The serial voting guidance in the Help Center should be consistent with Meta

Question

When a user suspects that they are the victim of serial voting, they should contact the site moderators with a custom flag which clearly explains what they think is going on (after 24 hours have passed, to allow for system to reverse the votes). While the CM team can be contacted directly using the "Contact us" form, this is generally not the preferred option, for a number of reasons.

This guidance is covered in many places on Meta Stack Exchange.

• What is serial voting and how does it affect me?

If the 24 hours has already passed and the suspicious votes have not been reversed, you can then flag one of your own posts and explain what happened so a moderator can look into it.

• What can I do if I'm the victim of serial downvoting?

If after 24 hours (give the script time to run) you still see a problem, then raise an "in need of moderator intervention" flag on one of your own posts and ask a moderator to look into the anomalous voting patterns. Be specific as to what you feel the issue is.

... If it still hasn't been corrected, contact the SE team via the "contact us" link at the bottom of any page on the site. Using this method is no faster than raising a moderator flag. In fact, people have reported that it tends to take longer.

• How to handle repeat voting fraudster who re-serial-votes after every reversal?

This is a problem for the mods to handle. Flag either account (the suspected serial voting account or the suspected target account) for the site moderators' attention.

... Using the Contact Us form will also work, but is overkill and less quick.

There are many more examples of similar posts on Meta Stack Exchange, as well as per site Metas, and as far as I can tell, they all consistently say that site moderators should be contacted first, and several don't even mention contacting the CM team.

---

On the other hand, the resource that is looked at by a majority of users is the Help Center, specifically the Help page on serial voting, and the guidance on that page is completely the opposite of the guidance on Meta:

What if I think I'm the victim of voting abuse?

If you see very unusual votes being targeted at your account, don't worry about it. You should wait at least 24 hours after noticing before becoming concerned, as the automated system should detect it and reverse it for you. If, after 24 hours, you do not see any fix to your reputation, please contact the team using the "contact us" form located at the bottom of any page.

This page says nothing whatsoever about contacting the moderators via a flag, much less that it's the preferred option to contacting the CM team. This appears to be an oversight, so unless I'm missing something, I think the wording in this Help page should be changed.

Site moderators have the ability to edit some of the pages in the Help Center for that site, and if this page is one of them, this might be the appropriate thing to do, especially if the guidance should be different on different sites across the network. But if the guidance is meant to be the same across the network, then I think all the Help center's pages should be updated in concert.

Answer 1

So this guidance is a bit trickier than you might think. While most posts here on Meta suggest flagging first, even the guidance on escalating past that or what happens after you've flagged is inconsistent here. None of the guidance on how to escalate voting irregularities matches other guidance and probably all needs rewritten, but how?

On the accuracy of the guidance

In this context, moderators have no power to do anything about suspicious votes aside from escalate cases they think are inappropriate voting to the CM team. By flagging for moderator attention, we are really just using moderators as a triage to weed out the clearly obvious "this isn't serial voting" cases in flags and only see the ones that really need our attention. That is... not particularly necessary. While there is nothing wrong with flagging for moderator attention, it is not any more or less effective than contacting us directly.

Contacting us could be considered an equivalent to flagging for moderator attention in many cases, but could actually be faster for simple and obvious cases of targeted voting. We have two different support queues: the main Public Q&A support queue that all contact form submissions feed into and a separate CoGro queue for escalations and trickier situations.

Working in support, I mostly perform the same "triage" functions as a diamond moderator would. If it doesn't look like a legitimate case of serial voting, I send an appropriate canned response explaining voting or pointing to relevant article and close the ticket. If it does need more attention, I may even have the bandwidth to quickly handle it myself (for super simple, straight-forward cases). But usually it just gets tossed into the same backlog queue as the moderator escalations would.

So there is a fit of a fine line there. I am capable of handling voting reversals and do so on occasion when it's easy to do. But I no longer am involved with processing CM escalations in the backlog and do not spend time diving deep into serial voting complaints. That makes it hard to draw any kind of line as to when contacting us directly would be more beneficial versus the same as flagging.

But there's a key point there that isn't obvious unless you look for it: nowhere in that did I say I ever looked to see that a user flagged for moderators before escalating to us or that I direct them to flag for moderator attention instead. That isn't guidance we ever send out. Because it frankly doesn't matter all that much in the end.

On the efficacy of the guidance

One of the things we get when you submit the contact form is a Referrer field, which tells us the last page you were on before you decided to contact us. It can sometimes be invaluable in pointing us to where an issue was occurring (because people often forget to include a link). But it can also be useful for general statistics on where people often are before contacting us about particular issue.

I have worked support for almost six years now, and cannot recall any times where someone contacted us about serial voting from that help center article or any of those Meta posts. I have seen it pop up in requests from users complaining that the reversal should not have occurred or is unfair.

Most of these users come to us either directly from their reputation history or from one of the posts that was downvoted. They don't go looking for guidance about what might be going on - most might not even know what to search for in order to find that guidance. They just go directly for either the flag button or the contact form to report that they are being targeted. And I end up closing a vast majority of these without response because the system has usually auto-corrected the votes on its own before I even look at the ticket.

Only a small subset of users affected by targeted voting in this manner are actually seeing this guidance: the users who ask about it on Meta and are redirected to them via duplicate closures. So while it is nice for everything to match up, I do not see it making much difference in the long-run. Nobody is reading it.

So what is the solution?

Honestly, I am more inclined to simply remove that section from the help center article. People clearly do not ever read it there. I would say that is more because it does not seem relevant to the article it is in, which is explicitly about a reversal that has already appeared on a user's reputation history (that history event is the only thing that links to it in the UI). Nobody is going to look at a help center article about why they have a voting reversed event in their history to find out what they need to do to make votes that haven't been reversed go away.

I'm not making this change immediately, but it is the change I am leaning towards for the case of the help center itself. I'm not inclined to spend a great deal of effort writing a shortened version of this long answer in a paragraph there when nobody is looking for this information at that location.