How can I encourage Stack Overflow to rein in the 'subjective' vigilantes? [closed]

Question

I'm on SO pretty much every day researching something to do with programming, and over the last month or two, I find I'm growing frustrated and a little disgusted at how many interesting questions have been closed as "not a good fit" or "too subjective".

I think there is a confusion between "subjective" and "calls for judgement". To me one of the most valuable things about SO is the tremendous breadth and depth of experience of the members, and experience translates into smart decisions and advice, and that advice cannot always be asked for, or expressed in terms of, bald facts and data. It's going to be 'subjective'.

I don't think SO should be Wikipedia. I think that if I could ask a question to a group of skilled programmers sitting around a table at lunch, and get interesting, valuable answers, then I don't care whether the question or the answers are 'subjective', I think that question should equally well be fair game for Stack Overflow.

I would like to see a mechanism whereby people who vote to close a question lose reputation when somebody expresses interest in that question or its responses.

How can I, or can I, influence this aspect of SO?

Answer 1

tl;dr

We already tried supporting those questions, we even gave them their own site. Sadly, it didn't work out. C'est la vie.

---

In 2010, a Stack Exchange site called Not Programming Related came out of Area51, the Stack Exchange staging zone. NPR was supposed to be a site where questions that were too subjective / broad for Stack Overflow would find a new home. The site was greeted with enthusiasm, and in theory it looked like a perfect solution: Stack Overflow would remain as laser sharp focused as possible, and NPR would host all those exciting and sometimes helpful (but not really answerable) questions.

However, as it usually happens, theory and practice are two entirely different beasts. NPR's promise proved extremely attractive to people who were more interested in posting joke answers, or just repeating earlier answers, or posting outright crap (Do you fart in the cubicle?). It didn't take long for everyone to realize that the site was not working, and most people just didn't bother with it. Here's what Quantcast tells us for the first year of NPR's existence:

Tons and tons of people visited the site after it went public, but very few decided to stick around. For over a year, the site experienced no growth to speak of. And even fewer people contributed worthwhile content. Turns out that while everyone loves those questions, very few are actually willing to spend any time to answer them (seriously), and maintain and moderate them.

Fortunately, Stack Exchange realized their mistake soon enough. First, Joel warned us that the site was "degrading into fairly stupid water-cooler nonsense" and then Jeff stepped in, and enforced the infamous subjective guidelines. It took more than a few months for people to realize that NPR's (by then already renamed to Programmers) scope had changed drastically. The site had been heavily advertised as Stack Overflow's toilet bowl, and naturally most people believed it was just that, long after the subjective guidelines were enforced.

To make matters even more complicated, there was a small - but solid - group of early members that were attracted to the site by its initial scope and were contributing high quality content. Unfortunately, they weren't enough to maintain a healthy site, and some - if not all - of them were (understandably) quite frustrated after the drastic change. They were doing everything right, and then suddenly the site they were originally promised was no more.

Then came the clean-ups. Changing the scope wasn't enough, we had to change the way we advertised ourselves to the world. Just saying that we were now a serious Q&A site about software design wasn't going to cut it, at least not until we cleaned up all the content that no longer fit the site's scope. After a few months of intense Meta drama, we deleted about 2K questions. And then some. A couple more months passed, but finally we started experiencing growth:

Whatever your opinion of the site's current scope might be, you can't overlook the fact that it only started growing when it became a vigorously moderated serious Q&A site. Growth aside, every other aspect of the site gradually improved. Hard questions that were once ignored in favour of the latest "let's build yet another list" question are now answered, correctly and relatively fast. More and more people have Programmers as their top site, the one site that they actively and regularly contribute. We are not there yet, but we are certainly on the right track.

The site you seem to want existed for more than a few months, and it failed. Horribly. I'm sorry but I see no reason to try it again.

Further reading

• What is the history behind the site scope change from NPR to “conceptual questions about software development”?
• Adding discipline to programmers.stackexchange.com
• Should Programmers.SE exist?

Answer 2

I think that if I could ask a question to a group of skilled programmers sitting around a table at lunch, and get interesting, valuable answers, then I don't care whether the question or the answers are 'subjective', I think that question should equally well be fair game for Stack Overflow...

Assuming it's like group sitting around "table at lunch" is exactly where you are mistaken.

