Moderator votes are "binding" - meaning that, while it takes 5 close votes to close a question, a single close vote from moderator will do it no matter how many other people voted before.

I would propose that moderators try (as much as possible) and refrain from performing actions that are binding - especially voting to close a question - unless it's one of 3 exceptional cases:

  • There's a VERY clear community consensus that the question ought to be closed (3+ existing close votes, many flags, negative comments, negative consensus on meta or chat).

  • The question is so 100% egregious that it must be closed immediately lest it cause severe immediate damage to the site (I'm having a hard time coming up with such emergency example, but let's say someone posts a clearly off-topic AND extremely offensive question - something you don't want to show up in Google search). Even in this case, I would posit that asking on META or chat first would be ideal, and waiting a reasonable (5 mins) time for people in community to pipe up.

  • The item was extensively publically discussed by the community (I'd propose the threshold of at least 4-6 people, for at least 1-3 hours; not sure of best limits) on Meta or chat; and after that discussion there's a clear disagreement with nowhere near strong majority, never mind consensus. As far as I understand, this case is specifically where moderators were meant to interfere - ideally, when the community agrees to bring it to a moderator decision. However, I strongly feel that usually the moderator action happens too soon before there's a consensus.

Please note that the above does NOT cover 2 cases where I strongly believe a moderator vote is a bad idea:

  • Moderator votes to close because as private community member, he feels the question is worth closing. I know one of our mods recently voiced this sentiment. It's a bit unfair to mods as people, but if you're the first/second guy to vote to close, please wait till community opinion is more clearly expressed and/or some discussion takes place. With great power comes great responsibility, and all :)

    Please note "well if the community wants it, they can vote to re-open" is an extremely annoying and very wrong cop-out. First, it's hard to find 5 people willing to bother to look into the issue, analyze, agree, AND vote to open, all in reasonable amount of time. Second, a lot of people would not vote to re-open seeing a moderator name on the closing even if they would have otherwise (I've seen that happen).

  • The question MAY be bad, and is controversial, and there's an active discussion.

    What would be a good action to take would be to lock/protect the question to avoid the discussion taking place on main site (if one is happening, or undesirable answers appear), and forcefully suggest to people to "take it outside" - e.g. chat or Meta.

    Now, if whoever wants to keep the question refuses to do that, this is a proper ground to close the question.

DISCLAIMER1: This post is NOT a reflection of any specific problem with current SciFi.SE moderators, though at times they have exhibited the actions that I will suggest above as "should be avoided". But the opinions expressed here have gelled back when I was a newbie on StackOverflow and had absolutely zero interaction of my own with moderator practices - merely reading moderation rules and observing mods in action, and are not a reflection on any specific moderator - rather, on overall rules.

DISCLAIMER2: The timing of this post was affected by 3 things (none of which was a specific action by a moderator):

  • Gilles' post in chat yesterday indicating that moderators should not set policy and only get involved when the community can't come to a consensus (did I get that right?)

  • the upcoming moderator elections. I want to have the policy discussed and set BEFORE we have new moderators (and can have input from current ones); AND allow voters who care about this issue to be able to gauge moderator candidate's views on the topic.

  • Just came to the front of my "questions to ask on meta" buffer queue :)

