13

Please read fully before blindly closing/downvoting/tagging/duplicating

The Problem

On SO, binding moderator votes work well. There is usually not a lot of grumbling, and those that do get a good answer. SO is very well defined and has a set purpose, goal, and vision. Therefor mods can confidently vote to close. Its a very good system.

However on beta sites, things can get a little iffy. Sites can be not well defined, controversial, and have a conflicting goal and purpose. For the benefit of discussion I'm going to focus on Programmers.SE

Programmers.SE has issues with being defined. Jeff says what the rules are, the community likes and dislikes. People on Meta.SO try to come up with a list, the community likes and dislikes. And if you look, the argument is pretty even on both sides. But that debate isn't the focus here. All you need to know is that the site is pretty controversial here.

Which leads me to this: On a controversial site where the definition isn't what you would call exact and is hotly debated, you have stuff like this:

alt text

No real explanation, just a link to a generic blog post about what might get a question closed. Which one does it violate? Why would it be closed? We don't know. For all we know bigown could of coughed and clicked it.

And you have the Programmers.SO classic:

alt text

To this day, we have no idea why it was closed. What was the only lesson learned? Jeff doesn't like religion. How did the community benefit from it? It didn't


So whats good? I consider this to be an extremely good close.

alt text

A comment and close vote. And there's even TomWij to back him up. I'm perfectly fine with the mod vote here.


Here's my point: On a site where its definition is still not clear, binding moderator votes become a huge issue. On SO everybody can (usually) agree on the site definition, and can vote to close confidently. However on a Beta.SE site, close votes come from how the closer interprets the rules. There is (not much) guidelines on whats off topic and whats on topic.

The Solutions

We have many ways this can be solved, all that would be specific to moderators on Beta.SE sites. These are only suggestions.

  • When there is only 1 or 2 close votes, force a comment - If a mod comes along and closes a question with no feedback, then everybody is left scratching their head wondering what the heck caused it to get closed, as I've said above. The comment clears this confusion. But once you get 3 or more people closing, then its more of a community decision and can convey to the OP that the community doesn't want this.
  • Limit the effectivness of a moderator vote - Yes this is unpopular among mods, but thats simply because they are loosing power. This comes from everybody interpreting the rules differently. A mod does the same thing, except has a more power. If they come along and think "This should be closed" when the community says it should stay open, then people get angry.

    So what if a mods effective vote could be limited. You could give mods the ability to cast a non-binding vote which sorta fixes the problem, or you could actually fix the problem and limit the number of votes a mod vote carries. Now they are mods so they opinion does matter, so how about 1 mod vote = 2-3 votes? Simple, easy, and completely solves the problem here.

The Effects

  • The community can learn - What does a single mod vote with no comment teach? Nothing! How do we know not to repeat the same mistake again? How do we get closer to defining whats off topic? How do we know whats good and whats bad? We don't. But if a closer commented, then we could answer all those questions and become a better community as a whole
  • Less resentment of mods - Binding mod votes are unpopular and can lead a moderator to be hated. Comments however explain whats going on and gives the OP and the community a change to debate about the close. Things could actually be discussed instead of abandoned.
  • Debates aren't dragged to other question comments or meta - We don't need a meta question for every mod closed question, yet there are many of them. Why can't that say in comments? Things should only be moved to meta when a lock/delete is in effect or the comments have grown so large or biased that it gets out of hand. And then you have some commenters who will try and say "Well this was closed, so why not this one?" trying to assume what the mod meant.
  • Many more

But what about reopen votes?

That argument works well for SO, but not on beta sites. There aren't that many people who have the power to vote, let alone vote at all. The community is small vs the masses on SO. Many people aren't going to spend the time doing maintenance outside of mods.

And besides, getting questions reopened is always a very hard thing to accomplish. Once a question is closed nobody pays attention to it, even the people that can reopen. I know I don't; I just watch what comes through the mod tools.

Relying solely on 5 reopen votes from a small community is asking a lot.


Thoughts?

1
  • Just to make this clear: I have no issues with Jeff or bigown, they just happened to of closed the most recent questions at the time of writing.
    – TheLQ
    Commented Oct 1, 2010 at 22:25

3 Answers 3

15

I'm fine with giving moderators the option to cast non-binding votes and I think that on betas, many of them would choose to do so on borderline questions.

But you seem to be proposing to essentially castrate the moderators, making them jump through a bunch of unnecessary hoops just to do their jobs.

Moderators aren't moderating because it's fun. They already have to constantly take flak from the community. They're moderating because they care about the site and the quality of questions and answers on it.

When your audience is programmers, negative actions are almost always going to be followed by hostile arguments. It's the nature of the beast, dealing with people who are not only naturally a little anti-social, but free to say and do as they please under the protective cloak of internet anonymity. But it's even worse on Programmers.SE, because the site comprises in large part the slackers and rabble-rousers from Stack Overflow (which was literally the site's raison d'être). Many of them are already in "argumentative" mode from the second they visit the site, and deeply feel that the site should be a free-for-all.

On the Cooking site, we almost always leave comments on closings. But that's a choice. It's done partly because the community there is fairly open-minded and polite, and several users are new to the entire SE system and literally don't know how things are supposed to work. That is simply not the case on Programmers.SE. If I were a moderator there, and was subsequently told that I had to justify every single action no matter how obvious or trivial and thus open myself up to a thousand pointless comment debates, I would probably quit.

