Jump to content
Search In
  • More options...
Find results that contain...
Find results in...
Buy OSRS Gold

Sell OSRS Gold
Sign in to follow this  
realidadk

Logic of Jagex Anti-Bot

Recommended Posts

Posted (edited)

7 years ago I botted an almost maxed account in RS3 24/7 until I got banned after botting in a high density bot location.

my hypothesis is that Jagex anti-bot system don't care about breaks or how many quests you've done.

botting has almost everything to do with statistics of patterns.

 

what is "possible" and what is "impossible" coming from a pattern logic.

the more you approach the "impossible" spectrum you get flagged and eventually banned.

 

for example, if your bot clicks on the same spot at the same time following the same synchronization, you are getting near the "impossible" spectrum (i.e. right clicking in the same location of the screen after moving the camera to the same direction twice in 10 minutes. or logging in and logging out at the same time twice).

 

this moves you towards an "impossible" spectrum of what is possible coming from a pattern.

 

a human would always make possible patterns because those patterns never fit with each other, camera movements, clicks, etc never match with each other.

 

the moment it matches, means that an "impossible" pattern is being created and that is associated to a bot. because only bots creates patterns that matches with each other. a human does not.

 

another thing is high density bot location. the more you bot in places that are currently used for bots, the more Jagex Anti-Bot watch will make an effort to unleash your "impossible" patterns (because these places are well known by Jagex staff).

Edited by realidadk

Share this post


Link to post
Share on other sites

Nothing much to add, but I think you're absolutely right.

Probably only time quests/breaks matter are in assessing the type of ban you get. If you've got a decent amount of quests done, and Jagex thinks you're a main, they'll likely wanna keep you around as a customer and issue a 2-day ban. Goldfarmers, are probably viceversa -- they know you'll make another account (with membership), so they're more likely to perm. 

I think the key to avoiding bans is this:

Keep your botting habits similar to how a legitimate player would play --

  • Don't bot more than 8-7 hours (i.e. don't bot substantially more than the average person plays)
  • Don't bot popular botting locations/methods (like you said above)
  • Avoid botting the same skill/method for too long (ex. do legitimate players ever log in, woodcut for 8 hours, and then log off?)
  • Use scripts with low user counts, ACB2, and recent release dates -- the less data Jagex has to create patterns (as mentioned above) the less likely you are to be connected to macroing

I think most of us know the info above, but it's good to write it out and demystify it -- Jagex can be beat, we just have to be smarter :) 

Share this post


Link to post
Share on other sites
Posted (edited)

That's why the Tribot API methods uses human mouse data to randomise clicking patterns and mouse movements. There's also things like how you open the bank, the way you drop items (horizontal, vertical, random? go back on misclick or continue?), how often your mouse leaves and enters the screen.

This data, I imagine, is then coupled with playing activity to create a profile for this specific bot, which can then be used to quickly identify other accounts running the same script. I believe this to be true from my experience running scripts I have written myself; even with low-to non-existent antiban (other than the human mouse movements provided by Tribot) and very long playing hours, if I only run a few accounts I rarely get banned. If I made any of these scripts free for the public my ban rates would increase substantially, because Jagex's software would have more data points to analyse and could more effectively distinguise my bots from real players.

The key to not getting banned is not using poorly maintained, free scripts on accounts you value. Those scripts are great for throwaway accounts/farms but don't expect you account to survive when its repeating the same patterns hundreds, if not thousands, of previously banned accounts have done.

Oh, and use looking glass lol. I've been banned when playing legit through the tribot client and world hopping as a bot would to test various methods. Should probably be careful with proxy providers too, not dealt with it much myself but I've heard the cheaper ones resell proxies that have already been used to farm and are already flagged by Jagex.

Edited by NeuroGenix

Share this post


Link to post
Share on other sites

Impossible spectrum or bot qualifiers, plenty of different ways to say the same thing. Their detection is ultimately adding up many different qualifying bot behaviors, and once you reach a 'beyond a reasonable doubt' threshold, you're banned. Their "beyond a reasonable doubt" methodology is probably pretty similar to a judge & jury. Different jury members have to be convinced of different things before saying guilty. Then the judge tallies up the votes and makes a decision.

