12

This is almost a verbatim copy of a question with the same title on Home Improvement Meta. It goes to the point and I don't see anything to be improved or expanded.

This has been tinkered with on Stack Overflow and several Stack Exchange sites. SO has now gone to 3 permanently and Shog9 commented:

Other sites can have whatever threshold is most appropriate for them

We are "other sites".

There's only a handful or so regular reviewers in general (of the non-mod variety anyways), and closure tends to be a pain: it takes days to have a question closed. Lowering the threshold to 3 would ensure faster community response time.

In short, let's set the close and reopen voting thresholds at 3 instead of 5. As this is a feature request, you can use the upvote and downvote to show that you are for or against the proposal.


As read in CM Catija's answer to Is it possible to reduce the number of close votes needed for certain types of close reasons? in Japanese.SE, this change is not straight-foward:

The number has to be the same for every close reason but we can change the number required to close a question to 3 or 1. We've run a couple of tests on other sites and are willing to test it on others. If you'd like to run a test here, have a discussion about it and then have one of your moderators ping me once you've made a decision.

I have a handful wanting to test in January already and I'm not sure how many I can do at once but what I generally do is a one-month test and then we compare some stats before and after to see what impact the change had. We're being cautious about making the change permanent from the outset because there's some concerns we want to avoid.

Here's some examples of sites that have tested this out or are asking whether they should on their metas:

25
  • 3
    (+1) I completely agree with this. I would write an answer, but the answer would look exactly like the question, because the problem could not have been expressed better.
    – wimi Mod
    Dec 14, 2019 at 10:13
  • 1
    And we recently lost one of our consistent reviewers which did not improve things.
    – mdewey
    Dec 14, 2019 at 11:56
  • Would this affect votes for reopening? I have rarely seen a need to vote to reopen, but sometimes it does happen. Dec 15, 2019 at 16:22
  • 2
    @aparente001 on SO the threshold for both closing and reopening was lowered to 3. See the announcement and the experiment from which this was concluded.
    – fedorqui
    Dec 16, 2019 at 7:02
  • Zero downvotes -- are we done? Can this be finalized? The more I think about this the more in favor I am, I don't know why this never occurred to me. Dec 17, 2019 at 8:31
  • 2
    @aparente001 let's run this question until the end of the week and then I'll contact the CM team to change the threshold value of this site.
    – Charlie
    Dec 17, 2019 at 9:06
  • 1
    @aparente001 such initiatives should be handed to the CM with quite a lot of support from the community, as well as some analysis. Just 4 upvotes seems a bit too little to me, so I would wait to have both more feedback and upvotes.
    – fedorqui
    Dec 18, 2019 at 10:35
  • @fedorqui - It's hard to get a lot of votes and comments at Meta. This seems like a bit of a Catch-22. If we had more meta-type participation we wouldn't have such a hard time getting the close votes. Dec 18, 2019 at 17:35
  • I agree with Aparente001 and I have to say that my reasoning for not wanting to lower the threshold yet is that we could very well do so for Meta proposals (and it seems we are uncomfortable about it) and approve with less (up)votes. We have complained for a long time about participation, and the main excuse was that we were a beta site. Now, no longer beta, we don't address participation, we just let fewer users (the few ones who care enough to get involved, in any case) decide about the management of these matters. If that's how we want to roll, fine, but let's be aware about it.
    – Diego
    Dec 19, 2019 at 17:56
  • 1
    @fedorqui'SOstopharming' no news... I have asked again, but I do not expect great surprises...
    – wimi Mod
    Apr 7, 2021 at 6:45
  • 1
    @fedorqui'SOstopharming' here are some news: meta.stackexchange.com/q/364007/643262 . Unfortunately, Spanish.SE was not chosen as one of the 12 sites in which three-vote closure will be initially tested. We will have to wait for the results of those tests...
    – wimi Mod
    May 3, 2021 at 22:04
  • 2
    I've updated this to status-deferred. While y'all aren't included in the test I started in May, the outlook is good that I'll be able to make this happen on more sites in Q3 2021. Thanks for your patience. :)
    – Catija
    Jul 1, 2021 at 14:33
  • 1
    @Catija I would like to point out that we have already lost one long-term, high-reputation user due to the continuous delays of this request (already 2 years), and the fact that many sites have already got it but we have not is probably making it much worse. I would like to have more Meta participation on this site, but if users see that their requests/discussions are "delayed forever", it is normal that they do not participate.
    – wimi Mod
    Sep 24, 2021 at 9:59
  • 1
    @Catija so, to improve the health of this community, it would be nice we got either "yes" or "no" to requests, instead of continuous delays, and if tests on "only some sites" were quickly followed by decisions, to avoid long terms of apparent unfair treatment. I hope there is a way to go in this direction (at least partially), because I am getting increasingly worried about the low participation on this site...
    – wimi Mod
    Sep 24, 2021 at 10:03
  • 1
    Hecho. A ver si esto ayuda a agilizar el proceso de cierre...
    – wimi Mod
    Oct 7, 2021 at 7:24

