3
$\begingroup$

Recently I've noticed quite a few low quality posts made by "unregistered" users whose user name is of the form "user3####". For example, this and this. I'm sure there are a couple others but I can't find them right now (maybe because their posts have been deleted). In my experience at Econ.SE, such unregistered users almost never made positive contributions, in the form of good quality questions and answers (with perhaps only a handful of exceptions). I wonder if we could/should require registration to post questions/answers on Econ.SE. Of course this is not going to completely weed out trolls or low quality Q&As, but at least it raises the cost of trolling/low quality posts by a little.

I am aware that there are legitimate reasons not to implement such a requirement network-wide. But I think the case for implementing it at Econ.SE is strong.


Edit (2020/10/23): Latest instances here and here.

$\endgroup$
11
  • 1
    $\begingroup$ I would be up for that but from the style of writing this seems to be the 'EROI' troll just with pivot to another topic. I dont think that this person is going to give up so easily and some of the accounts of that user were also registered. I already discussed the guy with EconJohn and KitsuneCavalry multiple times, they recommended flagging the activity so they can then delete/suspend the accounts but the person behind is just getting around the suspensions with VPN. I suppose here only thing that will help is if we completely ignore the troll and more promptly close those questions $\endgroup$ – 1muflon1 Oct 23 '20 at 0:17
  • 1
    $\begingroup$ I think issue here is that also EconJohn and Kitsune are in different time zones than the troll so they obviously cannot stay whole night to close up the questions/suspend the fake accounts. If elected I will try to help the guys with that but it is always good to flag these low quality contributions (low quality is valid reason for flagging), then downvoting which lets users to vote delete posts is also good way how deal with the troll (also when I said ignore in previous post - I meant to avoid interacting/responding to the troll). Hopefully, the user will eventually give up. $\endgroup$ – 1muflon1 Oct 23 '20 at 0:20
  • $\begingroup$ The volume of questions is low, so dealing with these posts is not yet that onerous. The question is how many useful unregistered questions have been asked asked that we would lose if we blocked such questions? $\endgroup$ – Brian Romanchuk Oct 23 '20 at 18:35
  • $\begingroup$ It seems this troll has been trolling other sites as well. According to the user ggmg-he-him this troll is already known at Politics and Skeptics SE. They also have meta post about the troll $\endgroup$ – 1muflon1 Oct 23 '20 at 20:15
  • $\begingroup$ @BrianRomanchuk: I can only recall seeing maybe 4 or 5 good quality questions coming from unregistered users. Of course this is most likely an underestimation because one has to click into a user's page (which I don't routinely do) to see if they are registered or not. I'm sure that statistic exists in some data dump somewhere. I'll try to dig it up when I have a bit more free time in a couple of weeks. $\endgroup$ – Herr K. Oct 24 '20 at 6:36
  • $\begingroup$ The terrible questions keep coming. If this is a single troll, has a wide range. economics.stackexchange.com/questions/40453/… $\endgroup$ – Brian Romanchuk Oct 25 '20 at 22:48
  • $\begingroup$ @BrianRomanchuk that is definitely the same (eroi) troll, in another answer the user again talks about that depletion nonsense. In this answer the user says "Rates fell because coal depleted and investment collapsed, therefore less bonds issued"... that is definitely the eroi troll - I would be surprised if there would be more users having such specific tinfoil hat ideas. But I agree that the repertoire of this troll is getting broader, but at this point I think you can tell from the way the person writes. $\endgroup$ – 1muflon1 Oct 25 '20 at 23:20
  • $\begingroup$ Yes, I saw that answer, and yup, EROI troll. $\endgroup$ – Brian Romanchuk Oct 25 '20 at 23:47
  • $\begingroup$ Again clearly the same eroi troll - see this question I seen the old account posting question dissing one of the answers of Alecos, and now it is directly referenced by this new account.. that is clearly the same user hiding behind new account. This time the account is even registered. Also the troll is trolling skeptics again, with Q about North Korea - maybe some cooperation between sites could help? $\endgroup$ – 1muflon1 Oct 26 '20 at 13:41
  • $\begingroup$ It looks like the person changed from “usernnnn” to using a name. Some troll questions/answers seemed to have changed changed username. Sigh. $\endgroup$ – Brian Romanchuk Oct 27 '20 at 23:13
  • $\begingroup$ @BrianRomanchuk that happened (to my best knowledge) because the fake accounts were merged together with original account of that user after being flagged and reported. $\endgroup$ – 1muflon1 Oct 27 '20 at 23:16
1
$\begingroup$