 

I imagine their detection processes include reverse script engineering where it builds a script based on your behaviors, and if it can qualify beyond a reasonable doubt, you're banned. Bot or human, there are certain things that are predictable & create patterns, but at some point you can say it's a bot or not. It's not a perfect system, but it's pretty damn good. 

  • Like 1

Share this post


Link to post
Share on other sites
Posted (edited)
3 hours ago, boatsb4hose said:

Impossible spectrum or bot qualifiers, plenty of different ways to say the same thing. Their detection is ultimately adding up many different qualifying bot behaviors, and once you reach a 'beyond a reasonable doubt' threshold, you're banned. Their "beyond a reasonable doubt" methodology is probably pretty similar to a judge & jury. Different jury members have to be convinced of different things before saying guilty. Then the judge tallies up the votes and makes a decision.

 

I imagine their detection processes include reverse script engineering where it builds a script based on your behaviors, and if it can qualify beyond a reasonable doubt, you're banned. Bot or human, there are certain things that are predictable & create patterns, but at some point you can say it's a bot or not. It's not a perfect system, but it's pretty damn good. 

the key is to create an spectrum and range of every possible click, movement of camera and choices.

this creates a chaotic and non-reproducible actions that might follow a large scale pattern (clicking a mob or item) but it doesn't fall into any possible low-scale pattern (same clicks, same movements, same velocities, same attempts, same logic for choices, without any range or spectrum of action. i.e. clicking randomly in a 20 coordenated location of a mob in a random mouse velocity between a range in a random linear movement between another range towards that click location randomly choosed).

 

in other words, creating random patterns within a range inside other random patterns within a range, inside other random patterns between a range creates the sense of a human behavior.

 

i don't think there's any judge and jury for this. Jagex AI determines if your actions fit a 0,0001% probabilities to be made by a human (checking your low-scale patterns) and you get banned once you access to that % automatically (there are no mods checking users, unless special cases).

 

i am not familiar with ABC2 10/10, but I supose it recreates some of these chaotic non-reproducible patterns.

 

ban from botting obviously comes from low-scale reproducible patterns.

Edited by realidadk

Share this post


Link to post
Share on other sites
7 hours ago, realidadk said:

the key is to create an spectrum and range of every possible click, movement of camera and choices.

this creates a chaotic and non-reproducible actions that might follow a large scale pattern (clicking a mob or item) but it doesn't fall into any possible low-scale pattern (same clicks, same movements, same velocities, same attempts, same logic for choices, without any range or spectrum of action. i.e. clicking randomly in a 20 coordenated location of a mob in a random mouse velocity between a range in a random linear movement between another range towards that click location randomly choosed).

 

in other words, creating random patterns within a range inside other random patterns within a range, inside other random patterns between a range creates the sense of a human behavior.

 

i don't think there's any judge and jury for this. Jagex AI determines if your actions fit a 0,0001% probabilities to be made by a human (checking your low-scale patterns) and you get banned once you access to that % automatically (there are no mods checking users, unless special cases).

 

i am not familiar with ABC2 10/10, but I supose it recreates some of these chaotic non-reproducible patterns.

 

ban from botting obviously comes from low-scale reproducible patterns.

whatever you believe in man :)

Share this post


Link to post
Share on other sites
Posted (edited)
1 hour ago, boatsb4hose said:

whatever you believe in man :)

my proposition is that we should leave beliefs aside (such as breaking habits or doing quests),

 

and try to approach a logic or an epistemological reasoning behind Jagex AI

Edited by realidadk

Share this post


Link to post
Share on other sites
Posted (edited)
20 hours ago, realidadk said:

7 years ago I botted an almost maxed account in RS3 24/7 until I got banned after botting in a high density bot location.

my hypothesis is that Jagex anti-bot system don't care about breaks or how many quests you've done.

botting has almost everything to do with statistics of patterns.

