102

tl;dr

Users that have retracted a close vote should be able to vote again on that same question.

I don't believe there is any danger in allowing this, provided some controls are in place (see below).

Long Version

As a 3k+ user, I want to look at a question at any point in time and determine whether it belongs on this site and, if not, state exactly why it doesn't in a close vote.

People who write questions are free to edit them at any time. As a result, close votes can become invalid and the new close vote retraction feature was introduced as a response to this.

While I welcome the new feature, I wonder why we have stuck with the "you can only vote once" paradigm. If I'm monitoring a question to the extent that I withdraw my close vote, why shouldn't I be allowed to vote again if subsequent edits render the question unsuitable? I propose that if you retract your close vote on a question, you should be allowed to vote again.

Note that by implementing this feature, one would also be able to change a close vote (by retracting and then selecting a different close reason). To combat willy-nilly vote changing (which may confuse other users), I propose one of the following defences:

  1. An edit must occur between retraction and re-voting. This ensures something has changed in order to make the user want to vote to close again.

  2. A time limit ensures that X minutes must pass between retraction and re-voting.

  3. No specific restrictions are enforced, but the reputation required to re-vote is raised to something higher than 3k. The idea being that more experienced close voters are unlikely to change close reasons lightly. Perhaps this could be a tantalising rep target at 7.5k (to fill the void between 5-10k).

A similar question was asked previously, but I'm hoping some of my defences listed above will negate the fears listed in the accepted answer.

Example Use Case #1

A poor question is raised and the OP receives several close votes and comments. Being a good user, the OP then adjusts the question to meet the quality standards of the site. One or more close-voters retract their vote now that the question is good enough.

But wait! A duplicate has now been found, which wasn't spotted originally due to the poor quality of the question. None of the retracting close-voters can vote to close as a duplicate and are forced to leave a comment. Hopefully passing 3K+ users will spot the comment and act accordingly, but unnecessary time is wasted. This question should be on the close queue and heading for death.

Example Use Case #2

Commonly on Stack Overflow, a user will present insufficient information to diagnose a problem. Perhaps their code example doesn't compile or they haven't provided definitions of all the referenced methods. I will leave a comment to indicate this and, after a small grace period, will vote to close with "This question appears to be off-topic because it lacks sufficient information to diagnose the problem.".

If the user edits their question and adds the info, I retract my close vote. But I often realise the question should be closed for another reason, e.g. a duplicate or the new(ish) reason: "This question was caused by a problem that can no longer be reproduced or a simple typographical error.".

Example Use Case #3

Perhaps the most common: you find a reasonable duplicate and vote to close. But then... you find the most wonderful, exact duplicate you could imagine. Too late, your vote is cast. You can retract it, but you can't then easily kick the ball rolling on a new closure with the real duplicate. Sure, a comment might cause things to start, but often it doesn't. Flagging for diamond attention isn't appropriate.

This can also happen when the OP refutes the original duplicate and provides some extra information that steers you towards a related but different duplicate.

10
  • related: Closure Vote - Changing closure reason Oct 11, 2013 at 13:21
  • 37
    I see a problem when I vote to close a question as "unclear what you are asking", OP edits the question, now it is clear, but offtopic/duplicate/whatever. In this case, I'd like to change my close vote. Oct 11, 2013 at 13:43
  • "if subsequent edits render the question unsuitable" you should roll back to the version that was suitable.
    – Mołot
    Jan 15, 2014 at 12:56
  • @Mołot Perhaps it wasn't great wording, but imagine if the changes to the question make it into a duplicate? That shouldn't be rolled back. That should be closed. Jan 15, 2014 at 13:00
  • @Duncan if it used to be a valid non-dupe, it should be rolled back :)
    – Mołot
    Jan 15, 2014 at 13:02
  • 3
    @Mołot Nope. I disagree. The important thing is that the question reflects what the OP wants to ask. If, as we discover more details, it's a dupe then that's a good thing and we should close it. Jan 15, 2014 at 13:03
  • 1
    @Duncan if there was details lacking, it should be put on hold as too broad or unclear. So in that case there was no valid revision to roll back to. And chameleon question by all means should be rolled back.
    – Mołot
    Jan 15, 2014 at 13:08
  • 6
    A more common failure mode in my experience: I vote to close, somebody else edits the question to make it suitable, I retract, and then the OP rolls back the edit. I'm very reluctant to retract close votes on small sites (where there aren't that many active users with VtC in the first place) for that reason. Jan 15, 2014 at 17:20
  • One use case that I don't see here, is voting to close for the wrong reason accidentally. This just happened to me on this question: blender.stackexchange.com/questions/32772/… It makes me quite angry that I can't correct my vote. For this reason, I don't see a reason why there should be a rep wall on it. Can anyone give me a reason why I shouldn't be able to change it? Jun 21, 2015 at 1:34
  • I too feel the need for this feature. I am not able to vote again after retracting. stackoverflow.com/questions/65444881/… Dec 27, 2020 at 5:04

