[ggm]

At a grossly simplistic level aside from simply not "cheating" at sampletime, are there changes to publication requirement which could ameliorate this?

I think a steeper hill to climb might be minimum sample cohorts and neutral party randomly assigned.

Does anyone offer a certified sample aggregation/distribution function?