Asking Questions Should be Easy

Context

We often come to the Internet looking for information we need. If that information is mostly generic and popular, we can usually find it through simple Google searches. However that’s not always the case. Sometimes we have questions that can’t be answered generically, because it’s based on a personal context. Such contexts can be hard to be translated into a search phrase. Besides, search engines are designed to link your search phrase to the relevant sites; and not meant to understand complex questions and find specific answers. It’s also possible that those specific answers do not yet exist in web. So search engines aren’t always the answer.

For such complex, nuanced or personalized questions; what we end up doing is looking for sites or forums that specialize in that subject. If we are lucky and skillful we might find such forums and be able to ask our question. But that’s not always the case. First, there are plenty of topics that do not have a forum dedicated for them. Either because there are not enough people to sustain such a site or simply because, no one took the initiative to do it yet. Second, the policies and moderation in these sites often create an atmosphere where genuine questions can get downvoted or blocked.

Whatever the reason might be, it’s been my personal experience that it’s generally hard to get answers to one’s personalized questions from subject matter experts within a reasonable amount of time and effort.

Out of the existing solutions that are popular today, there are two primary categories.

Sites that are meant for very specific questions. Example: stackoverflow, codereview and other stackexchange family of sites.

Sites that allow more general questions. Example: Quora.

Unfortunately, neither of these categories is serving the need of a big portion of the user base. Here is why:

What’s wrong with https://Stackoverflow.com?

Stackoverflow has a number of portals that specializes in various domains. Although Stackoverflow itself might be the most popular one which deals with programming related topics, there are other sites that go under StackExchange.com with specializations in subjects like Physics, Software Engineering, Design etc. But they all share the same kind of policies. And these policies have issues:

If you are familiar with let’s say StackOverflow you probably noticed a a lot of the questions it has are quite popular with a lot of up votes for both the question as well as for the answers and yet they are marked as off topic and subsequently closed. Many of these are off topic not because they are not programming related but due to the nature of the question. Anything that doesn’t have an objectively agreeable specific answer is off topic to StackOverflow. The rationale is that it would attract subjective opinions and therefore will be hard to maintain the quality of the content.

Problem 1: Too Restrictive Policy

But here is the problem. Many a times, people have questions that are broad due to the fact that they do not understand the context well enough to ask a series of specific questions. Remember they are still asking very technical question and not really looking for subjective opinion at the end, but they can’t quite form objective questions because they know so little. This explains why such a large portion of the questions are popular and yet fails to qualify for StackOverflow. I think this is a problem that needs to be solved. Stackoveflow is addressing a need, but there are other use cases that are not addressed. So they might be right in their business model, it’s just that we need something more. They are implementing their policy well, but the policy itself is not adequate to solve the issue I just described. So that’s the first problem.

Problem 2: Difficult to Use

The second problem is that a big number of the questions are closed because the user didn’t ask them in the right portal. May be it was a software architecture question and they asked in the programming forum. The culture seems to be that the questions gets closed and not just reclassified to the correct topic. I have seen instances where a user was bounced back and forth between groups with both moderators telling them to post on the other forum. It was crazy. And unacceptable. Questions shouldn’t closed because they are off topic. Instead they should be just forwarded to the right forum. So this is the second problem.

Problem 3: Unfriendly Culture

The third problem is actually much severe and wide spread. It is something that seems to have been implanted in the DNA of this site. I call this intellectual bullying. There is a very caustic culture of mocking or outright insulting attitude from a subset of the experienced users (with lot of points) towards newbies. Questions get downvoted and closed out left and right because the quality of the question was not good, because they asked something stupid or they didn’t do their homework before asking. I see the rational here. But more often than not, I sympathized with the asker. Many a times it’s hard to conclude whether they actually didn’t do homework or just genuinely lost. My question is, why should we not give them the benefit of doubt? Because if they are genuinely in need of the initial nudge to the right direction, it is something we could easily do. They need good answer to their stupid question, not insult. This current culture in Stackoverflow.com seems wrong to me on so many levels. And this is the most central motivation for me to make a better place for helping people with questions.

