3

By default on Stack Exchange sites, users with sufficient reputation (500 on our site) can vote to close a question or to reopen a closed one. Once a user votes to close or reopen using the link under the tags of a question, the question appears in the corresponding review queue and stays there until a decision is reached (five votes for or five votes against the action) or two weeks have passed since the last vote. Votes to close or reopen can be cast directly from the question without passing through the review queue. The magic number of votes for any such actions is five. Should we lower this number from five to three?

Stack Overflow made the switch. At least German and Literature have considered it. It should be possible for us too if we request it. It should also be possible to switch back if problems arise.


Update

The proposal has some votes in support and no comments or votes against, and it has bee featured on meta for a month. Moderators agree that this is a good idea. Thus we can consider it decided that we do indeed want to lower the threshold from five to three, and we are now waiting for staff to make the change.

If there is a need to return to this topic later for any reason, please ask a new question.

1

2 Answers 2

4

This request is live as of today.

I've reviewed your stats and, while there are few questions that get closed over the course of a month, nearly all of them are handled by moderators.

Graph showing the number of questions closed and reopened by users and moderators over time. The yellow line indicating closures by moderators is hidden by the purple line of all closures nearly everywhere other than in December 2021. In general, between 2-10 questions are closed per month with some months above and below that. More questions have been closed and reopened in the last year than the prior year.

That said, this change is unlikely to have much impact in the workload of moderators.

Graph showing which vote the moderator cast when voting to close a question. While there are some cases where moderators were casting the second or third vote to close, they were more commonly casting the first vote to close.

As seen in the graph above, moderators are casting their votes as either the first or second vote to close in the past. Moderators could opt to "slow down" voting to give community members the chance to weigh in but I leave that to you to determine.

There seem to be plenty of high-reputation members on the site who can take over this work if they opt to do so.

2
  • 1
    Many thanks for both the switch and the data to back it up! We'll probably post a separate meta announcement about this soon, partly to encourage people to vote.
    – Joonas Ilmavirta Mod
    Commented Jun 13, 2022 at 16:15
  • Sounds good to me :)
    – Catija
    Commented Jun 13, 2022 at 18:24
8

I think we should lower it from 5 to 3.

In regular voting of questions and answers all votes count. In closing and reopening they only make a difference if the threshold is reached, and this very rarely happens on our site. I would like those votes to become more useful and meaningful, as that would help build further our culture of community-driven moderation.

A vote by a moderator immediately causes the action. That's why I always hesitate to vote and often decide not to; I wouldn't mind giving my simple vote but I don't want to force my view on it. I feel easier giving the final vote if there are already two or three votes in support. This makes the community votes essentially advisory in nature, whereas they should be executive.

I hope lowering the threshold would make it easier for all of you to decide which questions should be closed. If the site grows and the culture for casting such votes grows, we could consider switching back. But at the current rate I see no risk in lowering the threshold. Closed questions can be reopened and in problematic cases you can flag for moderator attention or discuss the question at this meta site or chat.

We currently have 105 users with at least 500 reputation points; for current information on reputation distribution see the reputation league page. But what matters is not how many users have the privilege but how many are actively using it, and that is always just a fraction in cases like this.

You must log in to answer this question.

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