A reward curve which started as n^2 / exponential (thus flat), and then later changed into linear would work against self-voting as well as excessive rewards:
Or, even better, an idea of @clayop.
You are viewing a single comment's thread from:
Sort: Trending