4 Answers 4

4

Update

I've made the change - main site requires only 3 votes to close or reopen. I'll see if I can look at time to close as another element for this.


As of September 28th, the testing phase of the three vote close/reopen project is completed and the results have been posted. The next step is to reach out to the outstanding sites and explain the next steps and give an overview of the potential outcome of the test on those sites...

So, here I am.

Here are the two graphs from the queries I've been using for the duration of this project. Both of these graphs involve a period from 1 February 2020 until 27 September 2021.

This first one represents your close review completion percentages grouped by month:

Graph of percentage of close and reopen reviews completed between 2020-02-01 and 2021-09-27. The graph shows generally high completion percentages (between 80-100% for most of the earlier months but a decrease in percentage completed starting in January 2021 to between 60-80%.

In the grand scheme of things, these numbers are pretty good, even recent numbers that are closer to the 60% completion rate are better than many of the sites requesting this change. That said, it looks like there's been some struggle recently to keep the percentages high, so perhaps a slight boost by lowering the votes needed to close and reopen to three will be beneficial.

The second graph is the number of questions closed or reopened with moderator involvement vs user-only close and reopen. It also shows the total number of posts closed and reopened to more easily see the aggregate numbers.

Graph of number of posts closed and reopened on the site over time, by month from 2020-02-01 until 2021-09-27. In general, very few posts are closed and reopened in the course of a month - there's a peak of 15 in July 2020 but the votes to close per month since September 2020 has ranged from 4-10. While the community has been responsible for many of these closures in the past, the site has relied on moderators more in recent times.

When I look at the numbers here, I do see that the moderators have had to take over some of the closing on this site in recent months but, at the time of posting this question (which I've also checked) this site closed, on average, no more than one question every two days.

So, while I'm not saying "no" to the request here, I'd like to have some help seeing what I'm missing. I completely understand that needing to close posts can be challenging at times and in recent days there's certainly more need for this change - but I'm trying to see what's going on?

Are there many questions that should be closed that aren't being flagged for closure? If so, that's a separate problem that this won't fix.

Are there too few people participating in review to get questions closed?

This graph shows the number of people casting close votes in any given month (note the range on this graph is wider, going back to January 2019 - also, the bottom is at 6, not 0, which gives a very scary feeling to the totals).

Graph of "unique close voters" over time starting in January 2019. Between January 2019 and March 2020 there were between 12 and 19 close voters, with 12-16 on average. After that, the number of people casting close votes drops, ranging between 8-13 in any given month.

It looks like y'all have lost about a third of your voting participation and that is concerning. I'd be interested in understanding why those people stopped reviewing.

That said, in the recent leaner months, it's likely necessary for all of your voters to vote on nearly all questions nominated for closure - and that count of voters includes mods. Lowering the votes needed to close and reopen to 3 will reduce the strain on those who are still participating in closing and reopening posts but I think it'd still be good to understand the other concerns - if we don't take those into account, this may end up being a futile change since whatever drove those people away may also eventually cause the remaining reviewers to leave, too.

Barring an emphatic response, I'll change the votes needed to close and reopen posts here from 5 to 3 on Monday, 4 October 2021. If you feel like the notice period for this change is unnecessary, let me know and I'll turn it on sooner.

