A few years ago, a policy change was made so that unregistered users could no longer post questions, only answers. The reasoning given by Jeff Atwood for not requiring registration across the board makes sense...

I found in my years of blogging at Coding Horror that the value of one person happening by with some gold nugget of absolutely the right information you need vastly outweighs the cost of ongoing moderation of anonymous posts.

...for sites like Stack Overflow, Super User, and Server Fault, as well as a couple others. It can absolutely happen that a random person can drop by, see a question they know the answer to, write it out real quick, and drop it on the question to the great benefit of the asker.

As a moderator of Christianity.SE...this doesn't happen. The community's standards have evolved since the early days so that answers are expected to be substantial and have support and references. The VAST majority of unregistered-user answers that I have seen have some combination of the following problems:

  • not an answer (at all)
  • not an answer (wrong scope/perspective)
  • poorly written and/or formatted
  • no support or references
  • treating SE like a forum

It's bad enough that when I'm going through flags, if I see a user has 1 reputation, I look at their profile to see if they're unregistered. If they ARE unregistered, I am far more likely to just delete their answer (with a comment explaining why), especially if they haven't been around for more than a day, because I don't expect them to come back and fix the issues with their answer(s). If they are registered, I am much more reticent to take such actions.

In short, I understand Jeff Atwood's reasoning, and I acknowledge it works fine for technical sites. For a site like Christianity.SE, I honestly think that an unregistered user will practically never contribute a good answer. There's, well, pretty much no such thing as "absolutely the right information" on a site that is scholarly in nature. In theory, I do see that there could be such questions like "What work does this citation, <cryptic letters>, refer to?" that could in theory be answered perfectly by an unregistered user. However, in practice, such questions are extremely rare, and rarer still would be that a random person would just happen to know the answer.

My solution would be to enforce registration for questions and answers on some sites, like Christianity.SE. I honestly don't know how feasible this would be technically, and I understand if it's an unreasonable request. However, this issue has been bothering me for a few months now, and requiring registration would go a long way towards keeping average content quality high.

tl;dr: I believe that with some sites, the chance of some person happening by with a "gold nugget" of just the right information is practically zero, so there is a net detriment to allowing anonymous posts.

  • 3
    The site you're posting this on requires registration to answer... – Shog9 Aug 19 '15 at 5:10
  • 5
    @Shog9: Ah, didn't know that. Still, MSE could be considered a "special" site in a sense. That does address the technical aspect though, so thank you for that! – El'endia Starman Aug 19 '15 at 5:12
  • while i can totally understand and get behind the reasoning behind your post, i'm not sure it's actually attacking the problem from the right angle. I.SE also gets a lot of problematic first posts like you're describing here, but from registered and unregistered users alike: I'm not convinced that just enforcing registration would really make a huge difference. – goldPseudo Aug 19 '15 at 7:02
  • 2
    If MSE has this already, it seems like a no-brainer to me. It certainly won't solve the problem of bad first posts (nothing ever will), but it'll help at least a little and has no real downside for sites like this. – Ixrec Aug 19 '15 at 7:08
  • Yes, I think we can enforce users to signup for posting answers on religious sites. – Pandya Apr 29 '18 at 11:02

Jeff did it a certain way based on the circumstances at the time, but 4 years have passed by since then and so we have to consider if the sites and user base(s) has grown and changed to the point where there is no benefit from allowing anonymous answers.

Maybe this is very site specific based on niche rather than size. This is a valid consideration, as certain niches (such as sensitive topics) could attract anonymous users more than others.

tl;dr: I believe that with some sites, the chance of some person happening by with a "gold nugget" of just the right information is practically zero, so there is a net detriment to allowing anonymous posts.

This is speculation, but in the absence of facts/stats I agree.
It might have been the case years ago with a lot less users etc, but not necessarily the case now. Since 4 years ago, we have a lot more users.
We also have more sites in the network, and the sites can generate traffic to each other, from related topics, common/popular interest overlaps, the "Hot network" list, etc. More traffic means potentially more answers.


So a potential loss of an anonymous "golden nugget" is rare because it has to specifically be the scenario whereby:

  1. A user has an answer that would be a great loss if they could not post it
  2. And, as a non-user they understand what makes a great answer and so post a well written, well formatted (etc) answer
  3. And they would not otherwise sign up to post the answer
  4. And other registered users have not already posted a good or great answer which covers at least most of what the anonymous would have been

On busy sites, I think we have enough users now that it would be quite rare for the above scenario to play out.

And on smaller sites (including betas and Area51), and the rarity of this aside, I think we should be enticing users to register by not allowing anonymous posting, in the attempt to get multiple "good" answers rather than just one "great" answer from a passer by.

Need stats

However, without hard stats we're speculating and guessing too much, and stats would show if there has been any great answers posted by anonymous users recently, and potentially based on specific site(s) and niche(s).

Such as what percentage of total anonymous answers:

  • Are not deleted
  • Didn't need a substantial edit to make "great"
  • Have a positive score
  • Have a score greater than other answers on the same question
  • Is the only answer, is upvoted, and X time has passed by


I'm not the best person to get such stats, however.


Some good stats are very much needed here to provide facts, so we can remove speculative theories and weigh up the pros and cons.

It's possible, or likely, that the benefits of getting people registered and answering with their "golden nuggets" and then more activity, outweighs the low risk of any potential loss of a "golden nugget" because they only want to post anonymously.

  • 3
    Emphatically agreed on the stats. – Deer Hunter Aug 19 '15 at 9:46
  • The stat I would like to see is how many currently productive registered users started out their user life-cycle with a post as an unregistered visitor. – Caleb Jun 29 '18 at 13:00

You must log in to answer this question.

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