3

Looking at our closure stats with the new moderator tool, I notice the following:

  • Over the past 30 days, we've closed more than 40% of the questions asked. That's a worryingly large amount.
  • Of the questions closed, approximately 1/3 are either "too broad" or "unclear what you are asking", few are edited, and almost none are reopened.

Sampling a few of these, it looks like they are typically being put on hold in ~2 hours, which is way too fast for most OPs to respond. I thus have a concern that we are being too fast on the trigger. Many of these may have a good question hiding in there, and a fast close may cause OPs to become discouraged and abandon their questions, particularly because reopening is typically much harder than closing on this site.

I would thus like to suggest that, just like we give comments a 24-hour grace period before flagging for obsolescence, we should give "Unclear" and "Too Broad" questions a 24-hour grace period before voting for closure.

To be clear: I am talking only about "Unclear" and "Too Broad" - I'm happy to see most other closures happening quickly.

  • 13
    To me it is not the fast closing that is a problem, it is that reopening is difficult. – StrongBad May 30 '15 at 14:37
  • 2
    @StrongBad I'm not sure whether that would change the OP discouragement aspect. I also think that the culture may be difficult to shift on reopening. – jakebeal May 30 '15 at 14:39
  • We have a 24-hour grace period before flagging for obsolescence? – Wrzlprmft May 31 '15 at 9:58
  • @Wrzlprmft Grace period policy. – jakebeal May 31 '15 at 12:30
  • @Jakebeal: That’s for thank you comments, not obsolete comments in general. Most obsolete comments become obsolete because the person who really has to read them (the author of the post) has read them. There is no need whatsoever to keep such comments. – Wrzlprmft May 31 '15 at 13:23
  • 1
    We can potentially use the academia chat for this. Linking the article in question for a reopen vote if it is valid by the first vote to reopen should allow additional 3K+ to vote reopen if it is necessary. I'm not versed at making bots, but I assume that someone somewhere can make us a bot that just spouts out Reopen requests. – Compass Jun 1 '15 at 13:14
  • @Compass I like that as a step toward solving the difficulty of reopening problem. – jakebeal Jun 1 '15 at 13:35
24

I strongly disagree. While I do not deny the discouragement that putting questions on hold poses, this detriment is clearly outweighed by the advantages.

Closeworthy questions should be put on hold as fast as possible so that they are not answered in this state. Otherwise we run into the problem of answers that get invalidated by an edit, long discussions about opionon-based answers, partial answers and so on. Remember that those questions are problematic because of the answers they receive and putting them on hold has exactly and only this short-term effect: It blocks answers.

If you leave closeworthy questions open for 24 hours, you rely on the community detecting this problem and not answering such questions. I do not think this can work. In particular new users are likely to ignore this, answer the question and thus very likely face discouragement due to a changing question or the inherent problems that made the question closeworthy in the first place. And new answering users are somebody I really would not like to see discouraged.

I would like to mention that the above is backed up with experience from sites which are less active and close much slower than we do.


Another, different problem with your suggestion is that the system is not designed for it. Shifting the culture to your proposal entirely is difficult, so somebody will still flag or vote to close. Then, if I encounter a closeworthy question in the close queue that is younger than 24 hours: What should I do with it? If I close it, we are almost back to the old system. If I leave it open, such questions may never be closed. If I skip it, I leave the decision to those who do not adhere to your proposal and we probably have our close queues crammed like on Stack Overflow.


So, what can and probably should we improve?

  • When voting to close a question, explain to the asker what is wrong and what they need to do. The close notice does this to some extent, but only to some extent and the asker only sees this after closure (of which they aren’t informed).

  • When encountering a question in the close queue: Before voting to close, ensure that the problems that put the question there have not yet been resolved.

  • When you notice that an asker amended a question, retract your close vote.

  • Reopen questions quickly and whenever you think that a question is now in a state that would not be closed in the first place. Do not vote to leave a question closed just because it is lazy, trivial or has other flaws (that’s what downvotes are for).

  • If there is a straightforward way to make a question not-closeworthy, do not vote to close, but edit. For example, if a question’s only problem is that it is many questions in one: Do not vote to close but edit it so that only one not-closeworthy question remains (and tell the author what you did and why). Note in particular that one edit from the close queue removes a question from that queue.

  • 1
    Your last bullet conflicts with the ethic that we generally have on this site about tampering as authorial intent as little as possible. How would you suggest approaching that conflict? – jakebeal May 31 '15 at 12:30
  • 2
    @Jakebeal: To take the example I gave: I do not think that splitting a question (or at least intialising it) tampers with authorial intent. The authorial intent is to have the three questions answered as they are; having them ansered in one place is not a part of this intent (at least for any sane person). – Wrzlprmft May 31 '15 at 13:34
  • 3
    I also consider going against the authors intent to some extent acceptable, if that’s the only way to save the question. For example, if somebody posts a poll-like question (e.g., what do you think about X), I have no problems to replace this with the underlying concern (e.g., is X ethical?), if it is clearly identified in the question. That being said, if there are two ways to salvage a question and those depend on the author’s intent, I would not say anymore that there is a “straightforward way to make a question not-closeworthy”. – Wrzlprmft May 31 '15 at 13:36
  • 1
    @jakebeal, I can't agree with your characterization of the site's ethic. I think this answer is spot-on. Editing a question to make it more suitable is normal and to be encouraged. See, e.g., academia.stackexchange.com/help/editing: "If you are not comfortable with the idea of your contributions being collaboratively edited by other trusted users, this may not be the site for you." and academia.stackexchange.com/help/privileges/edit: "When should I edit posts? Any time you feel you can make the post better, and are inclined to do so. Editing is encouraged!" – D.W. Jun 8 '15 at 23:38

You must log in to answer this question.

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