First of all, let's welcome the new update log style with 2018.
The MachineLearning has received dramatic changes which will help it check more players with even faster and greater ways.
Logs:
Shortened and improved many of the MachineLearning utils.
Connected the MachineLearning check with the KillAura violations.
Implemented value rounding to the MachineLearning check.
Implemented cancelled-hit-checking to the HitReach check.
Added more verbose information to the MachineLearning check.
Fixed a Combat utils null-pointer console error.
Fixed a Speed LiquidBounce y-port bypass.
Removed a pointless null protections from some checks.
KillAura accuracy false positive against velocity hacks.
KillAura aim-accuracy false positive against velocity hacks.
Fixed an EntityMove skeleton-horse null-pointer console error.
Fixed a SelfHit velocity protection null-pointer console error.
Fixed a Velocity combined bouncing-block false positive.
Added a new developer event named MachineLearningEvent.
Optimized the detection thread of the Velocity check.
MachineLearningEvent:
Code (Text):
@EventHandler
public void Event(MachineLearningEvent e) {
Player p = e.getPlayer();
double default = e.getDefaultPercentage(); // Returns the default legitimacy percentage.
double average = e.getAveragePercentage(); // Returns the average legitimacy percentage.
Boolean combat = e.isCombatRelated(); // Returns a boolean depending if the player is in combat.
Boolean contact = e.isContactRelated() // Returns a boolean depending if the player is starting some combat.
}
This is not Spartan's MachineLearning check but simply the basis of it. Each developer can individually use these variables differently to create a custom detection. How Spartan generates the percentages and what it does next in the original check will stay as a secret. Feel free to make free or third-platform paid additions with this event.