ThoughtVNC Posted April 7, 2019 Share Posted April 7, 2019 Listen, even if mouse data had been tracked or not OSBot definitely has a higher ban rate than AHKs. I've experienced a lower ban rate using MM, but that's just my experience. Tracking mouse data in any capacity is very hard to believe, clicks per minute of a bot are very different however, Bots will make less mistakes than a human and click at consistent RPM. If anything, bots need to double click, spam click in some cases, become more delayed over time, use middle mouse button to change camera angle, etc etc. For those who pose the idea that "the API needs to be reworked", I can assure you this would not be necessary. If anything the API needs certain changes not to emulate human behaviors better but to evade ban detection. I can assure you that some scripts have a much lower ban rate than others, scripts that are very popular often have a higher ban rate; those that are outdated have a higher ban rate. If anyone wants to discuss this further with me, please PM and we can chat there. thoUghTVnC Quote Link to comment Share on other sites More sharing options...
ProjectPact Posted April 7, 2019 Share Posted April 7, 2019 10 hours ago, asdttt said: You said, and I quote, "I would have to discuss with the others about changing the mouse movement". You never confirmed whether the mouse movement would be improved..? I figured you and the other devs simply came to the conclusion that the mouse movement was fine, because.. What else was I suppose to believe...? Maybe your testing is outdated. I couldn't find a single miner bot on this entire form, paid or free, that bypassed 3 hours of mining. I had to make my own script to bypass, and only until I stopped using some of OSBot's API was I able to successfully bypass and still to this day I've yet to be banned. Even agility training is bypassing now, which is a very high banrate due to the amount of mouse movement required in most courses. It's got a lot of human-like behavior too like not being consistent, taking tiny breaks - like emulating a human moving the mouse off screen to click something in another window, ect. And yeah, it doesn't account for the majority of their detection's, but it still can and will lead to a ban - which in my case, was the only reason I was actually being banned. That's awkward.... 1 Quote Link to comment Share on other sites More sharing options...
Night Posted April 7, 2019 Share Posted April 7, 2019 11 minutes ago, ProjectPact said: That's awkward.... How dare you suggest it's his testing method which is flawed? 1 Quote Link to comment Share on other sites More sharing options...
Protoprize Posted April 7, 2019 Share Posted April 7, 2019 41 minutes ago, Night said: How dare you suggest it's his testing method which is flawed? How can you suggest that someone else is suggesting that their testing method is flawed? Quote Link to comment Share on other sites More sharing options...
asdttt Posted April 7, 2019 Author Share Posted April 7, 2019 (edited) Instead of assuming everything I'm saying is false, test for yourself I suppose. Does OSBot produce consistent moue patterns? Yes Can you scew these mouse patterns using OSBot API? Yes, although I'd personally recommend you just make yourself your own random mouse mover which will hide the OSBot consistent flaws and use OSBot to still click entities. It doesn't need to be advanced, hell you could even record your own small movements and apply it. Maybe there's a mouse mover utility within OSBot that doesn't produce these same flaws that I missed? Moving the mouse outside the screen DOES NOT produce patterns. The best way to see whether you're being detected by mouse is to check yourself whether patterns are being consistently repeated. Does Jagex detect stealth injection? No (I scavenged through the entire RS sourcecode) Does Jagex have code that can detect OSBot? Yes, although the server would need to manually send a check-class packet on one of OSBot's classes loaded in the VM instance - which they don't appear to ever do. But they have the ability to, so keep that in mind. Mirror mode, assuming it doesn't load any classes inside the RS VM would not be detectable by these means. Although I highly doubt they detect bots through this, and if they do, it'd probably be incredibly rare in order to preserve this check otherwise it'd be easy as fuck to bypass it. https://github.com/zeruth/runescape-client/blob/master/src/Client.java#L4315 Are mouse clicks detected from OSBot? No, OSBot does a fantastic job with pretty much all button related actions. Here's proof they use mouse time/location and ship it to the server: https://github.com/zeruth/runescape-client/blob/master/src/Client.java#L3371 Is moving the screen detected from OSBot? No, OSBot once against does a fantastic job emulating a normal user by moving the screen with keys. Just botted 15 hours of mining on a fresh level 3 straight through the night. 