Why is your char so hard for you to police? 1 person could literally handle all of this.
Why it isnt that simple? Because variation mostly.
So you ban those and the next would be adding more space between the chars, then you come up with trimming and the bots use other variations.
It sounds like a quick job but it is an endless back and forth.
But feel free to write an according regular expression and or list continuously
You really think it’s that easy?
Well, isnt really hard to fight… you just have to convert numbers and symbols to normal letters and check if some conditions are met. You can to allow a few warnings (you could mute that person instead of a direct ban) and you got your antispam, you just need somebody to check appeals.
Easier to limit characters below level 5 to area chat only.
Yes, you might do character transformation, add trimming to that and so forth. In the end it still isn’t a “quick fixing things”.
People are pretty creative so that one job will always be a cat and mouse game of updating the conditions that needs to be met which isnt making it easy.
Nonetheless a good idea with some muting functionality, even introducing a second job for checking
They can make it pretty easy. Players report gold spammer ( no matter how they type the site) report goes to live NW agent with the offending spam, Agent then simply closes the account. Jump on the spammers fast enough and long enough it becomes a cost expense (having to buy a new game each time). For the argument around gold spammers using hacked accounts, it is still a good solution since it warns the account that they actually have been hackked quite quickly and they can go arounf the business of getting their account back
Yes, yes I do. Parsing strings isn’t exactly hard.
Sadly there is an exploit where you can get free accounts. So all they would need to do is create a new steam account and then get a free NW account to spam with. All off it can be automated.
Amazon should have good enough algorithms to detect these and do the banning automatically for free. They are a leader in NLP tech after all. It would require just a few people to handle appeals if that. They probably just don’t care enough to bother allocating the resources.
There is a option ‘selling gold’ when you report some one.
It’s not happening much in my server. Saw it a few times since launch.
So all the reports from the servers could be checked by one guy right?
Today I saw one and he could work for a long time before it stopped…
And if they get banned.
They need to buy the game again right?
So the loose €40 or $40 a time?
Yeah, now that they’ve supposedly tracked down counterfeit keys, it should be costing them.
They should have volunteer moderators on the servers to mute them at least until they can ban them.
It’s actually really easy. This is literally textbook machine learning / deep learning solve. Super simple with modern algorithms. Amazon is literally at the forefront of this stuff too.
There should at least be a couple of GM’s checking on the servers in game. Or at least doing a routine check up every once and a while.
The easy way to solve it is for AGS to sell gold like GW2 does.
If you want it buy it, if you don’t then don’t but it sure worked in GW2.
Main thing is that when one variation of the message is banned, the sellers will find a workaround that isn’t caught by the checker. And if they try to make it too encompassing to check for any possible version, there’s a high chance that it catches people who aren’t selling gold. So either have a net with plenty of holes in it or a net that might catch unintended targets. I suppose the only easy fix is adding a human element here, to sift through the more fringe cases to double-check before banning.
Maybe it’s not happening in your server, but there is one going at all times of the day in my server.
The problem is steam family sharing lets the gold sellers evade every ban for free. Until that’s fixed every ban is useless.
Isn’t there an option to “ignore” ?
This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.