What's new
  • Visit Rebornbuddy
  • Visit Resources
  • Visit API Documentation
  • Visit Downloads
  • Visit Portal
  • Visit Panda Profiles
  • Visit LLamamMagic

Interesting Detection Related Readings from Ciggarc Forums

Status
Not open for further replies.

mephuser1000

Active Member
Joined
Jun 10, 2011
Messages
1,636
Reaction score
1
Just to speed up the process for who wants to read and act on them, since this forum is public and the Ciggarc is not.

redbulll said:
I found this article last night and thought it shed a lot of light on how bots are detected. Furthermore, it reinforces some already suspected assumptions on how blizz detects but goes into some real good detail on exact methods of bot detection. I will let all who are interested in learning to read for themselves, but some interesting highlights are as follows:

1.)Bots are detected thru repetive coordinate points as well repetitive deviations from set points...... so slight randomization makes no difference as multiple runs create repetitive "ZONES" that hit over and over even if your not hitting the exact coordinate points.
2.)Length of Profiles determines the ease in which these patterns can be detected. Shorter profiles allow the zones coordinate points to patterned quicker. Longer profiles take longer for them to determine patterns.
3.) Exact quote....."detection metrics trigger a bot alert whenever the number of average path segments reaches a certain threashold.
4.) Exact Quote....."one evasion vector for bots would be to use very long paths, which make the number of repetitions rise more slowly"
5.) Another point, used was the use of companies using public profiles and scripts to compare for pattern detection *DING DING DING*.

Anyways, I will not go on and on as I am no expert, the article is very informative and contains graphs of pattern recognition and discusses algorithms to detect it. I believe what is discussed is clearly how they detect. Bullet point #5 is enough for me to suggest keeping the Ciggarc profiles in a PRIVATE community.

Anyways, I found it very educational. Here is the link: http://www.iseclab.org/papers/botdetection-article.pdf Hope this helps or provides insight to developers.

Cheers,

Redbulll

bulzap said:
redbulll said:
5.) Another point, used was the use of companies using public profiles and scripts to compare for pattern detection *DING DING DING*.

Anyways, I will not go on and on as I am no expert, the article is very informative and contains graphs of pattern recognition and discusses algorithms to detect it. I believe what is discussed is clearly how they detect. Bullet point #5 is enough for me to suggest keeping the Ciggarc profiles in a PRIVATE community.
Even if you had to donate or apply do you really think that blizzard would be unwilling to do this if they thought it contained usefull information for them? You gotta remember they got multiply employees which only goal is to strike down on cheating and botting. So they couldn't care less if they gotta donate 10 dollars from a million budget. It will be similar to thoose who think they are safe on a big private torrent trackers but it have been shown multiply times that its possible to strike thoose. Imo the 5 bulletpoints more suggest that Ciggarc contineue to develop more versions (A and B waypoints and even more - know it takes time : ) ) since that would help against the recogniction.

redbulll said:
I agree. I think its possible for them to buy admission as you say. However, I believe they will take the easiest route to ban as many bottters as efficiently as possibly................which would most likely targeting sites where they can get profiles freely. As you say, I do not think it removes the possibility that we can get banned/detected, but I believe it reduces it. I completely agree with A and B waypoints as well.
 
what do you mean ciggarc forums are private? all you have to do is register to the forum.
 
honestly how many threads are you going to start?
agreed, not only that but look at the paper he linked, just because theres a write up on it doesn't make it true. that information is so out of date its almost hilarious. im cool with you guys making threads to talk about whatever you want, but at least have more then quotes from some other guy as your main content.
 
what do you mean ciggarc forums are private? all you have to do is register to the forum.
Yes, all you have to do to view the content on the CIGGARC forum is register :)

agreed, not only that but look at the paper he linked, just because theres a write up on it doesn't make it true. that information is so out of date its almost hilarious. im cool with you guys making threads to talk about whatever you want, but at least have more then quotes from some other guy as your main content.

I actually agree with you this time :D
 
This is silly, it's an old paper from some university students on how THEY would detect bots in games, with some theoretical tests they ran, with nothing linking it to Blizzard, Blizzard's actual bot detections, or even Diablo 3. They were likely writing a paper either hoping to be taken on by some big bot companies or purely to pass their degree/thesis/doctorate/whatever.

Hell, many of the maps Diablo 3 runs - they are such narrow corridors/areas, that their pathing mechanisms would break down and fail anyway. The few maps it MIGHT work on - Diablo is so packed with monsters that your bot goes off path loads by itself all the time. And there's no players to report you, something that paper even mentions being a bit of a "necessity" to cover the fact that bots can move randomly at times.

All that being said, it might be NICE to have a bit more randomization in the long, drawn out movements between waypoints - but with RadsAtom plugin, and profile makers having the ability to have a few different possible profiles of the same area randomly load by which point the entire map for any open areas would be covered in every which way anyway, I think even if this wasn't some random theoretical paper from some university students about how they'd like to detect bots in World of Warcraft we wouldn't need to worry ANYWAY :D

Threads like this do nothing but encourage scaremongering, and people can panic very easily when they see threads like this if they don't take the time to look into it themselves or read reasoned responses by others on the thread, so I'd really discourage threads of this kind being created in anywhere but a special "Conspiracy Corner" sub-forums, which doesn't exist, so I can only say I disapprove in general :P
 
Last edited:
This is silly, it's an old paper from some university on how THEY would detect bots in games

From like 2009 nonetheless :p

All that being said, it might be NICE to have a bit more randomization in the long, drawn out movements between waypoints - but with RadsAtom plugin, and profile makers having the ability to have a few different possible profiles of the same area randomly load by which point the entire map for any open areas would be covered in every which way

It's good to see more randomization in the runs like in radsatom & A+B profiles, if that could be mixed in with what the profile switcher does it would be even better.
 
i think we are pretty much all in agreement here. Thread Closed.
 
Status
Not open for further replies.
Back
Top