Would ELU like to start a trial of needing only 3 votes to close/reopen a question?

Question

Many sites in the network are currently considering whether they would like reduced close/reopen vote requirements, from 5 down to 3, for a one month trial.

As the close vote queue almost always has at least 100-200 questions, I thought this could be a site that could benefit from that. This would allow those who do use the review queues to get more done.

Note that both vote requirements would be changed together, so while it will be easier to close questions, it will also be easier to reopen them if the community thinks the closures were not warranted.

What does the community think?

UPDATE: Community Managers are working to implement this now.

UPDATE: The trial is now active

Answer 1

I think it is broke, and this would be worth at least trialling.

Currently there are over 160 questions in the review queue. I confess I don't use the queue much; I simply don't have the time and since my vote is a binding vote and designed for exception-handling, I would far rather that the community policed itself.

If there aren't sufficient members who are prepared to do that then reducing the threshold seems a good thing to do, or at least to try out. There are certainly a huge number of questions which are borderline in terms of closeworthyness, and if they are closed and then roomba'd that is probably a good thing.

It's worth adding here thanks to those who do use the queues and work on them. While the top close-vote reviewer is [now] a moderator, there are many who aren't, and we do now have nearly 150 Steward badges. Please keep up the good work. In the meantime, let's reduce the burden a little.

Answer 2

Would ELU like to start a trial of only need 3 votes to close/reopen a question?

Update April 30, 2021

Community Managers are working to implement this now.

No, thank you.

To close a question all that is needed are five users with at least 3,000 rep. Just five people out of hundreds or maybe thousands of regular users. Users can cast up to 24 votes per day.

I agree with Andrew Leech, the system is broken not because on January 14, 2020, there are 148 questions in the review queue, which is a TINY drop in the ocean compared to Stack Overflow's +3K, but because the site struggles to find twenty 3k users who feel invested enough to help moderate and keep the site clean.

Maybe the community should invest time in encouraging newcomers to stay, and start upvoting their efforts more often?

There are some EL&U users who only close posts in the review queues, it's very rare that they care enough to edit and keep a question open. I suspect these users will support the initiative wholeheartedly.

I do not.

The Practicalities

1. Banner

What if three close voters close a post for three different reasons, e.g. "opinion based", "off-topic": lack of research, and "needs more focus"? Which of the three reasons is shown to the author? Will all three reasons be displayed on the post notice?

Dan Bron, in the comments below, pointed out that the first vote is selected as the reason for closure when there is no clear majority. Here is the relevant excerpt, please note the significant drawback to this feature.

questions closed without a consensus

… And it’s even more problematic because of how we determine the displayed reason: we pick the oldest vote. So if someone picks a bogus reason, that gets discussed a bit and then two other people pick two better reasons… We show the bogus one. @Shog9 (source)

2. Reopen Queue

Lest we forget, the author of a question that has been closed by the community can cast a vote to reopen it. Currently when a question enters the review queue it drops out when three users have agreed to keep it open. Will that same number remain or will it be reduced to two?

3. Migration

In order to migrate a post to ELL there must be at least 3 votes in favour, by reducing the number of close votes to three you effectively reducing the number of posts that are eligible or suitable for migration.

A little history and an update

As of January 22, 2020, the number of questions that need to be reviewed on Stack Overflow is a staggering 4.8k, little wonder, then, the one-month long experiment promoted by Shog9 to reduce the backlog was greeted with overwhelming consensus and success in August of last year and then finally adopted on December 3, 2019 by Stack Overflow.

Source

The Review Queue on English Language & Usage does not have anything comparable to the above. In the ten days following the OP's proposal, the backlog has been reduced to 126 questions. This site needs more users willing to muck in that's all.

Answer 3

Although implementing this proposal carries obvious risks, it is worth trying because it promises to reduce the number of questions that are closed after being answered.

The purpose of closing seems to be to prevent the questions that do not belong to this site from being answered on it. When the concept of closing was built into the framework of the Stack Exchange, it was probably expected that closing would, in most cases, take place before anybody attempts to answer the question. On ELU, however, closing accomplishes that purpose only very poorly, because voting to close takes time, while answering doesn't. Many questions that eventually get closed thus get answers, sometimes several answers; quite a few of these answers even get upvoted, and some get accepted.

