@whotheff this is the core of the problem do, we want humans deciding what is and isn't acceptable on the network or do we want it to be an automated process using machine learning. At the moment we are doing both because the technology and models are still new but so far its caught stuff that we didn't and we haven't seen many false positives. I think in the long term it's probably better to have a system that is more automated as it scales, isn't biased, can't be sued (except maybe in America) etc. The suspensions only affects publishers that are sending high volumes of automated bot traffic to their sites. Self-miners, investors, publishers with normal amounts of bot traffic which behaves normally (googlebot for example) wouldn't be affected. I understand the concern but we are going to be rolling this out slowly and if it does a better job than we can without any potential bias then I see no reason not to use it.
Hopefully anyone with genuine traffic on a good website should now start to see their earnings increase from this update.
Hopefully anyone with genuine traffic on a good website should now start to see their earnings increase from this update.
THanks for the reply JSE. Imagine this: I post a very popular article on a blog of mine and suddenly i have 1000+ visitors more than usual.
So my JSE hit rate is increased atleast +100 times. The bot decides this is abnormal (based on the normalized traffic so far) and freezes my account. My question is, if that happens, is there going to be a manual, human check before an account is banned?
Because even a slow bot release with a long learning curve cannot predict 100% of the time a 100 fold increase in clicks.
