very complicated, it is not obvious what will happen and there is no way to know without deploying it. as long as it is iterated until we get a good solution, it seems as good as any other change at this point.

clearly things arent perfect now, so something needs to change. by observing the effect from this change (good or bad), then the next iteration should be that much better.

i did a whale for whale voting analysis and I didnt see any systemic pattern, so all the conspiracy theories are not based on the data, at least as of a month ago