The presence of a large number of such questions is awkward. While the regular, core contributors to this site attach great importance to closing, that importance is not obvious to casual visitors. Such a person may come to a particular page on this site from a general-purpose search engine, read the question and the answers to it, note which answer is accepted and which ones are upvoted, and not pay any attention to the annotation that the question is closed (which may strike them as some piece of bureaucratic boilerplate). In other words, a casual visitor's experience of reading the answers to a closed questions is not significantly different from the experience of reading the answers to other questions; from their viewpoint, these answers are all equally parts of what this site offers. The point that the core users are trying to make by closing a question is lost on the casual visitors, as long as some answers to it have been posted.

It gets worse, though. It sometimes happens that the answers that are posted to a question before its closing are all problematic. Closing then effectively protects these bad answers by preventing better answers from competing with them. Casual visitors who do not understand how closing works may be left with the impression that the problematic answers that somebody managed to post before closing are the best answers that can be made to the question.

So, if some question really deserves to be closed (and when that is the case is a separate topic), it is far better that it be closed before anybody posts an answer to it. Reducing the number of votes required for closing will probably significantly reduce the time that it takes for a question to be closed, and thus the awkwardness created by the questions that are answered before being closed (although it, of course, won't eliminate them completely).

Answer 4

I can't deny I've been remiss in not going through the review queue so often lately.

But when I do tackle the "Closevote" queue, I invariably hit my daily limit (20 VTC), because the queue is so long. And that really irritates me, because it means I can't VTC against current questions, as they arise.

Perhaps I should post my thoughts as a "Feature Request" here on Meta, rather than an "Answer" to this specific question. Whatever - my suggested solution is...

If a user runs out of closevotes while working on the review queue, award them extra votes.

---

As to the detail, perhaps they could be "special" VTCs that aren't valid for continued queue review (only to be cast against questions in other contexts). And maybe the number of such additional votes could be limited to the number of queue VTCs cast in the most recent session.

Just an idea.

Answer 5

UPDATE

After reviewing the stats, it looks like this change has been very much beneficial for you all. While you had been doing in the 40-80% completion range for questions gaining at least one flag or vote to close, you're now consistently in the 80-100% range, though it's been dipping a bit recently.

On top of that, while the mods were never doing a significant portion of the work, they're needing to participate less than they were, with the community doing most of the work themselves. Note, the numbers in the graph below are more volatile because they're quantity, not percentages - this means that they vary greatly based on the quantity of questions that are closed and reopened - it's also nice to see a slightly higher number of questions reopened after closure.

I'm going to call this a success.

---

Some of you may have noticed the Meta Stack Exchange post - Testing three-vote close and reopen on 13 network sites (it's linked in the featured on meta sidebar) - we've finally got this project under way and English Language & Usage is one of the sites we'll be running the test on.

Starting tomorrow, I'll be changing the site setting and closing and reopening will require only three votes. This test will run for 45 days and will be turned back to five votes to close and reopen while I review the data from the 13 sites. After we've seen the impact, I'll be posting results and, if there aren't negative impacts, we will change the setting to three permanently.

A few weeks into this, I'll be posting a question here on meta to ask for your thoughts about this change, so you will have an opportunity to discuss the impact.

Thank you so much for your patience while we got this prioritized and scheduled. There's a lot more information in the MSE post, so please review it.

Answer 6

I'm a frequent close voter, and I think it's worth a try.

There's a double-layered effect where misdirected questions often get answers that are of poor quality, which further increases the queue length. This will prevent some of those from ever getting in the queue.

I began focusing more on cleaning that answering when I noticed that google hits on nonsense material began overwhelming the useful content. I don't agree with all close votes. There are some legitimate differences of opinion on what constitutes an opinion question, whether single-word-/phrase-requests are valid in general or require more research, and what type of questions should be moved to ELL. However, there is a deeper issue on stackexchange that closing doesn't always mean a bad question. Duplicates are intended to push answers to the same question into the same pile. You don't have your question deleted, it's another possible entry point to finding an answer. The you-didn't-include-an-example-sentence is a call to clean or clarify a post, not eliminate it (and not something that I always feel is entirely necessary). I think welcoming new users would be easier if there were some clarity around the phrasing of "close"--even calling it "closed for new answers" would help, or having different "fix", "close" and "deleted" categories. I realize this isn't the question on the table, but any calls to reduce the number of close votes can be met with calls to give reviewers a sharper instrument to work with. There is a broken window syndrome that can be hard to recover from.

FWIW, stackoverflow has a similar issue I've noticed growing over time. Duplicate questions often rank higher on google search results than the original, and often contain more updated information (tech is evolving much more rapidly than language). They are closed, but not until people give fairly substantial answers. This has the negative effect of spreading the knowledge over multiple pages. Hopefully a 3-vote-close rule will reduce that.

