How to restrain improper "Not a real question" votes?

Question

Recently I've had the misfortune to come across a question that came very close to being closed due to "not a real question". Although in its original form it was missing important information (ie, that there was physical security in place, and the users don't have administrative control over the machine), four high-rep users jumped in on the "not a real question" close bandwagon before the OP could clarify. Once it was ascertained that the question was a desktop administration question, not a DRM programming question, these four votes made it impossible to migrate using the normal process; it's currently awaiting moderator attention.

While there are certainly situations in which a "not a real question" closure is warranted, I don't see this as one, and I'm a bit worried about how quickly people piled on there. Even if the OP did not clarify, "This is impossible" is a perfectly good answer to the question's original form. While some might not like the implied intent of the original form of the question, it does not justify the use of "not a real question".

In short, people seem to be using "not a real question" as a catch-all to get rid of questions where the OP is misguided about the possibility of their claims, hasn't had time to edit in necessary details yet, or where they simply don't like the question. What can be done to discourage this?

Some thoughts:

• It may be a good idea to break down 'not a real question' further. Asking people to say whether the question is 'ambiguous', 'vauge', 'incomplete', 'overly broad', 'rhetorical', 'unanswerable', or 'impossible' would probably help avoid people just using it as a catch-all.
• For 'incomplete' votes, it may be a good idea to add a cooldown, to give the OP time to respond to comments and edit.
• Perhaps undoing of close votes should be allowed after an edit, as with downvotes?

It should be noted that I'm not opposed to the close->reopen process in general. I just hate to see cases like this, where the OP gets four votes in a short interval, fixes their post, then later someone comes along with a (migration?) vote and gets the post closed, just as the OP is getting things back together. Adding some kind of mechanism to restrict this catch-all would, I think, help limit this a bit.

Answer 1

I've spent a fair amount of time going through problematic questions. In my experience, it is rather more common that when a user with < 250 rep or so asks a question that can't be answered in its current state, they proceed to mostly abandon it or fail to clarify appropriately. And while I will grant that closing was done too quickly in this case, the closing would have been fully justified otherwise.

The original question was:

• If taken at face value, impossible to solve
• If assumed solvable, badly incomplete

Given the later clarification added, I've since voted to reopen, but I also note that after being clarified the question is essentially off-topic and thus warrants closing anyway. Migration is all well and good, but there's nothing to stop someone from simply reposting the question in the appropriate location. The main purpose of migrating, as I understand it, is to preserve history and existing answers; in this case the answers are all useless (as they're responding to the original, invalid form of the question) and there's nothing else useful to carry over.

---

After sleeping on it I'd like to lay out my thoughts on how things should be handled and where it went wrong in this case:

• The question, in its original form, absolutely warranted closing:

  • As SO's user base has grown, it's been a consistent theme that the Q&A nature should be enforced rigorously, and to me that implies that questions should be clearly answerable. When a question "How do I do X" attracts two answers in 20 minutes saying roughly "you can't, X is impossible on multiple levels", that seems clear evidence that the question "cannot be reasonably answered".

  • Making the charitable (and in this case, correct) assumption that the poster does have a reasonable question in mind, this question was missing multiple critical pieces of information, later added at bdonlan's prompting. Prior to that, the question was objectively ambiguous/incomplete.

  • The suggestion in the comments that the close votes were initiated by linux users who disliked the idea is utter nonsense. The first close vote was from someone who doesn't really look like a linux fanboy; the second was from me, and I've voted to close at least one or two similar questions of the form "how can I subvert the user experience on Windows for unclear reasons".

• Closing a question before the OP has a chance to clarify is also absolutely fine, for essentially the reasons that Cody Gray gives. I don't see any value to a grace period past maybe 5 minutes or so--enough time for the OP to post, realize they made a mistake, and fix it immediately.

• However, as outlined in Cody's post linked above, there are two reasons for closing promptly, and neither really worked out in this case:

  • The question still ended up with three uselessly speculative answers based on incomplete information. Incidentally, a tip of the hat to bdonlan for keeping the site tidy and deleting his as soon as the clarifications obsoleted it.

