A sybil attack (voting with a rotating set of accounts) would be the easiest way to overcome this. So given the added complexity to implement the algorithm initially, and the relative ease of beating it, it's probably not worth doing. Plus it would incentivize a form of "bad" behavior (the sybil attack).
You are viewing a single comment's thread from:
Agreed. I didn’t come to that conclusion until I was mostly done brainstorming the idea though :)
Split the stake between multiple accounts and factor in the curation (loss) multiplier when considering votes.
It would discourage large stakes from frequently dolloping large votes in the same place so they'll at least have to cast their net wider.