Captcha to report and automagically disconnect botters

26/10/2012 10:24Posted by Andrai

Time to hire some new staff with the savings they made to actively watch the game for botters, exploters, other rule breaking (see trade chat)..

If you think of how many realms there are, then consider how many people you'd need per realm, we are talking literally hundreds, if not thousands of staff. Just to sit and observe the game, not actually helping with the large selection of other issues that Game Masters deal with. Do you really think that's a good investment?

26/10/2012 11:07Posted by Andrai

If you think of how many realms there are, then consider how many people you'd need per realm, we are talking literally hundreds, if not thousands of staff. Just to sit and observe the game, not actually helping with the large selection of other issues that Game Masters deal with. Do you really think that's a good investment?

There were never GMs per realm. You know that already.

But, I'm not sure how long you've been working at Blizz towers Takralus. But I assume you've maybe been playing the game at least long enough to remember a time where GMs would routinely "police" the game. I certainly remember seeing it on multiple occasions during BC.

Just like real world police, they cannot be everywhere all the time. But, if they were seen to be actively enforcing policies and rules by actively getting on servers it would reduce the number of people breaking them. Because, they'd be taking a chance that there MAY be someone watching.

The report system means action is often taken far too long afterwards. I also suspect that the non active stance taken now has led to the average level of abuse/policy breaking has risen, leading to a higher level before punitive action is applied by the GM team.

It's a self perpetuating problem. I remember it was never this bad, either with bots, or in game abuse. I don't even have my rose tinted glasses on!

So to be sure, I'm not talking about even one GM per realm. I am talking about a handful of people, watching trade chat on a realm. Moving along, moving around. Perhaps checking BGs from time to time. Having a presence, joining a conversation sometimes even. People seeing blue text in trade chat would almost certainly mean people would change their attitude, I'm sure of it.

Just an idea really.

Just having a GM hop on to a server to check if anything bad is taking place is simply a bad use of time, when you consider that there are tickets coming in constantly reporting—among many other things such as being unable to play the game—people actually breaking the rules. In the time it takes a person to do as you suggest above (and likely coming away having taken little action), they could have taken action against a large number of actual rule-breakers. Surely it’s better to have GMs address reported issues before logging in to randomly hang around in the hope of maybe catching someone?

When people cheat or break the rules, they’re always taking a chance, not that someone may be watching as with the old system, but that someone will use one of the really simple ways to report them. It’s not down to bad luck anymore. It’s much more a case of ‘when’ they're caught rather than ‘if’.

But why would Blizzard ban bots?
They have a active subscription and pay money.
Why would any company reduce their own income on purpose?
It doesn't make sense... Allthough I don't like so many bots either.
But for them to actualy ban most of them, and lose out on so much money
from the botter's subscription.. Would they realy do that. I don't know.

I think Nevalistis summed this up quite well: This is one of the biggest misconceptions we have, and I genuinely wish we could permanently clear it up. I'll provide a few hypothetical situations (mind you, these numbers are ENTIRELY made up).

Let's say 90% of botters were compromised accounts. This means that 90% of these botters aren't paying accounts; they're stolen accounts, which are generally fueled by stolen credit cards. These payments usually get disputed and taken back, which actually costs us money. If we're looking to make a purely fiscal observation, it makes no financial sense to let these continue (aside from the fact that we don't like compromised accounts to begin with - we want our players to be playing their own accounts safely and enjoyably).

Let's go on the other side of the fence and say 90% of these botters were otherwise legitimate players paying for their accounts, as you purport. When players bot, other players are inconvenienced by this behavior (and trust me, you guys outnumber the botters, even if you may feel it's the other way around). The inconveniences range from normal players having difficulty farming on their own to struggling to keep up with an economy that's being forcibly fluctuated via unfair advantage. When players are inconvenienced in this manner, they submit petitions.

Every petition submitted goes to a Game Master for review. A living, breathing person that is paid to provide customer service looks over it, does what's necessary for the situation (in botting cases, usually forwarding the info on to our exploitation/hacks team), and provides a response. Let's say 1-2 people are inconvenienced by a single botter (in all likelihood, we probably get many more petitions per botter than that). This would mean each botter is inconveniencing at least as many, and likely more, players that are positive to the community (the kinds of players we like and want to continue to play our game). For each botter we allow to continue botting, we potentially stand to lose more than we gain for a single subscription, just out of the sheer inconvenience it causes other players.

Even if you change those numbers around of legitimate players versus compromised accounts - we only stand to lose more if we don't take action on bots (which we do, regularly).

And that's probably the problem. If I can stand in front of a guy who's botting who has about as many achievement points as me, then the only reason why he dares to bot is because he does not feel there is a risk of losing anything by doing so.

And the worst part is that right now it's difficult to say that he isn't right.

Those who choose to believe they are above detection range far and wide, I’m sure. Possibly because they may have been getting away with it for a time. However, we investigate every single bot report that comes in. We don't usually deal with the reports immediately because the underlying issue would remain in place.

What we do is study the bots, see exactly how they do what they do, and fix it so that they don’t work anymore. Once we’ve applied the fix, then the ban waves take place. By doing it like this, we can also come up with better methods of detection, and devise systems that are more effective against bot detection and removal.

Having said that, it's an on-going fight. Botters develop new bots by working out what we detected and fixed, and it starts all over again. In the process, we’ve ruled out one way of cheating the system though. I can’t go into much more detail without providing botters with information that would help them :-/

26/10/2012 16:02Posted by Andrai

So there's nothing stopping the exploit team downloading and using them in a sandbox and learning EXACTLY how to detect it.

I'm not a coder, so I have no idea, but I think you're assuming that once we know how a bot works, we can easily stop it from working. I can imagine it may be very tricky to 'fix' a bit of code that currently makes the game run. I don't know though, as I said. I think detection is only half the battle though.