So I have just run across the fact that this question got protected, and I was wondering why that was done. After carefully looking at this, I meant to ask Daniel what was his intention — just to discover that I already did so 1.5 years ago...

I have often wondered what questions are protected for. Maybe I am only looking around in the wrong set of tags, but I have rarely found any "Thank you!!!" answers, while I have seen numerous questions protected either for reasons I cannot fathom, or in vain, since there's enough users who got a single upvote, but have no idea about how the site works, and slip in their useless answers right through the protection.

So am I totally wrong here, or is protection often applied either needlessly or uselessly?

  • 2
    I've seen some questions needlessly protected as well, and even unprotected them in certain cases. Questions that are protected by the community user, though, are almost always protected with good reason. Some questions may also be protected early when they get posted on sites like Reddit or Hacker News. – BoltClock's a Unicorn Oct 8 '12 at 14:19
  • stackoverflow.com/a/5249667/47064 While an example can sometimes help, the user didn't care to spent any words or code comments on what the various ways are used for (or when to use them) nor did he come up with any details. It's pretty easy to (copy and) paste a piece of code, but that doesn't make it an answer. You can imagine the quality of answers by other low reputation users, often much worse... – Tamara Wijsman Oct 8 '12 at 14:22
  • @TomWijsman The question is protected, not locked. Protection does not prevent from such answers being posted AFAIK. And I hardly think that's the reason to protect a question to begin with. – Bart Oct 8 '12 at 14:24
  • @Bart: Unless the user didn't have enough reputation at the time it was posted. – BoltClock's a Unicorn Oct 8 '12 at 14:25
  • I've protected questions that were getting extremely low-quality answers (even if it wasn't spam or "me too" answers). I've also protected something as a attempt to keep it from getting auto-wikied from 30+ answers. – Mysticial Oct 8 '12 at 14:27
  • @BoltClock: This one didn't, but that doesn't change the fact that Tom's original assumption was wrong. – sbi Oct 8 '12 at 14:27
  • @BoltClock'saUnicorn Ah, d'oh. That's true of course. There is the 10 rep minimum. Should have thought of that. – Bart Oct 8 '12 at 14:27
  • @TomWijsman Yeah, I forgot about the 10 rep minimum to post an answer to a protected question. I therefor assumed you were confusing locking with protection. My bad. – Bart Oct 8 '12 at 14:29
  • @sbi: Bart was confused, the original assumption (protection) was right; I didn't mention any form of "lock". – Tamara Wijsman Oct 8 '12 at 14:31
  • @sbi: Oh, the 10-rep requirement? Yeah, that hardly works on a site like SO anymore. – BoltClock's a Unicorn Oct 8 '12 at 14:48

In my experience it's done because a question is extremely popular on Google, the Stack Exchange Hot Questions list or some social network (reddit/hacker news mostly).

When this happens a flood of users join/view the site who have never used it before. Maybe they rush to answer the question with zero understanding of how to answer on Site X, maybe they don't know what Stack Exchange is and start posting Forum style replies, but the common thing is a flood of 0 experience users results in a lot of "answers" from people who don't know how to answer this question on this site.

It's worse on "bikeshed" ish questions where on Site X we have strict criteria and don't repeat ourselves with new answers, but Little Johnny Forumgoer doesn't know that and decides to share his opinion rather than an answer.

Protecting stops that, and it does it's job fairly well in my experience. Sure some bad posts may slip through from users of the site, but that's just day to day stuff. Protect keeps the post from drowning in noise.

  • Pretty sure you mean 1-EXP users. – BoltClock's a Unicorn Oct 8 '12 at 14:24
  • That explains why protection was invented and how it should be used. I am, however, questioning that it really is used that way. – sbi Oct 8 '12 at 14:28
  • @sbi I couldn't speak for SO, but I use it for that on UX and Workplace, and I see it done (and request it) fairly often on Gaming. Gaming in fact often gets protect automatically. – Ben Brocka Oct 8 '12 at 14:36
  • 1
    @sbi I'm not sure what answer you're looking for if not this. We can't know for sure why people are protecting posts; I often see non-mods protect posts that were linked from Reddit, but for all I know people were just bored and protecting posts at random – Michael Mrozek Oct 8 '12 at 14:37
  • I think he's asking about SO's situation. We have more users with more opportunities to do more things wrongly, after all. – BoltClock's a Unicorn Oct 8 '12 at 14:39
  • @MichaelMrozek: I am not looking for an answer, I am trying to get a discussion going. Hence the discussion tag. – sbi Oct 8 '12 at 14:44

