7
\$\begingroup\$

The serial voting checker script has reversed legitimate downvotes on the following questions:

These questions were all posted at the same time, and all had the same problems, that is, lacked details and information that we generally expect from homebrew review questions. Given they all lacked necessary details, these downvotes were legitimate, but were grabbed by the serial vote detector because it very obviously looks like targeted voting to a robot. See here for more details: How can I ask a good homebrew review question?.

Can this be undone? I know I can go and downvote them again, but that will just get reversed again as it will appear once again as serial voting.

\$\endgroup\$
1
7
\$\begingroup\$

Given they all lacked necessary details, these downvotes were legitimate, but were grabbed by the serial vote detector because it very obviously looks like targeted voting to a robot.

Let me clarify: this has nothing to do with robots. Those votes were grabbed by the serial vote detector because they were in fact serial votes. You voted a ton on the same person's posts in a small window of time, and the system considers that to be serial voting and to be inherently illegitimate. This has nothing to do with robots and everything to do with making sure one person does not unjustly target another person and mess with our systems. You didn't act with malicious intent, and each of those votes was individually legitimate, but the serial voting script doesn't know this and places no value judgement on vote legitimacy—it just looks at whether the votes are serial.

Can this be undone? I know I can go and downvote them again, but that will just get reversed again as it will appear once again as serial voting.

Moderators can't reverse the serial voting script's actions. Those votes are simply undone and gone.

Just ... don't re-do all these downvotes again. It'll just be caught and rolled back as serial voting again, and nobody can prevent that. I don't suggest trying to find a way to do it anyway since circumventing our anti-abuse mechanisms is generally a bad idea and done at your peril. There are other people beside you who will also be voting and helping to take care of things.

I know those questions all share the same flaws, but they're all closed now. Downvotes on a couple of those posts is plenty. Some of those might still warrant downvotes on an individual basis, but you can't do that without it counting as serial voting again. The situation is being handled though and things will be fine even without them being downvoted. Now's the time to let The Process™ do its work, and focus on next steps for taking care of things: we're engaging already with the author to provide feedback and get more detail.


This does raise rate limiting questions, such that maybe even a 2k user shouldn't be able to ask eight questions in the span of twelve minutes, so that the community has time to deal with problems that might show up in the first two or three and would otherwise show up in the rest, but that's a separate topic to the voting.

\$\endgroup\$
12
  • \$\begingroup\$ So if the posts were made by 8 different users, the votes are legitimate, but because they were the same user, they are illegitimate? I don't think I can get on board with that. The point of servial voting reversal is to protect users from being targeted by downvotes that are motivated by targeting the user. These downvotes were motivated by poor content - the user that posted them had nothing to do with it. The voting distribution would have been the same had they been posted by 8 different users or if the poster(s) were kept secret. \$\endgroup\$ – Thomas Markov Jan 18 at 13:56
  • 6
    \$\begingroup\$ I am simply telling you what the system considers to be serial abuse, and clarifying that it has nothing to do with looking like a robot. Kindly don't shoot the messenger? \$\endgroup\$ – doppelgreener Jan 18 at 13:59
  • 2
    \$\begingroup\$ I think what Markov is saying is that they aren't serial votes at all. I think he believes they simply look like serial votes to the system. Whereas you say that they simply are serial votes in reality, though perhaps there's an implied "to the system"? \$\endgroup\$ – Medix2 Jan 18 at 14:06
  • 4
    \$\begingroup\$ @Medix2 No, I mean they are serial votes. Even if they may be legitimate, they are serial votes. Serial voting means lots of the same kind of votes on the same person's material in a small time frame. There's no value judgement in there of whether the votes are legitimate or not, there's only judgement of whether they're serial. I am trying to clarify here what serial voting is such that the system catches it. I do think they're legitimate! They are also serial, and the system reverses that. \$\endgroup\$ – doppelgreener Jan 18 at 14:14
  • 1
    \$\begingroup\$ @doppel Makes sense to me, a difference of definition is commonplace as all else, +1 \$\endgroup\$ – Medix2 Jan 18 at 14:18
  • 3
    \$\begingroup\$ So wait a minute-- Speaking as someone who did not downvote in the first place, and really doesn't have a dog in this race: Because (a) someone posted a cluster of identically-flawed questions, and (b) people saw them as a cluster and downvoted them in good faith and otherwise appropriately as a cluster, and (c) these got flagged as "serial downvotes", therefore (d) there is no way for the original downvoters to re-downvote without violating the rules? In other words the serial downvote detector acts to protect serial bad questions? \$\endgroup\$ – Novak Jan 19 at 0:29
  • 2
    \$\begingroup\$ @Novak they can downvote some but not all or even most. You won't soon find me defending the system or its outcome in this situation, that's for sure. This situation is pretty exceptional, and in this case it's cushioning the author to get a fraction of the downvotes and reputation loss they might otherwise get. I've got no idea whether that's intended and desirable on the part of Stack staff for some reason, or genuinely an undesirable outcome. (I really am genuinely surprised there wasn't a rate limit hit sooner by the querent though.) \$\endgroup\$ – doppelgreener Jan 19 at 2:11
  • 1
    \$\begingroup\$ That is not important to make a fuss over (it is, as you say, a highly unusual situation) but that's a downright goofy overall result. \$\endgroup\$ – Novak Jan 19 at 21:27
  • 2
    \$\begingroup\$ Yeah the one place serial downvoting is legitimate is in response to poor serial posting, but the feature's not smart enough to know that, c'est la vie. \$\endgroup\$ – mxyzplk Jan 19 at 21:48
  • 1
    \$\begingroup\$ Yeah, I understand it would be tough to hammer out the logic for the algorithm to detect it. The inability (or unwillingess) to have a common-sense, human-intervention override is a little harder to understand. \$\endgroup\$ – Novak Jan 19 at 23:09
  • \$\begingroup\$ You mean the human intervention override that has resulted in all the questions being closed? Focus on outcomes, not if every pixel on the site can be twiddled. \$\endgroup\$ – mxyzplk Jan 22 at 17:19
  • \$\begingroup\$ @Novak It's mainly just low payoff for investment. Serial vote reversal is correct almost all the time; the times it isn't are exceptionally rare and low impact. Like, when's the last time we even ran into this scenario on RPG.SE? Plus, the situation's being resolved even without our downvotes, they're not an important factor right now. Meanwhile, staff would have to expend time on implementing the overrides and ensuring they're working correctly and ongoingly maintain them, which would be difficult to justify when there's other work to do. \$\endgroup\$ – doppelgreener Jan 22 at 18:23

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .