OFF TOPIC note: This question is hypothetical, therefore comments and answers disputing its hypothesis are off topic. That is, while the question of "Are there actually any trolls and moles here?" may be interesting, that is a different question.

Suppose for the sake of argument, that some minority of users were mostly here in the capacity of employees, such as those of of the Internet Research Agency of St. Petersburg (IRASP), or perhaps not that group, but one using kindred methods sponsored by other nations, corporations and organizations. Their jobs in part being:

  • Juggling as many identities using persona management software as seems strategically feasible.

  • Accumulating reputation, perhaps by merit, or gaming the system, or more likely both. A small network of personas might be used for logrolling.

  • Confusing or degrading communication contrary to their employers' wishes. Methods might include voting pools, time wasting questions and comments, red herring questions, or by skillfully offending and baiting emotionally sensitive partisans into some squabble which leads to their banishment or a question closing.

  • Promoting, whether directly or indirectly, the current slate of their employers' talking points.

  • Waging covert propaganda war against any other known trolls, if need be. (Or perhaps just pretend to, during budget reviews.)

This wouldn't be all bad. Sometimes a given employer's talking points would in fact be correct, and those users might naturally then provide good answers, or object to bad ones. Employees of opposed groups might cancel each other out.

But sometimes it would be bad. Misleading answers might be promoted, and good answers demoted. Ordinary users might be contemptuously manipulated. The site as a whole might be overly weighted down and founder.

What sort of policies on Politics.SE would help reduce those harms particular to paid trolls and moles?

Note: Answers should not obsess over mere suspicion or identification of specific individuals or groups. Think of it like pest control via sanitation -- a hospital full of stagnant puddles and dark corners provides a more habitable environment for pathogenic organisms, compared to a clean, well-lit hospital. (But it isn't possible for a hospital staff to find each and every pathogenic individual bacterium under a microscope.)

  • 4
    The network already has a good set of tools against unpaid trolls and paid spam. Is there anything special about these paid trolls that go above and beyond what a hobbyist one would do? Apr 25, 2018 at 18:43
  • 1
    @GGMG, It's like comparing a man with a guitar and an orchestra with a budget of a million dollars a month.
    – agc
    Apr 25, 2018 at 22:21
  • 5
    I'm voting to close this question as off-topic because the question does not serve any purpose other than to promote his political agenda. If hypothetically, this site was taken over by pink elephants (and you are not suggesting that it had been), but if it had been, what would the moderators do? Well, you aren't asking about pink elephants, because you are not promoting the idea that the site had been taken over by pink elephants. You are planting a different suggestion.
    – grovkin
    Apr 26, 2018 at 2:27
  • @grovkin, If the site was taken over by pink elephants, the moderators should drink something else, or less of it. Unless the moderators liked pink elephants, which are nicer than trolls. Conceivably the pink elephants might also attempt to game the system and perpetuate their kind...
    – agc
    Apr 26, 2018 at 17:39
  • You may want to re-ask the question. Something like what protections are in place to prevent trolling and false information seems what you really want to know. Where you have asked a what if question which is out of the usual guidelines. Apr 27, 2018 at 2:57
  • @user1605665, Thanks. It's not just about prevention however -- given the premise, it seems reasonable to assume under the current system that general prevention is impossible. Minimizing damage seems a worthy goal however, as well as discussing the best use of trollish data should its origins somehow become known after the fact.
    – agc
    Apr 27, 2018 at 17:32
  • 2
    I'd ask them why they'd think such a low-traffic site was a good use of their professional trolling skills.
    – user1530
    Jul 10, 2018 at 5:57
  • @blip, Maybe not, but we don't really know much about how efficiently managed Pro Troll outfits are. Perhaps busywork pays equally well. Or perhaps the goal is the prevention of high traffic, and low traffic is one indication of a job well done.
    – agc
    Jul 11, 2018 at 17:35

3 Answers 3


I don't think we could really do anything specific about it.

There is no way for us to tell a paid government agent apart from a loyalist citizen who just loves their government and considers it their duty to defend its interests. Trying to do so would just lead to a witch hunt. However, the damage caused by people who think that Politics.SE is an appropriate platform to spread government propaganda does not depend on if they get paid for spreading it or not.

So all we can and should do is what we do anyway:

  • Edit any posts which are useful, but have an obvious political slant
  • Downvote, vote to close and flag any posts which are not useful.
  • Flag abusive or discussion-provoking content
  • Upvote any constructive content
  • We certainly agree on the impracticality of witch hunting. But suppose after the fact a news story or leak reveals, maybe years later, that users X, Y, and Z were paid trolls working in concert, and their posts when later reviewed in that context, and taken as a whole, prove to have been far more effective as propaganda, (ingeniously so, diabolically so...), than the activities of three random partisans for the same cause would be. Would having such knowledge change anything with regards to their completed activities?
    – agc
    Apr 25, 2018 at 22:44
  • 1
    @agc I doubt that this would be useful. If the content is proven to be untrue propaganda, it can be erased right now. You don't need to wait until it is revealed that a user is a paid troll. And if nobody is able to discern those propaganda posts from quality answers, maybe there is quite a bit of truth in the propaganda...
    – Thern
    Jun 25, 2018 at 13:46
  • @Thern, Propaganda, (and many a resilient lie), often employs quite a bit of truth to help the lie blend in. A kind of Trojan Horse for the public mind.
    – agc
    Jun 25, 2018 at 15:19
  • 3
    Well, it is the task of sites like stackexchange to separate the true part of an answer from the false part. Cutting away the complete answer together with its true parts just plays into the hands of those that cry "censorship".
    – Thern
    Jun 25, 2018 at 15:54