Originally I was quite warm to this idea, as expressed in my comments, but after doing research on this issue I am not sure if it will help.

  1. According to the discussion on this software engineering meta post, their experiment with banning unregistered users was not successful.

  2. As already discussed in comments on this and on other meta discussions regarding this problem, based on the writing patterns and topics raised by the unregistered users (e.g. EROI/depletion Q - even if the users ask different Q or make different A they will usually have one post on this topic) they in fact seem to be all one user (Of course, we never can be sure of that but what is the likelihood that there are multiple people in the world having such specific EROI/Depletion questions and answers?).

    Furthermore, as brought to my attention by user GGMG-he-him in comments on this deleted post (you should have high enough rep to see it) this is also known troll at other sites (see this meta post). You might also notice that recently some of those accounts were merged into DJ Sims and you can see the user being active and causing problems at almost every SE site.

    Consequently, this would be a policy that would be made mostly for a single user. Of course, there might be more such users (I will try to deal with your data request in your other meta Q as soon as I will manage to figure out where to get the stats and if it's okay to share them - since the posts on other sites are old and rules on sharing/accessible data might have changed), but because this seems to be one user persistently creating sock puppet accounts with VPN even such statistics might not tell us much (given this troll persistence I would not be surprised if they created number of accounts counting in double digits). Although, eventually many of the sock puppets accounts are merged with main account of the troll user, some of them might have been just outright deleted/separately suspended which will bias the statistics.

  3. Even though it is likely just one (or at best few) user(s) I would not mind implementing any automated policy, but from my understanding from reading the meta posts linked in this answer this is not really making it much harder for the troll to be active. As pointed by @BrianRomanchuk in comments under your other question the user now registers the accounts and it does not seem that this creates much obstacles for the troll.

    Furthermore, from seeing discussions around this issues on other metas (see examples here or here or here) and discussing this in the past with other mods my understanding is that this kind of persistent troll requires manual handling. This is why it is very important to not just close/downvote such posts but also flag them so they can be promptly closed/deleted (when you flag a post moderators get notification about that - I am just mentioning that since some users might not realize that this is helpful).

$\endgroup$
4
  • 1
    $\begingroup$ Thanks. I agree that the case for banning unregistered users is less strong now. But I'd still want to wait and see what the stats on registered/unregistered user contributions are like before retracting my proposal. $\endgroup$ – Herr K. Oct 28 '20 at 4:08
  • 1
    $\begingroup$ By the way, congrats on being elected moderator! $\endgroup$ – Herr K. Oct 28 '20 at 4:09
  • $\begingroup$ @HerrK. thank you for the wishes, by the way now you got answer to the other question and see the statistics would you like me to escalate the issue and discuss it with the other mods, or do you think it is not worth while anymore? $\endgroup$ – 1muflon1 Nov 17 '20 at 21:47
  • 1
    $\begingroup$ Thanks for following up. Given the stats, I think the problem caused by unregistered users is less serious than I originally thought. Even among answers where the badness differential between reg and unreg users is largest, the absolute number is still manageable (only 13 in the past six months). So I agree with you that maybe we can leave this at the status quo for now. $\endgroup$ – Herr K. Nov 17 '20 at 22:25

You must log in to answer this question.

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