5
  • 2
    Thanks for this very detailed answer. One variable that I miss is the "average time to close a question", which I think is a relevant indicator: the faster we close a close or reopen loop, the sooner the OP will have relevant feedback.
    – fedorqui
    Sep 29, 2021 at 6:59
  • 2
    I agree with @fedorqui'SOstopharming'. Our close review completion is high because moderators end up doing the closing, as shown in the second graph. Moderators having to do more than half of the closing is not what this community wants, and three-vote close would improve that. The time it takes to complete the reviews is also a reason of concern: I do not have access to statistics but three of the last four close review tasks have been in the queue for 12 days, and are still there. By that time, answers have already been posted.
    – wimi Mod
    Sep 29, 2021 at 8:31
  • 1
    Thanks for the update and for applying the change on our site too, of course. I would say October 4 is fine, to allow others to voice their concerns, if any (although I do not expect much disagreement on this).
    – wimi Mod
    Sep 29, 2021 at 8:33
  • 1
    By the way, the "recent struggle" on the first graph is probably also due to the long time it takes to complete reviews: as September has not ended yet, there are still pending reviews that will probably be completed mid October. Our problem is not review completion percentage, but slow reviews that end up needing to be done by a moderator when they have been there for too long.
    – wimi Mod
    Sep 29, 2021 at 8:41
  • 1
    Thank you for the insights addressing the problem with participation. It really saddens me that the original discussion is from Dec 19 (almost two years ago) and we finally get this implemented, meaning that the root cause of the issue has been unaddressed during all this time. I completely agree that there is a separate problem that this [feature request] won't fix. To me it feels like a train crash in slow motion. Currently the site receives less than one Meta-Question a month (7 thus far in 2021) with very little participation on each, but hopefully something about this will be discussed.
    – Diego
    Oct 1, 2021 at 18:09
7

Update:

Testing of three-vote closing is starting on Thursday, May 6 on 12 SE sites, and we are not one of those sites. The test will last for 45 days, and then the results will be analyzed. After that (which will also take some time), hopefully we will get a chance.


The close vote count threshold has now been requested to be lowered to the CMs. The CM Catija has added Spanish.SE to the list of sites in which this will be tested.


Por mi parte de acuerdo con la pregunta y con los comentarios. Las colas de revisión se están eternizando, los flags de posts duplicados, offtopic o de poca calidad nunca llegan a nada, y al final como único moderador actualmente tengo que llegar yo para cerrarlos (o no) cuando eso debería ser tarea de la comunidad. Y no digo nada de los votos para reabrir, que incluso en el apogeo de actividad del sitio era misión casi imposible. Bajar el límite a 3 aceleraría las cosas y me daría menos trabajo, que esto de ser moderador único no es fácil (tampoco imposible porque esta es una comunidad tranquila, pero sus cosas tiene).

Por favor, nótese que estoy dando mi punto de vista, no quiero llegar yo a imponer nada.

5
  • Para que conste en acta (y evitar futuras quejas), esta respuesta ha recibido 5 votos, que es el umbral que en el pasado nos hemos marcado para aceptar propuestas de meta (debido a que por la escasa participación a veces no se sabe si la gente no apoya una propuesta o no ha tenido tiempo para verla/considerarla). Tras el edit, alguien ha debido de retirar su voto (ya que no puedes retractar tu voto de un post, salvo que el post reciba edits). Esta no es solo la respuesta más votada, sino que también cumplió las condiciones acordadas por la comunidad para considerarla un cambio consensuado.
    – Diego
    Jan 10, 2020 at 16:42
  • @Diego creo que en el caso de las feature-request lo relevante es la cantidad de votos que tenga la pregunta en sí, pues es la que plantea el cambio. Cambiaría si las respuestas perfilaran la propuesta o definieran nuevos elementos, pero no creo que sea el caso aquí. Por ello +9/-0 creo que es un éxito para nuestro sitio, creo que pocas propuestas (si no ninguna) habían obtenido tanto apoyo. Dicho lo cual, evidentemente también es relevante el apoyo a esta respuesta, tanto porque es apoyo del equipo de moderación como porque da más empaque a la propuesta
    – fedorqui
    Jan 10, 2020 at 19:50
  • @fedorqui'SOstopharming' gracias por ese update. Por alguna razón yo pensaba que ya estaba hecho e incluso iba a preguntar si ha mejorado ya un poco la cosa en las colas de revisión...
    – Diego
    Feb 11, 2020 at 18:59
  • @Diego le pregunté a Charlie y me explicó eso en la Tertulia, a ver cuándo los CM le contestan indicando el pistoletazo de salida de todo el experimento :)
    – fedorqui
    Feb 12, 2020 at 6:18
  • 1
    Acabo de poner el quinto voto de cierre para pregunta basica de redes routers y modems y (como en otras ocasiones) me sorprende que el límite no sea todavía 3 votos. No es que esté sorprendido ante la lentitud por parte de SE en atender la petición y realizar el cambio (o denegar la petición), pero después de invertir tanto tiempo y esfuerzo (por parte de nuestra comunidad) en animar a la participación en Meta, el que haya esta aparente falta de continuidad entre la propuesta de una idea o solución a un problema y su realización es frustrante.
    – Diego
    May 5, 2020 at 12:10
1