  • 3
    "It's hard to find 5 people willing to bother to look into the issue, analyze, agree, AND vote to open, all in reasonable amount of time" - this seems to be like a problem with attention. If a question gets closed and a select few people want it re-opened, a question should be asked on meta asking why it's off-topic, and discussing it's merits and relation to the scope of the site. If it remains closed after that, that seems to me like a consensus on the part of the community. Commented Jan 24, 2012 at 15:55
  • @IanPugsley - It's a problem with not-high-volume sites/tags. We had a MAJOR problem with that in [Perl] tag on SO - just not enough user flow to reliably get an obviously popular-to-reopen question to be reopened. Commented Jan 24, 2012 at 15:58
  • @IanPugsley - case in point: skeptics.stackexchange.com/questions/7668/… Commented Jan 24, 2012 at 16:11
  • There seems a fairly clear consensus that the mod was wrong to close, and so far only 2 reopen votes including mine. Commented Jan 24, 2012 at 16:12
  • 2
    @DVK 4 now. And its worth noting that the question was only asked 11 hours ago, and closed 8 hours ago. That's not a terribly slow response, particularly for a beta site.
    – Beofett
    Commented Jan 24, 2012 at 16:33
  • 1
    @DVK After a quick search, I see no relevant meta question about the closing of that question, and I don't really think it's realistic to expect it would be reopened without one. Commented Jan 24, 2012 at 16:34
  • @Beofett - my problem is PREISELY that mod action shouldn't be faster than equivalent community one. If it takes 8 hours to 5-vote close or open, a mod should unilaterally decide to close in 1-3 Commented Jan 24, 2012 at 16:41
  • 3
    If you have a problem with Skeptics mods, why don't you ask on their meta? Commented Jan 24, 2012 at 17:04
  • @EightDaysofMalaise - I don't have a problem with skeptics mods per se. I have a problem with giving ANY mods this power without checking it properly, which IMHO isn't done and is being abused - frequently unwittingly as they aren't aware of the issues - by ALL mods Commented Jan 24, 2012 at 17:56
  • 4
    @DVK: If you think that some moderator is abusing their power you can raise the issue with the team. Moderators have been removed in the past. That said various team members have voiced support for prompt action on the part of moderators in the past. Further, I can assure you that many (most?) mods spend a fair amount of time worrying about the correct use of their powers. We (I'm a mod on Physics.SE) have some private channels of communication and this is a common topic of conversation. Commented Jan 24, 2012 at 21:16
  • @dmckee - I think lots of it is unwitting. Think Superman baby accidentally pushing someone. The mod doesn't realize that (at least based on my observations) getting a Q to be reopened by full 5 votes once couple of hours had passed is nearly impossible in some cases for a variety of reasons, not the least of which is many people browse "newest first" order, myself included Commented Jan 24, 2012 at 21:49
  • 1
    @DVK If you have specific instances of mod power abuse, you should post them in a new meta question.
    – user1027
    Commented Jan 25, 2012 at 4:00
  • 1
    @Keen - As a new user, I (and many others) would hesitate to challenge mods. As a non-new user, it's LARGELY not worth my time (I successfully did that on here since whatever abuse happens is rare and not intended with malice - simply lack of sufficient contempaltion and the mods are open to being convinced; but I simply largely withdrew from Skeptics - in large part because of deep disagreement with moderator approach there) Commented Jan 25, 2012 at 14:36
  • @DVK This isn't Skeptics, so their mods aren't relevant here. If it's rare, then why do the mods need to change their behavior? We high rep people have the ability to reverse these rare occurrences.
    – user1027
    Commented Jan 25, 2012 at 15:24
  • @Keen - because the problem exists whether it's rare or not, and can be easily fixed by mods acting slower. Let the community to the work of policing where they can Commented Jan 25, 2012 at 15:34

5 Answers 5


I disagree with your proposal. The role of moderators is not to set policy (except very occasionally, when the community can't decide when left to itself). The role of moderators is to enforce policy.

If the site rules clearly say that a question is unwelcome, it's perfectly normal for a moderator to close it. If the moderator happens to be the first user seeing the question, so be it. Moderator closures aren't reserved for egregious cases; they're normal for clear cases. It's only in borderline cases which could go either way, cases where there isn't a clear policy, that it's better for a moderator not to intervene straight away. (If I see a borderline case, I generally leave a tab open and come back after a few hours, often also leave a comment, perhaps ask in chat.)

Moderator actions can be appealed, especially question closures (answer deletions are admittedly harder to appeal since fewer people get to see the deleted answer, but then answer deletions are rarer and only for egregious cases). Moderator actions can be challenged through comments, chat (sometimes), and meta. Reopen votes aren't a cop-out, they do work as intended; it's rare for a question to get enough attention for 5 reopen votes, but moderators usually will reopen a question if a meta discussion shows a dominant community wish to reopen, or if the question has been improved through editing.