0 ban. Don't give up so easily and accept your fate of a daily, or weekly ban. It's not a lost fight, it's just the people bypassing want to continue to bypass so they tend to hide their bypass methods from the public lol. There's no single method in bypassing either, so don't assume messing with mouse movement will suddenly make you bypass. I personally recommend you record yourself playing for 20 minutes, then apply your own timings, AFK-ness, reaction time, and mouse pattern (such as moving the mouse outside the window, or randomly moving it around) to your script. And keep in mind, none of this applies to manual bans. It's very possible a moderator will still find your botting ass and manual ban. How exactly they differ a normal player from a botting player is unknown. I doubt they just walk up to you and ask if your a bot, if no response, ban. That would be very unprofessional and would lead to the banning of foreign people who don't understand English lmao *Disclaimer: These are my own tested findings. Whether they're accurate or not is another story. Edited April 7, 2019 by asdttt Quote Link to comment Share on other sites More sharing options...
asdttt Posted April 7, 2019 Author Share Posted April 7, 2019 1 hour ago, Malcolm said: Patrick is right. I probably didn't mention this here earlier but I have mentioned it in other posts. Nobody knows how Jagex detects bots except for the employees at Jagex. Everybody could debate over it but its all speculation of what actually factors into Jagex's bot detection system. What I am certain about is that Jagex would never make their detection methods accessible to the public, especially their code for it and that you'll never know how they detect bots unless you either: Work at Jagex. A Jagex employee tells you. Both of these are very unlikely scenarios so we are left here just guessing. Well let's start with, we know what jagex takes from the client, and we know what Jagex sends to the client. We know that mouse movement is sent, so why not fix this flawed mouse movement. We know that clicks are sent, so why not have human clicks (Which we do, thx OSBot). Yeah sure there's guessing into what they do with the data collected, but there isn't any guessing as to what data they're collecting. And IMO that's more then enough. Quote Link to comment Share on other sites More sharing options...
asdttt Posted April 7, 2019 Author Share Posted April 7, 2019 4 hours ago, Night said: How dare you suggest it's his testing method which is flawed? ?????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????? I provided the code.. It's super basic math grabbing the differences between mouse movement on a 50MS tickrate. Are you saying java's sqrt function is flawed? Yikes... Quote Link to comment Share on other sites More sharing options...
Juggles Posted April 7, 2019 Share Posted April 7, 2019 So are you suggesting to add random mouse movements before/after interactions? mouse.moveRandom(); npc.intereact(); mouse.moveRandomFromCurrentMousePosition(); mouse.offScreen(); Is this what you are proposing? Quote Link to comment Share on other sites More sharing options...
asdttt Posted April 7, 2019 Author Share Posted April 7, 2019 (edited) 19 minutes ago, Juggles said: So are you suggesting to add random mouse movements before/after interactions? mouse.moveRandom(); npc.intereact(); mouse.moveRandomFromCurrentMousePosition(); mouse.offScreen(); Is this what you are proposing? My point is, for instance, when you click a moss giant legit. What do you do? I personally move my mouse off screen and AFk kill the giant. Introduce human behavior into your scripts without accidentally creating a pattern (In terms of hard data) is what I'm saying. Not going to go into anymore detail because I've said more then enough. Mouse movement's detected, now it's everyone's jobs to adapt their scripts if they're being banned, to scew mouse data heuristics. Not my job to tell you how to do it Edited April 7, 2019 by asdttt Quote Link to comment Share on other sites More sharing options...