Let me make a further suggestion.

Since many of us who post on this stack are scientists of one variety or another we should really do an experiment before making a permanent change. I propose that we ask for something like the experiment which ran on StackOverflow with similar metrics and then look and see what happened. I will repeat the link here https://meta.stackoverflow.com/questions/391832/we-re-lowering-the-close-reopen-vote-threshold-from-5-to-3-for-good although it is also in the original question which prompted this thread.

3
  • Agreed. In fact today I updated the question with the reasoning that the CM Catija provided in another Meta: apparently they run this experiment before lowering the threshold. Which makes sense and puts some numbers before a final decision is made, which is something we indeed need :D
    – fedorqui
    Dec 20, 2019 at 15:06
  • I appreciate your sharing that link. I took a look and here's what I think. There is an answer there that says, "For politics, skeptics, and similar sites often dealing with subjective topics this will probably have a bad effect.. even with the 5-vote requirement there are too many ideologically motivated closings, when people want to remove legitimate questions because they think they (or the answers given) don't support their views... and therefore there are lots of generic "not clear what you're asking" close votes even if the OP explained exceptionally when what is being asked." (cont.) Dec 21, 2019 at 17:14
  • (link is meta.stackoverflow.com/a/391969/445686.) Well, our site doesn't fall into that class. So, I don't think an experiment is needed. Dec 21, 2019 at 17:15
-1

Once more, I'm going to play Devil's advocate and disagree with the most popular idea and propose a different option.

I do agree with the assessment and feeling presented in the question. On the other hand, lowering the threshold makes me feel like we go with the easiest option (nothing wrong with that) and we just give up on our efforts of involving more community users in the moderation duties (I should say, in the "extended participation activities", since we have tried unsuccessfully to get more participation in Meta, as well as in the review queues).

While lowering the number of votes could be a potential solution, would we reverse this if the activity in the review queues increases? If the review pace gets even worse, will we lower the threshold again?

My proposal is this:

Let the moderation team participate in Close and Reopen queues.

(Meaning, let's make the mods not wait until it's blatantly obvious that a review queue/contribution became stagnated. Let's take advantage of the fact that the mods are wise, trustworthy and active contributors-reviewers. That's actually why we chose them and why we trusted them with mod superpowers and mod superresponsibilities.)

A long time ago we decided that moderators should stay away from these queues due to two main reasons:

  • We wanted moderators to use their binding vote as little as possible.
  • We wanted to encourage participation in the review queues.

Since not having enough people for reviews has become an issue (I see that close and reopen are specifically mentioned in the question, but not the other review queues) and the current moderation team seems to be formed by top contributors that have a pretty good understanding of what is on-topic and what is not, let's let them participate in these reviews again. Let's trust their binding vote.

I think that this is something that at least we could try for a couple months, see how it works, and if it doesn't or we don't like it we can go with the more upvoted option of lowering the threshold.

