From: msemtex
Hi guys,
I had been running a private script for OSRS on one account for the past three weeks with no issues. Within 4 days of using the script on a second account at the same time, they were both banned.
My main, which used the script a while ago, still remains unbanned and that one has membership.
Has anyone else been banned today? Also, would getting members be a form of antiban? I have heard that bots which are members only get 14 day bans.
Running the same script on several accounts (almost simultaneously) is most likely a horrible idea. This makes detecting bots extremely simple in a number of cases.
Botwatch does multiple passes to detect bots (or so they say).
Such algorithms usually works in a way alà this: For every stage a metric is computed based on some data gather from the character, if the result of this check goes over some barrier the character is flagged for the next stage, at each stage the complexity of the tests will (normally) increase.. Once the last stage is reached there will likely be some very computational expensive tests.. if this (last) test yields true you will be banned.
So for first stage even the slightest irregularity can send you to the next stage (
players that doesn't talk, players behind IP where bots have recently been detected, extreme activity, sudden increase in activity, doing the same ting over and over again, etc).
At the next stage somewhat more involving tests can be done: Say they
compare the basic data gathered in the first stage
to other accounts that has been "noticed" and to stored bot-profiles. They can also compare some basic characteristics both to other (noted) accounts (and stored bot-profiles) to see if there is some common behavior.
All in all they will do a series of (simple) comparisons to both other flagged accounts, and bot-profiles (accounts that has prev been banned), there might also be some tests to see if the char is repeating the exact same task(s) (without ever/or very rarely deriving from the pattern), such a test can also be done in the first stage, but probably not as involved.. Once that stage is over you will pass over to a new stage that does even more, and even more involved checks etc.. Anyways I've gotten to my point now.
So the point is: Your accounts was most likely sent up a few stages.. At this point the two accounts running the same script are compared to each other (as well as some other tests) and since they are in fact running the same script the behaviour of the two accounts will be nearly identical in all tests, so botwatch will quickly take both accounts to a higher/last stage in bot-evaluation process where some bad-ass tests are executed to verify if the character(s) are if fact bots.
Keep in mind that much of this is purely speculation, the actual tests they do at each stage is unknown, the number of stages are also unknown… And this is outside my area of expertise, and I am very tired atm.
The best antiban is to write the whole script so that it actually LOOKS human and ACTS human (to a degree). Some Wait(<random number>), a few hardcoded task executed every once in a while (usually after doing some exact task, so it's predictable) simply does NOT cut it, at least not if you plan on running several accounts using the same script.
