15
$\begingroup$

I've noticed an alarming increase in the number of highly-upvoted open-ended questions that just aren't good. Here's the problem I'm seeing:

  • Nobody learns anything from them. The answers are completely uninformative, and present no new concepts or ideas. Compare this to a site like code golf, where good answers incorporate paradigms of programming and require a lot of effort to shrink down, thus presenting new ideas to the reader.
  • Our job here is to optimize for pearls, not sand. These questions are failing this philosophical goal -- just look at the answer count, and try to figure out how many of them are actually useful.
  • Two syndromes appear: "Here's what I got, but I have no idea if it's actually optimal." and "Here's what I got, but I'm not going to explain how I got there."

What this results in is -- quite literally -- an unbounded number of possible answers that are low-effort. Here are some examples of what I'm talking about:

We need restrictions on these questions. These questions are generating some of the least informative, worst answers I've seen on any Stack Exchange site. The questions, though, are asking for answers like that, and everyone is upvoting them which only encourages it further.

We need to stop this problem cold, because it's growing increasingly worse.

Here's what I'm proposing to fix this:

  • Delete any answer which doesn't contain an explanation or justification as to why it could be the optimal solution. That doesn't mean that it has to be the optimal solution or that the justification has to be complete, but rather that you can't just arbitrarily guess "Mississippi gets a score of 14, I think?" and roll with it with no justification.

    As a user, flag these answers when you see them.

  • Put any question which actively encourages too many answers as 'too broad' until the author can limit the scope of possible answers to something reasonable.

    If the author of an on-hold open-ended question doesn't elect to improve their question within a reasonable time, delete it. It's just not useful to anyone in that state.

Feel free to answer as well if you have other ideas on how to fix this problem, or refinements on what the problem actually is.