Moderators are supposed to be the final arbiters on their respective sites; what they say goes, and if you have a problem with it, it's your responsibility to bring it up (i.e. on meta). This is the probably the one singular aspect of SE communities that actually works like a true democracy; moderators are elected officials, they don't need your consent to act and if they want your comments then they will ask for them. Without that, the system is paralyzed by bureaucracy and breaks down in a matter of days (maybe weeks).

As for the second option - making all moderator votes non-binding - what would be the point of having moderators at all, then? You might as well just add another reputation level that lets people vote on things like locking or suspensions. Of course it should be fairly obvious why that can't work in practice. Simply making moderator votes worth "a little more" than a normal vote isn't fixing any problem, it's actually creating a much bigger problem, which is that you fundamentally have a site without any moderation.

Moderators have to be allowed to do their jobs without interference from the community. That's the reason why we have moderators. If you have a problem with a specific moderator then either open a discussion on meta or send an e-mail to the team. That's how it's always been and how it should continue to be.

3
  • I don't have issues with a particular moderator (bigown just happened to be in the most recent closes) or moderators in general. And I don't have issues with most of the close votes that they do. However, there is a number of questions where the close was unclear, where we thought it met every guideline, but a mod just closed. You have to understand that a single comment by a closing mod would 80% of the time be agreed with. 10% would have the problem solved with 1 more comment. The rest are the controversial topics that need discussion. However, closed questions are a small percentage totally
    – TheLQ
    Commented Oct 1, 2010 at 22:06
  • @TheLQ: That's exactly my point. If you don't have issues with a particular moderator then you don't have a valid complaint. Moderators are all individuals and are all free to comment as much or as little as they please. If you have a problem with the amount of commenting a moderator is (or is not) doing, take it up with them, on meta, or with the team.
    – Aarobot
    Commented Oct 1, 2010 at 22:09
  • 2
    The issue is that the moderators as a whole don't comment much, and refuse to on the most controversial topics. If I had a problem with bigown then sure I would deal with it privately. But since its the moderators as a whole and its a controversial subject, then a meta topic is required.
    – TheLQ
    Commented Oct 1, 2010 at 22:25
8

Maybe it's just me, but the real problem beneath here seems less about binding votes and more about unclear closings. On Gaming, we get far more complaints about 5-person close votes than about the unilateral closing from the diamonds. Actually, the only real diamond incidents that I can think of were both for re-opening questions.

A recurring comment on our Meta is that it doesn't help when people aren't really clarified as to why people voted to close. And when the author requests a reason, they don't always get a response until they bring it up on Meta. Whether it's done by an individual or by 5 individuals, someone finds it confusing and opposes the decision which is often cited as not making sense.

Enough of that, let's address your proposed solutions.

Requiring comments of only diamonds solves a small portion of the problem, but it also has many of the same flaws that requiring comments of downvotes does. Sometimes someone else provides a perfectly good reason already (from experience on SO, this happens a lot with migration). Sometimes the reason is very clear from the close reason. A close with 5 people ending with a diamond is still just as capable of not making sense without a comment. And naturally, if a diamond simply isn't involved, there is still an issue. This basically adds a lot of excess paperwork to otherwise simple janitorial work in order to fix a small fraction of scenarios.

As far as removing the binding vote, the job of the moderator is in the upkeep of the site and immediate response to situations. They can delete and lock without needing community input, both of which are harsher and do not allow conversation. Removing the binding close results in a very strange conflict of power here - moderators can be soft only if the community agrees, but can definitively freeze all activity on a question. I think that alone is sufficient reason that making the vote "less binding" is not a good solution.

3

What does a single mod vote [to close] with no comment teach? Nothing!

Well, it does if you read the close reason:

closed as not constructive by Jeff Atwood♦ Sep 21 at 20:34

This question does not meet enough of our six guidelines for constructive subjective questions.

Just click that link! Right there! That explains the closure! You can learn ALL YOU NEED RIGHT THERE!

That said, I do agree that it is helpful to indicate which of the guidelines the question has run afoul of. But for a question about "why are programmers athiests?" I consider its .. lack of utility .. rather self evident to all but the most clueless of observers.

4
  • The issue that I have with that is the OP can look at the guidelines, say it meets 5/6 of them, but its still closed. When a closer says something like 1) No 2 Yes 3) Yes 4) Yes 5) Yes 6) No a discussion can start about 1 & 6 and the community now has an actual example. With no explanation though, we have no idea how far rules are being taken (especially #6), can't discuss anything, and can't the community can't learn as much as it could
    – TheLQ
    Commented Oct 1, 2010 at 11:20
  • @TheLQ: From what I can see, that question blatantly fails on #2, #3, #4, #5, and #6. #2 because most of the answers are one-liners or very short. #3 because it's not written in a neutral tone, it suggests a conclusion. #4 because even though it asks for "experiences", it actually quite clearly means "opinions". #5 because there's almost no way to back up an answer unless you went and did your own scientific study. And #6 because no answer to the question could possibly serve any constructive purpose. Satisfied? (And by the way, I'm atheist, so don't accuse me of being offended by it.)
    – Aarobot
    Commented Oct 1, 2010 at 14:12
  • I thought I was losing it, I even went to find one of the PSE posts closed as not constructive so I could click the link and make sure it went to the exact same blog post as the "really helpful comment" example. I can only assume when the close reason is "No real explanation, just a link to a generic blog post about what might get a question closed" but a comment pointing to the same place is "an extremely good close" that people aren't actually clicking the links Commented Oct 1, 2010 at 15:24
  • @michael the 6 constructive subjective guidelines are also in the programmers.se faq programmers.stackexchange.com/faq Commented Oct 1, 2010 at 20:17

You must log in to answer this question.

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