Example question: Strange job interview -- What were they really looking for?

In the example, the question is well voted for and has a good number of highly voted answers.

However, I would tend to agree that it isn't suitable/should be closed - as it's effectively asking people to guess what somebody else is thinking, in a process that is highly company specific. Most of the answers I'd argue could be seen as just opinions or speculation as well - even though there is a lot of genuinely sensible advice there too.

In situations where a question has a large reach already, with many answers and a lot of votes for both the question and answers. Should we still vote to close as we would normally (as if it was a fresh question), or should we take some consideration that a large number of people thought the question was useful?

  • 5
    This is a great question and I go both ways on this one. Being a moderator I only cast my close vote for what I believe is 100% necessary. Before I was a moderator if a question met one of the closing criteria I would strongly consider voting to close it regardless of a questions popularity. – Neo Feb 26 '20 at 17:21
  • 3
    Might make sense to define "a large number of people". For any question that gets pushed to HNQ, popularity instantly becomes a meaningless metric. HNQ visitors rarely downvote and most aren't able to VTC to begin with. – Lilienthal Feb 26 '20 at 21:01
  • 1
    Also, if you are after an answer to the general case of popularity versus topicality (which I swore we had addressed on meta before but I can't find it), it's generally recommended to not link to a specific example. As you can see answers then tend to focus on the circumstances of the example rather than the broader case. – Lilienthal Feb 26 '20 at 21:03
  • 1
    @Lilienthal you maybe thought of this? Also there is an older discussion of topicality vs popularity here (wrt closing questions with accepted answers it seems to be well covered here) – gnat Feb 27 '20 at 9:15
  • 1
    @gnat Hah, yes indeed. Another argument in favour of making general / best practice questions separate from specific instances because they are a pain to search for. :) – Lilienthal Feb 27 '20 at 9:23
  • Thanks all, will bear that in mind whenever using meta in future. – user81330 Feb 27 '20 at 10:07
  • When I read that question I didn't understand that it was looking for mind-readers, instead I understood it to ask whether anyone had seen that interview technique before and, if so, what is the point of it. In fact, re-reading the question now, I see that the asker specifically stated "Has anybody ever seen a similar exercise at a job interview?", which seems pretty clear to me. Maybe you didn't read the whole question before voting to close it? – Aaron F Mar 6 '20 at 16:06
  • @AaronF 3 of the 4 points, to me, were non answerable/required knowledge from the interviewers themselves. "What skills and traits are the interviewers really looking for?", "How does one do well?", "Is there a better way to address / prepare for such an odd interview?". As interview practices are highly company specific, especially an "odd" one like this, my belief was that all the answers will be nothing better than opinions and speculation. Either way, my concern was more with the vote-count than the specific reason (which I was comfortable with). But to be clear; "yes I read the question" – user81330 Mar 6 '20 at 16:17
  • Fair enough. Those other three points I personally thought were overly generic and obvious: "how does one do well in an interview?". For the first question I expected to find an answer saying "ah yes this is the famous planks of wood question, popularised in the late 1980s at IBM" or something along those lines. But yes, as you say, the specifics of that question are beside the point of this question :-) – Aaron F Mar 6 '20 at 17:34

Should we still vote to close as we would normally (as if it was a fresh question), or should we take some consideration that a large number of people thought the question was useful?

The close reason in this case makes no sense on the face of it. The reason given is:

This question needs details or clarity.

Add details and clarify the problem being solved. This will help others answer the question.

The question has 6 answers, including one that was accepted. And hundreds of votes among the questions. The question itself has 78 upvotes.

Seems pretty clear to me that the question had sufficient detail and clarity for the 6 folks who provided an answer before it was closed, for the hundreds who voted on the question and answers, and (most importantly) for the question writer to decide to accept one of the answers.

Sadly, I think this yet another case of having a few folks here who like to close questions more than actually help questioners.

I voted to re-open the question.

  • I think this kind of thing is inevitable now that the site has reached a point where any questions that aren't duplicates are likely to be extremely esoteric. One of my own questions was closed as "unclear" after it was upvoted and someone understood it well enough to provide the correct answer. Eventually it was reopened. I ended up closing it as a duplicate myself, once I knew the key words from the answer. – StackOverthrow Feb 27 '20 at 20:39
  • 1
    Questions that lack clarity often receive answers from people who are making assumptions or trying to answer the overly vague question anyway, and there's a reason why upvoting and downvoting is distinct from closing and opening. – Bernhard Barker Mar 3 '20 at 14:47

Numbers mean little in some cases.

