What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 134 Stack Exchange communities.

Every day, usually early afternoon my time, (that's about 0200UTC or 2100 EST), streams of really low quality questions start appearing on the site. These are frequently of the 'give me teh codez' variety, but they also include wondrously unanswerable examples such as this, or they're completely off-topic. They don't appear before that, and they usually stop by late evening my time (1200UTC or 0700 EST). Judging by the timing and the names that appear (where they are given) all these questions are originating in the same part of the world.

These questions are frequently posted by new users, and are usually quickly downvoted and closed. Rarely is there any response to requests for clarification.

As a regular contributor to the site I find that a lot of my time is consumed just fielding this junk. Is there some way that the quality filter can be raised, possibly on a geographical basis, to stop these questions appearing in the first place?

share|improve this question
I don't think any kind of geographical or country filter should apply. It should be sorted out by other moderation mechanisms. –  Szymon Jan 30 '14 at 1:11
Three downvotes already. I'd be interested to know the reasons behind those. –  Hobo Sapiens Jan 30 '14 at 1:16
@Szymon I don't quite agree - I think there's a good chance you're right and it would be toothless, but it is totally statistically possible that such a filter would be effective. I'm pretty sure it isn't a good idea, but it has the theoretical potential to work. –  Emrakul Jan 30 '14 at 1:16
@MikeW if you weren't aware, downvotes on meta sometimes mean disagreement with proposed feature requests. So I'd say people are disagreeing with your idea –  psubsee2003 Jan 30 '14 at 1:18
I downvoted it as I strongly disagree with your proposal of geographical filters. Also, I don't like putting hints without saying what you mean: for me you mean posts from India and you should have said that. –  Szymon Jan 30 '14 at 1:19
@Emracool I don't think SE should go down the path of dividing people based on where they live. That seems discriminatory, doesn't it? –  Szymon Jan 30 '14 at 1:31
I downvoted your suggestion, for the same reason @psubsee2003 wrote in the posted answer below (before I had scrolled the page to read that answer - time delay due to looking for the following link). In Raymond Chen's words: This is a global solution to a localized problem. A few low-quality posts doesn't make an entire geographic region or time period filter the proper solution. –  Ken White Jan 30 '14 at 2:15
@psubsee2003 I'm aware of the downvoting mechanism on Meta, but a simple downvote doesn't really spur a discussion. If people disagree that's fine, but I like to know why. It might (and in this case, has) modify my thinking. –  Hobo Sapiens Jan 30 '14 at 3:37
@Szymon I deliberately didn't specify India. There are a number of countries that fall broadly into the same timezone. Maybe my perspective is coloured by the timezone I inhabit - I don't see what goes on during my night, which is daytime for much of the Western world. –  Hobo Sapiens Jan 30 '14 at 3:41
While I don't believe geo based filters or time based filters are the best approach, I've noticed this exact same issue from a certain part of the world that I will not name and it is infuriating! +1 for having the courage to come forward and raise the issue. There has to be a way to come up with a solution that is more general in nature and can help users from around the globe improve their questions and thereby increase the quality of posts on SO without taking our time trying to filter through junk. –  Sly Raskal Jan 30 '14 at 4:34
I agree with this. People don't want a area filter, and that might be bypassed anyway. A time filter is pretty good, since the level of moderators goes up and down depending on time. A better solution, tho, is having every moderator be a anti-flag with a very low flag weight (0.000001 x normal). (Or the number of flags varying with # of moderators on, same thing) –  bjb568 Jan 30 '14 at 4:42
FWIW, here are recent examples of especially prominent crap spikes at Programmers, conveniently documented in the form of complaints in chat: 1) "...7 crappy questions in an hour (there was 8th one but asker removed it)" 2) "seven low quality questions landed in an hour..." –  gnat Jan 30 '14 at 4:58
@scrblnrd3 I set the title to what I wanted it to say. You are entitled to disagree, but your repeated edits setting the title to someone else's interpretation are unhelpful. If you have a contribution to make, please do so. Otherwise, feel free to butt out. –  Hobo Sapiens Jan 30 '14 at 6:14
Let's cool it with the edits - there's nothing deliberately inflammatory in the title (or question), Mike has his flame retardant suit on, so let's respect the point he's trying to raise. –  Tim Post Jan 30 '14 at 6:22
Just wanted to make a casual point that if we didn't have as many low quality posts, perhaps our close vote review queue wouldn't be so high. So there is some value in trying to improve how newcomers use the site initially. Cheers. –  Sly Raskal Jan 30 '14 at 6:34

4 Answers 4

You aren't the first one to bring this up.

