Meta Stack Exchange is a question and answer site for meta-discussion of the Stack Exchange family of Q&A websites. It only takes a minute to sign up.
Sign up to join this community
First, I'm already aware of this question: Should poll questions be Community Wiki?.
But I think the problem remains: Polls are like shopping lists, and I think they should be closed.
I would like to settle this topic once and for all, since it keeps coming up again as new Stack Exchange sites are created (for example, this question on Meta Sci-Fi).
Should poll questions continue to be allowed to exist as Community Wiki or should we be closing them?
This is a loaded question, and there are a lot of different issues to get through in a comprehensive answer.
First of all, as Arjan has pointed out in the comments, Community Wiki mode is no longer available for questions. Thus, one behaviour that should never be tolerated is comments being left saying "should be Community Wiki". Since this is no longer even possible anymore, I would suggest flagging those comments for a moderator to remove.
Of course, moderators can still make questions Community Wiki, and some people have taken to flagging their own questions immediately after creating them. Others still flag other people's questions for moderator attention to ask for Community Wiki status, which really is our collective fault, since we encouraged that so adamantly and frequently back in the "olden days" when CW was more prevalent. Since this is the only way to make a question Community Wiki now, the question then becomes: Does flagging for CW still make any sense?
Consider that:
The fact that it is so cumbersome to create a Community Wiki question and actually requires the help of a diamond moderator is a very strong hint that Community Wiki is the exception rather than the rule. On the other hand, anyone with sufficient rep can vote to close.
Poll questions have, in the past, been lambasted for their subjectivity. But there are specific guidelines for subjective questions now. So the question is really no longer "is it a poll", but "is it constructive?" If it meets the guidelines, then it does not need Community Wiki status. If it doesn't, then it's not constructive and doesn't belong on the site.
It's been stated again and again that Stack Exchange is not designed to handle questions with many answers. It doesn't need to be stated at all; it should be obvious from the design. The sorting, the pagination, the Accepted Answer concept; real questions have answers, not items or ideas or opinions.
Many of us even believe there to be an inverse correlation between the number of answers and their overall quality. There's certainly plenty of gray area here, but the important thing is that Community Wiki does not factor into this equation anywhere. Reputation is completely irrelevant to the issue, and the primary purpose of CW - encouraging others to edit and collaborate - does not apply to dozens or hundreds of effortless one-line answers. The exact same problem exists with or without the Community Wiki flag.
So I think it's obvious that Community Wiki for polls is not the answer. It never was. It was a horrible loophole, a gross misunderstanding of the purpose of Community Wiki combined with truth-by-repeated-assertion, resulting in mass abuse of the feature to justify some of the most useless "questions" imaginable.
But that only answers half the question. Even if we agree that they shouldn't be Community Wiki, does that actually mean that all polls should be closed?
That is where it starts to get dicey. For you see, the term "poll" is itself subjective, and whether or not a given question is in fact a poll is often open to debate.
For me, and I think I speak for many others, the only time you really know you're looking at a poll is when you see those one-line answers. But we can only close questions. Is it always the fault of the questioner that the answers suck? Do they deserve all the blame for the fact that people with nothing useful to say want to participate anyway and bring their reddit-style "tweets" into the answers?
Sometimes the questions really are bad. When someone asks, Which programming language do you really hate?, and doesn't bother to elaborate at all, that is just screaming for crap answers. It's hard for me to begrudge any of those participants their answers (okay, maybe I begrudge them a little) because the question itself was so ridiculously open-ended that any answer would technically be a "correct" one.
Then again, some questions that are worded as very obvious polls actually get reasonably good, well-written answers. See Best practices that you disagree with for an example. On Seasoned Advice (Cooking.SE), I can point to several examples of weak questions or even joke questions that, given an early, comprehensive answer, did not devolve into pointless blathering. On the flip side, I've seen questions that were definitively not phrased as polls that were still greeted by dozens of poor-quality answers; take, for example, one of Stack Overflow's oldest: Practical non-image based CAPTCHA approaches?
Recently I've actually taken to using Metafilter's guidelines as a rule of thumb. To summarize, the types of poll questions that are very difficult to salvage generally have some or all of the following characteristics:
The question provides no criteria for evaluating answers. "What's your favourite X?" or really almost anything with the word "you" in it seems to uncritically welcome answers from anyone and everyone - it values opinions and stories over facts.
The question does not seem to have any practical purpose, implied or explicit. Perfect example: What do we developers have in common? Uh huh, and your point is...?
The question starts off with an "example". This is the mating call of the serial procrastinator: hey, let's play a little game! I spy with my little eye...
Another variation that MF doesn't mention are those of the form "blah blah blah me me me blah blah blah my life sucks blah blah blah, any advice?" These don't always sound like polls, but they call for so many assumptions and so much speculation that they inevitably turn out like polls.
I believe that these types of questions should be closed - not because they are "polls", per se, but because the author hasn't made a good-faith effort to make his/her questions practical and relevant. I love the Not Constructive close reason on Programmers for this; on other sites, I would close as Not A Real Question, except for case #4 which is rightly closed is Too Localized.
However, if a poll-like question does appear to be hinting at a more practical problem then it is better to edit the question as early as possible. If it's already started to attract low-quality answers, then downvote those answers and/or use the new improved flagging system to flag those answers as "low quality".
Poll answers are broken windows. Just as you're more likely to get imitators when poll questions sit around festering, you're more likely to get the me-tooers when low-quality answers clutter up an otherwise decent question. There's definitely a "quality" movement afoot on SE and at least some of the moderators will help with cleanup when it is reasonable to do so.
All of that being said, the reality of the situation is that every site is different, moderators and even team members disagree on how much open-endedness is reasonable, and nothing I've written here is going to guarantee you cooperation from anyone else. These are my thoughts on the matter and mine alone, and although I feel that I have a rational approach to this issue (backed up by plenty of evidence), it is completely up to individual communities and moderators to make the final decision.
What you seem to really want - a firm directive from on high - is probably not going to happen. But don't let that stop you from doing your part, and keeping the sites you frequent as clean as you can.
