8
$\begingroup$

The Context

I have seen many questions being voted for closure as being primarily opinion based. Now, if you think about it for a moment, aren't all questions opinion based (except the hard science questions of course)?

If there was one and only one correct (and legitimate) answer to a question, why don't we delete all the other answers once an answer is selected by OP? Take for example this question:

How large would a world have to be for a sailing ship to never make it past the horizon?

There are 12 answers to this question, with one marked as accepted. The accepted answer is at +36 score while the runner up questions have also scored +17, +15, +7 scores (the top 3 respectively).

Now considering that one answer has been accepted, shouldn't we delete all the remaining answers? Logically speaking, we should, because there is a question and there is an accepted answer to it. Keeping other answers intact would mean that we are consciously or unconsciously confusing anyone else browsing the site with a similar (or the same) question in mind. On the other hand, if only the accepted answer is available, it would provide the person with a direct, straightforward answer for his/her question.

Now I agree that this would be a very bad (and resembling a dictatorial society) practice because everyone has their own mind and someone with a similar (or the same) question in mind might find other answers more useful than the accepted answer. But the logical decision of a non-opinion-based society would be to delete all non-accepted answers for a question, since they represent nothing more than the opinions of the answerer-posters about the topic.

The Suggestion

Maybe we should not shun opinion based questions. As it happens, everyone is unique and even questions marked hard science get different answers (logically speaking, only one answer should be voted up for such questions), many of which get voted up.

So I suggest that there should be an opinion-based tag available for question askers. Of course no-one is obliged to answer a particular question and knowing beforehand that the question is primarily opinion based would save the time and mental effort of all contributors.

Let there be idea generation and asking people's opinion about things. However I do admit that having a proliferation of such question may be detrimental to the core concept of the site. Opinion based question which are too broad or require too long answers can and should of course be discouraged and closed. However asking people's opinion about precise matters in short/medium length replies should be allowed.

$\endgroup$
11
$\begingroup$

There can be multiple correct answers. Acceptance doesn't mean "this is the best (or most-correct) answer"; it means "this is the one the person asking the question liked best". Sometimes it isn't even right. So acceptance is orthogonal to other considerations.

So, why is opinion-based bad? Because even though there are often multiple correct answers to a question, we do want to be able to evaluate the correctness of answers. If the question is a broad "so, what do you think I should do?" or "what's the best X" or the like, there's no way to do that. The Stack Exchange Q&A format works well precisely because it's Q&A -- not discussion threads.

Answering many questions will involve some amount of opinion, especially on a site like ours. Note that the close reason is "primarily opinion-based"; it's not saying "no opinions at all", but, rather, that this question will mostly (or entirely) produce opinions and discussion.

See the blog post Good Subjective, Bad Subjective for a discussion of what makes good not-purely-factual questions.

$\endgroup$
  • $\begingroup$ ^_^ agreed. Would it not be a good idea to offer some viable pathways for a person who asks "what should I do now?" Not a detailed, in depth answer, but an outline of options with brief description for each. Yes? No? Under consideration? $\endgroup$ – Youstay Igo Mar 13 '16 at 4:48
  • $\begingroup$ "It isn't even right" actually seems to be the case for the question linked in the OP... TL/DR: there are several interacting effects, and one of them is probably by far the strongest, but the asker didn't initially notice its existence, and so far hadn't received an answer that led them to reconsider (I hope mine will be it). The accepted answer does not consider the probable strongest effect at all, resulting in a ludicrously large figure that makes no sense with the other effect but is correct if reading the question as given literally. $\endgroup$ – January First-of-May Mar 13 '16 at 10:41
  • $\begingroup$ @YoustayIgo the notice on a question closed as primarily opinion-based links to this page in the help center. Does that cover the guidance you're looking for? $\endgroup$ – Monica Cellio Mar 13 '16 at 18:22
  • $\begingroup$ Thanks for the link and reply. I think I should delete this topic because it looks like the idea isn't well received by the community and I don't want to create a controversy. $\endgroup$ – Youstay Igo Mar 14 '16 at 16:03
  • 1
    $\begingroup$ @YoustayIgo no need to delete anything. If you were left wondering then chances are somebody else will be in the future. Meta provides a useful record of past discussions. $\endgroup$ – Monica Cellio Mar 14 '16 at 17:20
5
$\begingroup$

So I suggest that there should be an opinion-based tag available for question askers. Of course no-one is obliged to answer a particular question and knowing beforehand that the question is primarily opinion based would save the time and mental effort of all contributors.

TL;DR: No "opinion-based" tag.

It seems neither answer here so far has addressed this part, so if you'll indulge me.

  • Tags are used to categorize questions based on either their content, or the subject matter expertise needed to answer them. This is why we get tags like , , , , , and so on. Tags allow filtering to questions about subject matters of interest. Generally speaking, it should be possible to say "I am an expert in that-tag's-name" without making it sound too weird. ("I am an expert in biology" is a perfectly reasonable thing to state. "I am an expert in shapeshifters" has some odd implications, but doesn't sound totally outrageous. What does it mean to say "I am an expert in opinion-based"?)

