This post was inspired by this question.

Suppose a new user asks a question, but it isn't clear what the question is asking. Within minutes, it will start attracting down-votes and votes to close.

At that point, is it even worth the poster trying to improve the question? I get the impression that the question is permanently doomed, with no hope of ever getting voted back up to at least zero, or of getting a useful answer. But if that is the case, what is the point of allowing other users to leave requests for clarification?

The poster would be better off deleting it, and writing a new version of the question from scratch.

  • 1
    IMHO your question is valid in general, but the example you linked to is IMHO not a good one for the described situation. I think the problem with that linked question is not "it isn't clear what the question is asking", but that it is based on a wrong assumption we have seen way too often here: that the differences in efficiency between these different coding variants (assuming there are such differences, which is probably wrong, too) have any major practical relevance. – Doc Brown Apr 15 at 15:00

Unfortunately, I don't have any hard data on this. I'm not sure if it's possible to build something in the Data Explorer to get some actual data on questions that have been turned around by editing, but it would be important to realize that not all questions are salvageable, even if people try. Getting data may be difficult.

I will address the editing versus deleting and starting from scratch. There is some risk to this strategy. Software Engineering (and other Stack Exchange sites) have automatic blocks if users ask low-quality questions. The algorithm is not known, even to moderators. However, factors include downvotes and deletions. Deleting and reasking the question, especially if it isn't improved and results in more downvotes, may just result in a user who is making a good-faith effort being blocked.

My recommendation would be for users who are running into these experiences post here on Meta. Link to the specific question and get some feedback. There are some higher rep users who are active here and have a good insight into how the general community thinks. Moderators also receive a notification for new questions. It's totally within the scope of Meta to get help formulating a question. In some cases, the right solution will be to make very significant changes, which perhaps should result in deletion and a new question (that shouldn't be downvoted, although we can't control individual behavior). In other cases, there may be edits made by either the community or additional things brought up that the original asker should address.


There are several reasons why questions may attract downvotes and close votes. In my observation, the reason you mentioned ("it isn't clear what the question is asking") fits only to a small subset of them. For the majority of questions it is mostly clear what they are about. That does not mean, however, downvotes and close votes to them are unjustified.

Questions getting lots of downvotes and close votes are often expressing a certain misunderstanding about the communities' expectation about a valid question, or what this site is about. Those questions can rarely be saved, at least not without editing them so much they turn into a completely new post.

So when is it worth to edit an already downvoted or close-voted question? I think there are some preconditions to be met here:

  • The question must not violate blatantly the topicality of the site (like being a coding question, or a research request, or a homework dump, or an educational request, or a request for legal advice).

  • It should not be a request for a discussion, a question about a purely speculative issue, a request for a list-of-things, or a general "I am just curious" kind of question. Ideally, it is a question about a real, practical, focussed software engineering problem the OP faced in their work or studies.

  • It should not be based fully on obviously wrong assumptions (though some of the best answers on this site are ones which debunk some maybe-not-so-obvious wrong assumptions in certain questions - here is one of my favorite).

  • The OP should have tried to find an answer on their own before (especially on this site, so there is no obvious existing duplicate).

If all those preconditions are met, but a question gets downvotes either, which kind of edits may be helpful to convince the community to convert their downvotes into an upvotes?

  • clarifications, of course, if the question contains unclear parts;

  • background information, examples, motivations and context, if that is what missing.

  • references to prior research of the OP, telling why it did not help them, or links to material which give further information;

  • removal of problematic phrases like request for 3rd party resources, or some biased assumptions (if those are only minor parts of the question and can be removed without changing the core);

  • improvement of spelling, grammar, wording and structuring and formatting;

  • removal of rants, colloquial speech, or any other kind of distracting phrases which might trouble readers.

Questions which meet the preconditions above are definitely not "doomed". I have seen several times questions being saved and upvoted after an edit of the mentioned kind, even if they were downvoted first. And in case the original downvoter does not respond to the edits, other community members usually will give enough counter-upvotes to balance the initial downvotes.

Let me add a final note. The site's system allows to cast anonymous downvotes without any comments (for good reasons - there is no way to change this). Unfortunately some of our community members seem to abuse this possibility on a regular basis for questions which (in my eyes) meet the former preconditions. That leaves askers often in the situation of a guessing game which kind of edits would satisfy the voters' expectations (if any). But hey, any online community has it's trolls, and my best recommendation here is to ignore them. When in doubt how to improve a certain question, try to talk to the more constructive part of the community, either through comments, or (as Thomas Owens already wrote) here on Meta.

You must log in to answer this question.

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