Speaking in terms of fps games I believe it to be very easy for any human to detect aimbots when reviewing gameplay footage. Wallhacks are a bit trickier to detect, and probably requires longer reviews, but I'd estimate that above average players could quite accurately detect this hack as well. This leads me to believe this should be quite a good fit for machine learning, no? It just seems like a problem that lives in that space of "It's hard to define but I sure can tell when I see it" - which ML seems to excel in?