Stack Overflow tends to work differently simply because of scale. SO gets about 1000 new posts per moderator per day, SFF a little more than 10. So a vast majority of posts on SO are never even seen by any moderator.

  • 1
    "they're normal for clear cases" - that's the problem I'm having. I have seen WAY too many of the closures (if you want a specific example, the most recent one that involved me was the "identify wizard of oz" one, but many more NOT involving me) where it was anything but clear, with plenty of people agreeing the question should be open or at least has a lot of merit and being on the fense Commented Jan 25, 2012 at 0:48
  • 1
    In other words, I'm not disputing that the moderators should close in clear cases. I'm disputing the assumption that moderators unerringly know which ARE clear cases without letting at least SOME community input to accumulate - and I don't mean 1 close vote and 2 flags and zero discussion allowing the poster to explain themselves and try to convince people where the criticism is not necessarily correct) Commented Jan 25, 2012 at 0:51
  • Heck, I can probably dredge up at least 3 closed questions of my own where - after much effort - I actually managed to convince the mod that the question wasn't worth closing without ANY participation from anyone else at all. And sorry, but those 3 cases can't be used as "well, see, it's OK to let mods close" - that's like saying "it's OK to let people be jailed without due process as long as they can appeal" Commented Jan 25, 2012 at 0:53
  • 1
    "it's rare for a question to get enough attention for 5 reopen votes" - you just made my strongest point for me. Yes, I know that theoretically process can work. Practically, it is iffy to be successful on low trafic site - AND, the questions DO get majorly damaged by those closures. Many people don't look at older questions. Almost nobody removes the downvotes they gave because of moderator's opinions. Commented Jan 25, 2012 at 0:57
  • 2
    @DVK I dispute your assertion that the closure was opposed by the community in the Oz review quote question. I forget how the chat discussion went, but you were the only commenter defending the question in its original form. Closing a question and reopening it after it's been editing is perfectly normal. Be careful with that comparison with legal due process, you're moving really fast towards Godwin's law here.
    – user56
    Commented Jan 25, 2012 at 0:58
  • 1
    @DVK I don't think anyone downvotes because of moderators' opinions (and for any that do, I'm sure an equal number of others vote the other way out of contrariness). Please read what I wrote: in practice moderators really do step in to reopen when the situation calls for it.
    – user56
    Commented Jan 25, 2012 at 1:00
  • I can't give your the graph for Oz question, but to the best of my recollection it went this way: 3-4 upvotes, zero downvotes. Then 1 downvote. Then a mod posted that "this is trolling" and closed - nearly instant 4 downvotes (1/2 since removed). I can't for a second believe that it was a random coincidence. People LIKED the question on its merits, and started hating it when the mod openly stated with his authority that it was trolling. BTW, now it has 8 total upvotes, so people CLEALY saw merit in it once the brouhaha ended and disparaging comment was removed allowing Q to stand on merit Commented Jan 25, 2012 at 1:03
  • 2
    @DVK The timeline shows +8/-3 on the first day, it doesn't distinguish pre/post closure or pre/post edit. Note that upvotes and close votes don't have a strong correlation, bad questions often get highly upvoted.
    – user56
    Commented Jan 25, 2012 at 1:05
  • As far as the only one defending - I recall at least 1 more person but need to peruse the chat. Commented Jan 25, 2012 at 1:07
  • I agree with the latter, I was discussing whether mod opinion can sway downvoters. And given the time pattern, I'm convinced it did. I saw pile-on downvotes on SO before - right after a mod closed the question (whereas before that it was 1 DW). Don't recall seeing that on Skeptics. Commented Jan 25, 2012 at 1:08
  • LOL!!! Speaking of people NOT defending the question, here's a chat quote from some shady character I haven't seen before: ["Gilles @MarkRogers] I don't think we should be judging questions on the asker's intent. We should be examining whether the question (or rather, the answers that it's likely to elicit) are useful to future visitors and searchers And on this basis, it is a useful question" Commented Jan 25, 2012 at 1:12
  • 1
    I promise I won't call the mods any Godwiny names :))) Commented Jan 25, 2012 at 2:26

Flags will bring a moderator to the scene and if they agree with the raise, will act on it as suggested. But if you're not a moderator, you won't see that and might think they're trigger happy.