I once had an answer deleted after it had over 30 upvotes/

So if it warrants closing or deletion in your opinion, just go ahead.

  • In other words, "You SE your way, I'll SE mine" but with a vastly different weight of opinion given to those with close votes. – StackOverthrow Feb 28 '20 at 16:41
  • 1
    Yep, we don't make the rules. – Kilisi Feb 29 '20 at 12:14

Personally, I see this as difficult to answer in the general sense, without context. In other words, it depends.

For me, it depends on the close reason I'm considering, and the "type" of activity the question is getting.

I think questions asking for legal advice are a good example of why the context matters, and specifically why a question with a lot of activity should remain open even if it's borderline. There's a relatively well established precedent to leave "legal" questions open, if they're about legal topics you'd expect most competent employers/managers/HR people to know the answer to. I might not personally know the answer to a legal question, so I might consider closing such questions, but if lots of well-liked answers exist on the question, I would take that as an indication that it passes the "you'd expect people to know this" test, and it should be left open.

On the other end of the spectrum, some questions end up attracting lots of attention, and even lots of highly-voted answers, where the answers don't even really answer the question, or there isn't even a clear question and the answers are getting voted because people "like" them, or there are other contextual clues that the question is off topic, and the existence of many popular answers doesn't really change that. In fact, it may even be an incentive to close the question, in the hopes of getting attention from the OP to clarify or edit as appropriate.


Upvoting and downvoting are not really part of moderation (at least not in the same way closing and reopening are) and are not the deciding factors of what belongs on the site. There's a reason why they're entirely distinct from closing and reopening.

You can start upvoting at 15 reputation, which is a pretty low bar. It's not too difficult to get that with a single question or answer, at which point you may have no idea what belongs on the site and what doesn't. You may even have gotten that reputation through the 100 reputation association bonus by coming from another site without having had any activity whatsoever on this site.

You need a whole 3000 reputation to cast a close vote (you can flag it before that, but flags can't close a question). That requires a bit more effort and hopefully the activity on the site to get there would've taught you what belongs here and/or you care enough to find out.

So yes, take the score of a post into account, because a high score may mean you're wrong in thinking it should be closed, or a negative score may mean you're wrong in thinking it should be open. But don't refrain from casting a close or reopen vote you're sure about only because of the score of the post.

If questions regularly manage to get high scores before getting closed, it may mean (a) we don't have enough moderators to deal with the volume or (b) there's disagreement over what should and shouldn't be closed among moderators.

We don't have that many questions, but whether we have too many voters that manage to spike up the score before it can get closed is an open question.

But there is definitely some fundamental disagreements about what should and shouldn't be closed between significantly sized groups of moderators on this site, which is a huge problem.

Note: by "moderators" I mean official and unofficial ones, who are able to cast close votes, i.e. anyone with more than 3000 reputation.

The bar for posting an answer is even lower, but answers have even less to do with moderation than upvoting or downvoting.

Although having too many answers may in fact indicate that a question is too broad.

Also keep in mind that a high score may indicate a question has potential, but it's missing something before it is a good fit for the site or it can be properly answered (meaning it should be closed and then reopened after being edited to fix any problems).


That depends on the answer to another question:

Is this site based on a planned, top-down design, or is it a grassroots, bottom-up community?

If the former, then sure, close away.

If the latter, then upvoted questions are good questions by definition. If there are many such questions, then the site description needs to adapt to the questions.

  • 1
    And then you have Reddit for Professionals™ – Mars Feb 28 '20 at 8:26
  • @Mars In which case? – StackOverthrow Feb 28 '20 at 16:34
  • 2
    dwizum mentions in their answer that often, questions/answers get "likes" rather than quality upvotes. without control, this would eventually lead to no difference between SE and Reddit – Mars Feb 29 '20 at 1:01
  • @Mars Reddit sucks for other reasons, but this is precisely why SO is hemorrhaging its most valuable contributors. – StackOverthrow Mar 1 '20 at 17:58

You must log in to answer this question.