49

If you vote to close a question and it later gets reopened, you are not allowed to vote to close it again.

If you vote to reopen a question and it later gets closed again, you are not allowed to vote to reopen it again.

But if you vote to delete a question and it later gets undeleted, you are allowed to vote to delete it again, and (I assume) the same goes for undeleting. Why the inconsistency?

I can think of two possible reasons:

  1. When the no-close-twice rule was implemented, it was to solve the problem of open-close wars. There have not to my knowledge been delete-undelete wars, so it was deemed unnecessary to impose the same rules on deleting/undeleting.
  2. When the no-close-twice rule was implemented, they simply forgot to apply it to delete/undelete votes as well.

Is one of these correct, or is there another reason?

Note that this has resulted in delete/undelete conflicts, even between moderators and users:

Screenshot of the post history

3
  • Huh... I'm not sure i've ever even tried voting to delete a question that i'd previously voted to delete. goes off to scan "recently undeleted" list...
    – Shog9
    Jul 24 '09 at 23:21
  • 1
    Can we have a link to the question that the screenshot is referencing? I cant seem to remember that one just from the deletion names. Apr 10 '13 at 19:16
  • 2
    I agree with Behaviour--there's a need for this to be revisited in some sites.
    – apnorton
    Dec 28 '14 at 1:45
24
+50

I forget why we didn't apply it :)

Seriously though, we haven't seen widespread delete/undelete wars. Before adding additional rules/complexity, we like to see how the community uses a feature.

If it ain't broke, don't fix it!

Status as of April 10, 2013:

There have only been 111 questions on Stack Overflow where someone's cast more than one delete vote.

This includes questions where a delete vote was cleared by re-opening and then later re-applied after re-closing, but excludes situations where folks deleted/undeleted/redeleted their own questions and those where moderators did the same.

Many of the questions involved are extremely long-lived poll questions that were controversial well before we even started tracking this stuff, so the actual numbers might be slightly higher; still, there does not appear to be any broad pattern of abuse.

Why hasn't this been abused?

Well, closing and deletion aren't really separate systems - deletion requires closing, and because of this there really isn't much opportunity for it to be abused. A close-reopen war can occur on any question, but for a delete-undelete battle to occur, you need:

  • A post with enough supporters to undelete it after it's been deleted once, but...
  • ...Without enough support to be re-opened, and...
  • ...Without enough controversy to get moderators involved (as their delete votes are binding and they can lock a post to prevent it from being deleted).

Undeletion is pretty rare, but over half of all questions that do get undeleted by vote are then re-opened, invalidating any pending delete-votes and preventing any new ones. Because delete-voting requires that a question be closed first, you'd need a fresh set of close-voters to even have a chance at casting a second delete vote on a re-opened question.

31
  • 1
    Fortunately, that one stayed deleted the second time around (sorry, mmyers). No war, just emphatic action...
    – Shog9
    Dec 8 '09 at 21:10
  • 4
    I think this could use a revisit
    – Ben Brocka
    Jul 19 '12 at 20:43
  • 2
    I just raised another MSO question about this. Do you think this should be re-evaluated? Personally I think it's "broke" :)
    – Rachel
    Apr 10 '13 at 20:25
  • 3
    @Rachel I've brought it to the attention of the community team - it's not a difficult change, so I have no objections there. Apr 10 '13 at 21:35
  • 3
    @Rachel: See my update to this answer.
    – Shog9
    Apr 10 '13 at 22:54
  • 1
    No, the plan is to wait for evidence that this even can be abused. Close-warring was an actual, visible problem at one point, but the mechanics for how delete works are already considerably different and in four years there's not been any indication that this part of their design isn't working just fine. Also... You're really stuck on that hot question thing, @gnat - and that hasn't been declined, it just doesn't have the priority you think it does.
    – Shog9
    Apr 11 '13 at 13:55
  • 1
    @gnat: deleting questions that are under heavy meta discussion is kind of a separate issue; even so, that particular question had enough moderator involvement before it was even deleted once to make it a poor example of anything (beyond, perhaps, the need for moderators to be more proactive when they identify a controversial question) - for various reasons, I don't think it's feasible to ever limit moderator votes in this manner.
    – Shog9
    Apr 11 '13 at 14:18
  • 1
    @gnat: I know - that's why I excluded moderators (current and past) from my data. However, there's a ton of moderator involvement in Rachel's example, which I tend to think invalidates it to a degree - in a certain sense, most moderator-votes are multi-votes.
    – Shog9
    Apr 11 '13 at 15:26
  • 1
    @gnat: there's a diamond on the undelete line... And that's after it'd been closed, twice, by two different moderators. Also not shown: two subsequent undeletions and a re-deletion by a moderator. I think they have it under control. (should I invite you to chat since we're already in chat?)
    – Shog9
    Apr 11 '13 at 15:56
  • 2
    @Shog9 To be fair, the first undelete/delete by George was solely to mod-delete the question and prevent it from getting undeleted. The 2nd undelete was to lock the question instead of deleting it after I raised some objections to that. I'm OK with "we don't see any pattern of abuse so far so this isn't a priority for us", but I'd like to be sure it's checked on every once in a while. I think if this was widely known, it may get abused a lot more by small groups of users, and moderators would have to keep stepping in :)
    – Rachel
    Apr 11 '13 at 16:03
  • 1
    @Rachel: it's easy enough to check, but right now... Undeletion itself is rare; undeletion without re-opening rarer still - there's just not that much opportunity for abuse here.
    – Shog9
    Apr 11 '13 at 16:10
  • 5
    @Shog9 I'm more concerned about repeated deletes than undeletes, especially since undeletes are harder to get attention for and there's no mod-undelete the way there is a mod-delete to stop them. I've had users tell me they vote to delete solely because there's a reopen attempt for the post and they don't think it should be reopened, and often feel delete votes get incorrectly used in place of editing or explaining the problem. I hate seeing a user coming to meta asking why their question was closed and for help editing/reopening it, only to see the post get deleted instead.
    – Rachel
    Apr 11 '13 at 16:23
  • 1
    Just the moderator votes, @Rachel.
    – Shog9
    Apr 11 '13 at 16:38
  • 2
    @Shog9: on mathematics,SE at least, this does seem to be an issue - although I don't have a query at hand to count the number of multi-votes. I had posted another question about this on meta.SE, which unfortunately didn't get much attention: meta.stackexchange.com/questions/250433/… Mar 15 '15 at 12:40
  • 5
    FYI, this is now changing.
    – V2Blast StaffMod
    Oct 26 at 18:21

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