Hypothetically, If we knew for certain that someone was posting here as a paid troll, then they would be banned.

Sowing confusion is contrary to what we want to do here, and someone here for only that purpose is not likely to change their ways due to getting punished.

Understand that's a very big "If". We are unlikely to actually be able to identify for sure that someone as a paid troll.

That also applies to a fairly narrow definition of "paid troll" who is someone paid specifically to misinform us. A social media representative openly posting on a company's behalf, might not qualify for instance.

  • Cosmetic... A hypothetical agency could switch to a new identity, (or another undiscovered old one). Please clarify what should be banned, i.e. further posts by said user, questions, comments, votes, answers, or what.
    – agc
    Apr 26, 2018 at 2:24
  • @samiam Would your answer be the same if the question was asked about "paid political operatives" rather than about "paid trolls"? Both are paid to push an agenda. Only one is a dysphemism. If the question was asked neutrally I think it would ask whether paid political operatives, who didn't identify as such, would be prohibited from posting questions and answers.
    – grovkin
    Apr 26, 2018 at 2:54
  • @grovkin I think what matters most is the content. A political operative sowing confusion with misleading information becomes a troll (in the colloquial sense). It is not just another term for political operative. A political operative that posts good content is a 'contributor'. I agree, though, that reasonable people can disagree on what is 'good content'.
    – user5155
    Apr 26, 2018 at 13:13
  • @JeffLambert, on 'good content': It's not so much about content -- the hallmark of trollery lies in their efforts to manipulate metadata and strategically sow confusion through fraudulent means. Any "operative" who does that is a troll. If some Think Tank wishes to post its opinions in the name of that Think Tank, that'd be fine... but if the same Think Tank hires 50 outsourced typists to repeat the Think Tank's opinions and upvote each other, that's trollery.
    – agc
    Jul 9, 2018 at 22:21

Given the premise that sufficient proof existed of hireling propagandist trolls, (let's say there was something like an auditable Wikileaks troll list such that it could be confirmed), I'd suggest:

  1. The users in question should be banned from future activity, barring some future evidence that exonerates them.

  2. No deletions of troll accounts, comments, answers, or questions unless the posts are offensive on some other grounds. A known troll's postings would be valuable in of themselves, keeping them around would:

    • Show that SE doesn't hide problems.
    • Show users what troll posts look like, and how they may differ from ordinary user posts.
    • Provide meta-data, if any, on what a troll's employers are interested in.
    • Have value as humor.
    • May preserve some otherwise unobjectionable and useful answers.
  3. Keep their posts public mostly as before, with certain allowances:

    • All votes from non-Trolls should remain as before.
    • Votes, favorites, and answer acceptances from actual trolls should be printed alongside the legitimate votes, with some clear indication that those votes should be regarded as fraudulent.
    • Upvoting, downvoting and favoritizing of troll texts should be generally suspended, but it should still be possible to change an accepted answer from, or even to, a troll answer.
    • All troll posts could be in a conspicuous troll-font, perhaps like one of these Halloween fonts, perhaps with a conspicuous background color, and accompanied by some kind of troll-ish icon.
  4. For those who'd prefer to forget, or don't want any static, some kind of user option, perhaps in on a config page, to disable the display of any of the above, (votes, text, etc.). For those users, SE would look the same as it does now, minus any known troll posts.

  5. The user page of a known troll should make it clear that the user was a known troll, (special fonts and icons would help), with as much general data as can be had, perhaps using special troll tags.

    An example of a troll tag would be if a troll was known to have worked for IRASP or whoever, there could be an IRASP tag, or whoever, which would link to a list of every other known troll with that employer. Or a more general tag indicating, if known, the political power bloc that employer is allied with.

  • Is anti-troll education a goal of this site? This seems like a lot of features to add to achieve it.
    – user9389
    Jul 11, 2018 at 20:29
  • @notstoreboughtdirt Given the Q.'s premise, adding some features seems somewhat preferable to pro-troll obliviousness. IMHO for an interface used by (possibly) millions, interface programmer time is at least 1000 times less important than user time.
    – agc
    Jul 11, 2018 at 21:55
  • Stipulating all your facts, I'm not clear why anti-troll education is an expected focus here.
    – user9389
    Jul 12, 2018 at 15:57

You must log in to answer this question.

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