$\endgroup$
17
  • 2
    $\begingroup$ Would "lateral-thinking" questions that become a cesspool of poor answers or answers specifically trying to circumvent the correct answer also fit in with this category of questions you're looking at, or are they another matter entirely? $\endgroup$
    – Bailey M
    Commented Jul 17, 2015 at 19:27
  • 2
    $\begingroup$ @BaileyM I'm not sure, and I think it would depend on the question. In general, though, I don't think so. I think those count as story questions, which should also be closed imo, since nobody can seem to find an example of a good one. $\endgroup$
    – user20
    Commented Jul 17, 2015 at 19:30
  • 2
    $\begingroup$ I think the comment "I joined Puzzling just to upvote this" says a lot about how those story puzzles are received by the community, haha. $\endgroup$
    – Bailey M
    Commented Jul 17, 2015 at 19:37
  • $\begingroup$ "Delete any answer which doesn't contain an explanation or justification as to why it could be the optimal solution." Could you please give an example of how such an explanation might look? I'm having trouble imagining one in cases where it's not provably optimal. $\endgroup$
    – xnor
    Commented Jul 17, 2015 at 23:58
  • 3
    $\begingroup$ Also, are there any statistics we can get about how many of the bad answers and votes come from people coming in from HNQ? I'm not sure if the questions with many answers are themselves bad, or if they feed the flawed hotness algorithm on HNQ and bring in masses of users outside the community. $\endgroup$
    – xnor
    Commented Jul 18, 2015 at 0:05
  • $\begingroup$ @xnor Unfortunately no stats from HNQ. As far as an explanation... it's going to depend on the question, but even something minimal like "I started with the longest word in the dictionary and worked from there" would be helpful. $\endgroup$
    – user20
    Commented Jul 18, 2015 at 1:12
  • 1
    $\begingroup$ @Emrakul In that case, doesn't requiring an explanation fall under the existing policy? $\endgroup$
    – xnor
    Commented Jul 18, 2015 at 1:21
  • $\begingroup$ @xnor Yeah, it does, but I wanted to make sure people were on board with it before going through the previous open-ended questions. $\endgroup$
    – user20
    Commented Jul 18, 2015 at 1:24
  • 1
    $\begingroup$ I think a lot of the puzzles you list could be closed as duplicates of each other - taking an arbitrary sequence and asking for the insertion of operands is a superficial change, and the form of the answers don't change much due to such changes. (The trouble is that I can imagine writing such puzzles in a way to suggest particular solutions, allow arguments restricting the form of optimal solutions, or yield easy proofs of optimality - and these might not look too distinct - though the author could give a hint which could convince us of the puzzle's uniqueness) $\endgroup$
    – Milo Brandt
    Commented Jul 18, 2015 at 15:36
  • 2
    $\begingroup$ Funny that you make the pearls versus sand argument considering a pearl starts out as nothing more than a grain of sand before it slowly grows into a pearl. I think that the site benefits from having more voices and more questions, even if I am irritated by some of these questions which are little more than cryptography homework. It's not my place to tell others which puzzles are valid and which are not; and if I dared to tell the obvious majority of a site that their opinions and thoughts were so, I'd be an embarrassment. $\endgroup$
    – Kingrames
    Commented Jul 22, 2015 at 17:20
  • $\begingroup$ @Kingrames The difference between Stack Exchange and a site like Yahoo Answers is quality control. This isn't about whether anyone likes these questions (clearly people do) as much as it is whether they're a good fit and of high quality. It's possible to ask both basic and complex crypto questions in both a high and low quality way. Moderation -- both by us and the community -- says far less about the content of the question than about its quality. That's what the metaphor is trying to communicate: we're given a bunch of sand, now optimize for the ones that become pearls. $\endgroup$
    – user20
    Commented Jul 22, 2015 at 17:25
  • 2
    $\begingroup$ I do not believe that deletion is an appropriate form of pearl growing. You claim that you're optimizing for the "ones that become pearls" but that's just what happens to any grain of sand that is given room to improve and the conditions for success. Those questions should not warrant deletion, but rather more feedback. It is far more effective to water a garden, than to remove dead brush alone. It's a given that weeds are removed; You haven't convinced me that you're targeting exclusively trollworthy posts. $\endgroup$
    – Kingrames
    Commented Jul 22, 2015 at 17:30
  • 2
    $\begingroup$ @Kingrames I'm going to drop the metaphors because I don't think they're clear enough. It's the job of the community to determine what does and doesn't work here. When we find a category that doesn't work well on our site, it's our job -- the community as a whole -- to mark it as such through closure until it is improved, and maybe deletion if it can't. Question closure is feedback, and if the user doesn't come back to improve their question, it's not going to get any better. There are automatic deletion scripts in place to find such questions. (Responding here because chat fell through.) $\endgroup$
    – user20
    Commented Jul 23, 2015 at 5:06
  • 1
    $\begingroup$ Considering that the comparison being used here is with Code Golf, I find it curious that the example of "All those questions about filling in the operators to make a number" was given as a bad thing, considering that it's pretty much the puzzling equivalent of a Code Golf question. There might be multiple answers, but there's probably only a few good ones, and the concept of "best" is well-defined. And if people are providing low-quality answers, then that's on the answerer. Especially my one, which was making pi, the answerer was providing ideas on how to make numbers. $\endgroup$
    – Glen O
    Commented Aug 9, 2015 at 11:08
  • 1
    $\begingroup$ "fill in the operators" is a common kind of puzzle (not just on the site), it's absurd to complain about it being on here just because some people give bad answers. $\endgroup$
    – Glen O
    Commented Aug 9, 2015 at 11:09

2 Answers 2

Reset to default
6
$\begingroup$

An alternative proposal

We should be encouraging questions that capture the creativity that is "Puzzles". This doesn't mean that we should allow any puzzle, no matter what... but let's not throw the baby out with the proverbial bathwater.

As such, here's the metric I propose for which questions should be deleted:

  1. If an answer doesn't provide either reasoning or creative alternatives, it should be deleted. This means that a solution that looks at the problem from a different angle is kept, whereas informing the questioner of the standard answer without explanation is removed.

    This means that, if the question is "What is black and white and red all over?", you'll delete "A Newspaper", but accept both "The traditional answer is 'A Newspaper' because of 'red' and 'read' being homophones. The question is meant to be spoken" and "A Zebra with sunburn" as answers (although explanations should still be encouraged for creative answers irrespective of whether they are 'obviously' satisfying the puzzle). None can deny that the latter is a valid answer, that it (was, when originally given) creative, or that the question itself is very broad if you allow creative answers. Which brings me to point 2...

