![]() I’m not a botter, but I’m a software engineer and architect that enjoys finding bugs in public RuneScape bots and breaking them in game. ![]() When one gets flagged for botting, that fingerprint can be used to flag all other bot accounts that have similar fingerprints. If every account runs through an ML model to generate a fingerprint, then bots running the same scripts will have very similar fingerprints. The super interesting stuff comes in when you start fingerprinting. Things like rate of mouse speed, acceleration time, distribution of clicks, click repetition, accuracy, offscreen time, delay between actions, etc can all be tracked and fed into an ML model to determine if it’s human input or machine. Many websites do this to minimize bot traffic. Runelite and the base Java client both track mouse movements over the canvas, but I doubt Jagex has built a complex ML tool yet, but it’s super common nowadays. In player inputs, the distribution is not a normal distribution, and varies greatly depending on game. Some forms of financial fraud are super easy to detect due to something called the Newcomb-Benford law, which has many similarities. Fun fact: normal distribution of random inputs is actually a leading indicator of botting.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |