Is it time to add restrictions to be fulfilled in order to ask questions?

To me it seems like there are individuals that (mis)use the forum using different user names asking similar questions, often without sharing their own thoughts. One reason for doing so might be to "solve" homework problems without leaving traces of that here (my speculation).

Perhaps, it should not be possible to ask anything unless the user have say 3 upvotes on some answer?

  • 1
    $\begingroup$ If you suspect a user is using different names in order to repeatedly ask the same question, or to get around daily/monthly question limits, then you should flag the question for moderator attention. This violates the rules of MSE and deserves moderator intervention. $\endgroup$ Aug 20, 2013 at 20:16
  • $\begingroup$ @AlexBecker Thanks, I will do that, but can the moderator see things like IP-addresses or how could (s)he conclude such behaviour? $\endgroup$ Aug 20, 2013 at 20:22
  • 5
    $\begingroup$ 3 upvotes on some answer are not really easy to get. I think this problem can be handled well by (down)votes and flagging. $\endgroup$
    – gofvonx
    Aug 20, 2013 at 20:22
  • $\begingroup$ @gofvonx Sure, perhaps 3 upvotes in total would be a better choice. $\endgroup$ Aug 20, 2013 at 20:24
  • 6
    $\begingroup$ With a site dedicated to all levels of mathematical studies, it is to be expected that the first interaction of many will be a question and not an answer. So requiring this as an entry criterion sounds unfair to me. Moreover, such a feature would probably have to be buried deep in central SE code - and would prevent the opening of new SE sites in the first place (no answer without questions). $\endgroup$ Aug 20, 2013 at 20:25
  • 14
    $\begingroup$ A drawback of 'forcing' people to answer something, is that this likely will have as a side-effect an increase of low-quality, repetitive, or even wrong answers. This effect seems worse to me than any peoblem that might be solved by having such a restriction in place. $\endgroup$
    – quid Mod
    Aug 20, 2013 at 20:25
  • $\begingroup$ @AD. I can't comment on what exactly moderators can see, but I can say that we have ways of confirming that this is going on in many cases. $\endgroup$ Aug 20, 2013 at 20:25
  • $\begingroup$ @AD. The details are not public, but moderators have certain tools to detect if accounts belong to the same user. I think the tools are rather effective, and often it is very easy to confirm that a user is just a sock puppet of another user. $\endgroup$
    – user9733
    Aug 20, 2013 at 20:28
  • $\begingroup$ @quid You have a point. This is probably harder than I thought. $\endgroup$ Aug 20, 2013 at 20:30
  • $\begingroup$ @Hagen von Eitzen I agree with the first point, not with the second: SE developers can enable policies per-site, e.g. daily/monthly question limits are enabled on 4 sites out of 105. $\endgroup$
    – user90090
    Aug 20, 2013 at 22:24

2 Answers 2


As a different way of (slightly) reducing the degree of "openness", StackOverflow requires users to register before asking a question. (One can still answer as an unregistered guest). Seeing as Math is the second SE site by the rate of questions per day, it would not be unreasonable to enforce the same requirement here.

Possible benefits:

  • the extra minute that goes into registering an account will lead to more users glancing at the help pages before posting.
  • users will be less likely to lose their credentials after asking a question, and consequently more likely to accept and upvote the answers they receive. (Unregistered users can't vote regardless of reputation).
  • $\begingroup$ +1 That might be a good idea, perhaps the pre-user should answer some random question on how to ask a question too. $\endgroup$ Aug 20, 2013 at 20:38
  • $\begingroup$ @AD. That's unlikely to be implemented for Math, since there is no such thing for SO (which gets many more low-quality questions than Math does). $\endgroup$
    – user90090
    Aug 20, 2013 at 20:42
  • 1
    $\begingroup$ Something should be done as I see it. $\endgroup$ Aug 20, 2013 at 20:44
  • 2
    $\begingroup$ This is not a good idea. I recall an essay (I think by Paul Graham) about the success of some [early] online stores over others. The conclusion was that people didn't want to register just for buying once or twice. So stores who would allow purchase without registration had more sales. Requiring registration would decrease the number of question. Whether that is going to be a good thing or not, is a whole other question. $\endgroup$
    – Asaf Karagila Mod
    Aug 20, 2013 at 23:01

Yes, it is:

Mathematics has reached traffic levels where this sort of rate-limiting and outright blocking of questions makes sense, so we've enabled quality blocks here as well. These blocks are based on the community's collective rating of posts from both a user's account and other accounts on the same IP address, thus limiting the ability of folks to abuse the community by just creating new accounts.

Requiring registration, as 90090 suggests, has proved useful on Stack Overflow for other reasons - primarily the overhead of dealing with account issues caused by unregistered accounts coupled with SO's insane traffic levels. But we've not seen it have much of an effect on question quality, and as Asaf notes it can end up having the unintended side-effect of driving away folks who you'd prefer to welcome in.

  • 1
    $\begingroup$ Could this be done on Physics too ? We are getting too many effortless one line/one sentence questions and corresponding one line/one sentence answers, etc too ... :-/ $\endgroup$
    – Dilaton
    Aug 21, 2013 at 9:00
  • $\begingroup$ Physics is growing nicely, but I think the volume of questions (and problematic askers in particular) is still at a level where it can be handled by the moderators without too much trouble, @Dilaton. Do your part - vote often, flag particularly egregious examples. $\endgroup$
    – Shog9
    Aug 22, 2013 at 17:24
  • $\begingroup$ I have deleted a few comments. Please keep discussion on topic for meta math SE. $\endgroup$
    – Alexander Gruber Mod
    Aug 25, 2013 at 16:42

You must log in to answer this question.

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