8

We currently have over 800 protected questions. Some have been protected since 2011.

Isn't it time we unprotected questions that have been protected for longer than, say, a year?

Rationale

I see protecting a question as a temporary measure, to let a question cool down a bit. Once it has cooled down and is no longer likely to attract low quality answers by new or drive-by users, there's no need for it to be protected any longer.

| | | | | |
  • I knew there were a lot of questions that didn't need protection anymore, but 8k?! Wow. – Gallifreyan Jun 2 '17 at 17:29
  • 2
    Ah, it's 800. Phew. Gotta clean my glasses. – Gallifreyan Jun 2 '17 at 17:35
  • 3
    That's like 2% of total questions...seems pretty reasonable to me. – Paulie_D Jun 2 '17 at 17:42
  • I've unprotected scifi.stackexchange.com/a/11621/19561 and scifi.stackexchange.com/q/17705/19561 just for the heck of it (I wanted to test unprotecting). – SQB Jun 2 '17 at 17:46
  • 5
    What will we really gain by unprotecting them at this point? Spot checking them, many of these were protected long after the questions were posted, and the low-quality answers were spaced apart. They weren't all just from short floods of LQ answers. – phantom42 Jun 2 '17 at 17:49
  • 2
    It takes 1 upvote to clear the protection. With nearing 40k questions on the main site, they should be able to find something else to answer first. – phantom42 Jun 2 '17 at 17:53
  • 11
    We have plenty of new users grabbing old questions and posting very silly answers. Honestly I've seen more of that kind of review than new question reviews, with less to show for it. It feels like they're just finding it by google search and then contributing random thoughts when they're not familiar with the site - and older and more popular questions are going to continue to be a target for that kind of use. I see no reason to unprotect questions like that at this point. – Radhil Jun 2 '17 at 18:07
  • 6
    "Once it has cooled down and is no longer likely to attract low quality answers by new or drive-by users, there's no need for it to be protected any longer." going down the list of the protected questions and looking at the deleted LQ answers, this does not seem to be the case. LQ answers a year or more apart in many cases. – phantom42 Jun 2 '17 at 18:13
  • 1
    I think there's a FR on MSE requesting that protecting questions due to HNQ status be time-sensitive... and if there's not, there should be. meta.stackexchange.com/questions/105097/… – Catija Jun 2 '17 at 18:13
  • 2
    @phantom42, Paulie_D, and Radhil all make good points. PLEASE don't start unprotecting questions en masse, at least not without further discussion here first. – Rand al'Thor Jun 2 '17 at 23:26
  • @SQB And sure enough, the first of those questions you unprotected did attract another unsourced answer from a new user after unprotection. – Rand al'Thor Aug 8 '17 at 14:32
25

Most protected questions shouldn't be unprotected ...

... for exactly the same reason they were protected originally: they attracted low-quality answers before and would be likely to attract them again. E.g. this question with 10 deleted answers over the course of three years: if we unprotected it now, it would probably just keep on attracting bad answers until it got protected again.

There are rare cases of questions protected because they're the target of concentrated troll attack (like this question with 11 near-identical answers in one day), and those can be unprotected once the attack is over. Or, perhaps slightly less rarely, questions protected because they attract many bad answers in a short period while they're on HNQ (or maybe for some other reason - HNQ isn't necessarily the root of all evil). Again, if the low-quality answers are tied to a particular time period rather than to the nature of the question itself, feel free to unprotect.

But in most cases, a question is protected because it's the kind of question that attracts bad answers, for whatever reason (of course this doesn't necessarily make it a bad question), regardless of the passing of time. These questions almost certainly shouldn't be unprotected.

(This answer is mostly based on an explanation I already gave to @Gallifreyan in chat.)

| | | | | |
  • Certainly you're not saying that every protected question should stay protected? And certainly you don't mean to go against the official policy (heresy!) which says questions shouldn't stay protected forever unless it's really needed? – Gallifreyan Jun 2 '17 at 23:27
  • 10
    @Gallifreyan 1) No, not every, but most. (See the first word of this answer.) They were protected for a reason, after all - what you need to do is to look at that reason and see whether it still applies, and from my experience on this site, the answer is usually yes and occasionally no. 2) What official policy? The one that says "Do unprotect questions that aren’t currently attracting a lot of attention and don’t have a long history of unproductive answers"? I'm not going against that - I'm telling you that most protected questions do have a long history of unproductive answers. – Rand al'Thor Jun 2 '17 at 23:35
  • 1
    @Gallifreyan if the official policy is to unprotect questions then someone should tell Stack Overflow. (If you can't see that yet: 20,890 protected questions dating back to 2008). – KutuluMike Jun 6 '17 at 13:09
  • HNQ isn't necessarily the root of all evil. Uh, well, I talked to some of the Old Ones, and they say they spawned from HNQ, and that it was really HNQ that tempted Eve in Eden. So, you're clearly wrong, there. – user31178 Jun 16 '17 at 12:24
  • 1
    Your thoughts here? interpersonal.meta.stackexchange.com/q/2214/345 – NVZ Dec 5 '17 at 7:54
  • @NVZ Sorry, I'm no longer interested in putting much time into that site. Feel free to copy anything you want from my answer here over to there though. – Rand al'Thor Dec 5 '17 at 9:46
  • Okay, I've written something similar to yours. I wish you would stay on IPS, though. – NVZ Dec 5 '17 at 10:58
6

It sounds like you'd rather make your job easier than allow potentially valuable content to be added to this site

To add to what's been said, a lot of your perspective on Protected changes after you hit 10k and you can see the deletions. When you go to a popular Protected question and you see the wreckage... you begin to understand why it exists.

