8

Some network sites, including StackOverflow, have changed the threshold for closing questions to three votes instead of five:

https://meta.stackoverflow.com/questions/391832/we-re-lowering-the-close-reopen-vote-threshold-from-5-to-3-for-good

This also comes with a corresponding change to the reopen threshold, so it's easier to both close and reopen questions.

How do people feel about requesting it for this site? I'm not sure if there would first be some analysis from StackExchange staff or whether they'd just do it, but either way it probably won't happen right away.

I'll post my own view in an answer but of course please write a competing one if you disagree.

2
  • 1
    I had some general thoughts on this over on Anime.SE, if anyone needs some clarity.
    – Machavity
    Oct 23 '20 at 14:08
  • We don't have any close queue right now, nor have we the last three or four times I've visited the site in the last two weeks. Is this really a problem needing solving?
    – Joe
    Oct 31 '20 at 22:57
7

Personally I feel this would be a good change. A lot of questions that should be closed do get the votes in the end, but it takes a while. In that period they acquire answers or draw attention.

On the flip side sometimes questions are fixed or closed inappropriately but struggle to (ever) attract enough reopen votes.

The obvious downside of making this change is that we could get more questions flipping between closed and open repeatedly if different people take different views on them. My feeling is that this would be quite rare.

5
  • The "think different" part of me says if we're worried about excessive "flipping", we could potentially ask for "4" (unless there's an unknown-to-me reason for it being "3 or 5"). However, if StackOverflow – who I assume have much more traffic and voters – aren't seeing "flipping" as a problem (complete assumption, based on nothing other than they haven't reversed their "Permanent™" decision), then I suspect we won't.
    – TripeHound
    Oct 21 '20 at 22:39
  • @TripeHound StackOverflow had to do this for a very different reason I think - simply far too much needing closing. They're also far more strict - and I'd say too strict, honestly; it's extremely common that in my small tag I see a question closed from review that shouldn't have been (the reviewers are either just plain wrong, too strict in my opinion, or don't understand the language to know what a good question is).
    – Joe
    Oct 22 '20 at 23:18
  • @TripeHound there is a known reason for it not being 4. Nov 10 '20 at 21:57
  • @RobertLongson That post doesn't give a reason for not choosing an even number, other than the fact that the documentation for the setting (at least as it was 5 years ago), said "Keep this an an odd number." Are you aware of a reason for that restriction, and is there any possibility that that restriction has been lifted over the last 5 years? Nov 11 '20 at 15:31
  • @BenMiller-RememberMonica That comment made the current developers reluctant/unwilling to try 4. They don't know why the comment is there. Nov 11 '20 at 15:38
6

I think it would be best to require 5 votes to close, and 3 votes to reopen.

Many people put much effort into their questions (especially some non-native English speakers), and having them closed would be more appropriate as a last resort, not an immediate response.

On the other hand, if 3 community members feel a post is worthy of being reopened, that seems sufficient. It also will allow posts edited after closure (or inappropriately closed) to be reopened faster.

4
  • I believe that isn't an option on offer, though we could ask if it's very popular. Oct 23 '20 at 16:13
  • @GS-ApologisetoMonica Thank you. Much appreciated. Oct 23 '20 at 16:14
  • 2
    I was thinking the opposite, 3 to close, 5 to reopen, as there should be a higher threshold to undo a closing, in my opinion. Nov 1 '20 at 18:48
  • 3
    @JTP-ApologisetoMonica I disagree, but I certainly respect your opinion. I find there are too many people quick to close anything they don't instantly understand. Honestly, I don't think it's as bad on this SE site as many of the others, but I've noticed it as an overall problem plaguing SE. I've even witnessed people bragging about how many posts they have closed, like that's somehow something to brag about. Helping people write better questions, in my opinion, needs to be the priority, not encouraging individuals on a power-trip to quickly close as many questions as they can. Nov 2 '20 at 0:31
3

I don't see a problem getting close votes on this site.

I participate in https://webapps.stackexchange.com/ where a year ago we started using the three close votes limit. The issue was that many questions that should have been closed had trouble attracting 5 close votes because there were few users that would look at any question that wasn't related to their area of expertise.

If the question was a obvious duplicate, for example I can't remember my gmail password; it might never get 5 close votes.

I don't see that we have that level of problem. So I don't think we need to make the change.