It's incredibly difficult to ignore patterns that are so resoundingly repetitive - we do get a lot of crap questions when the sun is shining on the other half of the world. It's not a language barrier, but generally a heightened sense of urgency that contributes heavily to this problem.

However, it's not confined to the other half of the world, it's simply more noticeable because a fundamental lack of competence is more easily obfuscated by someone's ability to more elegantly articulate themselves, or (for the blessed few) to show some evidence that they've read over our customs and culture prior to asking their question.

What you're seeing is a shift in the craft.

People are trying to become programmers not because they discovered a love for programming, but purely for the economic and/or social status gains.

Sure, these people might live on the other side of the world - but you're not seeing the patterns in the pattern, or the forest for the trees --

  • They don't possess clever problem solving skills
  • They don't possess the curiosity that kept most of us awake into the wee hours of the morning working on a problem for the pure satisfaction of figuring it out
  • They don't look at problems as a programmer because they simply are not programmers, they frankly lack the natural talent for the craft that many of us have. They're never going to be good programmers.

Location is pretty much incidental, we're talking about talent.

That's not something we can fix with a quality filter, however complex. Stack Overflow tends to mirror the state of the industry as it is or was at any given time since we became popular, that's why so many researchers study it. If you stop and think a bit, I'll bet you agree that a good portion of the strong objection you feel toward these questions is actually a bit of angst about the future when it comes to the craft.

While I'm not a huge fan of our automated blocks, they do give people a few chances to show that they can think like a programmer by asking questions demonstrating that they know what a programmer would need to answer it. I don't like these because they don't sufficiently account for how quirky our engine and conventions can be for new users, and that it can take months for people that do have talent but not much skill to dig out from under a bad start.

We're already doing what we can about this problem, we're forcing people that are capable of getting better to get better, and those are the only people we can help. It's them, the pearls, that we must optimize for; algorithmic tweaks just treat the symptoms. In democratic Overflowistan, diamonds mine you.

We've taken so many surgical strikes at this in the past in hopes that something small would result in a big gain - and gained a bit of ground in the process. However, I think we need to face the reality that we're looking at holding back the tide with a bottomless bucket, and change our strategy to find the real talent in that pool and help it flourish.

More stringent quality filters aren't going to help in the long term, they're just going to allow us to ignore the cause for a little while longer.

To your actual request:

I think we should work on improving the question blocks and just-in-time help instead.

share|improve this answer
This answer brought to you from the other half of the world. –  jmac Jan 30 '14 at 6:47
"Location is pretty much incidental, we're talking about talent." - I fully agree. I work with a lot of people who moved from India to develop software here, and several who are still in India. Most, if not all, of those specific people are very competent, skilled programmers. Just because there are a lot of not-competent programmers in that area doesn't mean we should blot out all questions coming from there. –  jadarnel27 Jan 30 '14 at 6:58
This explanation of root causes makes a lot of sense, particularly if you think about localizing not by geographic region, but also by time or subject matter. In the dotcom boom, you saw a lot of the same behavior in the U.S. from people getting into programming for the "wrong" reasons. The mobile development tags on SO also have a lot more of these problems than other tags, due to the gold rush in Android, iOS, etc. When programming is just a job, and you're only there to punch the clock, it should be no surprise that you'll not put the same amount of effort into your questions. –  Brad Larson Jan 30 '14 at 18:28
@BradLarson It's not just a question of effort, I think some people really do try, they just don't have the kind of abstract problem solving skills required to determine what they really need to know in order to move forward. To a programmer, it's a question of just filling in blanks. To someone that simply doesn't think and operate well in the abstract, it's just one big freaking blank, and their job / grade is riding on it (hence the heightened sense of urgency that compels them to continue trying the same thing over and over until they luck into an answer). –  Tim Post Jan 31 '14 at 7:23
Maybe a tiny IQ test should be added, instead of a captcha before posting questions in SO, Programmers or similar. Like solving a tangram or something. –  mavrosxristoforos May 2 '14 at 10:13
@mavrosxristoforos IQ is simply a rough measurement of your ability to learn. It's more of how you can think, not if you can think (though, arguably, if is a prerequisite for how). I'm a very competent programmer and I stink at logic puzzles. –  Tim Post May 2 '14 at 11:05
@TimPost Then maybe a small programming test. Like "iterate from 1 to 10 to post this question". I believe for-loops should be a prerequisite for SO :-) –  mavrosxristoforos May 2 '14 at 15:09