7 Answers 7

20

In theory this sounds useful, and quite consistent. What you're proposing is that we treat close voting as we do normal voting, where you can change your vote one way or the other once a post has been edited, as many times as you want for however many edits it receives.

I worry about the following things, however, and the thought of them leaves me sufficiently concerned that I'm willing to say this might be a bad idea.

Consider:

  • Easy to just edit a post so you can vote to close it again
  • The same group of users hovering over a question to decide its ultimate fate, now consider this group at odds with each other. The system would have to allow this only on substantive edits.
  • Suggested edits approved by people other than the author seem really weird here, even if substantive.
  • How often do you really think this is necessary?

Now, we could work around the potential for abuse, but that leads me to my last point:

  • This would not be easy to implement

I'm not marking this as declined because you've tickled my consistency bone and I want to see input from more folks. In other words, this is just my opinion on the feature, and not an official response to it.

In the meantime, if you see a question flip-flopping like the scenario you describe, it's probably time to have a moderator take a look at it; they aren't limited on the number of times they can close or reopen a question with a single vote - and there may be other oddities they should have a look at.

15
  • 2
    Maddeningly, I can't think of more examples right now. I just know that I've felt the need to do this on several occasions since the new retracting feature arrived. Your points about edits are excellent and valid. Perhaps I put too much faith in the idea that people of a sufficient reputation (i.e. my 7.5k idea) would not behave that way. In the meantime, I'll try to consider a simple method to counter those problems because I agree that the feature isn't worth it if the implementation isn't fairly painless. Oct 11, 2013 at 13:41
  • @DuncanJones I don't think many of our higher rep users would abuse this on their own, those that would are surely a small minority. When you get into a group scenario, with snark showing and tempers climbing, people can behave somewhat differently. My concern is, contention predicates the use of this; I've seen even some of our top closing users arguing with each other in comments as if the question author wasn't even there, or had a say in it. This is why we don't let them use magic outside of Hogwarts. I may be over-thinking, more examples if you manage to find them would help.
    – Tim Post
    Oct 11, 2013 at 13:48
  • Ah, excellent. Johannes has just thought of an example which has certainly happened to me before. Although I suppose this is only a small variation on what I already suggested. Oct 11, 2013 at 13:50
  • What if any edit was considered valid, but not if made by the person trying to re-vote? Fairly easy to implement, but less open to abuse. Oct 11, 2013 at 13:56
  • I'm still worried about the 'hovering' thing, which we see somewhat often coming from chat. Intentions there are always good, but when someone's question becomes the room topic briefly .. I need to chew on this some more.
    – Tim Post
    Oct 11, 2013 at 14:09
  • 5
    To avoid the hovering issue, it could be changed to only 1 successful close and open vote Oct 11, 2013 at 14:25
  • 2
    I stated this in the comments of my original request for the current feature: allowing people to re-cast a retracted vote also allows them to keep bumping the question into the Close Votes review queue.
    – animuson StaffMod
    Oct 11, 2013 at 16:01
  • 4
    @animuson so, far, the only argument that I've found for not allowing people to freely cast and undo their close votes as much as they want, is because it would allow them to bump the question in the close votes queue. Are there any other arguments that I'm missing Oct 11, 2013 at 19:04
  • @animuson And it might also be useful to understand why people might be driven to do that? I struggle to see the benefit... Oct 12, 2013 at 12:45
  • I've added a bounty to hopefully draw some attention and provide that community feedback you seek! Jan 15, 2014 at 12:06
  • @TimPost You put me in strange position, as I agree a lot with both you and OP ;) I hope my answer answers "hovering" and other issues? If this was a forum site, my answer would really be a reply to you. If I omitted some risk, please tell me.
    – Mołot
    Jan 15, 2014 at 13:14
  • 2
    I've just encountered another case today. This question (revision 1 at least) I voted to close as "unclear". However, it's quite plausible that once some additional info is added it will become "too broad" or perhaps even a duplicate. Jan 16, 2014 at 9:27
  • @TimPost I've edited my question to include two new use cases. I think the recent adjustments to closure reasons (on SO at least) have made this problem more common. Apr 30, 2014 at 8:49
  • 3
    @Duncan The use case for duplicate is quite interesting. I'm going to talk to a few folks on the team about that, because we're eyeing duplicates in general. Gonna touch back in a bit regarding the rest that you added, thanks for caring about this so much.
    – Tim Post
    Apr 30, 2014 at 12:10
  • 1
    @animuson Why not only have it get bumped if the question has been edited, or it's the first close vote? But seriously, what would the point of continually retracting the vote to bump the question? I feel like that's more of a situation for mods to take care of instead of the SE system. If a user is doing that type of thing, chances are, they're probably not very good users, and would most likely abuse other parts of the system as well. Jun 21, 2015 at 1:46
12
+25