Koschei Posted April 7, 2019 Share Posted April 7, 2019 Some people refer to me as the Chad of Runescape botting. What do I do to get so swole you may ask, it's simple. I hover over my woodcutting skill no matter what I do, all these people are trying complex calculations to figure out the secret, but they've all left that one method, which has flawed all their research. 1 Quote Link to comment Share on other sites More sharing options...
asdttt Posted April 7, 2019 Author Share Posted April 7, 2019 21 minutes ago, Malcolm said: Ok I'll bite and go off the basis that mouse movement is sent and analyzed. Now you are left guessing how the mouse is analyzed, which aspect of the mouse is analyzed and you're playing cat and mouse on how to "fix it". Dude you literally called out a developer of osbot suggesting that his testing is outdated and flawed. You've also said that you have produced evidence of a fix where really you haven't because nobody knows what the problem really is except for the employees at Jagex. So unless you're actually Mod Weath you really don't know how to solve any sort of detection issues, you like us, are guessing. Why do you think they collect mouse data...? To waste their bandwidth for fun? This is really annoying me. Please stop being ignorant lol Which part of the mouse is analyzed? Huh? Did you not even read my post? Wtf? Uh yeah do you know when the mouse movement was first introduced...? It's absolutely outdated and his tests were from like 2 years ago wtf? Evidence? I'm not banned after changing my mouse movement. There you go.. Here let me just go to jagex's datacenter and yoink one of their servers so we can view their exact code /s. Enough wasting my time. You're in denial, not my problem. OSBot's detected. Get over it. They can fix it and even the current developer admitted that the mouse is flawed. Quote Link to comment Share on other sites More sharing options...
Juggles Posted April 7, 2019 Share Posted April 7, 2019 11 minutes ago, asdttt said: My point is, for instance, when you click a moss giant legit. What do you do? I personally move my mouse off screen and AFk kill the giant. Introduce human behavior into your scripts without accidentally creating a pattern (In terms of hard data) is what I'm saying. Not going to go into anymore detail because I've said more then enough. Mouse movement's detected, now it's everyone's jobs to adapt their scripts if they're being banned, to scew mouse data heuristics. Not my job to tell you how to do it There's already a mouse.moveOutsideScreen(); method on OSBot. I use it in all my scripts to act as if I am AFK watching Netflix 1 Quote Link to comment Share on other sites More sharing options...
asdttt Posted April 7, 2019 Author Share Posted April 7, 2019 1 minute ago, Juggles said: There's already a mouse.moveOutsideScreen(); method on OSBot. I use it in all my scripts to act as if I am AFK watching Netflix Yup and like I said, it doesn't contain any flaws from my testing. That's why I recommended it from any other "randomization" from inside the Mouse API. Still though, they compare large/small/medium movements against different checks, so you still should probably include some small/medium randomization that doesn't include OSBot's API (Assuming you don't have an API method that doesn't produce the same "endpoint" flaw I've discussed). Even keeping the mouse perfectly still is better then moving it randomly around with OSBot's mouse API. Just now, Malcolm said: @asdttt I don't think you get it. You're analyzing the osbot mouse and telling the scripters that they make a shitty product and they should do all these things which I know you haven't mentioned it but I will for you, to make the bot undetectable. I'm not responding to you. You wont even bother reading anything I've said or testing yourself. Mouse is flawed, get over it 1 Quote Link to comment Share on other sites More sharing options...
Molly Posted April 7, 2019 Share Posted April 7, 2019 (edited) I think the important thing to take away here is that they do at least send the mouse data to their servers, it doesn't look very human, and if they wanted to they could use this as one of many metrics to detect bots. That alone should be enough to encourage a change, whether it be in the way OSBot itself handles mouse movements or by us scripters to handle movement ourselves. Edited April 7, 2019 by Molly 5 Quote Link to comment Share on other sites More sharing options...
asdttt Posted April 7, 2019 Author Share Posted April 7, 2019 1 minute ago, Malcolm said: Are you dense? I haven't denied this at any point in time LOL Then why the hell are you arguing with me LMAO... Quote Link to comment Share on other sites More sharing options...