To find out why, consider studying Shirky's article A Group Is Its Own Worst Enemy. The article is fantastic, it could profoundly teach the reader about the difference between a lunch group and large community, but for the sake of brevity I'll just quote the key point relevant to the mistake in your assumption:

"human interaction... doesn't blow up like a balloon"

---

Simply put, when there's a handful of colleagues in a group, a subjective conversation can go like this: first person asks a question, second answers it, third one answers from a different perspective, fourth sparks a joke, fifth adds a side note... and that's all.

• In a setting like this, you get a reasonably limited amount of information, an amount your brain can handle... you've got something you can learn from.
  

At Stack Overflow, you better think of something like 1000 people sitting around the... world.

In such an environment, your subjective conversation will probably go like this: first person asks a (subjective) question, second answers it (so far so good huh?), twenty more add answers from different perspectives, fifty more attempt all imaginable kinds of jokes, hundred more add all imaginable kinds of side notes... and so on and so on, over and over again, until your brain explodes.

• Trying to learn anything from this would be like drinking from a firehose
   
  

Answer 3

I think that if I could ask a question to a group of skilled programmers sitting around a table at lunch, and get interesting, valuable answers, then I don't care whether the question or the answers are 'subjective',

That's fine. We care.

We care because subjective discussions will destroy this site. We care because subjective discussion is exactly why many of us are on this site instead of on many of the thousands of programming forums on the Internet.

Stack Overflow is not intended to be everything for everyone. You come here to ask objective, practical questions that have real, objective answers. If you want to have a subjective discussion about some programming issue, that's wonderful.

Don't do it here just because there happen to be a lot of smart people here. You'd be subverting the very reason why those smart people are here.

What you're asking is no different from wanting to read a book at a movie theater; what you want is at odds with why the place exists.

I would like to see a mechanism whereby people who vote to close a question lose reputation when somebody expresses interest in that question or its responses.

This is one of the most ridiculous suggestions I've ever seen someone seriously suggest for this site. The potential for abuse of this is so massive that nobody would ever close anything again.

How can I, or can I, influence this aspect of SO?

This is one of the founding aspects of SO. You can no more influence this than a fish can change the flow of the river. The anti-subjective bias is not merely part of the community, it is at the very core of why Stack Overflow was invented.

Again, it's like being annoyed that movie patrons ask you to turn off your reading light at the theater. You're not going to get them to change their views either.

SO is lenient in some respects to different opinions. This is not one of those cases.

Answer 4

So far the only site that I've ever been on or heard of that removed questions such as those is the SE network, so to answer your question of:

Where should this information be exchanged if it's not on stackoverflow?

Literally anywhere else on the entire internet. Any forum, q/a site, or interactive site that allows one to ask programming questions. Virtually none of them will refuse to allow questions like this. There are lots and lots of places where people can go to ask these questions.

You might say that you're unlikely to get good answers on those sites, since SO has so many experts who may have more valuable information to give. That's not a coincidence. SO tends to have higher quality content, more experts, etc. specifically because it is restrictive in the questions that it allows. By prohibiting questions that have shown, over an extensive history, to cause problems and result in lower quality content, the remaining questions end up being (largely) of great quality, and it creates a site that experts want to be active on, as they don't have to deal with a lot of the lower quality questions, the bike sheds, and the arguments/flamewars/ranting/etc. that come along with questions like those you suggested.

Answer 5

I don't think SO should be Wikipedia. I think that if I could ask a question to a group of skilled programmers sitting around a table at lunch, and get interesting, valuable answers, then I don't care whether the question or the answers are 'subjective', I think that question should equally well be fair game for Stack Overflow.

Some questions aren't really answerable questions, they are discussions.

Stack Exchange is a Q/A network, not a discussion board.

---

Interestingly, a Wikipedia type site seems to me the exact format you are wanting - because you can have more discussion of pros/cons to different things which is more applicable to an article than an "answer."

Answer 6

I understand why you and others ask this question. Sometimes questions with a fair amount of subjectivity do receive useful answers, possibly get heavily upvoted, then get closed.

The main goal of SO is that someone can use a search engine to get an answer about a topic, and have a really good chance of getting useful information. So, why sometimes close questions that have elicited useful information?