"if subsequent edits render the question unsuitable" you should roll back to the version that was suitable, not vote to close.

On the other hand, changing a close vote reason without retracting it would help in the situations when question turns from "unclear what you are asking" to "off topic" or "too broad". And I see it happen pretty often.

Allowing close vote change would help to close question with a proper reason, as Duncan wanted, but without flaws Tim Post♦ mentioned. No option to cast another vote prevents "hovering" and constant re-closing by the same group, and edit-to-close approach, at the same time allowing questions to be closed with the reason most proper at the moment of actual closure, if the reason changed.

11
  • Can you expand on why this approach would negate Tim's concerns? Jan 15, 2014 at 13:04
  • 3
    Because once vote is deleted, the same person would not be able to vote again. So no way for "The same group of users hovering over a question".
    – Mołot
    Jan 15, 2014 at 13:05
  • I agree that edits that break a question should be rolled back, but we also tend to say that the author gets the final say on edits -- so what do we do if the author is the one who rolled back to the unsuitable form? Jan 23, 2014 at 22:00
  • @MonicaCellio we treat it as an chameleon question - current consensus seems to be: rollback to what it meant when answers was posted, if it invalidated them, and comment to the OP about not chameleoning, and keeping questions in good shape. If in doubt, or if OP still disagrees and it starts to cause problems, flag for ♦'s attention describing it as chameleon and let them worry.
    – Mołot
    Jan 24, 2014 at 7:41
  • Oh, once a question has answers I agree -- an edit that invalidates answers needs to be rolled back no matter who did it. I meant before that. Thanks for the tip to pass the buck to the mods. :-) Jan 24, 2014 at 13:53
  • @MonicaCellio even without the answers if OP insist on turning a question into something unsuitable, passing it to ♦ is best bet. They are either paid to handle this, or they were chosen as most trusted ones to handle this. Both unsuitable questions that can be edited to be OK, and edit wars with OP, are bad things to happen.
    – Mołot
    Jan 24, 2014 at 14:08
  • 1
    If the OP insists on an unworkable question it should be closed, not replaced in place. (Someone else is free to ask the workable question.) But that brings us back to the original feature request: once one of these edits occurs now, you can't recast the close vote you retracted on the basis of the good edit. If it has to be handed over to a mod then so be it, but this is something the community ought to be able to do. On large sites it may not be a problem (ok, so five other people will close it), but on small sites this can be a hinderance. Jan 24, 2014 at 14:44
  • @MonicaCellio I came from Drupal Answers - it is not a problem there. Moderators are pretty responsive. And "community" should be able to re-close bad question, OK, but at the same time the same 5 people should not be able to re-close it again and again. If community is too small, well ♦s are exception handlers, so it is their job to step in when there are 2 needs impossible to meet.
    – Mołot
    Jan 24, 2014 at 14:59
  • On one site I use (I hesitate to call out a bad example for public scrutiny), most moderators like to say that the community should handle things that it's very difficult for the community to actually handle because of its small size. I guess the takeaway here is that there's a lot of variation in the 100+ SE sites, and we need to remember that what works on some doesn't on others and, conversely, that what's a problem on some isn't on others. It's a pity we can't do more per-site tuning. Jan 24, 2014 at 15:03
  • @MonicaCellio You can always ask to change moderators.
    – Mołot
    Jan 24, 2014 at 15:06
  • 1
    Oh, they know how to stay in the community's (and SE's) good graces on average. It's just that some functions work better than others. Jan 24, 2014 at 15:11
12

