In addition to Splat, Activision's Team RICOCHET is using Machine Learning to detect known cheat behaviors and identify abnormal actions in both client and server data. The team reviews around 700 gameplay clips daily to combat cheating, and with the introduction of the 'Replay Investigation Tool,' the Machine Learning model can autonomously assess up to 1,000 clips per day. Despite these efforts, Activision acknowledges that the battle against cheaters is still a work in progress.
Key takeaways:
- Activision is introducing a new feature called 'Splat' to its Ricochet anti-cheat system in Call of Duty: Warzone. This feature causes detected cheaters to freefall to the ground at the beginning of a match, rendering their game unplayable.
- The 'Splat' feature is designed to only activate for verified cheaters and won't accidentally engage for other players. It remains impervious to player reports, ensuring it won't be triggered unintentionally.
- Team RICOCHET is leveraging Machine Learning to detect known cheat behaviors and identify and challenge abnormal actions in both client and server data. This aids human operators in discerning unusual or malicious behavior amidst the vast volume of normal data and legitimate player activities.
- Activision has acknowledged that the battle against cheaters is still a work in progress, despite the introduction of innovative features like 'Splat' and the use of advanced technologies like Machine Learning.