22

There are a set of questions that are a bad fit for Stack Exchange. Every site has a /help/dont-ask page that describes what SE sites are looking for and what kinds of questions aren't a good fit for SE sites. Sometimes, the questions that aren't a good fit become very popular - they are shared and linked to on various sites around the Internet or otherwise generated a large number of views and answers. However, deletion may not be appropriate because they have useful content that should be preserved because it doesn't exist elsewhere or deletion may cause 404 errors to unregistered or low-reputation users coming from other sites. We could use better tools for managing these "bad", but very popular and useful questions.

Below are a number of suggestions to address this:

  • Change the text of the "historical lock" reason. Right now, this reason says that the lock exists for questions that have "historical significance". However, some questions can slip through the cracks for a small period of time (6 months - 1 year). To me, this isn't enough time to be "historically significant" to a site. The idea that this question is not a good question or an on-topic question is good (and links to the /help/dont-ask and /help/on-topic pages), but not the historical reason. This reason should reflect that the question and its answers add value to help make the Internet a better place.

  • Autogenerate moderator flags when popular questions are either closed or deleted by the community. Popular could be determined by a number of factors such as views on the question, votes on the questions, votes on the answers, number of answers, answer score, and/or average answer score.

  • An additional lock type that prevents deletion or reopening of a question, due to its value, but allows people to leave comments to note issues with the posts (possible reasons for commenting that would be valid: URLs that have gone dead and new URLs, revised information that makes part of an answer out-of-date, or links to related questions or postings on the site or other sites). It may be worth discussing other permissions that could still be granted, such as editing the question or answer, voting on the question and answers, or voting to delete low-quality answers. Per the comments, this could be a repurposing and update to the wiki answer lock reason. (Personally, I feel that repurposing the historical lock would be more appropriate).

What else can / should we be doing to balance the desire to close questions that aren't a good fit for a site and prevent people from using old, open and unlocked questions as evidence that they can ask their question, but also ensure that content lives on appropriately and is accessible to others, even if it's not the best match for the current definition of an SE site?

  • 2
    I guess new type of lock would be the ideal solution, or "partial lock" that allows voting and comments but block deletion of the question itself. (since many times users are eager to delete closed bad questions) – Shadow The Princess Wizard Jun 1 '15 at 15:05
  • 1
    @ShadowWizard I incorporated the idea of a new lock type into the answer. – Thomas Owens Jun 1 '15 at 15:11
  • 2
    You'd want to both prevent deletion and also prevent it from being reopened. – Servy Jun 1 '15 at 15:15
  • We perpetually get questions on Movies&TV about the images in our header... which is really a Meta question. There's a lot of discussion going around about it but we don't seem to want to use the historic lock because it's not actually a good question, so we don't want to encourage it. If we could have a similar lock we could either customize the text of or something similar, that would be very helpful. – Catija Jun 1 '15 at 15:27
  • @Catija In that case, couldn't you migrate to Meta? Or are they too old to migrate at this point? – Thomas Owens Jun 1 '15 at 15:28
  • @ThomasOwens Migrating to Meta doesn't prevent future questions. If we could lock it on the main site, we could mark subsequent questions as duplicate and the existing question would show up when people search for it, so it could likely reduce the chance someone would ask. – Catija Jun 1 '15 at 15:29
  • consider editing to explain what's the point of allowing people to leave comments. If it is meant for meta discussion, this can be done at site meta, or did I miss something? – gnat Jun 1 '15 at 18:33
  • 1
    @gnat This is done. – Thomas Owens Jun 1 '15 at 18:35
  • 2
    The third bullet point should be covered as far as functionality by the "wiki answer" lock although the intended purpose is completely different (and I'm not sure that it allows comments on the question itself). – jscs Jun 1 '15 at 18:35
  • 2
    The ideal solution would be not holding a site and its mods (diamond and citizen both) hostage to these questions in perpetuity just because of their popularity. I've said it before and I'll keep saying it: if this stuff is so damned valuable, then those who think so should take some initiative in preserving it, not throw fits when others vote to not have to deal with it anymore. If the content is so stellar, search engines will find it wherever it goes: that's their entire purpose in life. – jscs Jun 1 '15 at 18:36
  • @JoshCaswell I think it's slightly different than wiki answer. Although maybe revising wiki answer would be suitable. – Thomas Owens Jun 1 '15 at 18:36
  • 2
    Yes, I should have been clearer: semantically it's totally wrong, but the functionality is there. – jscs Jun 1 '15 at 18:37
10

Lemme see if I understand this...

  • You have questions that don't fit the most strict interpretations of the current definition of a good question.

  • These questions have answers that are valuable enough to others that you don't want to lose them.

  • The questions aren't "dead": the answers are still worth maintaining, or potentially even augmenting with new or updated information.

  • You don't want new users to model new questions on these existing questions.

Sounds like you want to have your cake and eat it too. I'm afraid you're not going to be happy with this answer...

Hard decisions

There's a certain bias that tends to afflict decisions regarding a site's scope as it is refined and focused over time: folks are much more likely to discuss problematic questions than they are those that are simply answered without much fuss. Much hay is made from questions that spark bitter, opinionated arguments, attract trollish rants, or simply become overwhelmed by vapid commentary; in comparison, questions that garner a few well-thought-out answers tend to become an issue only after the fact: whether they're quietly closed and deleted or ignored until used as justification for a trainwreck, these questions then become the dry tinder ignited by someone else's spark.

Whether it happens within 2 hours or 2 years, sooner or later someone has to make the hard decision: is this question problematic in and of itself?

If you can answer that, then you can decide what to do with the question. If you can't, then you should do nothing; there is no middle ground. There is not, and cannot be, a special "this question makes me vaguely uneasy" status.