@TimPost makes some strong arguments against treating close voting like regular voting (i.e. clearing the 'I haz voted' flag after each edit). But I think the real issue is a subset of this, and is that retracting a close vote ought to allow another close-vote. There are many reasons for this. In my personal experience, I have experienced at least the following reasons, in addition to the excellent reasons in the original question above:

  1. I VtCed as duplicate to the wrong question, and I want to change the duplicate link.

  2. I VtCed, then the question was edited, to make a different close reason more appropriate (from Off-Topic to Too Broad, for instance)

  3. I VtCed, the question was edited to bring it within scope, I retracted my VtC, then the OP rolled back the edit. Now I want to VtC again.

Even if nothing else is changed in the VtC/Retraction system, a voluntary retraction (as opposed to the question being closed and reopened, for instance) ought to clear the 'I haz voted' flag.


EDIT Feb 9, 2015

Another case I happend upon today:

  1. OP asks a question.
  2. Question gets closed as Too Broad
  3. OP re-asks question.
  4. Users VtC new question as a duplicate of the old.
  5. OP deletes the old question.
  6. Dupe VtCs are deleted, and now these users can't VtC as 'Too Broad' on the remaining question
7

If you want to prevent hovering, then change the vote mechanic so that rather than forgetting the vote after reaching 5 and changing the status of the question, the votes, both to close, and to reopen, remain in place and the status changes when one exceeds the other by 5.

In other words, if 5 people vote to close, and later 5 people vote to reopen, then the question is reopened, but those first 5 people's votes are still recorded so they can't add a second close vote to close it again. If 5 other people vote to close, then it is closed again. Unless one or two more people vote to keep it open first, which you should be able to do before the question is actually closed.

Either way, retracting your vote should not prevent you from changing your mind again later.

6

This question and its status tag as of writing this answer () seems to have caused much confusion over the years. As such, I'm removing the tag because... while it might be completed in some cases, the actual request here is not. Whether we would change this feature, I don't know at this point - like Tim, I do appreciate the limitation of requiring an edit or, in particular, an edit by someone else - but I also see room for gaming this and I'm uncertain how frequently the hypothetical in the question occurs.

Before I dig into that any more, for those who want to know when we actually allow recasting close votes, read this paragraph... if a question that you cast close votes on doesn't get closed, you can cast a new close vote 14 days after the vote aged away (for the full details read this giant post). This change occurred in March 2015, just a few months before animuson tagged this as completed.

You can not recast a close vote if:

  • You retract it.
  • The post is closed using your vote to close and then reopened.

