Many of your proposed filters do not really address Eugine_Nier’s point about Goodhart’s law.
If there is any structural bias in the first generation of vote filters, there are many reasons to be concerned that those who do not like the measure will not be sufficiently powerful to cause changes to the vote filters going forward.
Many of your proposed filters do not really address Eugine_Nier’s point about Goodhart’s law.
If there is any structural bias in the first generation of vote filters, there are many reasons to be concerned that those who do not like the measure will not be sufficiently powerful to cause changes to the vote filters going forward.
Wait, I thought Goodhart’s law was the one about “teaching to the test”?
Yeah, not all of those are equally good. I suspect they may all be better than the current criteria, but don’t hold me to that.