To understand that you have to understand the effort it takes to make sure answer quality is reasonably high. Even with the current, pretty strict, guidelines (which do sometimes rule out useful content) a lot of people volunteer (a very few get paid) what is collectively an enormous amount of time just trying to maintain the current standards - and are struggling to even do that (the backlog in the review queues is enormous).

So, like many matters in programming, you have to drop the ideal (in which hordes of knowledgeable and benevolent moderators and users spend the time needed to curate any number of subjective questions, deal with the many people who are upset that their question has been rejected while a similar one has not (because, you know, that's pretty much inherent in the whole subjective thing), and generally spend the large amounts of time needed to make things ideal) in favor of the possible (in which smaller hordes of overall surprisingly knowledgeable users and moderators spend an enormous amount of time curating the site but often come up with heuristics to save enough time to make completing the job possible).

Some of the time-saving heuristics include closing categories of questions that historically have taken up large amounts of curator's time and produced a disappointing quantity and quality of good content. I agree with you that some good content is lost, and certain categories of very useful content are nearly absent from SO. I think after a little time on the site this problem is readily apparent.

What is less readily apparent is how very hard it is to have that content without burying it under content of far lower quality. I would love to see some method of solving that. I'd complain too - it seems a shame that certain things are missing - but, unfortunately, I don't have a feasible solution that scales to the level of this site and uses the resources available. I don't think anyone has figured that out.

Answer 7

Even if they do become a bit of a discussion

Well, that's the crux of the matter, right?

We find discussion to detract from the purpose of the site - to be a repository of questions and answers.

Popularity and usefulness don't come into that equation.

Answer 8

Edit 5/16/2019:

I've realized this really just comes down to a misapplication of the original Atwood article. People applying it not recognizing how the nuances and abstract nature of programming make it a different case than cameras and graphics cards. I'm not saying Jeff was wrong. I'm saying the people that are taking an article written in the context of "superuser.com" with all hardware examples, are doing a disservice to stackoverflow.com by insisting programming components are the same thing.

This is a misapplication for a number of reasons:

There's nothing abstract in the camera example

If somebody asked "I need to capture a highly accurate image of a scene in a very low amount of time, would anybody balk if somebody said you need a need a "Kodak"? And before you say yeah, but that's a brand name, what if it's 1889? There's a time lag between when something goes from abstract to common concept. Just because we can take for granted that everybody knows what generalized component concept of "camera", does not mean we have the same thing in programming yet.

He asks a specific question about cameras. But in the case of programming libraries, we're either asking a general question, or we are asking a specific question with context.

For example, when somebody asks "What component can I use to dynamically create a PDF file?", they are really saying "How do I write a Excel document from this data I have?". The fact that they had the notion that the answer is likely an existing library and phrased it as "what component" does NOT make it a shopping question, but instead simply shows a very basic understanding of programming.

This misapplication is dated

In an age of EVERYTHING being modular in a NuGet or npm, "component" as universal as an "if" statement, even what would once have been considered "native" libraries.

**Rewriting "component" questions does not change the answers"

In Atwood's example of "which camera should I buy for low light photography", his rewrite of the question "which features should i look for for low light photography" will result in DIFFERENT answers.

However, in the case for programming components, if somebody says "How do I write data to an Excel file?", they are still likely to get the same references to components. So the guidance here should be to simply suggesting a rephrase of the question rather than closing it.

Asking what we should learn is EXTREMELY subjective, especially in programing

Jeff states: "Thus, when it comes to shopping questions, don’t ask us what you should buy — ask us what you need to learn to tell what you should buy"

I think this ok when a concept is well established. Cameras might be a good example. A question like ASP.NET vs. JSP is well established as dynamic web technologies. But what exactly is the conversion for most component questions? In the case of writing an Excel file, do I really need to learn the openXml standard? In the case of a reading a web response, do I really need to learn the whole TCP/IP protocol?

Dated answers are handled by the voting mechanism Atwood states that specific answers are only valid for a year. This decidedly NOT true for many component answers. Even if it was, the voting mechanism will take care of that for us, buy letting old answers sink over time.

Follow up questions don't really apply Jeff gives the following questions that would need to be asked in a true shopping question, but I can answer them all right now for all component questions. As further evidence, I challenge anybody to find any of these questions in the comments on ANY question on Stack Overflow:

What is your budget?