 

what is "possible" and what is "impossible" coming from a pattern logic.

the more you approach the "impossible" spectrum you get flagged and eventually banned.

 

for example, if your bot clicks on the same spot at the same time following the same synchronization, you are getting near the "impossible" spectrum (i.e. right clicking in the same location of the screen after moving the camera to the same direction twice in 10 minutes. or logging in and logging out at the same time twice).

 

this moves you towards an "impossible" spectrum of what is possible coming from a pattern.

 

a human would always make possible patterns because those patterns never fit with each other, camera movements, clicks, etc never match with each other.

 

the moment it matches, means that an "impossible" pattern is being created and that is associated to a bot. because only bots creates patterns that matches with each other. a human does not.

 

another thing is high density bot location. the more you bot in places that are currently used for bots, the more Jagex Anti-Bot watch will make an effort to unleash your "impossible" patterns (because these places are well known by Jagex staff).

The impossible conditions you've described are, first of all, very possible for a human to do occasionally. If you don't click the same pixel sometimes twice in 10 minutes, then you are a robot.

Second of all, Tribot and all of its premium scripts do not contain any code that would ever make it deliberately do this. All camera movement, mouse movement, and mouse clicks use pseudo-random algorithms to generate random values along a distribution curve we believe adequately matches human data. Each distribution curve is slightly different per account. Every account botted with tribot uses different values to determine these actions, meaning that no two bots will have the same patterns, even over long amounts of time.

 

This is because even human patterns can be botlike. Just because your bot does humanly possible things, it will still look like a bot if it shows the same patterns as 15,000 other accounts. 

 

9 hours ago, realidadk said:

the key is to create an spectrum and range of every possible click, movement of camera and choices.

this creates a chaotic and non-reproducible actions that might follow a large scale pattern (clicking a mob or item) but it doesn't fall into any possible low-scale pattern (same clicks, same movements, same velocities, same attempts, same logic for choices, without any range or spectrum of action. i.e. clicking randomly in a 20 coordenated location of a mob in a random mouse velocity between a range in a random linear movement between another range towards that click location randomly choosed).

 

in other words, creating random patterns within a range inside other random patterns within a range, inside other random patterns between a range creates the sense of a human behavior.

No.

Humans aren't random. In fact, they are very predictable. The more random the clicks, the less humanlike they are.

 

For example, when it comes to mouse movement, humans are all pretty similar in the way they do it. This was theorized in 1954 by Paul Fitts, and then later proven on computers by Stuart Card, William English, and Betty Burr in 1978. Since then, plenty more studies have gone on to confirm that Fitts Law remains valid, and studies as new as 2012 have used this law for further research. If you read the math behind it, you'll see that human mouse movement is not random: https://en.wikipedia.org/wiki/Fitts's_law.

In fact, humans are so predictable that there is an entire career path dedicated to making software better at adapting to human patterns. It's called UX Design. And it pays very well.

 

Reaction times are similar. Take a look at this study: https://www.frontiersin.org/articles/10.3389/fnhum.2015.00131/full

That study looks at reaction times to computer stimuli as a function of age. As you can see, we can predict rough reaction times to certain stimuli based on the age of the person. And even in each age group, it's not random. Take a look at figure 2. Even at each age, the reaction times tend to flow very close together at a bit below center. If we were to model it as a distribution curve, most of these would be considered Inverse Gaussian Distributions (Visualization).

 

All of your ideas are about 13 years behind where we are at now in botting. Try to keep up.

Edited by wastedbro
  • Thanks 1

Share this post


Link to post
Share on other sites
2 hours ago, wastedbro said:

The impossible conditions you've described are, first of all, very possible for a human to do occasionally. If you don't click the same pixel sometimes twice in 10 minutes, then you are a robot.

Second of all, Tribot and all of its premium scripts do not contain any code that would ever make it deliberately do this. All camera movement, mouse movement, and mouse clicks use pseudo-random algorithms to generate random values along a distribution curve we believe adequately matches human data. Each distribution curve is slightly different per account. Every account botted with tribot uses different values to determine these actions, meaning that no two bots will have the same patterns, even over long amounts of time.

