Thanks @abh12345! On the selfvotes: Yes, I was looking into this some time ago, and we saw 6% direct self votes and 12.8% of all voters giving 80% or more of their rshares to 6 or less authors back then. So yeah, there are plenty of ways to obfuscate self-votes. I should probably revive this approach to check if that number decreased as well. You are of course right, the direct self-vote rate alone is not necessarily conclusive. But hey, it looks promising :)