8
$\begingroup$

I've been thinking about the problems at the heart of recently, and something finally came to mind. Story questions!

What I'm going to call a "story question" is a question that asks you to fill in a story to solve a problem. For instance, this question asks you to write a part of the story to describe how the police knew someone was a murderer. Other questions have been:

Many of these questions have been well received - but mostly by people on other sites. If a question doesn't hit the Hot Network Questions list, it is, for the most part, downvoted.

Here's the problem with these questions as I see it: there are an unlimited number of answers. Anyone can create an answer involving... pretty much anything. People have written long stories as answers, and people have written joke answers - that end upvoted. For instance, "the dog was hit by lightning and is now on fire" is obviously a joke, but by the text of the question, it isn't wrong.

Not only this, but these questions tend to lead to an enormous number of edits to invalidate answers that go beyond the bounds of the original question.

In other words, no matter how much these questions are restricted, there are still an unlimited number of answers. Not only that, but most of those answers are terrible. Sure, they're briefly entertaining, and that sadly tends to get upvotes, but that doesn't make them good answers. Well-received ≠ good.

These seem to be the body of our low-quality posts at the moment, at least to me. What should we do about these questions? (I lean new close reason, personally.)

$\endgroup$
7
  • 2
    $\begingroup$ "unlimited number of answers" practically screams "Too Broad," so I'm not sure that a new close reason would be necessary. Just a change in policy (if we do decide that we don't want these). $\endgroup$
    – Doorknob
    Jun 17, 2015 at 3:55
  • $\begingroup$ As far as I've seen, such questions have been closed and downvoted fairly consistently, but it takes time. The problem is racing against HNQ on a site with a fairly small dedicated userbase. Not sure what we can do better here except with a change in HNQ policy to exclude/remove questions with significant downvotes and close votes. $\endgroup$
    – xnor
    Jun 17, 2015 at 6:31
  • 1
    $\begingroup$ "no matter how much these questions are restricted, there are still an unlimited number of answers" - I don't think this is true. With a question that's carefully enough phrased, often the intended answer really is the only one that works - it becomes more like a [logic-puzzle] than a [lateral-thinking] question. We just need to encourage people to think like pedants, and anticipate and invalidate all the silly 'stab the oracle' answers that people might come up with. $\endgroup$
    – Rand al'Thor
    Jun 17, 2015 at 9:14
  • 2
    $\begingroup$ @rand Do you have a concrete example of a question that was sufficiently restricted (in its first revision, before terrible answers come in) such that there is no possible answer but the right one? $\endgroup$
    – user20
    Jun 17, 2015 at 11:36
  • $\begingroup$ @Aggie The reason I specify the first revision is because, once you're in the second revision, edits would now exclude existing answers specifically, which leaves dangling obsolete answers that no longer work. $\endgroup$
    – user20
    Jun 17, 2015 at 21:55
  • $\begingroup$ @Emrakul I guess I am at a loss then, because I am consistantly amazed at some of the off-the-wall answers that people come up with. I'll admit there are shortcomings, but with the brilliant minds that are part of this community, it is impossible to catch everything on your own. $\endgroup$
    – Aggie Kidd
    Jun 17, 2015 at 21:56
  • 2
    $\begingroup$ @AggieKidd This is sounding like a good argument for having a "sandbox" for such questions before they get posted up. Of course, everyone involved in the sandbox would then have to refrain from answering. Hmmm ... tricky one! I believe they have a sandbox system on PPCG, but I've no idea how that works. $\endgroup$
    – Rand al'Thor
    Jun 20, 2015 at 10:12

2 Answers 2

Reset to default
3
$\begingroup$

I think the question raised by Emrakul is excellent and I support his point of view. However, Rand al'thor has a point that there are - or in absences of an example can be - story-questions which are really "puzzles" in the sense that they need a very specific answer which one can not "at once" see. I would therefore think that the proposed tag should indeed be created/used.

This tag should in its tag-description specify, that questions of this type need to be well-restricted allowing for a narrow set of answers only, and that logical answers not loop-hole-fun-jokes are expected.

Adding the tag does not solve the issue of bad questions of this kind. I think we still should down-vote and close questions, if they are not good and their answers are likely to be too broad. This may very well be the majority of them. Nevertheless, it is something different than a 'full-blown-ban' of such questions and therefore less hostile and more open-minded.

I do believe that good story-puzzles are possible, so give them a chance...

Similar to the tag, a tag would have a high yield of "to-be-closed" questions, but eventually the higher quality content will remain and educate new people how to create such puzzles...

$\endgroup$
2
  • 1
    $\begingroup$ As with my reply to rand's comment: does anyone have an actual example of a question where the first revision sufficiently restricted the possible answers? (Even if we think it could work, we still need evidence to support it.) $\endgroup$
    – user20
    Jun 17, 2015 at 14:11
  • $\begingroup$ @Emrakul No we don't. My suggestion was to "carry on as current" (i.e. close/delete bad "story" questions) but introduce the story-puzzle tag (with explanation of what would be okay) nevertheless. $\endgroup$
    – BmyGuest
    Jun 17, 2015 at 14:43
-1
$\begingroup$

We've discussed this type of question before - they're called situation puzzles and we decided to disallow them for a while before allowing them again.

$\endgroup$
1
  • $\begingroup$ The meta post you link to was a LONG time ago, way before all the changes this site has undergone. $\endgroup$
    – Rand al'Thor
    Jun 20, 2015 at 10:09

You must log in to answer this question.