I was reading this answer, and it included the following:

Stack Exchange doesn't make it too hard for someone to create an account just to post crap. So far though, the benefits far outweigh the troubles, and although at times it gets a bit tedious for us moderators, let's just concentrate on the positives.

I know that Stack Exchange has had a longstanding policy of supporting anonymous participation, both in terms of not requiring a login for most posts and in terms of not requiring a real name to be provide as part of registration.

With respect to this policy, is any of the following documented?

  • rationale for establishing the policy in the first place
  • evidence in support of continuing with the policy
  • arguments against such a policy (or in support of a different policy)

To be clear, I'm not asking about the policy of enabling anonymous read access. I'm asking about the policy as it relates to write access.

  • 4
    How would you enforce real names? There's not much stopping me from saying my name is "John Doe", and I don't quite think it's appropriate for a non-financial website to ask me for government ID (which, irritatingly, Facebook does, but never mind that).
    – waiwai933
    Nov 24, 2013 at 22:09
  • I would like to authorize writing (at least voting up and answering) with a logged in account but with javascript disabled. I personnaly use the *SE answers through Google, in noscript+requestpolicy restrictions (namely, my day-to-day brower is lynx). And so I cannot vote up what helps me, without fireing up firefox. Nov 24, 2013 at 22:09
  • 1
    Facebook asks for an official ID in the UK @waiwai933. Are you sure? I'm fairly certain it doesn't (although I signed up a long time ago and barely use it). I'm fairly certain it can't as there's no single official ID and no requirement to carry one. Nov 24, 2013 at 22:12
  • @benisuǝqbackwards For names that their algorithms don't recognize as real names, they sometimes request ID to check. There was a little controversy earlier this year after their algorithms went a bit wrong. (That said, in the vast majority of cases, no ID check happens).
    – waiwai933
    Nov 24, 2013 at 22:16
  • More clearly: I would like the creation of the priviledge to upvote and answer without javascript. Set it at 1000 points, or lower. Objective: more participation. Nov 24, 2013 at 22:17
  • 4
    I don't understand this question because it makes allowing anonymous participation look like a choice. What serious alternative is there to it?
    – Pekka
    Nov 24, 2013 at 22:26
  • 7
    Many people are opposed to requiring real names in order to participate on a site, with some good arguments against such policies. Accounts are already required to post questions, and I don't think we want to put additional roadblocks in the way of answerers. Note that the troll referenced in that question had created an account, and made it fairly easy to figure out who it was. We knew exactly who our most prolific SO troll was, too. Nov 24, 2013 at 22:44
  • 2
    @probablyPekka I'm no expert on this, which is one of the reasons I asked the question, but as noted elsewhere in this thread, the policies of Facebook (required) and Google (encouraged) (support.google.com/plus/answer/1228271?hl=en) would be examples of serious alternatives. Of course, any policy can be violated, so this is as much about what culture you're trying to establish as anything else. Nov 24, 2013 at 23:17
  • 2
    @BradLarson Thanks for the reference; it was enlightening. FWIW, my interest is not so much around tracking down trolls, but on the general accountability/protection tradeoff. For SO, at least, I think a lot of this is related to where on the WoW <=> IEEE spectrum one things SO "is/ought" to be. Nov 24, 2013 at 23:26
  • I'm sure this has been talked about in one of the podcasts, but it's going to be tough to find that info without listening to them all... I remember several instances in which even requiring users to create an account was viewed as a barrier to participation. (That has since been established on SO, though, because of question quality)
    – Pekka
    Nov 25, 2013 at 0:03

1 Answer 1



As Pekka mentioned, this was first discussed fairly early in the site's development on the podcast:

We will never make you log in and create an account to do things. That's one of the things we kinda took out of Wikipedia, was that we wanted to make it sort of egalitarian.

Another sorta subtle difference is, on every question page, there's a box at the bottom inviting you to participate. You don't have to click through to another page, or do another thing, there's just a text input box and it's inviting you to participate by answering a question.

That is the unit of work on our site - asking and answering questions. You can log in - with your OpenID - or, there's your traditional "blog comment" name, email, homepage, and you can just type in whatever you want there and post an answer without having created an account.

(edited - full transcript available here)

Jeff goes on to reference Jan Miksovsky's classic blog post Hurdles at the entrance to a site, which describes the ridiculous amount of resistance most sites impose upon people who don't even know if they want to use the site yet.

With all these hurdles, it's a small miracle some web-based services end up with any users at all. Each hurdle constitutes an opportunity for the user to leave. The site is effectively asking the user, "Are you sure you want to use us? Are you really sure? How about now? Are you sure you're sure? Hmm?" Some users are going to take one of these opportunities and leave. People are growing increasingly leery of starting down the hurdle-strewn path of a new site. They've been down similar paths so many times that they've concluded the experience won't be worth their time unless they're already confident the site will provide substantial value.

