. So this solution would harm RMT not at all, while forcing players to type out wretched Rogadyn names by hand.

[Niwashi]

Weeks? Hardly. Your average spam-bot lasts only a few hours before getting banned, and pretty much never lives longer than a day.

Well, the few times I've checked, every bot that spammed me lived a lot longer than that. If I forego blacklisting them, and instead make a note of their names to compare with later spam, then the same bots will continue spamming me throughout the day and the day after. A couple days at a time is as long as I've actually kept track of it that way before giving up and blacklisting them, so I can only say definitively that their normal lifespan is more than a couple days. (Including "more than" in that assessment is because of how trivially low the chances would be for every bot I encountered having been just created right at the moment I started keeping track of them, then staying active throughout that time only to be banned immediately as soon as I stopped tracking it.)

Plus, we know from the messages SE themselves put out that they only ban a few spam bots each week, nowhere near the number who are active. According to their last message about it, for instance, they banned 616 spam bots over the course of an entire week. That's less than 10 per server. At best guess, it looks as though it would take them a couple weeks or so to clear out the spam bots that are active at any given time, meaning that's about how often the bots need to be replaced. I think your "a few hours" estimate is just wishful thinking. The numbers just don't support the idea that they could be clearing them out multiple times per week, let alone multiple times per day.

You're correct, though, in that this system would reduce this lifetime from hours to just minutes. I still believe that RMT have the resources to continue to proliferate their spam-bots at this enhanced pace while still remaining profitable. So, a placebo it remains.

If you mean they'd still remain profitable overall as an RMT company, sure. But the spam bots themselves would no longer be profitable to them. In order to be worth using those spam bots, each bot has to attract enough new buyers who wouldn't otherwise be buying from them, for those buyers to add more to their net profit than the cost of the spam accounts and effort to set them up.

[I'm going to be making up all the numbers and percentages below, but they only need to be reasonable enough to illustrate the point, and I think they're close enough for that purpose:]

The vast majority of players won't buy gil. There's various reasons, whether it's because they're afraid to break the ToS, too principled to break agreements or do something that harms the game's economy, or the most common reason, that there's simply not much incentive to spend extra money you don't need to just for extra gil in a game where most playstyles provide plenty of gil in rewards to cover any costs you'll come across. So lets say around 1% of the population are potential gil-buyers.

That wouldn't mean that 1% of a spam-bot's messages bring in revenue. Gil buyers who have already looked into the RMT market are generally going to buy from wherever they'd already decided to buy from, so spamming them is either unnecessary or ineffective. The real target market for these spam messages is new potential gil buyers who can be persuaded to use the spam-bot's site for it. And persuading them generally takes a lot more than one message. It either takes inundating them with enough spam that whenever they think of gil-buying, your site is what immediately comes to mind, or else it takes messaging them at the ideal moment when they just decided they'd be interested in buying gil but haven't yet looked into where to do so, and are willing to just take the advice of the next message about it that they see. Since trying to hit a perfect moment like that is also achieved by repeated messaging, in either case a spam bot needs to send lots of messages to potential buyers before they'll get a sale out of it. So let's say it takes a hundred such messages to get a sale from them. (It's probably more than that, but a hundred is a nice round number to make the point with.)

Having 100 messages reach the 1% of players who might respond means the spam bot needs, on average, to send out 10,000 messages in order to generate a sale. But it still wouldn't be profitable yet at that point. To be more attractive to customers, RMT sites have a lot of low-cost deals, relying on volume rather than high prices to raise profits. If a spam bot only attracted one new customer and that customer spent $3 or $5 once and was never heard from again, it wouldn't cover the cost of the spam bot's account, let alone any of the other overhead costs. To achieve profitability, they have to either bring in a whole bunch of these small-fry customers, or reach one of those rare customers whose appetites for gil is insatiable and who will keep coming back to buy more and more of it, spending more on that than on the game itself.

So if we figure around 10 new customers is the tipping point for making a spam bot worth the cost of setting it up, that bot needs to remain active long enough to send over 100,000 messages before it gets silenced or banned. Currently, they last long enough to do significantly more than that, so they're a good deal for the RMT companies, who will continue making more of them whenever their current ones finally get banned. If, however, we could report a spammer as easily as blacklisting them and after 20 or so such reports the spam-bot's account was silenced, that wouldn't work. Even if a few people are a bit slow to report them, so they manage to eke out 30 messages by the time 20 of them get reported, that's still only 30 out of the 100,000 they would need in order to be worthwhile to the companies creating them. They're unlikely to even reach a potential gil-buyer, let alone convince them to make a purchase.

The RMT companies are in it for the money. They're not going to keep spending money on spam bots that don't bring in more in profits. As soon as SE fixes the system to prevent a bot getting out enough ads to be profitable, the RMT companies are going to move their advertising budget elsewhere (to web sites that won't kick their ads off after the first few page views). Getting the spam out of the game wouldn't end RMT (though it would cut it down a bit), but at least we'd be able to play in peace without that constant annoyance.

Gather-bots and level-bots ... Getting these guys regularly banned would be much more harmful to RMT

That's true, but it's also a lot harder to do. Spam is clearly identifiable at a glance. But the activities of a bot that's gathering gil aren't all that different from those of a legitimate player who's gathering gil. SE has to watch them long enough to verify if the patterns and timeframes involved make it clearly a bot. (And then they monitor longer after that to watch where the gil goes in order to identify buyers.) It takes a significant amount of time and effort that can't be streamlined away to something that can be done with a click, the way getting rid of the spam-bots can.

It makes sense to take the low-hanging fruit, the part that can readily be eliminated. That's especially true when, as you pointed out yourself, the spam is the part that's directly harassing players and which players therefore care the most about. Spam is a bigger detriment to the game than the actual RMT that it's advertising. (What's more, getting rid of the in-game spam would cut down somewhat on the RMT itself. The reason RMT companies go for this relatively expensive form of advertising is because it's effective at bringing in more sales. The corollary is that without it, there'd be fewer RMT sales.)



(Oh, man. I really didn't intend this message to grow so frickin' long. )