> Maybe some questions are just not "too broad" but simply broad, and that one may wish to have a number of possible solutions to a problem and pick the one that fits one's case best

I think there are two kinds of "broad":

1. There are multiple ways to solve a specific problem with a specific expected outcome. "How do I sort an array?", for example, is a broad question in that it can be solved in many ways, but you can definitively say whether any given solution is correct or not.

2. The problem is not specific, and there is no clear way to determine whether any given answer is actually a "solution". For example, a single question with only the text, "How do you write a game?"

Most "broad" questions fall on some spectrum between these two, but the closer a question is to the latter category, the less likely it is to generate any sort of helpful feedback. The community does sometimes label the former as "too broad", but 95% of the time the given close reasons are correct, and when they are wrong the site provides documentation on how to get your question re-opened.

It's hard to maintain a proper balance between moderating the inundation of terrible questions and avoiding the false-negative closing of legitimate questions.

I'd love to see the SO community's answers to "How do you write a game?".

I think this is an interesting and common response to the policies. As if the "Too Broad" close policy is the only thing stopping volunteer contributors from writing a 10 page thesis "How To Write A Video Game" for every questioner that asks for it.

Isn't this alleviated somewhat by marking duplicates? If nobody ever writes any good answers on the original question, oh well! It's just annoying when nobody can write a good answer to a broad question.

But hey, it's their site, and I'm just some rando on the internet - they get to make whatever policies they want.

My emphasis was less on the replication of effort, which duplicate tagging definitely resolves, and more on the idea that broad open-ended questions raise the bar for a well-qualified high-quality answer far above and beyond that needed for a simple and concise question of limited scope.

Allowing such questions effectively lowers the percentage of questions that get good answers, thereby training users NOT to expect an answer to a posted question, and leading to decreased engagement.

On the other hand, by disallowing such questions, the questions that do get asked are more likely to be answered, and this encourages repeat visits and increased engagement.

Hey, by the way, in case you happen to come back and see this after the fact: I've found your perspectives in this thread compelling, and while I'm not completely convinced, I'll definitely be incorporating your arguments into my thinking in the future.

Just thought I'd let you know you haven't been shouting into the wind!

My personal experience with such questions is that very often, the process of engaging in dialogue to determine what kind of answer(s) the asker actually wants is difficult and time-consuming. This typically complicated by askers who don't actually show up to the dialogue, leaving people who want to be helpful with no way to figure out what specific answers are helpful. There tends to be no adequate way to helpfully answer such questions, so people try to filter them out as poor uses of time.

The scenario you describe is somewhat different, in which the asker is well aware they have a broad question. Such people are advised to make clear that they know full well it is a broad question and are seeking broad comment on multiple viable approaches.