 

This is because even human patterns can be botlike. Just because your bot does humanly possible things, it will still look like a bot if it shows the same patterns as 15,000 other accounts. 

 

No.

Humans aren't random. In fact, they are very predictable. The more random the clicks, the less humanlike they are.

 

For example, when it comes to mouse movement, humans are all pretty similar in the way they do it. This was theorized in 1954 by Paul Fitts, and then later proven on computers by Stuart Card, William English, and Betty Burr in 1978. Since then, plenty more studies have gone on to confirm that Fitts Law remains valid, and studies as new as 2012 have used this law for further research. If you read the math behind it, you'll see that human mouse movement is not random: https://en.wikipedia.org/wiki/Fitts's_law.

In fact, humans are so predictable that there is an entire career path dedicated to making software better at adapting to human patterns. It's called UX Design. And it pays very well.

 

Reaction times are similar. Take a look at this study: https://www.frontiersin.org/articles/10.3389/fnhum.2015.00131/full

That study looks at reaction times to computer stimuli as a function of age. As you can see, we can predict rough reaction times to certain stimuli based on the age of the person. And even in each age group, it's not random. Take a look at figure 2. Even at each age, the reaction times tend to flow very close together at a bit below center. If we were to model it as a distribution curve, most of these would be considered Inverse Gaussian Distributions (Visualization).

 

All of your ideas are about 13 years behind where we are at now in botting. Try to keep up.

This was beautiful to read, as an actuary, this made me tingle.

Share this post


Link to post
Share on other sites
4 hours ago, wastedbro said:

The impossible conditions you've described are, first of all, very possible for a human to do occasionally. If you don't click the same pixel sometimes twice in 10 minutes, then you are a robot.

Second of all, Tribot and all of its premium scripts do not contain any code that would ever make it deliberately do this. All camera movement, mouse movement, and mouse clicks use pseudo-random algorithms to generate random values along a distribution curve we believe adequately matches human data. Each distribution curve is slightly different per account. Every account botted with tribot uses different values to determine these actions, meaning that no two bots will have the same patterns, even over long amounts of time.

 

This is because even human patterns can be botlike. Just because your bot does humanly possible things, it will still look like a bot if it shows the same patterns as 15,000 other accounts. 

 

No.

Humans aren't random. In fact, they are very predictable. The more random the clicks, the less humanlike they are.

 

For example, when it comes to mouse movement, humans are all pretty similar in the way they do it. This was theorized in 1954 by Paul Fitts, and then later proven on computers by Stuart Card, William English, and Betty Burr in 1978. Since then, plenty more studies have gone on to confirm that Fitts Law remains valid, and studies as new as 2012 have used this law for further research. If you read the math behind it, you'll see that human mouse movement is not random: https://en.wikipedia.org/wiki/Fitts's_law.

In fact, humans are so predictable that there is an entire career path dedicated to making software better at adapting to human patterns. It's called UX Design. And it pays very well.

 

Reaction times are similar. Take a look at this study: https://www.frontiersin.org/articles/10.3389/fnhum.2015.00131/full

That study looks at reaction times to computer stimuli as a function of age. As you can see, we can predict rough reaction times to certain stimuli based on the age of the person. And even in each age group, it's not random. Take a look at figure 2. Even at each age, the reaction times tend to flow very close together at a bit below center. If we were to model it as a distribution curve, most of these would be considered Inverse Gaussian Distributions (Visualization).

 

All of your ideas are about 13 years behind where we are at now in botting. Try to keep up.

I believe you intentionally skipped the "range" and "spectrum" part of my analysis. absolute randomness it is impossible to recreate without strings attached to it (because it creates infinite variables and probabilities that cannot be reproducible).

 

that is why every pattern must be random WITHIN a range/spectrum in which that randomness it is executed.

 

obviously this is purely theoretical because your positivistic determinism was proven to be flawed within AI because researches and scientists do not have a clue about how consciousness works to mimic it within an AI system.