I think when people ask questions, they implicitly admit their lack of knowledge. Sometimes that lacking might be big. I also think, asking a good question on any topic is quite hard. Given this, isn’t it actually counter-intuitive to expect high quality questions from people who are asking about something to understand it better? I sure think so. I think stupid questions are the most productive ones. Without asking them one sort of risks remain to be stupid. I don’t buy that everyone starts as a genius and asks all the right questions from day one. I believe we should flip our mindset on this. Don’t downvote stupid questions. Rather try to really help that person. It’s about helping one another at the end, and not establishing intellectual superiority over one another.

I think a culture of helping is severely lacking in the StackOverFlow community today. And this is the strongest reason for me to say, Stackoverflow is good, but not enough.

What’s wrong with Quora?

Quora is much better. It doesn’t have a lot of the problems that StackOverflow.com has. Yet it has its own list of issues making it fall short of being a place where questions can be asked and answered effectively.

Here are the major ones:

Problem 1: User’s lack of Control over their own Content.

Users do not have enough control over their content. If you ask something, the Quora automated moderator bot could mark the question as “need improvement” resulting it to be not available for others until you make changes. There have been many instances both with me personally and with other people when the question actually was completely grammatically correct and meaningful and yet marked as requiring change. It was obvious the Artificial intelligence algorithm failed to make sense of a sentence that a human could do without any problem. This is a very annoying problem. Users should have the ultimate say on what the language of the question should be. If it’s really bad, it will receive less attention organically. Moreover, even with using AI, there should be a way to bypass it when people reconfirm that the question looks good. Quora currently doesn’t do that. It leaves the situation at the mercy of their automation which clearly isn’t smart enough.

Then there are cases where the questions gets automatically merged with an existing question. This can be also very undesirable. Many a times questions that were asked before might need to be re-asked with a newer context. A question might not get the same answers if it were asked three years ago vs today. So having an existing question that is similar to the one asked shouldn’t force the question to be merged with it. One can also ask it to get a fresh new look at the topic, because may be the older question didn’t get many good responses anyway. And the algorithm itself that judges similarity often fails to detect subtle but important nuances to the questions that make them practically different questions but get labeled as the same. This again happens all too often and without any consent of the author of the question whatsoever.

Problem 2: Questions can’t have any details!

This is a very strange change made by Quora last year. For a site where the primary use is to ask and answer question, it is an understatement to call this weird. Yes you can’t add any details to your question. Everything must fit in the couple of lines of text. Forget about adding any images. The only thing you can add is a link.

They do have a rationale behind it. Like StackOverflow’s policy, it has a meaning, but not really the most helpful one. They want “canonical questions”. What it means is, questions must be generic and objective enough to not require too much details. They somehow want to formalize the knowledge content in the questions and answers in a way so that everything is captured only once in one place. Variations of the same underlying questions can’t be accommodated. It’s a nice idea. But not practical. Not at least for people who have specific questions for their specific situation. Because most of the time the devil is in the details. We can also say, things that are so generic, can perhaps be found by just doing a simple Google search.

So, while I understand the thought process behind this rather strange decision, I again think it leaves a vast portion of the questions unsuitable for it.

Problem 3: Doesn’t have the goodness of StackOverflow

With all the issues I have listed above, I didn’t say it was all bad. It is quite good at few things. One of them is that, it has a lot of answers that are very specific answers to specific technical problems.

With Quora the context is much more fuzzy. All sorts of questions are asked there. If I ask a question like “How can I implement a Breadth first search algorithm using javascript” I won’t expect many clear to the point answers. Quora is good for higher level, and less technical questions. The more technical it gets the less likely is to get good to the point answers. It’s too wide open.

What I want is the best of both world. We should be able to filter our questions and answers based on how objective and technical we want them to be. The system should allow such kind of filtering and not mix up too high level topics with very domain specific ones.

Problem 4: Lack of Good Answers to Questions

It takes real effort, time and energy to write good answers. Often times, there are people with credentials who can write good answers but just don’t have enough incentive to do so.

Quora’s “canonical question” is also an attempt to address this. If questions are very generic, it can appeal to a much wider spectrum of the user base. So writers have a motivation to answers those for the reward of popularity. Popularity is a strong magnet. Rightly so. And this strategy of canonical question works too, at least to some extent.

But what happens to those questions which are important to a few hundred people or may be a thousand but not to hundreds of thousands? Or what if it’s very important to only a few people but they really want an answer – to a question that is very answerable? How can they get the attention of the target audience with the right qualification? In Quora, they simply can’t.