52

Ahem, this is going to be a rough one, so it's gonna be a long one too ala Pascal's famous quote. First, some context; we have

How to deal with robo-reviewers?

Should we have a Grace period to avoid premature Voting to Close and Reopening Cycle

(and more)

This thing is bothering me for some time already, especially after the migration of many users who do more or less exclusively reviewing on the site but not much of actual content contribution. Not that I am dismissing or underestimating the effort that goes in but there is a fine line between having a lust-driven relationship with review buttons and actually reviewing the questions properly (which means once in a while clicking that $%$%^#& Skip button if you don't have any idea about the subject).

Fun fact; I have tried a few identification experiments such that I found a recently closed question and then voted for reopening (i've chosen closed-with-valid-reasoning ones to maximize the effect) and some users who actually voted for closing also voted for reopening. Now no need for calling names for this but, if you excuse my French, this is bullock-cart-pulling-bullshit (emphasis is mine).

Now, I don't know why you hang out here but I like being borderline useful while learning a lot. Also, as a recent-beginner of Python and Linux, I really feel the pain of some of the users that ask quick questions with haste because sometimes you really don't know what else to do and having copy/pasted every damn solution on SO or any other place, it quickly becomes a mess. Last thing you need is to put n-propeller-powered-jetillion nuances/rituals about what the site rules are or whatever the forum owners' girlfriend implemented somewhere. With a little bit of effort and gentle nudges, most new users give in without any sour feeling and make a MWE, it takes time and patience and we have a lot of it in here. Or had. Now all I see is dictating comments on how things operate here and if you don't obey my warning, I'll call my mod brother and he'll whoop yo ass. attitude.

Hang on; another thing needs to be out-of-the-way; There are these things called Stackexchange network rules, nobody knows who come up with them and why but in their own environement they make sense (I hope). And through some law-abiding-never-law-questioning citizens of that network we are introduced to these things.

For a recent one that claims that I'm not informed well about a site that I'm not even registered see Consensus vs accepted answers on meta

However, if this means that powers that be actually dictate what we are doing here then, first, it goes against the notion of community-driven Q&A site, second, I am out of this place, in the first instance (which has no impact of course but I'm out anyway). It bears no threat value just my opinion. The reason is that I didn't sign up for this, and truly I don't care what owner of SE network thinks about TeX-SX. All I know is that, there are users who really dedicate their time and patience to come up with the wittiest solutions that we all see and also there are these review-people who click on the buttons like crazy in the name of some obscure nonsense that is maybe applicable to a humongous site that is called Stackoverflow that bares no resemblance to this site, I mean, in any possible statistical aspect.

My argument goes like this;

  • Our number of question intake is not changing dramatically, so we were OK with not closing every question in 2 hours --> We will be OK without closing them in 2 hours
  • Our number of active users is amazing including those review-junkees, even spams don't last longer than a few 5-minutes --> No need to fear about the unanswered/asked ratio, we are doing pretty good without closing them like crazy.
  • New users were getting confused already with stock comments --> Now they are more pissed off in average (based on my observation so no facts here) But this question Why is \rotate useless in deluxetable? believe it or not closed twice! Not once. We reopened it David wrote an answer and we closed it again. (The OP mentioned closing already but no we have to be faster, modern world doesn't wait, so in your face noobie, don't come back again)
  • The close voters don't read the comments or any discussion going on under the question or answers. To be honest they don't read at all as far as I understand.
  • This sucks.

These close-voting-users however become the unofficial saint-of-the-noobs when they vote for suggested edits. Suddenly every edit ranging from very good ones to oh-it-was-my-screen-having-dust-not-an-actual-comma type of microscopic edits. All pass with flying colors.

Closing remark

If you are so eager to contribute, answer dem questions properly.

In the spirit of one of the pillars of TeX-SX : Vote early and often!, Review Late and Rarely!

So, after all this tasteless rant, here is my actual problem; I'm tired of keeping track of wrongly closed questions or questions with threatening comments under them. I don't know any other way than actually pointing fingers to users which would cause even more bitter taste so question is how can we (assuming that I'm not the only one feeling this strange transition) reach out to these users without creating a primary school level snitching-to-mod game?

I really want to get the attention of those frequent-voting users because I think we can arrive to a better understanding (though it seems difficult if they are high on badges and euphoric on their name up on the The Most Awesome Reviewer list).

Afteredit on an important detail The reason why I seem to have a tone that targets individuals is because I don't expect the powers-that-be to unimplement this review system anytime soon. So the only way to change this situation is through reviewer behavior. Had this system not in place, they wouldn't be able to vote this quickly and erroneously. Hence while it is true that I speak mostly about reviewers, the target is indirectly and essentially these buttons that amplify and encourage careless reviewing behavior.

