Hi, @Shog9, I was kind of curious about some issues with the too broad closing reason, and was curious if you had a second at some point to give me some guidance.

I would like to pursue the theory that the too broad closing reason is not working as intended, in that it closes questions which are in fact not overly broad, which leads to not only being counterproductive but also being not welcoming (even though it was designed to be more welcoming).

I was curious if you could perhaps give me an idea of what type of metrics I should research as far as evidence of my theory.

Anyway, feel free to at your leisure, or not. I will see the ping or check back in.

@TravisJ ...I guess I'd have to hear more about your theory

I have Thoughts on the nature of Too Broad

but not really sure if they're aligned with what you're looking for

Sorry, but my thoughts are kinda rambly. Please don't feel the need to respond immediately if you are otherwise disposed.

@Shog9 I am okay with being wrong if that is what you meant. Perhaps your thoughts will enlighten me, so if you think that might at least be on topic where the topic was "is too broad functioning", then I am all ears. By relation, I can ask on mse if you want as well, but I was kind of hoping for something a little simpler at this point.

As for my theory. I will only briefly touch on where too broad came from, since I am sure you are intimately familiar with it. NARQ and Too Constructive were seen as "not welcoming" as they didn't provide enough concrete feedback for users who had their question closed; as well as being seen as being used improperly from time to time.

Too Broad was supposed to therefore provide concrete feedback, as well as being properly used (in addition to no MCVE which I will also address here). However, while that was a good goal, I do not not think that is the way it is being used.

I see far too many questions get shut down as a result of needing an answer longer than a paragraph. "That's too broad", the closers say, "it takes longer than a paragraph". Moreover, if it is code based, and doesn't have an entire running example, it is either "too broad" since it might require some expanded explanation or "off topic" because the code is not enough or "wrong exchange" because some users are even convinced that code is not longer on topic at Stack Overflow.

Bite-sized, refactor, fix this small thing, must have an mvce which is also bite-sized - this mentality... this had led us into a position where questions are so pigeonholed that it is nearly impossible to provide a question of substance. "Boring", or "too long, nope" is a common response to questions which are not bite-sized.

There is a pervasive outlook that somehow providing code or explanations is not what we do at Stack Overflow. In my opinion, this outlook has been encouraged by the "too broad" and "no mcve" reasoning. This has led to users shutting down situations which require perhaps partial tutorial with code. We have 30,000 characters to answer, and there is clearly a demand for code.

Not just "here do my work" or "here do my homework", but "here is a legitimate problem that I am facing that many other people will probably face too". The third category is in strong need of help, and is being failed miserably by Stack Overflow.

The greatest and easiest example of this being fact is the example of GitHub's growth. It is just pure, donated code. An unbelievable amount of code. All placed there for the public to see and peruse. 29 million users producing code. The idea that there isn't a supply of users that are willing to write code to solve problems is clearly an idea that needs to go. There is. It's right there.

We need to allow it to flourish at Stack Overflow. In order to facilitate that happening, we need to be able to allow questions which may take some code or explanation that is in depth to answer; and which may be generated by a question which is not entirely specific on the exact solution required so much as which defines a problem that needs solving.