Protection is a bit of an odd feature to give to regular users, and I'm not convinced that it is really needed. The later introduced auto-protect when three posts from new users are deleted pretty much removes the need to involve the community in protecting questions.

When I protect qestions, it is usally triggered by some "low quality" or "not an answer" flag on a question that is getting a lot of views from somewhere. So there is at least one deleted non-answer from a new user when I protect.

Protection is sometimes overused, and there is a bit of a lack of guidance for users that gain this privilege. It has a cost that is often overlooked, as it can prevent valuable answers from outside users.

  • 1
    +1 Since the auto-protect got introduced the manual protection is indeed somewhat useless except for some exceptional occasions where a really high end reputation user or moderator deems early on it is a benefit to protect the question. – Tamara Wijsman Oct 8 '12 at 14:39

I'm probably an example of what you would call misuse. When I see a single first/late answer (e.g., a user whose first or near-first contribution is an answer to an old question) that adds nothing new, I protect the question. (I often also vote to close the question, since this often happens to old poll/opinion questions.) I reason as follows: protecting does no harm, and if, in fact, the question has been called out somewhere, I'm protecting it from the stampede.

But I don't wait for actual evidence of a stampede.

As I see it, there should just be a higher rep bar to adding an answer to an old question with plenty of existing, upvoted, answers. Then this wouldn't come up.

  • +1 I do the same except that my threshold is 2 low quality answers. – Mysticial Oct 8 '12 at 14:35
  • -1 because I disagree with the first part of your answer (there usually is no reason to believe that a question will be picked twice for a late-coming newbie answer), but +1 because I agree with the last paragraph. So it's +-0 from me. – sbi Oct 8 '12 at 14:42
  • 1
    Now that you mention it, most (> 95%) of late first answers to old questions with massively upvoted answers are complete crap (usually worth of deletion). – Mysticial Oct 8 '12 at 14:46
  • @sbi: Multiple late-coming newbie answers can pretty much be expected on the share of questions that cause a high influx of new users. There are always some users between them that pretend this is a forum and just feel the need to post an answer regardless its quality. But well, that's basically what protection is for: Prevent answered famous questions from needless bumps and noise injections. – Tamara Wijsman Oct 8 '12 at 14:50
  • @TomWijsman: As has been said before, if there's a wave of crap answers from new users, the question gets auto-protected, so your argument is moot. – sbi Oct 8 '12 at 14:51
  • @Mysticial: And what's the percentage of questions that's afflicted by this? If it's high, old questions should be auto-protected. If it's low, protecting doesn't help, because usually it's unlikely that some other newbie will do the same. (And if, auto-protection kicks in anyway.) – sbi Oct 8 '12 at 14:52
  • @sbi Not much. I don't have statistics, but it's enough for me to notice. The loop question that I answered back from December picked up multiple low-quality late answers before I protected it. (and it isn't even a bike-shed question) – Mysticial Oct 8 '12 at 14:56
  • 1
    @sbi: It is not moot, because it is part of the paragraph you decided to disagree with: protecting does no harm, and if, in fact, the question has been called out somewhere, I'm protecting it from the stampede. Famous questions get picked out easily, hence if we see such question with one answer we choose to protect it as we know more answers are likely to follow. – Tamara Wijsman Oct 8 '12 at 14:56
  • @TomWijsman: We're running in circles. This has already been argued against: it does harm, and you don't need to protect from stampede. Why do you keep bringing up the same old arguments despite? – sbi Oct 8 '12 at 14:59
  • @sbi: Can you please tell me where you have read my first comment before? If not, we're not running in circles and you are just drawing a straw man argument. – Tamara Wijsman Oct 8 '12 at 15:03

You must log in to answer this question.

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