  • 6
  • 3
    @ClaudioFiandrino Numbers are irrelevant. lockstep practically retagged the whole site without causing any controversy or annoyance. So he deserves all kind of authority or badge or whatever. But the issue is about whether we review as meticulously as him? Or randomly clicking buttons because the system gives us the chance to do so in the name of all active users. Exhaustion is a fair point I didn't think about that much. Let me chew it for a while. – percusse Sep 15 '14 at 11:53
  • 2
    Very good point! Honestly, I always thought that the differences in voting reflects somehow human nature of seeing things from different angles. Though, it's true that we need to keep in mind, and add to previous reasoning, the knowledge level of each user, which is obviously different. The point I totally share with you is the ridiculous amount of time in which questions are closed. – Claudio Fiandrino Sep 15 '14 at 12:37
  • 2
    @JosephWright Not only encourage but also isolate them so that there is absolutely no way to avoid a handful of users taking over the entire community behavior with a few clicks of a button. Fixing the damage they cause is a major task; finding the questions, voting for reopening, even comforting the user that it was a mistake etc. We are much better off without this automated review screening. Note that auditing won't help here because we don't even see the review items anymore due to this weird selective redirecting system to particular users and it is self-reinforcing. – percusse Sep 15 '14 at 14:45
  • 5
    ...perhaps another point to remember here is: There is no shame in using “Skip”. In fact, a Skip incentive has been mentioned on Meta Stack Exchange, but it doesn't seem to be of interest in a larger community. Again, perhaps because it comes with another incentive... – Werner Sep 15 '14 at 16:05
  • 2
    @Werner We were doing just fine before this system arrived. And if you remember it was a much fluent practice both closing and opening the questions. Because more than 5 people were involved no matter which user it was. SO model is actually hurting many many satellites not only TeX-SX. Plus people can vote without even reading the question which is sheer nonsense. It might happen that someone understands the actual problem while the majority cannot see the culprit. These nuances are all swallowed by this automation. – percusse Sep 15 '14 at 16:14
  • 3
    @AdamLiter As long as you invest time and care, you don't need to stop at all. I'm not trying to discourage people, quite the contrary please do but you don't need to rush. If it happens that you were not fast enough to mark so be it. That's what I'm doing. – percusse Sep 15 '14 at 16:18
  • 4
    @tohecz As I understand it, the 'big picture' issue here is not the behaviour of a specific individual but the fact that the review queue system tends to encourage rapid reviewing as it's 'easy'. Thus it's a system issue, not a personal one. – Joseph Wright Sep 15 '14 at 20:30
  • 8
    @ClaudioFiandrino As a fairly active user, I would say that the one thing that will discourage my participation is if the site becomes perceived as being an unfriendly place, not whether there are too many questions that maybe should be closed. More and more I'm finding myself seeing closed questions and asking "Why?" (like percusse) and also seeing far too many "Please add a MWE" to questions that strictly speaking can be answered without. (There, I said it. :) ) Such behaviour does not contribute positively to the site IMO. – Alan Munn Sep 16 '14 at 15:13
  • 4
    @percusse -- i'm willing to 'fess up that i have, on occasion, voted to close a question, and later to reopen the same question. in my defense, i note that in all cases i can think of, either the question was revised to add a new wrinkle, or someone provided enlightenment in a comment that made it obvious what was different about the question. (and i'm perfectly willing to skip questions whose topic i know nothing about. also to hunt for possible duplicates, to justify my reason for closing; but sometimes finding an item i know is there isn't easy. i love the "frequently asked" list!) – barbara beeton Sep 16 '14 at 19:28
  • 2
    @Werner If this is happening with the relatively older users, I don't have any difficulty addressing them right away (valid also the other way around). But we can't reach to the reviewers anymore. They are like the old dudes in the balcony in Muppet Show. They dislike almost everything :) – percusse Sep 17 '14 at 20:46
  • 7
    I ask for MWEs a lot. But I think that the addition of an MWE improves a question in many cases even when it is not strictly required to answer it. First, it may well be that people here could complete code to produce an MWE but providing an MWE in the question makes it easier to answer and, yes, requires the questioner to make the effort required to create one. But, more than that, I think that an MWE makes many questions clearer even to those not 'in the know'. That is, the fact that a question is clear to an expert doesn't make it clear to everyone. An MWE can help greatly with that. – cfr Sep 18 '14 at 22:37
  • 3
    I don't see any reason not to ask a new user for an MWE provided some information is provided about what that means, of course. But, then, I think knowing how to create an MWE is the single most useful thing one can learn about TeX so perhaps I'm biased. Effort is ill-defined but questions which post an image and say 'please draw this for me, preferably in 3D' certainly demonstrate very little effort. In that kind of case, I don't think it is unhelpful to suggest somebody should have a bit of a go themselves. Also, an MWE often tells you what knowledge a user already has. – cfr Sep 18 '14 at 22:52
  • 3
    I've obviously visited the wrong (or right) forums, in that case, as that's not been my experience. But I profoundly disagree about MWEs. I don't mostly think people should post MWEs for the benefit of this site. I think they should create MWEs because they are extraordinarily useful. If you know nothing about TeX, you can use an MWE to troubleshoot even though the cryptic error messages and log file mean nothing to you. That was so long before TeX SE existed and it works even if you know zilch about programming and almost nothing about TeX. The existence of sites like this makes MWEs less... – cfr Sep 19 '14 at 0:33
  • 8
    @percusse -- "..., if everyone read the manuals or even TeXbook there wouldn't any need for this site anyway." i disagree! i've been at this gig longer than anyone else on this site, and i find it mighty useful! maybe i don't ask a lot of questions, but that's largely because other people have the same questions, and there are a lot of knowledgeable participants who are able to answer (most of) them before i even look at the site in the morning. and i'm a big fan of mnwe's (i'm facing a bug now that requires at least 4 variations to demonstrate). onward with the "good bad example"! – barbara beeton Sep 19 '14 at 15:42
29