  • Tags are discoverable. While a newcomer to the site might very well not get all the tags right, if their question is on topic they should be able to find at least one already existing tag that accurately categorizes their question. Tag synonyms can help here, because they allow people who might know one term but not another to find tags that accurately describe the question they are asking.

  • Tags, and in particular tags used on a large number of questions, imply that questions about a specific subject matter are generally accepted by the community.

An "opinion-based" tag fails #1, is doubtful in #2, and goes against the Stack Exchange format so implicitly fails #3. ("Primarily opinion-based" is a network-wide close reason, not up to each individual site to decide about. We even get that one on Meta sites, where almost by definition every question tagged discussion is primarily opinion-based.)

Also consider what we say at What topics can I ask about here?:

Note that questions must be specific as well as answerable. If you are looking for discussion, brainstorming, or an overall process rather than specific question-and-answers, the Worldbuilding Stack Exchange might not be a good place for your question. Such questions may however be welcome in Worldbuilding Chat. Also see Good Subjective, Bad Subjective for why this type of question is hard, and some guidelines on how to post good subjective questions which are less likely to end up being closed.

Now, opinion-based questions aren't necessarily bad. There are valid situations that call for discussion rather than question and answers. That, however, isn't what the main site is about. As pointed out in the help center article, that kind of questions are often a better fit for chat, which allows a lot easier back-and-forth than the main site's Q&A format does.

Lots of the questions we do get on Worldbuilding are, to some degree, opinion-based; there is no single correct answer. Even on Stack Overflow (which is probably where, on the network, the Q&A format strictly speaking works best), with well-asked questions, you rarely get questions that have a single correct answer and every other answer is incorrect; answers tend to focus on different aspects, or suggest alternative solutions to the problem. The distinction is whether a question is primarily opinion-based or, as explained in the close voting/flag dialog (my emphasis),

Many good questions generate some degree of opinion based on expert experience, but answers to this question will tend to be almost entirely based on opinions, rather than facts, references, or specific expertise.

That is the type of questions we do not want; to borrow Monica's wording, we want to be able to evaluate the correctness of the answers posted in response to the question. Questions that are "primarily opinion-based" do not allow us to evaluate answer correctness, causing the Stack Exchange implementation of the Q&A format to break down because voting no longer works; hence, that kind of questions are not welcome, and we should not encourage them by having a tag that implies they are accepted.

It's similar to Wikipedia's advice of verifiability, not truth which follows from their policies of no original research and verifiability.

As an aside, that is also why we tend to prefer longer answers to single-sentence answers; a longer answer is more likely to explain its reasoning, thus being easier to evaluate the correctness of as well as more likely to be useful to people other than the person asking the question being answered. An answer merely being longer is no guarantee it will be better received, but all else being equal, an answer that explains its reasoning will be longer than one that doesn't, and the answer that explains its reasoning will be easier to evaluate the correctness of.

$\endgroup$
  • $\begingroup$ What does it mean to say "I am an expert in opinion-based"? - it means I can generate solutions to make situation viable. Sorta engineering, no science, or encyclopedia or google skills , but rearrange set of block to make it's alive. There is reality-check tag which have to no to be used alone. $\endgroup$ – MolbOrg Jun 25 '16 at 20:30
4
$\begingroup$

I see two reasons to keep non-accepted answers:

  1. The question was vague, and people had to make assumptions about what was being asked.

  2. The question was too big, and there are things that previous answers missed.

Now, problem 1 is caused by 'too broad' or 'open ended' questions, which have their own problems, but here those problems manifest with answerers who don't really know what you want. People who answer these kinds of questions have to assume what you're asking for, so a lot of them are going to get it wrong. However, their answers are probably still good answers to questions that don't exist yet, so anyone looking for answers to that question may be able to eventually find them. As it stands, this is a bad system, so instead of answering too-broad questions I think we should just close them until the OP can be more specific.

As for problem 2, this can be solved by community wiki answers, which can be edited with new information. However, due to the point-based SE system, this would rob contributors of their Internet points, so sometimes we have to let it slide. Either way, multiple answers of this form are just different pieces of the same one answer, thus you can think of them as essentially the one answer that is not deleted. Other answers that are simply wrong or don't contribute anything new to that 'one' answer should be deleted.

So, yeah. Opinion-based is bad, I think, because it's another piece of the problem 1. Answerers can't really know what you're asking, so they make up a fantasy question and answer that. If someone wants to know the answer to that fantasy question, it'll be hard for them to find this answer, and overall the fantasy question that may have helped the most people may not be the same as the fantasy question the OP thought it was asking, so its answer won't be put to the top of the list. If people asked properly limited questions with one best answer, I think you could probably delete all other answers.

$\endgroup$
4
$\begingroup$

In addition to the other answers, you should also keep in mind that even if three answers are explaining the same thing in the same correct way they may all take different approaches to that explanation. For some people one explanation would be more helpful than others.

Look at this question for example:

https://stackoverflow.com/questions/21692193/why-not-inherit-from-listt

I've not checked them all but most of the top answers are all actually saying exactly the same thing from the technical point of view. However they are explaining it differently. Some of those explanations will work better for some people than for others.

$\endgroup$

You must log in to answer this question.

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