17
  • Just when I was asking JNat to help me lower the threshold you come with another answer and I had to retreat my words... :-D
    – Charlie
    Dec 19, 2019 at 14:41
  • If we look at the principal contributors to the Close queue we see that there are five who have voted 100+ times one of whom has withdrawn from the site. Realistically if we keep the threshold at five then @Charlie is almost certainly going to have to intervene for anything to get closed. Reducing the threshold to three means he may find we have done the work for him. Since we are currently short of diamond moderators I think we should try to limit what he has to do.
    – mdewey
    Dec 19, 2019 at 17:17
  • @mdewey Why? Charlie has a binding vote as a mod, which we didn't want him to exercise, so he has stayed away from the queues. His workload would lessen, since he wouldn't need to track stagnant revisions, and he'll be able to intervene as as he does mod duties. We lower the threshold to 3 but then we have a top contributor that we don't let intervene?
    – Diego
    Dec 19, 2019 at 17:46
  • I have also to ask: what is the main goal here? Reduce the mod workload? Close questions sooner?
    – Diego
    Dec 19, 2019 at 18:05
  • @Charlie - Do you have a preference? I've always been concerned about the moderator(s) having to shoulder too much of the load. // Diego, this is an interesting point of view. Hmm. I guess that if we abandon the practice you and fedorqui had, of avoiding a hammer-close (I hope I didn't mis-use that term) when possible, in other words, if we go with your proposal here, then questions might routinely get closed with one or two opinions; whereas if we go with fedorqui's proposal, then we would typically be closing with three opinions. I guess I prefer three, over one or two. Dec 19, 2019 at 18:19
  • @aparente001 for me it's difficult to say because some recent posts have been closed with 5 votes. So if a post gets 2 or 3 close votes, does that mean that the community is not participating or that the community has divided opinions? Should I then hammer close the question or not? It would be my decision in the end.
    – Charlie
    Dec 19, 2019 at 18:26
  • @Charlie - How would the following be: for now, we lower it to 3. I am assuming that in January or February there will be an opportunity for the moderating team to be expanded. So, if and when the team is expanded, this could be revisited. // Another thought: maybe the threshold to give people hammers could be petitioned to be lowered. But I doubt this would happen overnight. // My main concern is that I don't want you to burn out while you're waiting for more moderators. You could Dec 19, 2019 at 18:33
  • ... burn out by covering the site 24/7 on your own, or you could burn out by having to decide whether to use your hammer, etc. I confess I'm not clear whether we have some high-rep users, or active ex-moderators with a hammer-close ability currently Dec 19, 2019 at 18:35
  • 1
    @aparente001 users can get hammer powers if they get a gold medal in any tag, then they can directly close or reopen questions with that tag. But no gold medal has been granted in this site yet.
    – Charlie
    Dec 19, 2019 at 18:48
  • 1
    Thank you both for this interesting conversation. This is what I aimed to with my answer (sorry that comments are a poor format for this exchange). To complicate matters further, questions could be closed/reponed with 2 votes + Charlie's. Since his vote would be the last vote needed, that would "remove" the binding effect of the vote. How do we feel about that? I will go with whatever the community decides and I will also commit to visit review queues more often, but I can't help feeling that we are treating a symptom (lack of participation), instead of the root cause of the problem.
    – Diego
    Dec 19, 2019 at 20:21
  • @Charlie note the hammer power is just to close as duplicate. For the rest of close reasons, the gold tag holders do not have a binding vote.
    – fedorqui
    Dec 20, 2019 at 6:57
  • 2
    @Diego I think it is a key point what you say: what is the main goal here? Reduce the mod workload? Close questions sooner? To me, the goal is to close loops faster, so the feedback to users is given earlier. If a new user needs two days to know how to improve their question, they will probably gone. If we reduce it to few hours, we still have time. Also, closing fast prevents from getting bad answers.
    – fedorqui
    Dec 20, 2019 at 11:32
  • @fedorqui I fully agree with that, but maybe new users are just going to get sooner a message saying "your question doesn't meet our standards" and no further help about how to improve it. As you know, in the Lista de comentarios útiles para el sitio we have some suggestions to remind users to try to be kind and helpful instead of snarky or indifferent to newcommers. I think we can/should give feedback to those users via comment while their question is in the queue. I don't think they get much help or guidance once their posts get a flag.
    – Diego
    Dec 20, 2019 at 15:59
  • My two cents: I think we should do the three things that have been mentioned in parallel: close more quickly (I'm on record being impatient about this!), give people gentle but helpful feedback (I do that a lot, and I try to tailor my feedback to their level of savviness about the workings of the site), and work to increase participation. Actually, I think that a collaborative process, working with new participants to help them get the help they seek, is an important part of what can be done to increase participation. (In general, I'd like to see more dupe closures.) Dec 21, 2019 at 17:19
  • @Charlie - Is this proposal still in limbo? Could we have a status report when you have a moment? Dec 30, 2019 at 3:38

You must log in to answer this question.

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