1
  • 1
    I agree we do mostly get the votes but it feels slow. I've edited my answer to clarify that point. Oct 22 '20 at 14:10
3

I think that far too many questions are closed on Personal Finance where I primarily participate, sometimes foolishly. Many questioners put a lot of effort into their question(s) and they are seeking a solution to a legitimate problem, yet the rigidity of the system/participants reacts adversely.

I also think that participants have a greater propensity to close questions and far less to reopen them.

Should we ask to lower our close and reopen votes threshold to 3 (from 5?)

What's best? I have no clue. Lower the threshold to 4 and reevaluate over time. I would err on the side of encouraging visitors and members to participate rather than closing questions quickly and not reopening valid ones, dissuading future participation.

4
  • I don't think you'd be allowed to pick 4 because it's not an odd number. Nov 10 '20 at 21:56
  • What's the need for it to be an odd number of close votes rather than an even number? Nov 10 '20 at 23:16
  • Per the link in my comment the current developers don't really know why the original developers warned them that was so. Nov 11 '20 at 4:17
  • err on the side of encouraging visitors and members to participate... That works best when moderators are also encouraged to continue moderating. I've observed that as the main contention with your viewpoint on Stack Overflow, but don't know as much about this site.
    – jpaugh
    Nov 11 '20 at 20:48
2

We could lower the threshold to 4 instead, and see how it goes.

I agree with TripeHound's comment:

The "think different" part of me says if we're worried about excessive "flipping", we could potentially ask for "4" (unless there's an unknown-to-me reason for it being "3 or 5"). [...]

GS mentioned in a comment that "we do mostly get the [necessary close/open] votes but it feels slow". Lowering the close/reopen vote threshold to 4 would slightly increase the speed of closing/reopening while preserving the existing moderation quality on this site.

1
2

To put some data to my assertion in comments, from this query, close votes per question closed:

1: 60
2: 64
3: 74
4: 48
5: 207

Out of 453 questions that were closed, 207 (46%) were closed with either no moderator vote or the moderator being a 5th vote (so, acting as a regular citizen). 124 (27%) were closed with a moderator casting either the first or second vote. 122 (27%) were closed with a moderator casting either a third or fourth vote.

Changing the policy would, then, affect perhaps 27% of questions (the questions closed with 3 or 4 votes), and perhaps 122 questions over the past 300 days, for less than half of a question per day.

This assumes that nearly all questions that should be closed, are, but I think that is the case here - we don't tend to have questions sitting in the close vote queue for months.

Close dates are fairly good on this site. Of the 423 questions opened in 2020 that were subsequently closed, 174 questions were closed on the same day they were opened (41%), and 292 (69%) were closed within one day. 383 (91%) are closed within a week. Only 40 questions in total were closed in more time than one week, which seems entirely reasonable for a site of this size.

Edit: Per GS's question, most questions that are closed do have answers. Of the 434 questions (slightly more due to later runtime of the query, note) that were opened in 2020 and subsequently closed, 123 (28%) have zero answers, while 205 (47%) have one answer, and the remaining 106 (24%) have more than one answer.

5
  • I guess for me the interesting question (which I was too lazy to dig into) would be how many closed questions get answers. Nov 1 '20 at 10:25
  • @GS-ApologisetoMonica Meaning, how many questions are closed having answers? Or closed->reopened->answered.
    – Joe
    Nov 2 '20 at 2:02
  • I updated the answer with the former (questions are closed having answers).
    – Joe
    Nov 2 '20 at 2:08
  • Something we need to consider while looking at this data: It only shows questions that have not been deleted. This skews the data, because closed questions without answers get automatically deleted by the Roomba in 30 days. This explains why you see that most closed questions have answers; the ones that didn't were deleted. It also may have an effect on the other conclusions that you drew, because the questions that were closed quickly are more likely to have been deleted than the questions that took longer to close. Nov 9 '20 at 22:46
  • @BenMiller-RememberMonica Good point - had forgotten about the roomba! I'm not sure what can be done about that, though; 30 days isn't really enough to have very useful data points I don't think.
    – Joe
    Nov 9 '20 at 22:46
0

It seems there's a fairly wide diversity of opinions and not enough data anyway. I won't do anything further for now, and if I propose it again I'll look for better data first.

You must log in to answer this question.

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