And why it's rarely removed.

Understand that a lot of new people come to the site (especially when they're in the Hot list) and they want to just gush on the topic. Hot simply exacerbates the problem. You want to at least raise the bar to keep the totally new out, lest they make for moderation work. Wanting to reduce this is NOT a bad thing. I'm continually impressed that SF.SE keeps the moderation to a minimum. I'd like to see it stay that way.

| | | | | |
  • It looks like the user the left that remark has plenty of rep on other stacks, so I'm not sure this is terribly constructive. (... but I hear you. My personal most-inspired-headache wreckage had the exact same answer posted three very different times over the course of a year by the exact same user. I'm still confused how.) – Radhil Jun 3 '17 at 19:12
  • 3
    Add to this that IMHO the slightly more "trivial" and "fun" topic of the site enhances issues like that. Not only do most people have something to say about their favourite work of art, they might not always be used to discussing it in a little more focused and professional way either. – TARS Jun 3 '17 at 19:56
  • 1
    Hot questions aren't really the main issue here, because hotness is temporary. Questions which are only protected because they're HNQ can be unprotected after a while. – Rand al'Thor Jun 4 '17 at 15:36
  • I’m editing in a reference, since it’s not obvious who you’re quoting. – Adamant Jun 4 '17 at 22:16
2

Yes.

Just in case - here's a query to find all the protected questions, even if you don't have 10k powers yet.

This FAQ main meta post reads:

  • Do unprotect questions that aren’t currently attracting a lot of attention and don’t have a long history of unproductive answers.

And to quote some bits quoted in my other answer on this topic:

Judicious use of this feature is critical to allowing these sites to handle large amounts of external attention, but over-using it breaks the system: Stack Exchange sites depend on a constant influx of new blood, both to answer new questions and provide updated information on old ones. When in doubt, err on the side of letting new users prove themselves before locking them out.
Changes and guidelines for the Protected Question status on Stack Overflow Blog

It seems to be more or less official policy that protection is not meant to be permanent, and that it should be used to deter only multiple bad questions in short periods of time.

There have been plenty of reasons given (in linked posts) that protection is in a sense against the spirit of Stack Exchange, which relies on new users and their answers.

I sincerely doubt all ~800 questions have to stay protected to this date. Protection, in my opinion, should be used when there's an influx of bad answers or non-answers that is easier to block altogether than to handle via review queue. In this case I think we are more than capable of handling new answers to old questions, since we are not expecting them to flood the site.

Note also that bad answers are still answers (which has been discussed in chat recently and prompted this question by Mithrandir: When should 20kers vote to delete an answer?). They can be improved, if not by the poster, then by us, and they would still be valuable.

Which ones to unprotect?

The age of the question, the number of deleted answers, and the nature of the question is what I think should be taken into account:

  • Questions from the beginning of time (2011) should probably be unprotected - their age of glory has passed, and it's unlikely we'll experience a flood of bad answers for them.

  • Questions with few deleted or low-quality answers should probably be unprotected - if they didn't attract bad answers then, why should they do so now?

  • Questions with touchy subjects, or questions that are prone to "enthusiastic" fan answers, a la the "Cap vs. Spidey" question, which got a lot of answers comparing the two characters instead of listing occurrences of them fighting.

| | | | | |
  • 3
    If a question is unprotected and then starts attracting more crap, it's easy enough to re-protect it. – Catija Jun 2 '17 at 19:38
  • 6
    @Catija But I would strongly discourage people from unprotecting questions en masse. Finding them all to reprotect would be a hell of a job, unless we keep a list somewhere beforehand. It's also a lot of unnecessary effort, especially if we can already tell they shouldn't be unprotected because they've been attracting bad answers regularly for years on end. And in the meantime we'd have to be deleting new bad answers on those questions, thus creating even more unnecessary work. – Rand al'Thor Jun 2 '17 at 23:24
  • 1
    @Randal'Thor I'm surprised at that attitude... It sounds like you'd rather make your job easier than allow potentially valuable content to be added to this site. We're talking about hypothetical bad answers that may or may not ever be added. You don't need to keep a list of questions that are unprotected because if they don't get more bad answers, you don't need to protect them again. – Catija Jun 2 '17 at 23:45
  • 6
    And you're talking about hypothetical good answers that may or may not ever be added. Again, the bar to get past the protection is a measly 10 rep. I'm not against unprotecting some questions, but my unscientific check of the first 20 or so in the list didn't have any that stood out as really "needing" to be unprotected. – phantom42 Jun 3 '17 at 4:30
  • Having read through some of these answers and the last meta question a bit more I guess I see the opposing point, but this just seems to be a point of ... conflicting pressures, I guess? We want to encourage new users and answers - but we also want good answers, and we are given a lot more tools to manage that end of the deal. Rand's thoughts seem like a good compromise between the two. – Radhil Jun 3 '17 at 11:25
  • I must say that i disagree — leastways in accordance to quasi-bureaucratic agreement with the validity of the reasons against your proposal as have been expressed on the Meta Stack Exchange site. (cough.) If there is reason to unprotect a question, then yes: most certainly. Our present placement in the temporal spaces can certainly be part of that reason; the amount of time elapsed since the question was initially protected, however, is not. – can-ned_food Jun 4 '17 at 7:39

You must log in to answer this question.

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