16
$\begingroup$

I understand the idea of a question ban - if you keep making bad questions you can't do so anymore. But what's to stop someone from simply creating a new account and continuing to do so?

$\endgroup$
  • 25
    $\begingroup$ Moderators have tools for detecting that. They are somewhat ... secret. Actually, such askers are often too lazy, and keep reasking the same (or at least similar) questions. Many eagle-eyed regular users can spot those even without the tools diamond mods have. They then summon a moderator to the scene. Lately I have been somewhat irritable, and not very lenient. $\endgroup$ – Jyrki Lahtonen Feb 18 '18 at 16:33
  • 44
    $\begingroup$ We could tell you, Zachary – but then we'd have to ban you. $\endgroup$ – Gerry Myerson Feb 18 '18 at 21:19
  • 4
    $\begingroup$ I'd also like input on whether there are mechanisms to stop people from simply creating a new account if they've reached there daily, or weekly, or monthly question limit, but they insist on asking more? Users create new accounts merely because they want an answer to a question of theirs that was closed or deleted, so they create a new account to do so. $\endgroup$ – Namaste Feb 18 '18 at 21:51
  • 2
    $\begingroup$ I asked a related question in the math mods' office chat this week Friday. $\endgroup$ – Namaste Feb 18 '18 at 21:54
  • $\begingroup$ Just when I was thinking that Math.SE doesn't have Wikipedia-style drama... $\endgroup$ – Bill Thomas Feb 23 '18 at 21:48
1
$\begingroup$

The question escalates to the point, what can the SE do with wrongdoings by the same person, from continuously re-created (and banned) accounts.

The answer is, that doing this "correctly", the SE has no way to avoid the behavior.

However, doing this "correctly", is harder as it seems on the first spot. And the wrongdoers have no way to know, exactly what is the SE doing to detect them.

What the SE can detect:

  • IP (or network) of the wrongdoers (it can be falsified, but not easy)
  • browser version string and other fingerprinting (most parameters can be easily falsified, but not all, and we have no way to know, exactly which ones are watched)
  • if the attacker, because from this point they are attackers, is coming from the tor network
  • wording / typical grammatical mistakes of the wrongdoers (the SE probably doesn't utilize any algorithm for that, but the mods probably can detect the easier cases with experience)

On the meta SE, there are sometimes mentioning from previous or current "fights". The main threat of the SE system is the spamming and the voting circles, and not the LQ questions. Of course the the exact details are not public.

As far I know, the main strategy of the SE in the case of the LQ posts is to enforce them to post also useful content, and then filter out the rest.