PS--Two clarifications. The queue counts you're seeing do not include items you've already voted on, but haven't yet reached the close limit. Also, you can cast up to 20 close votes per day on any given queue.

Answer 7

The evident purpose of this proposal is to make it easier to close bad questions quickly. To the extent that it would promote that goal without producing undesirable side effects, I have no objection to it—and I can't imagine that any other EL&U participant would object to it either.

The problem is that EL&U's built-in options for voting to close a question do a very poor job of distinguishing bad questions from good ones. The primary culprit here is the "show research" close reason, which was instituted several years ago to make it easier to close bad questions quickly (sound familiar?) but which in practice makes no distinction between a cruddy question of no interest to anyone but its poster and a question of potentially widespread and enduring interest whose only supposed flaw is its failure to provide background research. Now we have another attempt to make closure easier. If adopted, it will surely increase the number of interesting questions that get kneecapped for a reason unrelated to their quality as questions.

Responses to "show research" closure vary. In some instances, the original poster tries to satisfy the "show research" requirement; in other instances, other site participants attempt to do so (and thereby incur the wrath of certain close voters who consider such efforts to be an exercise in shamelessly presumptuous audacity); in still other instances, reopen voters simply decide that the "show research" close reason shouldn't prevail against a question that they approve of; and in many cases, the question is simply abandoned to its fate. But whatever the outcome, the closure experience is likely to have a discouraging effect on the poster. The natural result of having a fundamentally interesting question closed for a reason unrelated to its inherent quality is to be less inclined to post another question on the site—a point worth considering the next time you wonder why EL&U has so much trouble attracting good questions.

EL&U has an earnest and vocal contingent of site participants whose main interest in the site is in maintaining or raising its standards for acceptable questions and answers. In pursuit of this interest, they seek to disqualify as many bad questions as possible as quickly as possible. But their desire for quick disqualification of bad questions has led them to adopt standards that don't distinguish truly bad questions from good ones. The current proposal accelerates this process without improving the qualitative discernment of the site's approved close vote reasons.

---

My interest in this site does not primarily involve keeping out the barbarian hordes. Rather, it involves trying to answer good questions. For that reason, I am more concerned about protecting good questions from closure than about closing bad questions quickly.

If (as seems likely) EL&U is poised to make question closure easier, without simultaneously replacing the "show research" close reason with a close reason (such as "too localized") that actually goes to the fundamental badness of the suspect question, I would like to see it also take a significant step to protect good questions from quick closure.

I can think of two simple ways to accomplish this. One would be to link the close voting privilege to demonstrated interest in the area(s) that a question asks about. That is, we could adopt a requirement that, to be entitled to vote to close a question, a site participant must have some baseline score—say, 50 upvotes—on at least one of the tags associated with that question.

A second option would be to give site participants who do have a substantial interest in the area that a question asks about the power to protect it from closure. That is, we could permit site participants who have accumulated some baseline score—say, 100 or 200 or 400 upvotes—on at least one of the tags associated with a closed question the power to reopen the question with a single vote.

Either of these changes would give site participants who have demonstrated an interest in a particular category of EL&U questions a greater ability to weigh in on the merits of posted questions related to that area and to preserve good questions that are at risk of being caught up in the crude dragnet of "show research" closure.

Answer 8

The requirements for close-voting have only been reduced recently, but already I have experienced expedited closure for off-topic questions, and I can say it's already an improvement. Much better user experience from a moderation perspective.

Answer 9

I for one will be very ready on shortened reopen tools.

Note that that tool page looks at all possible reopens, not just those that are on the most recent queue (notice the different UI). Also, it allows you to look at the delete queue (and undelete queue).

Use the tools that are available! (but also don't abuse the tools) Just because there is a button to press doesn't mean you should press it (and thankfully there's an 'unpress' button).

Answer 10

I'd like the close threshold to stay at five for any question that gets more than 100 views (an arbitrary threshold, pick one). My concern is that there are some subjects that just attract close votes like magnets, often as a result of a quote that has a political angle, religious angle, or deals with some other sensitive subject. There was a time when no question that mentioned the military, no matter how tangentially, stood a chance, but that seems to have relaxed a bit.

Another option would be to close after three close votes from the review cue or five total (review cue + close button). This might just encourage a few more to use the cue and ought to be fairly simple to implement.