> So are very good players, very bad players, players with weird hardware issues, players who just got one in a million lucky…

They will all cluster in very different latent spaces.

You don't automatically ban anomalies, you classify them. Once you have the data and a set of known cheaters you ask the model who else looks like the known cheaters.

Online games are in a position to collect a lot of data and to also actively probe players for more specific data such as their reactions to stimuli only cheaters should see.

Valve has already tried this with VACNET if I am not mistaken. Judging by how big the cheating problem still is, they were not very successful.