The problem of cheating the game makes a good game. Many of the games become boring game, because the rogue can not handle it. The game has a lot of players like CS: GO is on this issue for a long time, but Valve also revealed artificial intelligence for managing players cheated in the latter half of the name VACnet
VACnet a network deep learning that capture behavior. cheating game By capturing bot for specific behaviors that aim (aim bot) makes it possible to limit the data for input to have Deep learning.
Data input is the movement of VACnet aiming 0.5 seconds before the shoot. That x y movement, however, and the effect of the shot is. accurate, miss, or headshot , the data is a series 140 series. To hold that one player is cheating or not.
When VACnet caught cheating. The system will send the information to investigators consider the following. As well as reports from other players. However, flat rate of reporting by คนนั้นอยู่ that 15-30% only. As a flat rate from a report by the VACnet in 80-95% .
Now VACnet run on the server, 3,456 core include 8TB of Ram and is expected to add about processing power is two times in the future.