The heart of this question is the retraction feature and aged away close votes aren't retracted by definition. While I'd say that it's relatively common for a question to need a different close reason, I don't know that it's particularly common for a question to follow the pattern as described in the question:

  1. Get a close vote
  2. Get edited to improve it thus making it a good question
  3. The close voter becomes aware of the edit, agrees that the question is in scope and then retracts the vote to close
  4. The question gets another edit (or more likely a rollback) that makes it again out of scope for the site
  5. The same close voter is aware of the edit and wants to recast the vote to close.

I could probably just pull data on retraction counts and get a very low number without even looking into the other steps... but I have no clue how to query SEDE for retractions. 🤣

For use case 1 (Close votes are cast and then retracted after an edit and then a duplicate is found), there's not really a solution here other than finding other people to vote to close and leaving a comment pointing at the duplicate. That said, this example doesn't meet the question's own requirements - if the retraction happened after an edit and a duplicate was found - there wouldn't necessarily be an edit between the retraction and wanting to close as a duplicate.

For use case 2 (edits cause close voters to identify a different reason the post should be closed). While this could be addressed by allowing retracting and allowing voting for a new reason, this has the same problem as the last example - there's not an edit occurring between retracting the vote and wanting to cast a new vote to close. Also, I'd say the better way to address this situation is to allow a close voter to change their close reason - no retraction needed, no second close vote required.

For use case 3 (a better duplicate is found and user wants to change duplicate target), this could be solved with the same change for use case 2 (allow changing close reason) but it is also at least partially addressed by the 2017 feature allowing gold badge holders and moderators to edit duplicate targets.

So, I guess, in the end, even if we opted to build something like this, I think that other changes would meet the needs better than this specific request and some features released since this question was created actually address some of what it was trying to solve.

1
  • So, to be clear, this request is implicitly declined, because use case 3 is solved by the dupe target changing feature, use case 2 is better solved by another feature request, and use case 1 is extremely rare? Apr 10 at 11:30
3

I have a recent example where I think it's very frustrating that I can't retract my original close vote, and cast another (or simply change my close reason, in order to make the site working as intended).

This is the post: Ubuntu 22.04 Bluetooth does not work

I originally voted to close, because the post lacked detail. Now the user has edited the post with more detail - and in the meantime I have found out that this is actually a bug, that should be reported, and thus isn't within the scope of Ask Ubuntu.

Now it would be completely natural for me to retract my first close vote (lack of detail) and replace this with another close vote (this is a bug).

But I can't do that. Can anyone explain why the above shouldn't be possible? Actually here I would rather have that I can change the reason for the close vote - so why isn't this possible then?

-2

I would agree with this if it were limited to duplicates. If you find a better duplicate it makes sense to be able to update it. If, as you say, one finds a duplicate after closing it for another reason it makes sense to be able to update to help the OP more, however poor their question.

I don't really see that it's fair to the OP if it were done for any other close reason.

6
  • I think duplicates would be a common use case, hence why I included it in the question. But why would it be inappropriate (or unfair) to vote again to close if a question has been edited to the point of unsuitability? Other passing 3k users could pull the vote trigger - why do we specifically want to prevent 3k users just because they have voted before? Oct 11, 2013 at 13:21
  • 1
    Because you can roll it back @Duncan and then ask for a moderator lock if the OP keeps reverting. Oct 11, 2013 at 13:23
  • Sorry, I didn't quite understand your last comment. Would you mind rephrasing? Oct 11, 2013 at 13:26
  • If someone messes up their question then roll it back Oct 11, 2013 at 13:32
  • I'm kinda liking a 'Suggest better duplicate' feature on its own, to be honest. It could allow input of another question then churn it back through review. The repetition there seems a bit off, though.
    – Tim Post
    Oct 11, 2013 at 13:37
  • I think rolling back the question would be a bit rude in certain circumstances. Perhaps the user has altered the question significantly, but just strayed off the edge of acceptable. Simply rolling back that change would be rude. But a comment, plus a close vote (in case the comment is ignored) would be appropriate. Oct 11, 2013 at 13:42

You must log in to answer this question.

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