Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 157 Stack Exchange communities.

What is meta?
Here's how it works:
  1. Any Stack Exchange user can ask a question
  2. The community provides support, votes on ideas, and reports bugs
  3. Your voice helps shape the way Stack Exchange operates

Users have the ability to self-delete their questions, as long as no upvoted answer exists. There are two major patterns of misuse of the ability to delete your own questions:

  • Deleting a question immediately after it was answered, the answer has not enough time to receive an upvote and prevent deletion.
  • Deleting closed or heavily downvoted questions and reasking them mostly unchanged.

These patterns are harmful to the community as they can delete useful content or hide that a user is reasking a bad question multiple times. The only mechanisms against these currently are the automatic question ban and manual intervention of users that observe these patterns.

I'm proposing some new methods or adjustments of existing ones to prevent these harmful self-deletion patterns:

To prevent the immediate deletion of questions after they are answered, there are two possible solutions:

  • Only allow deletion after a specific delay from the creation time of the answer, so that the answer has enough time to receive some votes and the associated protection from deletion.
  • Prohibit deletion of questions with answers of a non-negative score. This would immediately protect questions that were just answered and still allow deletion if the answer is downvoted. Could be gamed by the asker downvoting themselves, if they have the privilege.

To prevent the misuse of self-deletion to reask the same bad question I propose a new automatic community flag "Consecutive self-deleted posts", similar to the consecutive closed questions. I don't have any hard data, but I would suspect that benign self-deletion of two questions in a row is very rare, and there wouldn't be too much noise caused by this flag.

share|improve this question
Deletion of downvoted questions is a self-correcting problem. Doing so is a big factor in automated question bans. – Charles Nov 16 '12 at 19:54
@Charles Automated question bans only exist in a handful of SE sites (the trilogy, MSO and ProgSE) – Yannis Nov 16 '12 at 19:59
@Charles I know, but the question ban is the very last line of defense and I'd like to catch that stuff earlier. This would allow the automatic question ban to weigh self-deletion less severely. The current implementation discourages users to clean up their own questions, which I think is less than ideal – Mad Scientist Nov 16 '12 at 19:59
@MadScientist The question ban tends to kick in at around 3-4 bad questions. I don't see how a person could do so much quicker than that. Would you want to ban someone just for doing this once or twice? Also, any user who's a serious contributor to the site really doesn't need to worry about being banned for self deleting their content. Once you have a handful of moderately acceptable content getting a ban at all is pretty hard if you're not actively trying. I don't really think there's a problem here. – Servy Nov 16 '12 at 20:07
note asker with 125+ rep can downvote all the answers (if all are at score <= 0) to hack around your "non-negative score" requirement. Upon question self-deletion they will get back all the rep lost in that – gnat Jan 14 '13 at 7:28
If a question got heavily down-voted, deleted and re-asked with little or no changes won't it get heavily down-voted again? If it was changed even slightly and it doesn't get down-voted again isn't that a sign that the problem with it was fixed and there is no issue? – Joe W Jul 15 '13 at 13:44
up vote 11 down vote

We have other time delays in the system (for voting to delete, accepting an answer (longer for self-answer), and for protecting a question), so a time delay on self-deletion would be consistent. The tricky part here is that the delay is based on the time of the first answer, not on the post itself. This seems possible (the data is there), but I'm not an SE developer. One would hope that self-deletions are infrequent enough that an extra server call (if needed) to determine its legality wouldn't be too burdensome.

As noted in the question, a restriction that a question with a 0-score answer can't be self-deleted can be gamed by a user who's picked up some rep. If he got the rep the slow way you would hope he'd be past such bad behaviors, but (a) we've all seen cases where that isn't true and (b) an association bonus carries you most of the way to the downvote privilege.

Alternatively, we could require that -- just as with votes to delete from other users -- a question must be closed first. Closing requires assistance from other users and/or moderators, so the asker wouldn't be able to close and then delete his question on his own.

For cases where a question really ought to be deleted quickly, there's always flagging.

share|improve this answer

What if a self-delete question with any answer gets translated to a (non-binding) delete-vote instead (similar to how self-close votes work). However, the further delete votes on this question aren't subject to the 48hr delay or -3 score rules normally required.

This'll mean that 10k+ users can clear the question if upon review the OP isn't trying to do a hit and run with an answer. (nb: I suppose upon deletion it would need to be considered a self-delete vote rather than a community delete vote though...)

After n many hours/days the delete votes (either all, or just the OP's) expire, and the OP's delete vote becomes binding again (subject to normal rules).

If it really needs to be deleted, then the OP's last resort is to flag for moderator attention and explain why they think the question and answer aren't of value (or something).

share|improve this answer

I agree with the problem that this proposal attempts to solve. However, I only really like one metric.

Prohibit self-deletion of questions with answers of a non-negative score.

I think that this encompasses the 0 score, an answer was present, scenario. This seems to be the largest issue raised with self deletions (that a valid answer was provided and then subsequently deleted).

The point made as a disadvantage (that the OP will just downvote and then delete) could be gamed. However, that does require the reputation hurdle of 125 which I believe is a significant amount. By then, users probably will not follow the pattern of removing their content. The benefit is that all the users below 125 reputation (which probably encompasses a decent swath of the user base following this deletion pattern) will now no longer be able to quickly delete their question if an answer is posted.

share|improve this answer

You must log in to answer this question.

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