 

this happens because consciousness do not follow any predeterminated or reproducible function, and researchers contradict each other in any empirical evidence trying to disprof it. mostly because positive empiricism do not have the theoretical tools to approach such a complex phenomena that quantum mechanics or theoretical physicists do (i.e. in quantum computers a number can be overlapsed, same happens with an electron behaving as a wave and particle at the same time in quantum physics).

Share this post


Link to post
Share on other sites
17 minutes ago, realidadk said:

I believe you intentionally skipped the "range" and "spectrum" part of my analysis. absolute randomness it is impossible to recreate without strings attached to it (because it creates infinite variables and probabilities that cannot be reproducible).

 

that is why every pattern must be random WITHIN a range/spectrum in which that randomness it is executed.

 

obviously this is purely theoretical because your positivistic determinism was proven to be flawed within AI because researches and scientists do not have a clue about how consciousness works to mimic it within an AI system.

 

this happens because consciousness do not follow any predeterminated or reproducible function, and researchers contradict each other in any empirical evidence trying to disprof it. mostly because positive empiricism do not have the theoretical tools to approach such a complex phenomena that quantum mechanics or theoretical physicists do (i.e. in quantum computers a number can be overlapsed, same happens with an electron behaving as a wave and particle at the same time in quantum physics).

Yikes lmao

  • Haha 1

Share this post


Link to post
Share on other sites
16 minutes ago, realidadk said:

yeah. try to keep up.

You're quite far off base mate. As you said you botted 7 years ago, and you information is outdated. Try botting to max now and I think you'll have a very different experience. @wastedbro is right with what he has said, humans are predictable in their movements - there are trends in clicking patterns and people will have a preferred style of interaction etc. By making these completely random will only increase suspicion of an account. I think you should look more into human behaviour to understand better the patterns which occur. 

Share this post


Link to post
Share on other sites
2 minutes ago, flamekiller999 said:

You're quite far off base mate. As you said you botted 7 years ago, and you information is outdated. Try botting to max now and I think you'll have a very different experience. @wastedbro is right with what he has said, humans are predictable in their movements - there are trends in clicking patterns and people will have a preferred style of interaction etc. By making these completely random will only increase suspicion of an account. I think you should look more into human behaviour to understand better the patterns which occur. 

main problem between me and him is what we mean by "predictable movements".

 

let's suppose an OSRS screen has 480.000 pixels (800x600).

 

the probability one has to click twice in the same pixel after a mouse drift is almost impossible (unless there is no mouse drift and you are clicking twice in the same spot)

 

when these "impossible" patterns add and match with each other, you are flagged.

 

it is not about "predictable movements" in what you do (i.e. killing same mobs for hours in the same location, using the same items and food) but what is the randomness between these predictable movements (not clicking the same 1/480000 pixel after mouse drifts, not moving the camera the exact same ammount of time, etc)

Share this post


Link to post
Share on other sites
Posted (edited)
11 minutes ago, realidadk said:

main problem between me and him is what we mean by "predictable movements".

 

let's suppose an OSRS screen has 480.000 pixels (800x600).

 

the probability one has to click twice in the same pixel after a mouse drift is almost impossible (unless there is no mouse drift and you are clicking twice in the same spot)

 

when these "impossible" patterns add and match with each other, you are flagged.

 

it is not about "predictable movements" in what you do (i.e. killing same mobs for hours in the same location, using the same items and food) but what is the randomness between these predictable movements (not clicking the same 1/480000 pixel after mouse drifts, not moving the camera the exact same ammount of time, etc)

No, the difference is that I'm working from years of research, development, and trial and error, while you're spouting what sounds like a 14 year old's interpretation of a magazine about artificial intelligence that he skimmed once.

 

PS: Your logic right there only makes sense if playing Runescape meant literally randomly clicking an 800x600 square. Since very obviously it's not, you're logic is thoroughly stupid. It's so nonsensical that I feel like an idiot trying to explain it because I'm still unsure if you're trolling or not.