That's it. End of answer. I warned you that you weren't going to be happy with it...


Appendix A: broken windows

When discussing questions that are not inherently problematic themselves but which can be made to match a problematic stereotype, a common lament is that they will encourage copy-cat questions, which will not all be so benign.

This can be true. I've watched it happen; a savvy newcomer always looks around first, and tries to model his contributions on those that appear to have been well-received in the past. If the top questions on the site set a bad example, you will get more of them - and often from the very people you would hope better of. I experienced this first-hand: years ago, I joined a site where the front page was perpetually filled with vapid, opinionated (but entertaining!) write-ups, and immediately jumped in to write my own vapid, opinionated (but not very entertaining) post.

That said, this is an argument against leaving problematic questions around; it is not an argument against good questions that fit a bad template. You want to avoid sending the message that carelessness and outright vandalism is expected; do this by removing questions that were answered carelessly or in bad faith, not by removing the counter-examples! The oft-cited broken window theory doesn't apply to perfectly-good convertibles.

Most of all, avoid letting "clever" askers lead you by the nose when you observe problems with their questions; them pointing out a superficially-similar question does not compel you to remove it before dealing with the problem at hand. I'm fond of the response, "Earn 3K so you can vote to close it if you find it troublesome; I'm busy dealing with the problem at hand." - but I'm sure you can find a more polite way of saying roughly the same thing.

Appendix B: historical locks

The purpose of this tool is simple: get broken windows, real ones, off of the list of top questions without breaking The Internet. It was introduced because we'd built up a sizable archive of questions that were, quite frankly, terrible - but which were also heavily cross-linked, both within the network and on the 'Net at large. Removing them sparked an outcry; keeping them encouraged more of the same.

Therefore, Lock for Historical Significance was implemented to be the deletion-without-removal. Questions thus locked disappear from the homepage, from the lists of top questions and tag lists. They can't be voted on or modified. They are the closest we could get to just replacing popular deleted questions with a redirect to archive.org, and should be used only when your other option is to delete a persistently popular question. Don't get hung up on the name; "historical significance" is a euphemism for "lots of inbound links" and nothing more.

If a question still has lasting value to your community, then Historical Locking is the wrong solution, and a work-alike lock with a different name would be no better.

Appendix C: autoflagging for popular questions

This already exists, sorta: there's no flag, but popular questions are automatically put into the reopen queue when closed.

Moderators can always search the list of deleted questions and sort by score or filter by view-count to get a list of other potentially-lost treasures.

  • 1
    If you can answer that, then you can decide what to do with the question. If you can't, then you should do nothing; there is no middle ground. The problem is I don't know what I'm (community member or moderator) supposed to do. My options are to do nothing and leave a possibly broken window on a top questions list, close the question and hope people don't delete it, delete it and destroy inbound links (those 10000+ views in 12 months probably came from somewhere popular), or historical lock it. (1/2) – Thomas Owens Jun 2 '15 at 0:18
  • 2
    It feels like the right thing to do is to lock. It has a lot of views that came from somewhere. Maybe a really popular blogger linked to it. Or it's a top post on Reddit. Somehow, a lot of people found it, and if it is extremely well up voted (the question and answers), then people also found it useful. I don't want to destroy someone's post and lose the content that was found useful. Maybe I'm caught up on the word "historical", because a question asked 12 months ago is generally held to the same standards as one asked today, yet historical lock implies that it was an acceptable question then – Thomas Owens Jun 2 '15 at 0:21
  • 2
    If you don't see a clear problem, then there's no harm in doing nothing, @Thomas. If there is a problem, then edit or close or delete or lock. In my experience, problematic questions are generally pretty easy to observe, especially once they have answers - "the weed of crime bears bitter fruit", as The Shadow used to intone. If you aren't seeing it, it probably isn't there. And yeah, 12 months is ages in Internet time... – Shog9 Jun 2 '15 at 0:28
  • @ThomasOwens one can argue that a question that attracted many views becomes part of site a history, no matter if it's old or recent – gnat Jun 2 '15 at 21:00
-6

I don't think this is a problem.

Questions that have historically been good, or that have a strong significance to the site, but that no longer match the site's purpose, should be locked. That question was asked during a point in time when it was valid to ask, should be left alone.

Some questions like this are quite good, and serve as an example of what can and cannot be posted on a site, regardless of quality. And they reflect an important distinction that all SE sites should keep in mind.

A Close Vote Is Not A Super-Down Vote

Questions that are low-quality can still be relevant to a site, just as very well-written questions can be good, but inappropriate for a site. The former should be down-voted with suggested edits for improvement, the latter should be close-voted with suggested edits to make it fit the site, and sometimes the two will meet, but they serve different purposes, and should not inherently cross.

An Alternate Lock

An alternate default lock for this type of question might be a good idea though - but that seems like something that each individual site should handle, because each site will have its own reasons for locking certain posts.

  • 4
    As a moderator on Programmers, I can tell you that it is a problem. We have many questions that are 1-2 years old (written after the change in scope for the site), but they are too broad, primarily opinion based, or off-topic. For these questions, "historical lock" doesn't make sense. The only option is closure, but that allows the community to vote to reopen these "bad" questions or vote to delete some questions that shouldn't be deleted. – Thomas Owens Jun 1 '15 at 15:23
  • 2
    It's also a problem for "good" questions during a site's beta, which become out of scope through the beta phase of a site's development. – enderland Jun 1 '15 at 15:24
  • 1
    @enderland: That does seem like a good match for historical lock, though. – Nathan Tuggy Jun 1 '15 at 19:24

You must log in to answer this question.

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