First off, have you really worked with so many of these components that you can give me multiple answers? In any case, I'll take the cheapest one.

Where do you live?

Cyberspace.

What are your preferences?

See tags.

Which alternatives will you consider?

Working code.

When do you want to buy?

Yesterday, I'm already behind deadline.

----End Edit---

My original answer:

My discussion on this was closed (Too many closed questions) referring me here to this question.

This is not an answer, but further support for the asker:

It would seem that if there is even the possibility of the question being subjective, it is closed. These days there must be a dozen ways to code something, yet those questions are allowed to stand. Yet the instant somebody asks a "what is a library that will do 'X' it is closed. The irony is that these questions, even in their closed and decaying state, are often the most valuable to me. Sometimes I'll test three or four third party components before I commit to one. This is a time consuming process that I would gladly augment with the vote-validated answers that others might have.

Examples include:

• https://stackoverflow.com/questions/744966/any-open-source-c-sharp-ocr-library
• How do I create an Excel (.XLS and .XLSX) file in C# without installing Microsoft Office?
• https://stackoverflow.com/questions/408212/best-net-blog-engine
• https://stackoverflow.com/questions/361271/best-free-controls-for-net/361331#361331

I'll concede that the last one is too general but I left it on there because even if it was free control that does 'X' it still would have been closed.

If these questions cannot be asked on Stack Overflow proper, which subsite should be considered (keeping in mind that the other sides hardly get anywhere close to the SERP rankings that Stack Overflow gets).

Does the asker need to explicitly ask the "subjective" guidance questions to avoid getting closed?:

• Please explain “why” and “how”
• Please elaborate
• Please only share items that you have personal experience with.
• Please post any data you have relating to your experience

Isn't that a mess? Shouldn't we be able to rely on the answerers to follow the guidelines instead and if they violate them make effective use of the down vote button beyond simply being wrong? Why does all the burdon fall on the asker when it's the answerers that should know the rules and how they can best serve the asker without turning them away?

Are these questions even subjective? It is a fact if a component meets a need. "Best" is subjective, but isn't that inherent anyway? Is anybody going to list the item they've already discarded for themselves? Are we really going to let a single adjective decide a questions fate? Isn't this easily cleaned up with comments?

I understand we're trying to avoid the wild west of a forum, but we're not debating politics here. Many of these questions could be answered with "here's what I did and here's why". We could restrict answers to things that are pulled directly from the examples on MSDN, but why exactly do we want to do that?

What is the guiding principle here? If we want to help programmers save time and create better software, then things that stop them from getting to page 17 on Bing should be included.

EDIT:

For those that chose to dismiss my examples as "shopping" questions please explain the difference between asking "How do I create an excel file from .net" (which there are dozens of un-closed examples of and "What is the best component for creating an excel file". Doesn't one question simply acknowledge that the best candidate for an answer is likely going to be a component?

If there is no place for subjectivity on Stack Overflow, what is the point of the voting mechanism?

Answer 9

Design questions call for judgement

Very often the kinds of questions of the "how do you design X?" variety call for judgement. Often, that judgement comes with experience. Providing a satisfactory answer (possibly among many such answers) usually requires the kind of background that comes from having solved a similar design problem in the past, and seeing the aftermath.

There are two wrinkles in this.

First, there can be many satisfactory answers to a design question. Choosing among those answers can, in fact, be subjective. But that doesn't mean that the satisfactory answers, as a group, don't add value. It also doesn't mean that the question itself is subjective to the point of lacking merit.

Second, the asker usually doesn't recognize their case as belonging to a given design pattern. The only people who see the pattern are those who are acquainted with one or more of the satisfactory answers. So the question usually lacks the definition that will make it obviously belonging to a pattern.

The result is that design questions are often marked as subjective and closed prematurely. Those of us who see a pattern in the question need to respond, even if there are votes to close in progress. I'm not saying this applies to all questions with votes to close. Just the ones where there is a non obvious pattern at play.

Answer 10

Some of these questions are very old. Over time, Stack Overflow has become more strict on what is allowed.

And these strict rules are there for a reason. There are enough places where people talk about which tool is "the best". Stack Exchange sites aim at giving definite, objective answers.