$\endgroup$
  • 4
    $\begingroup$ tor is not used only by bad guys. Maybe someone who is already a regular on some SE site wants to ask an on-topic question that reveals something embarrassing. Or an illegal immigrant wants to go home without confrontation. I could go on and on (and in fact, I did, but deleted some for brevity.) $\endgroup$ – WGroleau Feb 23 '18 at 6:02
  • 3
    $\begingroup$ "The main threat of the SE system is the spamming and the voting circles, and not the LQ questions" On math.se? Says who? $\endgroup$ – Did Feb 23 '18 at 6:43
  • $\begingroup$ I use Tor. I would be mildly perturbed if SE starts preventing me from contributing just because I use it. $\endgroup$ – J. M. is a poor mathematician Feb 26 '18 at 7:41
  • $\begingroup$ @J.M.isnotamathematician Non sequitur. And then? $\endgroup$ – peterh Feb 26 '18 at 7:41
  • $\begingroup$ @Did I say, with my SE experience. $\endgroup$ – peterh Feb 26 '18 at 7:42
  • $\begingroup$ I am more commenting on WGroleau's comment and not your answer. $\endgroup$ – J. M. is a poor mathematician Feb 26 '18 at 7:43
  • $\begingroup$ @WGroleau I didn't say that Tor is used only by bad guys. I am surprised, is it really the MathSE? You don't really look very logical. $\endgroup$ – peterh Feb 26 '18 at 7:44
  • $\begingroup$ And I didn’t say anything about math. You implied that use of tor is an indicator of abuse. I see I’m not the only one who thinks that is a bad idea. $\endgroup$ – WGroleau Feb 26 '18 at 12:01
  • $\begingroup$ @WGroleau You are one of the 2 people, giving something into my mouth what I didn't say, and then attacking me on this. This is the style of the darker layers of the Meta Stack Overflow. $\endgroup$ – peterh Feb 26 '18 at 13:35
  • $\begingroup$ You call that an attack? If you didn’t mean what you appeared to mean, it would be sufficient to say so. $\endgroup$ – WGroleau Feb 26 '18 at 13:58
  • 1
    $\begingroup$ "The main threat of the SE system is the spamming and the voting circles, and not the LQ questions" / "I say, with my SE experience" Any evidence supporting this claim? Spamming and voting circles are easy to detect and to eliminate, but both are small potatoes when compared with the somewhat prevalent misled ethos which pretends that any kind of quality control is by essence fascistic. That is a threat to the site, if you ask me (I mean, to the math site on SE, some other sites act quite differently). $\endgroup$ – Did Feb 26 '18 at 14:29
  • $\begingroup$ @Did Spamming are easy to detect, but there is a lot. Voting circles are only easy to detect if they are primitive. Detecting a well-worked voting circle is hard. Imagine if there are many sockpuppet accounts, all using different fingerprinting, and the directed graph of their votes are intentionally tuned to harden their detection. | Another hardness in the picture is that the "good old boys" of the sites tend to vote for each other, thus a well-worked voting circle detector would find the "good old boys" of the sites in the first line... $\endgroup$ – peterh Feb 26 '18 at 21:57
  • $\begingroup$ I don't buy it, @peterh. SE is usually a few steps behind the users gaming the system with alternate, additional accounts which are created for the express reason of gaming the system. $\endgroup$ – Namaste Mar 3 '18 at 23:53
  • $\begingroup$ It seems more likely that MSE refers to secret tools of Mods and of the SE team to detect abuses I refer to, but I worry the claim is mostly a claim, and is an attempt to deter users from misbehaving, but not very successful. Of course I don't know. But I've seen too many users creating separate accounts for a whole host of nefarious reasons. $\endgroup$ – Namaste Mar 4 '18 at 0:02
  • $\begingroup$ @amWhy Ok, then I say all what I think I know. The SE probably logs all the IPs from which voting or user logins happened. They also know if somebody is coming from tor (it is easy to detect, for example even on the wiki, if you are from tor, you can't edit). They are probably not enough smart to utilize any tricky voting chain detector, it would require a sophisticated graph algorithm, and the SE is surprisingly weak in implementing sophisticated algorithms. Furthermore, these graph walking algorithms would find the "good old boys" of the sites in the first line. Because not the voting chains $\endgroup$ – peterh Mar 4 '18 at 0:30
0
$\begingroup$

There are technical measures in place to make this difficult. As posted on Meta Stack Exchange,

Can I simply create a new account?

No. The automatic ban is at a lower level than account.

There isn't much of any information available about what "lower level than account" means. The details are intentionally kept secret by SE, to make it more difficult for people to game the system. However, it seems likely that the system takes into account IP address among other factors.

Besides the technical measures, there is also some amount of human oversight. As Jyrki Lahtonen posted in a comment,

Moderators have tools for detecting that. They are somewhat ... secret. Actually, such askers are often too lazy, and keep reasking the same (or at least similar) questions. Many eagle-eyed regular users can spot those even without the tools diamond mods have. They then summon a moderator to the scene.

$\endgroup$
  • $\begingroup$ Those supposed "moderator's tools" don't work very well, or else they fail to use them enough. The very "eagle-eyed regular users can spot them even without the tools diamond mods have..." (then we flag). And in doing so, there are repercussions for the flaggers. $\endgroup$ – Namaste Mar 3 '18 at 23:43
  • $\begingroup$ What about tools to detect a new account by a user whose question was closed, in order to repost the question? What about tools to detect users who delete a heavily downvoted and closed question, create a new account to ask it again? What about tools to detect the interactions between two or three or four accounts owned by the same individual. SE is many steps behind in detecting these sorts of issues. Eagle-Eyed users can only do so much with the few tools available, and likely a small percentage of users "gaming the system" by creating new accounts are ever detected. $\endgroup$ – Namaste Mar 3 '18 at 23:47

You must log in to answer this question.