That's the same as if they came across it on their own. If they spot something and are in tune with the community, they don't have to always sit on their hands waiting for someone else to vote to close. If it's there and they see it, why shouldn't they act on it?

Moderators should understand the site and community. That includes being able to know when to close or re-open questions they read.

If you feel like moderators should wait for everyone else to do the work, to vote or act and only turn up once the rest of the community has had their say, you don't trust that moderator to know what they're doing or what the site is about.

  • 1
    Damn right I don't trust the moderator more than 4-5 (or even 2) people. They are human. Commented Jan 24, 2012 at 16:50
  • 6
    @DVK Then I hope you're voting in the upcoming elections.
    – user1027
    Commented Jan 24, 2012 at 17:03
  • @Keen - I'm writing in Santa as a protest vote! :) Commented Jan 25, 2012 at 14:39
  • @DVK A Timelord as mod?! Madness, I say!
    – user1027
    Commented Jan 25, 2012 at 20:23
  • @Keen - Fine, I'm writing myself. I am not a Timelord. Commented Jan 25, 2012 at 20:28

I think that this is a reasonable philosophy, and I agree in general with the sentiment (these are very similar to the personal guidelines I use on parenting.se, although the community is much less active there so I am a bit more aggressive in using my close vote there than I otherwise would be).

However, I do not think that this needs to be formalized as a policy.

The idea behind electing moderators is that they are individuals that the community trusts to do the best job possible. Setting up rules for when moderators can and cannot close questions implies that moderators should not be able to use personal discretion when acting in that role, which is contrary to the trust we, as a community, have expressed by electing them.

In actuality, there are checks and balances in place that protect against perceived "abuse" of moderator powers, or even simple disagreement between moderators and the community.

Having multiple moderators is one of the primary examples of this. Our moderators should be working together, and discussing controversial issues between themselves (as well as with the community at large). Ideally, they should come to an agreement (or at least a compromise) on any issue of moderation.

Also, as was mentioned, members of the community can vote to reopen. Based upon my recent experiences, I disagree with your assessment of the "vote to reopen" system. I do believe it works.

By way of example, I recently had a question that was closed as general reference by a moderator (although the moderator was not the only one who voted to close). However, I disagreed with the reasons for the closure, and we (yourself included) discussed why it should/should not be closed, citing various meta discussions on the subject. In the end, the community voted to reopen. I have absolutely no problem with the way the situation was handled, both by the moderator, and the members of the community who voted to close or open. I would also have been perfectly fine if the question had remained closed due to lack of community support for reopening it. However, I do not feel it was in any way a problem that a moderator closed the question even though it was clear (in hindsight, at least), that there was not a clear community consensus that the question should be closed.

In short, I don't see any reason to really have these rules (or guidelines) for moderator activity defined here. The moderators will do what they think best for the site, and we have to trust them to use appropriate discretion. However, if they make mistakes, the community has the power to fix them (within reasonable limits). Rather than trying to avoid this ever happening, I think that the situations where there is a disagreement between a moderator decision and community consensus is healthy and beneficial, as it helps to further define and clarify the boundaries.