'Executive summary' of the following answer:

  • The key aim of reading questions is to help the asker
  • Voting to close is thus (usually) something to do after comments or similar to improve questions have failed
  • Rapid closing is ideally aimed at clear cases such as questioner agreeing, spam, massively off-topic, etc.
  • Minimal working examples (MWEs) are a useful part of this process but are not a 'requirement' nor always needed to understand a question
  • Asking for a MWE is partly a way of helping the questioner help themselves: once they can do this, they are likely to be able to solve more problems themselves
  • Keep the questioner in mind: we were all beginners once

Quite what a 'real' answer here would be I'm not sure, but I think a summary of the consensus from the comments would be in some way useful. Before I do that, though, I'd like to observe that the question has a good number of votes (23 at the time of writing) and so presumably the concern raised is shared by a spectrum of 'active' users. (We know of course that votes on meta are a bit tricky to interpret.)

There seem to me to be two parts to the question: one part about closing votes, a second about comments on questions, and both have interesting comments here. I'll therefore try to each both of them and summarise.

In terms of voting to close, the general feeling seems to be that there is no need to rush to vote to close questions. Over time, some questions can indeed be closed, for example when clarity is never achieved, the problem clearly resolves to a minor issue (typo, etc.), or for many other reasons. However, that doesn't mean that they have to be closed quickly. The latter action is something that the review queues do seem to encourage: they tend to pick up questions not long after asking, and it's then rather easy to vote to close without giving things time. In particular, and perhaps one that we might get the dev team to consider, it's notable that the buttons to vote to close questions in the queue come at the top of the page, making it rather easy to vote without reading the question, any comments, and answers. Moving the button might be a good idea!

As with many things on the site, there are badges for reviewing, and there is at least some likelihood that a certain amount of motivation to review quickly comes from this. That's not just been raised as a concern here: it's come up on the main meta site too. At the moment, I can't say I see any change by the Powers being likely, so we have to adopt a sensible approach to reviewing (see below).

On the comments front, the main concern is about asking minimal working examples (MWEs). Experience suggests that most of the time a MWE is useful in dealing with a question: they help track down errors, give clear demonstrations of what's happening and so on. At the same time, there are perfectly clear questions that don't require a MWE. Importantly, asking for a MWE is a request to the questioner to help us to help them: it's certainly not a requirement for asking a question. Hopefully the Text building blocks get this about right, but we can always revise them. (Of course, everyone is free to write comments however they like: the pre-written text is just there as a handy reference, really.)

On this area, it's perhaps worth noting something that happened when the close reasons got revised to drop the old 'too localized' and bring in 'unclear' plus custom 'off topic' reasons. The mod team can add a small number of custom 'off topic' reasons to the standard list, and as part of the roll-out the developers added some suggestions. One they put was 'off topic as there is no MWE' (or words to that effect), which the mod team didn't adopt. Over all, lacking a MWE is fine if the question can be understood, so the mod team reasoned that questions which need a MWE but never get one are 'unclear' and not 'off-topic'.


The overall approach that's been taken to date on the site to trying to improve 'sub-optimal' questions rather than closing them: after all, the idea of reading questions is to try to help the person asking. As such, the approach most users have taken is that voting to close is something to do after you've tried to guide the questioner to make things clear, and only when that seems to have failed. Of course, there are questions which are massively off-topic, where the OP agrees to them being closed and so on, and those can all get dealt with quickly. For a lot of questions, though, what is best is a bit of patience in helping the questioner to help us to help them!