Note that eventually Stack Overflow reached a level of traffic that made it prudent to impose this extra bit of resistance on people here just to ask questions. As you can probably surmise, one of the downsides of making it really easy to participate is that a lot of folks end up doing so when they clearly have very little interest in doing so positively - but the "cons" don't end there. If you listen to the podcast, you'll see a few of them called out early on, but here's a more comprehensive list:

  • Because the site is so usable with so little effort, many folks forget that they didn't create an account. When they try to log in later, with whatever credentials they thought they might've used, frustration results.

  • If you don't enter a valid email - or lose access to it later - there's no good way to regain access to an unregistered account if you lose your cookie.

  • If you don't care about reputation or the privileges that come with it, it might be easier to just keep creating new unregistered accounts by participating anonymously and clearing cookies. This can be done for malicious reasons, or just out of laziness... But even when the participation is largely positive, it can be confusing to other readers. We don't focus too much on social interaction here, but it is important - it's nice to be able to recognize someone who often posts good or bad answers.

  • Some privileges depend on having a stable account. You can always comment on your own questions (or on answers to them), but if you lose the account you used to post the question then you lose this privilege. Same goes for editing and accepting answers.

  • Speaking of privileges, unregistered users can't vote. This is intentionally done to prevent abuse, but it does tend to reduce the signal generated by folks who could otherwise be providing valuable feedback on the posts they're using.

Even on Stack Overflow, we still see the potential benefit of encouraging people to answer questions sufficient to outweigh the downsides of transient accounts. We're always working on making the transition from one to the other easier though.

"Real" names

First off, very few people use their "real" names on a regular basis, where "real" means legal first name + last name. In some cultures that isn't even a normal construct, but even in the US how often do you really identify yourself or address others in this manner? The truth is, you probably travel, conduct business, volunteer in your community, worship and hang out with family or friends on a regular basis without using your full, legal name... And much of this you probably do without using a name at all.

So you're attacking this from the wrong angle, assuming that a "real name" policy would somehow be the default and a system that allows anonymous or pseudonymous participation is a conscious deviation from this. It's not. Implementing and enforcing such a policy is difficult and can easily cause more problems than it solves - particularly when using an externally-identifiable name is likely to be of more potential value to the author than to others on the site.

###See also:

  • +1 and thanks for the thoughtful response, Shog9. I particularly appreciate the transcription effort and the references. I'm a little taken aback by your "Real names" section, though. I don't see how I'm "attacking" anything and I don't see how I've "assumed" any of the things you're suggesting I'm assuming. I really felt I was just seeking information. :-) Finally, I agree that people conduct a lot of business without using any name at all, but I honestly don't know of any non-virtual situations where people use some completely made up names. And most folks share their name if asked. Nov 27, 2013 at 8:01
  • I'm using "attack" as a synonym for "approach" or "address" in this context, @Peter. Yes, I'm making some assumptions as to your assumptions - I think this is justified given the way you stated the question: a "real name" requirement is never a default option, but something that must be intentionally built into a system and enforced on top of more mundane data validation rules - even when legally obligated to do so, this is a very hard problem to solve.
    – Shog9
    Nov 27, 2013 at 17:01
  • I don't see how this has anything with "default behavior". SO was "designed" and every design decision was a conscious one. I assume some choices involved considerable work to implement and others were "easy". I simply asked for information on rationale/alternatives associated with this particular decision, which you provided. Also, simply removing the "optional" from the registration form would be a straightforward change that could be made which would most likely increase the number of real names provided (for better or worse), so not all steps towards a "real name policy/culture" are hard. Nov 28, 2013 at 3:11
  • Not requiring registration was a conscious choice - I've provided explanations for why that choice was made. Not enforcing real names may have been a conscious decision as well (if you listen to that podcast episode you'll hear Joel talk briefly about the concept of "verified" identities) however without any evidence to support such a conscious decision I'm reluctant to call it one. By the same token, I would be reluctant to call the lack of unicorn-themed easter eggs in comments a conscious design choice; there may well be good arguments against them, but I doubt they were ever even needed.
    – Shog9
    Nov 28, 2013 at 3:19
  • Wow. So you're equating (say) the option to not to include the text "(optional)" after the "Real Name" field as comparable to the decision not to include unicorn-themed easter eggs? That seems like a very odd analogy. Not only is there no precedence for unicorn-themed comments in real of physical worlds, but not implementing them requires zero effort. By contrast, there is a substantial precedence for requiring real names in physical and virtual registration and making them optional requires extra effort (i.e. inclusion of the "optional" text). Nov 28, 2013 at 4:37
  • When's the last time you walked up to someone you'd never met to ask them for assistance in solving a problem, but first demanded that they provide you with their full, legal name (offering your own in exchange, of course). I remain unconvinced that such a requirement is either obvious, natural or traditional, outside of bureaucracy. The real question - to date unanswered - is why this field even exists at all...
    – Shog9
    Nov 28, 2013 at 4:43

You must log in to answer this question.

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