Even if it was a good idea, geographic filters simply won't work. Localization via IP address is too easy to get around so it won't be reliable, and will become even less so as people learn that there is a geographic filter in place.

But ultimately, I object to the idea in principle. Instead of designing a filter to target specific problematic individuals, you are effective deciding an entire region of the world should be filtered out because some people in that same country are subpar programmers.

Don't get me wrong, I have no problem if a reliable filter can be designed to prevent poor quality questions, but you don't throw out everything because of several bad apples, you try to save what you can first.

Similarly, I'm not a fan of time-based filters either, although they are less objectionable because you are theoretically filtering based on when you have inadequate moderation available to handle poor quality questions. The problem is the quality filter is not perfect, has too many false positives, and does not always provide adequate feedback, so if you turn it up a bit, you'll end up filtering more reasonably average posts.

share|improve this answer
Re your first paragraph you could argue that if someone is too clueless to figure out a really simple programming problem, they won't be able to figure out how to use a proxy while using SO. :) But I agree with you on the rest - improving quality filters overall should be the way to go. –  Pëkka Jan 30 '14 at 4:17
@Pëkka, you severely underestimate the ingenuity of today's common idiot. –  Charles Jan 30 '14 at 5:16
@Pëkka it's a simple browser extension - this is how I remain addicted to Hulu & Netflix. One click, it's stupid easy. –  Tim Post Jan 30 '14 at 8:16
@Pëkka my point there was when it becomes common knowledge, someone is going to advertise methods to do it, and everything will be able to do it. –  psubsee2003 Jan 30 '14 at 10:03

I understand your frustrations, believe me. Try hanging out in the review queues between 0100 and 1600 or so GMT, and you'll see much of the same thing - a lot of users that don't understand how to communicate and how to ask questions reviewing submissions from other users that don't understand how to communicate and how to ask questions. It's infuriating, and it almost makes me want to ragequit the review queues. Almost.

This is not a problem unique to the users that are awake during these hours. Bad askers exist in every language and every timezone. Have you gone through the hoops that SO makes you go through when you ask your first question now? There's an outrageous amount of hand-holding. Any reasonable, responsible person is going to take notice, pay attention and do it right.

Those types of questions happen when people that don't care and also don't know how to explain their problem try to ask. They're helpless, clueless, and probably can't be saved.

You only notice them more because of two reasons:

  1. When they're awake asking questions, a lot of other users aren't on the site to help edit and close, and those that are on seem to have lower standards.
  2. A poorly-formatted, grammatically incorrect bad question looks initially worse than just a normal bad question, even if it'd be just as unanswerable if perfectly formed.

Outside of the real-world implausibility of geo-IP filtering without huge side-effects, any sort of geographical or time-based filtering mechanism is more likely to cause problems than not. Also, to be very frank, you weren't here before there were quality filters. Be glad that it's as good now as it is. There were also a lot of false positives, and we still see them weekly.

The bar does need to be raised on questions with offensively bad logical quality, but an algorithm isn't likely to help here.

If the question can be saved, edit it. Hell, even if it can't be saved, edit it until it's not a mess. Then downvote, optionally closevote and move on. That's all we can do for now.

share|improve this answer

I quite strongly disagree with an idea of geographical filters or different requirements based on geographical location of users. This is a discriminatory policy and, in my opinion, goes against the spirit of SE communities.

Bad posts are added by users from all countries. The only thing that is commonly different is the level of English.

Bad posts should be sorted out by the moderation mechanisms and there's no need to single out users from some countries here as well.

share|improve this answer

You must log in to answer this question.

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