Questions also need to show prior effort. This prevents the site from being overrun by "help vampires". But also, when I have a problem and look on most other sites, I see a list of things that I've already tried myself. When I look on Stack Overflow, I find someone else who has also tried these things - and failed. Then (hopefully) an answer follows that shows a new insight, that I have not yet tried myself.

So, even though there may be informative things in the answers, questions asking for "the best product" or questions that show zero effort should be closed. It may seem harsh, but it is necessary if we want to keep a high signal-to-noise ratio.

Answer 11

SO has some standards for questions to be welcomed... the FAQ has detailed information about what you can ask in here and what you cannot ask.

• JPA Implementations - Which one is the best to use?: SO expects questions to be answerable. That question is opinion based and also the answer can change according to your requirements. This is not a discussion site, you can ask your questions and get direct answers.

  Also probably it will not offer much as a future reference since requirements differ and choices change too. If some application is the best choice at present, that does not mean it will be the best in 3 years time.

• How to read a large text file line by line in java?: SO is not a place where you can make other people do your job. You must try and search before you ask something in here. You must have some specific problem to be solved, not a whole concept to be written for you.

  Closing an answered question is necessary because if it is not closed, then it will encourage people keep doing the same thing. If SO do not close an answered question which do not fit its format, then people keep asking them since they will think "No problem, we got our answers".

• Are weekly status meetings necessary?: SO is for programming questions. SX has many sites for many Q&A formatted questions and this is not the right place. Visit programmers for such questions.

And one more thing, this is a community moderated site. If there are no strict lines for moderating, then everything will be a mess. There are thousands of people in here who can access moderator tools and there must be guidelines for them to follow.

Answer 12

The simple answer from reading many of the replies here is that the current infrastructure wasn't made to accommodate subjective discussions. The best way SO could reign in the "subjective vigilantes" is to provide infrastructure to support subjective exchanges. In this way the subjective and objective questions could be dealt with differently.

Answer 13

I personally sort of agree with you, and I searched this question up before I was going to ask it myself. To me it seems like some questions are unnecessarily closed for the sake of being stubborn about enforcing the rules, because I think they can be very useful questions, and others undoubtedly believe this too. For example, what if I wanted to ask "What is a good Java dictionary library?" and I couldn't find the answer to this elsewhere? Should I use a different site just to ask this? I know there won't be as many responses on different sites and they won't be as "good," so I'm naturally inclined to ask them on Stack Overflow. But don't interpret this the wrong way, I'm new to Stack Exchange and these are only the truthful thoughts from a naive perspective, and I'm sure there are reasons for closing subjective responses.

Answer 14

As of 2017, there's a few sites, both established and in beta where answers are somewhat subjective, and have covered subject material one might consider a little too subjective on older sites. In a sense, the trick's really to propose, and build up sites that can handle subjective topics well, rather than try to shoehorn a traditional forum structure of "anything goes"

The advantage of the 'traditional' Q&A site has been that the format, for most part has been designed to minimise noise. Quite a few original design decisions, were focused around that, and site norms have evolved over time - we have Product Recommendation sites (for software and hardware) - which are arguably subjective, that probably wouldn't have existed before.

We also have fairly dedicated sites like workplace and IPS that cover subjects that by nature that are subjective.

I find in many cases, that these sites often have fundamentally different dynamics from more traditional Q&A sites. There's a little more drama, and a few issues we'd never face.

Many of the sites often need to find their feet as far as scope goes - SR had the "You're doing it all wrong" post for example. And while I find "what's the best tool" to do "bar" makes for a terrible question - there's little loss when you're getting answers that include the process and not just the tools on more traditional, purely non subjective sites.

In both cases - there's never been a total ban on subjective posts- there's been rules of the road for these since 2010, and many of these work brilliantly even on sites which are purely subjective.

Answer 15

Stack Exchange is by now a failing product that doesn't serve the needs of those asking questions, but rather a handful of 20k+ rep members gaming the system to feed their own egos. It's like Facebook and the toxic social media only 10x worse.

I've been calling on the site to fix itself for years and only received downvotes, closes/on-holds in exchange.

The best way to "encourage" Stack Overflow to change is to create an alternative, preferably one where there are no downvotes and moderators are held to account for every single mistake.