Don't know if it's perfect but I think the phrasing you'd used downthread "Downvoting all of a user's nodes with prejudice" would read as clearer / less vague than just "Massive downvoting" (and then giving e.g. "Exclusively downvoting a given user across multiple days" as an explicit example of "with prejudice"). Then again I might could see just "Massive downvoting" so long as there were a couple of clear, tangible examples provided afterwards as to what's considered such an abuse.

I will think about how #1 could be worded better, but I also think it might suffer from overcomplexity if one tried to cover every case. Plus, I have a dim memory of reading of a existing policy that already states basically the same thing as #1, but I'm having trouble finding it right now.

For #3, I meant what is described in that Site Rules link, plus for example posting something anonymously, then logging in and upvoting it.

Thanks. Note that "nodes you (the human) wrote" says nothing about user accounts. If you upvote nodes you wrote — however you manage to do so — that is contrary to the spirit of PerlMonks. Imagine if, for example, sundial started posting anonymously and then using his account to upvote each of those, to try to offset the negativity they'd inevitably attract, that would be an abuse. Again, we're just looking for large-scale, systematic abuses. One or two here and there is not something we'd be concerned about.

Having thought about it a bit more and having read the rest of the thread, here are the updates on my thoughts above:

The node I was thinking of that already discussed the systematic downvoting was in fact History now influences voting. So it seems like a policy like #1 has already been in place?

As for the wording of #1, I think using the various explanations you gave in this thread, like here and here, to expand upon the first point would probably be enough to make clear what the policy is intended to combat. Considering the mentioned punishments are fairly severe (XP is much harder to come by these days...), I also think it would be useful to clarify in the policy:

• whether there will be one or more warnings
• whether there will be a discussion with the user
• what is the statute of limitations
• whether users have to report that they think they are being downvoted, and/or whether there will be an automatic system like you mentioned here, and if the latter, what its rules are and how often it is run (one concern I would have is what this system's tendencies for false positives and false negatives might be...)
• what the potential punishments are; you already mentioned a few in the root node, but I imagine there could be others like deleting the downvotes?

So since this is getting kind of complicated, an alternative suggestion would be to make the policy "abuses will be judged on a case-by-case basis after talking to the affected individuals" - which is pretty much the system I thought was already in place.

Codifying such things is not so easy! ;-)

well ... there is an edge case! :)

I'll rather send you a /msg ...