  • 4
    I was going to basically post this, so thanks. I agree with the sentiment of this meta question, but having an official policy or "rules" for the moderators of when they can and can't cast close votes seems like a bad idea.
    – thedaian
    Commented Jan 24, 2012 at 16:44
  • @thedaian - it's more like guidelines than official policy. Meaning, when one disagrees with moderator decision that was too trigger happy, it can be pointed to when requesting to reverse that decision Commented Jan 24, 2012 at 16:51
  • 1
    Good answer. I guess the reason you and I disagree for the need of this is precisely because I so severely disagree with the magnitute of power discrepancies. Any question that was closed by a mod with subsequent 2-4 reopen votes to me is a testament of an overall system failure, since I estimate from my fairly long experience that such question would NOT have stayed (or even got) closed without moderator action. Also, as I replied to the other answer, I may trust the moderators to have the best interests at heart, but being human, I don't trust them to be flawless. Commented Jan 24, 2012 at 17:55

Yes, moderators should be cautious with their binding votes. But that doesn't mean they should wait for a "community consensus" on each question before they act on it. Or at least, not an explicit one; the "community consensus" has already been reached, it's the topics and scopes that are on and off-topic. If a post is clearly out of bounds, meaning on a topic that the community has decided is off-topic, or of a scope the community has already decided is too narrow, the moderator should close the question, or migrate it if another site is willing to take it. If the community disagrees with that decision, we need to go to meta and discuss how we should alter the community consensus on acceptable topics. When it is ambiguous whether a question is off-topic, however, I think a mod should wait and see what the community makes of that edge case. For the record, I am strongly in favor of giving people the benefit of the doubt, so for any post not clearly off-topic (or otherwise in need of closure), I will wait to see what the community thinks.

  • But my problem is precisely the fact that mods make mistakes in whether a specific Q is within the accepted rules and bounds. They don't make them any more than average user (and likely less), but being human, they DO make them. Commented Jan 25, 2012 at 14:40
  • 1
    @DVK The community can make mistakes / disagree on a post too. If one person votes to close a post, often others will vote to close it either without considering it thoroughly, or if they were considering it but not quite there. A community decision can be and has been overruled just like a mod decision.
    – Kevin
    Commented Jan 25, 2012 at 15:25
  • 1
    And recall that mods are notified when a meta post is made. If anyone cares enough about a question being closed, they can bring their concerns to the mod's attention immediately, and let other people weigh in on it there too.
    – Kevin
    Commented Jan 25, 2012 at 15:27
  • correct. But this process has major costs - see my replies to Gilles's answer. Commented Jan 25, 2012 at 15:35
  • 1
    @DVK I don't see what's a "major cost" - having to convince a mod? If a question is worth reopening, it's worth taking at least a bit of effort to do so, at least ask in chat, there are frequently at least the five 3k users you need in there.
    – Kevin
    Commented Jan 25, 2012 at 15:54
  • this changed - chat used to be a lot emptier. And my point is that preventing mod mistakes shouldn't be upon the user, it should be upon the rules that mods follow. Commented Jan 25, 2012 at 16:00

My philosophy of moderation is pretty much as follows.

  1. If I see a flagged message, and it is specific to a moderator-only thing, I'll make a judgement call. This could be move a question, convert an answer to a comment, etc.
  2. If the question/answer is terrible, I'll close it.
  3. If the question is bad, but not terrible, I'll wait a while to see if the community wants to do something with it.

I do get impatient with people who flag to close a question, and have reputation, but don't vote to close the question...

  • 1
    A “doesn't belong here” flag is now converted into a close vote if the caster could vote to close.
    – user56
    Commented Jan 25, 2012 at 0:40
  • Didn't know that. Good to know. Commented Jan 25, 2012 at 0:41
  • @Gilles - Thanks! Very good feature! Did i miss that on the blog or was that a stealth change? Commented Jan 25, 2012 at 14:42
  • @DVK Only some major changes are announced on the blog. It's mentioned in Recent feature changes to Stack Exchange dated 2011-03-23.
    – user56
    Commented Jan 25, 2012 at 20:04
  • If a question is terrible, shouldn't you downvote it (or edit it)? Closing is for off-topic, subjective, general reference, etc, not for terrible questions, right?
    – Tony Meyer
    Commented Jan 25, 2012 at 21:14
  • Well, by terrible I mean obviously ready to close. If a question is very poorly asked, or not specific, it's worth closing. Commented Jan 26, 2012 at 0:23

You must log in to answer this question.

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