The issue:

 

The most common mistake people make, is to ensure that their script gets the most xp an hour, sacrificing reliability, for small-term gain. STOP IT! Scripts should be made to appear human like, this means 80% of the time, lacklustre. Humans generally start out quick and paying attention, then they tend to slow down a bit, relax and chill, "have fun" so to speak. Then randomly they will regain attention, pick up and focus on what they're doing, this lasts on average for around 20-30 minutes and then they go back to lacklustre for around 40-50 minutes.

 

The potential fix:

 

Something you could do in your script, is store a few variable, such as;

 

 

So i started thinking about a way i could improve the idea then allow it into my script.

 

So i came up with this:

http://prntscr.com/5hbzi5

 

What do you guys think of the concept? Im allowing my variables to become randomize. 

Test out two versions of the script in my opinion.

One with the varying attention simulation, and one solely focused on best yield.

Across multiple accounts with attempts at controlling as many variables as possible to give most accurate results.

Doesn't matter how you do this, it will form a pattern over time and it'll be easy to tell that it's a bot

Doesn't matter how you do this, it will form a pattern over time and it'll be easy to tell that it's a bot

Yes, but the problem here is how to stretch out the time before anything repeats :3

Doesn't matter how you do this, it will form a pattern over time and it'll be easy to tell that it's a bot

 

While true this can easily be countered, if scripts are made to simulate actual human behavior as opposed to maximum efficiency they will generally be harder to detect. You will still have the guys that run a plank maker for 20 hours a day and get banned because of it but it will help the people who bot intelligently. The people who change up what they bot, and don't bot for extensive periods of time will become far harder to detect as those patterns will take significantly longer to form. To take this one step further, if one were to develop scripts where a profile could be loaded in, such that the human behavior that is simulated is unique per user then detection would decrease even further.

Sounds interesting, i think that the mouse movement is the most important part though.

I don't think it matters, but you can give it a try.

 

Yes, but the problem here is how to stretch out the time before anything repeats :3

notrly

 

 

While true this can easily be countered, if scripts are made to simulate actual human behavior as opposed to maximum efficiency they will generally be harder to detect. You will still have the guys that run a plank maker for 20 hours a day and get banned because of it but it will help the people who bot intelligently. The people who change up what they bot, and don't bot for extensive periods of time will become far harder to detect as those patterns will take significantly longer to form. To take this one step further, if one were to develop scripts where a profile could be loaded in, such that the human behavior that is simulated is unique per user then detection would decrease even further.

if such a system was unique per account then it's possible

For everybody: this class isn't really a antiban system it's more of an intelligent mouse handler. Meant to help the scripters implement an on and off switch for these different type of mouse events.

Doesn't matter how you do this, it will form a pattern over time and it'll be easy to tell that it's a bot

Key word over time. This system is meant for those who don't run their bots for over a year. It's meant for the average joe to bot like 2-3 hours.

Yes, but the problem here is how to stretch out the time before anything repeats :3

i created my own randomize timer. So the delay will always be random. The update timer too.

While true this can easily be countered, if scripts are made to simulate actual human behavior as opposed to maximum efficiency they will generally be harder to detect. You will still have the guys that run a plank maker for 20 hours a day and get banned because of it but it will help the people who bot intelligently. The people who change up what they bot, and don't bot for extensive periods of time will become far harder to detect as those patterns will take significantly longer to form. To take this one step further, if one were to develop scripts where a profile could be loaded in, such that the human behavior that is simulated is unique per user then detection would decrease even further.

I mean for those who bot for 20 plus hour shouldn't be using this type of system. It's meant for the average job. We could always add in the option of if they want the system in or not

Sounds interesting, i think that the mouse movement is the most important part though.

To bad I couldnt think of any other thing i could add in. Maybe miss clicking but idk.

I don't think it matters, but you can give it a try.

It kind of does for me. Becuse I'm my AIO crafter I have some script modes that allow hovering. But if the person runs the script for like 3 hours they will be hovering for the whole three hours. With my system they will be allowed to hover, stop, have a reaction delay, support multiclicking, and as well as having all three different mouse type turned off so you could have a normal bot.

if such a system was unique per account then it's possible

this is just the beginning stage of the class. it could be improved and it's possible to add in a profile.

To everybody: I'll probably release as open source.

Test out two versions of the script in my opinion.

One with the varying attention simulation, and one solely focused on best yield.

Across multiple accounts with attempts at controlling as many variables as possible to give most accurate results.

Can I ask you why I would do two version test.

Edited December 16, 201411 yr by josedpay

For everybody: this class isn't really a antiban system it's more of an intelligent mouse handler. Meant to help the scripters implement an on and off switch for these different type of mouse events.

Key word over time. This system is meant for those who don't run their bots for over a year. It's meant for the average joe to bot like 2-3 hours.

i created my own randomize timer. So the delay will always be random. The update timer too.

I mean for those who bot for 20 plus hour shouldn't be using this type of system. It's meant for the average job. We could always add in the option of if they want the system in or not

To bad I couldnt think of any other thing i could add in. Maybe miss clicking but idk.

It kind of does for me. Becuse I'm my AIO crafter I have some script modes that allow hovering. But if the person runs the script for like 3 hours they will be hovering for the whole three hours. With my system they will be allowed to hover, stop, have a reaction delay, support multiclicking, and as well as having all three different mouse type turned off so you could have a normal bot.

this is just the beginning stage of the class. it could be improved and it's possible to add in a profile.

To everybody: I'll probably release as open source.

Can I ask you why I would do two version test.

Sorry to say this but it won't help at all if its only meant for botting for 2-3 hours - in fact, it'll do nothing but reduce the gains

Edit: ok I'll actually say something helpful

Every task a bot does is bot like, whether it be interacting with an Npc or hovering over a skill. The less tasks a script does, the less likely it'll be detectable - it gives Jagex less information to detect you, and 'antiban' only adds on to the information which jagex can use to detect

Edited December 16, 201411 yr by dog_

inb4stolen from cory m8

http://osbot.org/forum/topic/63478-mouse-handler/#entry699439

Can I ask you why I would do two version test.

 

To observe whether it can deliver less bans.

'antiban' only adds on to the information which jagex can use to detect

I've been saying this since the beginning.

 

To observe whether it can deliver less bans.

You'd need thousands of tests to get any slightly accurate data.

Good stuff mang, keep at it. If this test fails continue to innovate! Not sure if you saw my threas before but I detailed a similar problem with scripts in the past. They're in-humanly efficient. This is because scripts nowadays execute a certain part of code as SOON as it realizes a certain condition is met (Task-Based scripts). MY proposed solution was to break a Task into two states: Needed() and doable(). When a task is needed, it will be activated very quickly, but when it's only doable it would simulate a player's reaction time.

 

To observe whether it can deliver less bans.

Yeah, it's just like the scientific method mang.

 

-First form hypothesis.

-Then conduct a test which tests your new variable. As Arctic said you need to observe the effects of both using your anti-ban and not using it in order to gain accurate results.

-Finally, collect your results and tell us what you learned.

 

 

 

As to everyone else, unless you've formed your own Hypotheses AND tested them then you're talking bullshit. At least Jose has come up with a solid hypothesis based on logical reasoning and given a possible test case.