  • Getting the user to notice and prompting them to fix the question was unnecessary, because Skkard was already doing so. The question was edited 7 minutes after posting, then again 24 minutes later with the significant info, at which point it had already accumulated 4 no-longer-accurate votes to close. I cannot imagine close votes being in any way a motivation for a user without enough rep to see them.

In short, Skkard was being exactly the sort of "particularly conscientious user" Cody brushed aside in the post giving reasons to close promptly. It is important to remember that this is very much atypical. Far more common when I'm reviewing questions with close votes, particularly from low-rep users, is to find an ill-formed question asked X hours ago, where X is also the last time the user was active on the site, and no indication of when they'll return. Questions like that should be closed with extreme prejudice.

In my view, the mistake in this case--by which I mostly mean my own mistake, since I can't really speak for the other four who voted to close--was hastily concluding that it was another instance of a fire-and-forget bad question, rather than one the user was actively improving based on feedback.

However, I don't think it would be helpful to split NARQ into more specific reasons, and I really don't think it would be wise to add any significant limitation on closing. No amount of editing could save a question like "I want to write an operating system in PHP, is using Zend appropriate for this, also please give me some example code quickly" and it would make me very sad if I couldn't vote to close it. Very, very sad.

---

Given the above, a concrete proposal: Display, in a location easily visible when adding a close vote, how long ago the question owner last viewed the question. If a question is salvageable, or somewhat borderline, seeing something like "Question owner last seen: 35 seconds ago" would probably be enough for me to stay my hand and return later.

Answer 2

I strongly support the idea of having to give a reason when you vote to close a thread. I can't begin to describe how frustrating it is as a new user to find your question closed with no suggestions on how to improve it, and you're too close to it to see the problem. That is probably why, as someone pointed out, most new users don't bother editing the question to have it reopened - they have no clue what they need to change. According to the FAQ when you downvote you should leave a comment explaining why, and it doesn't seem like very many people know that.

It's also too bad there is no way to tell if the user currently has the edit window open. I could easily imagine a situation where someone with ADHD like me would be too quick with the post button and immediately open it to edit while someone else is already voting to close it.

And maybe for those people who might vote to close just because they don't like the question or user, or for other non-valid reasons, there should be a limit to how many votes per day each user can place or have close votes impact the voters rep. Then they would be more likely to only use it when it's really applicable.

I'll close this with a quote from a comment on another question on this site, hopefully the original author doesn't mind.

"The problem isn't Stack Overflow. The problem is the hoards of people itching to close and/or delete questions and answers they don't even understand." – Jon Harrop

Answer 3

In short, people seem to be using "not a real question" as a catch-all to get rid of questions where the OP is misguided about the possibility of their claims, hasn't had time to edit in necessary details yet, or where they simply don't like the question. What can be done to discourage this?

The problem is not the description of the closing reason, but it's probably the fact that some users don't read the description; they read "not a question" and they select it. Even splitting a closing reason in two, there could be users who select the wrong closing reason, or select a random reason, as for those users it's only important the question is closed.

Perhaps undoing of close votes should be allowed after an edit, as with downvotes?

Users who can vote to close a question can also vote to re-open it, which (in some way) is like undoing the closing vote, but it doesn't even require the same users who voted to close; it's a more generic mechanism that can be used from both who voted to close the question and who didn't vote.
It already happened that closed questions have been voted to be re-opened: In some cases, the question has been re-opened because it should not have been closed; in other cases, the question has been re-opened after the OP edited it.
An automatic removal of the closing votes (which is not what you are talking of) is not something that it would be possible. SE software cannot automatically decide when an edit is sufficient to make the question a valid one; users still needs to read the question as it has been edited, and decide if what has been closed as not a question now can be classified as a question.

Answer 4

It's not like this was another badly formed question written in broken English and not making any sense from user82312341 who only has 1 reputation point.

But as camccann suggested, there's a balance between giving someone the benefit of the doubt based on their reputation score. Skkard has 115 points, but he has an otherwise good track record of answering and asking questions.

Outside of this particular question, there's still a culture of "robo closures" that I think comes down a bit too hard on many users. I like bdonlan's ideas - and perhaps the "cooling off" period could scale based on the reputation of the user posting the question.