Valve believes that the only efficient way to curb the menace of Counter-Strike cheaters is to have a machine do it for you.
Responding to a thread on Reddit a couple of days back, the developer explained why it cannot simply hard-code the system to detect cheaters on sight. According to Valve, doing so would simply be entering an “arms race” with cheat creators.
This means that every time the anti-cheat detects an illegal third-party injection, the cheat developer simply evolves the code to fool the system. This in turn forces Valve to discover how the cheat is bypassing security measures, resulting in an endless loop of frustration and annoyance.
Taking into account the massive Counter-Strike community and the fact that millions of games are played every day, the task of accurately and timely detecting cheaters is not humanely possible.
Here is where machine-learning comes to play.
“The process of parsing, training, and classifying player data places serious demands on hardware, which means you want a machine other than the server doing the work,” said Valve. “And because you don’t know ahead of time who might be using this kind of cheat, you’d have to monitor matches as they take place, from all ten players’ perspectives.”
What Valve has achieved in creating is a new AI system that observes and learns data taken from every player in every Counter-Strike match. This also means that the company is utilizing considerable resources, such as a data-center with thousands of CPU cores in its midst, to take the battle against cheaters on a whole new level.
For now, the new AI anti-cheat system of Counter-Strike is being used to detect spinbotters. The hack gives players a 360-degree field-of-view, allowing them to head-shot anyone who approaches from behind. It is something that even Valve has struggled against. But no more.
Hopefully, Valve has a kill-switch for the AI in case it becomes too self-aware in the future.