  • 1
    Great summary and answer! – Werner Sep 19 '14 at 15:24
  • i agree with Werner. but i would suggest making an "executive summary" at the top, listing the essential points, followed by the commentary, so that "casual" readers have the option of not having to read through the whole answer to determine what's important for "adjusting" their behavior. – barbara beeton Sep 19 '14 at 15:52
  • @barbarabeeton Would that essentially mean moving the last paragraph to the top? – Joseph Wright Sep 19 '14 at 15:55
  • @JosephWright -- moving the last paragraph to the top might do. but maybe an alternate approach, presenting those points as a list. (as much as i hate such "executive summaries", they do have their uses, and capturing the attention of readers with short attention spans is probably the most valuable.) – barbara beeton Sep 19 '14 at 16:01
  • 1
    @barbarabeeton I've gone for a list: hope it helps. – Joseph Wright Sep 19 '14 at 16:08
  • 1
    @JosephWright -- super! the only thing i might suggest is a brief comment to the effect that, in appropriate cases, it might be worth pointing out to a user that constructing a minimal example can help them isolate the problem and its cause on their own -- a skill-building experience. only in fewer words. (i always use too many words; i can only defer to mark twain and his comment about how brevity takes more time.) – barbara beeton Sep 19 '14 at 16:23
17

As we have seen from the comments, every detail leads to another full blown discussion and I am not the TeX-SX correspondant so I'll try to be a user again.

Because there is user involvement that might lead to confrontation and since we are extremely avoiding it, I have to be very careful with the sentences I am using. Otherwise it starts to get dissected and my point gets lost. Since it has some upvotes and zero reviewer answers here.

Wait actually there is a cute one but not so useful; I initially mentioned that instead of chasing down close votes I would rather prefer having my usual TikZ porn and this is a response

I have no time to discuss with you, sorry. I will not answer your question on meta. Good luck with your TikZ porn. I am not deciding in the name of everyone... I am rewiewing in my proper name, of course.


Lovely. So coming back to the essential reason of this whole long, long post is this last bit in the response. But anyway no need to make it personal;

  1. Reviews are not our personal toys, they are, no matter how terribly implemented, here for site moderation and consistency. They define what is proper and suitable and what is not. They are also exemplary for future visitors to give an idea why a question is closed or why something is edited, retagged. You can't be on your own clicking buttons and not taking null responsibility. But from the look of it, we have no incentive to handle this so this opens for a possibility of everybody hanging around according to their leisure. This is a stronger version of the subtle question I've posted initially. So now we can put it on the table more suitably. This would undoubtedly cause a lot of trouble. If we are smelling the smoke, it won't help if we keep saying that it's still not that hot around here.
  2. Asking for MWE is, with all my respect, getting super-annoying. Here is a simple test; ask yourself this Am I going to answer this question if the user actually includes a MWE? If your answer is no then don't bother commenting. The potential answerer can do the same. You are really really not helping. And it is not making anything easier. Because if you see it under every question, it loses its effect (already lost but nevermind).
  3. The questions are unclear if the OP responds to and still doesn't include information. If two users are interacting, it's not your call to decide on the question. Stop marking every question as unclear. They are not. There is no time limit for a question it can be open for years.
  4. If you will invoke a I'm keeping the site tidy argument, then start joining our Answer the Unanswered sessions. There you will have ample amount of happy-trigger satisfaction. No need to slam the questions to the users' faces.

I hope this clarifies a bit about my point. Please come to the chat instead of commenting to this one because apparently this post didn't make any impact so no need for further surgery. Of course you are free to do so but I'm a little reluctant to reply to avoid extensive commenting (see above).

  • 5
    I am so with you on that MWE point; some people seem to be asking for MWE's without any intent of answering the question. – morbusg Sep 21 '14 at 6:39
  • 8
    on MWE as I've commented elsewhere I'm not that fond of any of the "text blocks": posting boilerplate comments never seems that useful. If I do ask for a MWE I usually state that every question should have one, as that's a reasonable approximation to the truth, but I answer questions without a MWE if answer doesn't need a test (or making one isn't too much work) – David Carlisle Sep 21 '14 at 14:48
  • 5
    ++++1 for point number 2. I think all of us understand the value of MWEs, but they are most relevant when someone is getting an error or an output they don't understand. If someone is asking things like "How do a make a table a single cell shaded grey?" or "How do I read a CSV file in LaTeX?" asking for a MWE at best gives everyone some code to copy. Such questions are perfectly clear without a MWE even if adding one might make it slightly faster for someone to answer. – Alan Munn Sep 21 '14 at 16:48
  • 3
    TikZ pr0nZ FTW! :) – Paulo Cereda Sep 29 '14 at 12:19

You must log in to answer this question.

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