Edited by wastedbro
  • Haha 2

Share this post


Link to post
Share on other sites
16 minutes ago, wastedbro said:

No, the difference is that I'm working from years of research, development, and trial and error, while you're spouting what sounds like a 14 year old's interpretation of a magazine about artificial intelligence that he skimmed once.

 

PS: Your logic right there only makes sense if playing Runescape meant literally randomly clicking an 800x600 square. Since very obviously it's not, you're logic is thoroughly stupid. It's so nonsensical that I feel like an idiot trying to explain it because I'm still unsure if you're trolling or not.

❤️ u wasted

Share this post


Link to post
Share on other sites
1 hour ago, wastedbro said:

No, the difference is that I'm working from years of research, development, and trial and error, while you're spouting what sounds like a 14 year old's interpretation of a magazine about artificial intelligence that he skimmed once.

 

PS: Your logic right there only makes sense if playing Runescape meant literally randomly clicking an 800x600 square. Since very obviously it's not, you're logic is thoroughly stupid. It's so nonsensical that I feel like an idiot trying to explain it because I'm still unsure if you're trolling or not.

programming AI basically consists in meta-models of behavior.

 

so, if a human does actually click on a screen, it is a model of behavior.

 

what's your point? just to ensure that your fragile ego doesn't get touched or to have a discussion?

Share this post


Link to post
Share on other sites

The fact of the matter is Jagex will always have more data to work with than us. This is because they have the entire player base to gather data from.

Talking from experience if you're coding your scripts without factoring in human data you'll most likely get banned quick.

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Sign in to follow this  

  • Our picks

    • Over the past few months, I’ve been working diligently on a new project - TRiBot X. Everything has been written from the ground up, with all of the best practices of software engineering. Every aspect of TRiBot has been re-imagined to support three main goals: flexibility, useability, and reliability.
        • Like
      • 30 replies
    • Come give us feedback on the next version of TRiBot!
        • Thanks
        • Like
      • 74 replies
    • TRiBot is looking to improve a lot of its customer relationship management, customer on boarding process, customer experience, design elements, community engagement and pretty much everything else you can imagine when it comes to marketing.

      Our goal: To ensure that the marketing done TRULY reflects the experience and does not shine an inaccurate light on what TRiBot is lacking in.

      So I ask, what do you love about TRiBot and what do you hate about TRiBot? What does O S Bot, Rune M8, PowR Bot and Dre amBot do better? (yes I purposely didn't spell it right 😂).

      Love, 

      RileyZ
        • Like
      • 23 replies
    • Over the last three weeks, I've been working on upgrading our server infrastructure. It's finally ready and is now live!

      Why?

      Increased reliability - less server errors


      Increased availability - less downtime


      Increased security - keeping us and you secure


      Increased capacity - ability to serve you better


      Increased speed - less waiting for things to load


      Faster development - server and service updates will come faster


      What are the changes?

      Move from a single AWS EC2 instance to AWS ECS (Elastic Container Service)


      Distributed computing


      Load balancing


      Git management of server files and filesystem


      Redis caching


      How?

      AWS ECS (with 10 EC2 instances)


      AWS ElastiCache (Redis)


      AWS Load Balancing


      AWS EFS (Elastic file system)


      Please bare with us as I continue to tune the server for maximum performance. Slow loading speeds may occur temporarily. I thank everyone for their patience.

      Please post on this thread if you experience any issues other than slow loading times.
        • Like
      • 51 replies
    • This release will:

      Fix prayers and world hopper API (Thanks @JoeDezzy1 and @erickho123)


      Improve banking API (Thanks @Encoded)


      Adds methods for returning and using Java Lists, rather than arrays


      Slightly randomizes some hardcoded behaviour


      Removes sleeps from waitConditions; the efficiency saving potential is negligible in these use-cases, therefore cleaner code is preferable


      Other back-end improvements





      Note: If you are using LG, please restart both the RS client and TRiBot.
        • Sad
        • Haha
        • Thanks
        • Like
      • 90 replies
  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...