  2. If the question is broad because of creative alternatives, it should be kept. If it is broad because of poor design, it should be deleted. The way to tell the difference is that the "right" answer should be obvious when compared to a "wrong" answer. This doesn't mean that one has to know that it's strictly the "right" answer.

    Note that the "right" answer may actually be given in the question itself, thus providing the obvious solution off the bat. A great example, that was closed and, in my opinion, shouldn't have been, is this one. Answers provided demonstrate a wealth of creativity that can be of use to people who find themselves in the position the questioner found themselves in. Various answers provide examples of loopholes that one could plausibly exploit, and that is of value to readers.

    By comparison, look at this one, which deserved to be closed due to broadness. Yes, there's a "right" answer, but it's not especially creative, and alternative solutions aren't, either. There's nothing to really be learned from either the question or the answers.

  3. Trust the users to know particularly good or bad questions. Questions that are highly upvoted are upvoted for a reason. The question serves a purpose to those people. They judge it as a quality question, for whatever reason. The above metric only applies to those in the midrange - a few net upvotes or downvotes, but not a lot. Note that this isn't to say that we shouldn't close popular questions... just that broadness isn't a justifiable reason to do so.

Using this metric, the concept of a "pearl" is a little different from that proposed by Emrakul. But I do believe that, like with Codegolf.SE, this site should prioritise creativity, rather than formality. Broad questions aren't inherently a problem for sites such as these, because they permit people to show alternative approaches to the same problem. On Codegolf.SE, the top-voted question has 156 different answers! It was the very definition of broad - all it said was "write code that produces the number 2014 without having any digits in the code". But that broadness allowed the question to produce a great variety of answers, all of which were capable of teaching people something about the respective language and/or programming in general.

Note that Codegolf.SE have had arguments in their meta about categories of challenges that tended to be more about people using well-known exploits and loopholes, and they dealt with those as specific issues. They didn't try to blanket-ban questions that were broad because of some people misusing broadness as an excuse for laziness.

My point, with all of this, is that if a question can inspire hundreds of answers, but they're all interesting and thoughtful, then such a question should be welcomed, not rubbished for being too broad. This isn't like math.SE, where questioners are seeking concrete facts, and broad questions aren't sufficiently well-defined.

$\endgroup$
9
  • 2
    $\begingroup$ +1 right, creativity and diversity are a good thing. $\endgroup$
    – A E
    Commented Aug 10, 2015 at 13:58
  • $\begingroup$ In bullet point one, I'd recommend deleting an answer that is simply "A zebra with a sunburn" for the same reason we delete "A newspaper". And instead keep "A zebra with a sunburn because zebras are black and white, and the sunburn makes it red" or something like that. It seems inconsistent to keep an alternative answer without an explanation. $\endgroup$
    – Brian J
    Commented Aug 12, 2015 at 17:39
  • $\begingroup$ @BrianJ - I meant that it was an answer that obviously fulfills the question (without needing any further explanation). I would think that anyone could understand a zebra is black and white, and that a sunburn is red. Explanations are needed where it's not obvious. Of course, in practice, you'd expect more than just "a zebra with a sunburn", but mostly because you wouldn't expect such a simple question... and because "a zebra with a sunburn" isn't really a creative answer any more - it's a standard answer. $\endgroup$
    – Glen O
    Commented Aug 12, 2015 at 18:29
  • $\begingroup$ @BrianJ - note that I've edited in a bit of clarification, that we should still be encouraging explanations in that case... just that we shouldn't be jumping to deletion for the creative, obvious answers, as it drives away users who are creative but aren't used to the expectations of the site. $\endgroup$
    – Glen O
    Commented Aug 12, 2015 at 18:32
  • $\begingroup$ @GlenO I think that covers my concern. Thanks! $\endgroup$
    – Brian J
    Commented Aug 12, 2015 at 18:38
  • $\begingroup$ What's black and white and red all over? A nun who was pushed down the stairs. xD $\endgroup$
    – moonbutt74
    Commented Aug 20, 2015 at 23:18
  • 1
    $\begingroup$ @moonbutt74 - I hope you're not in the habit of pushing nuns down stairs. :P $\endgroup$
    – Glen O
    Commented Aug 21, 2015 at 5:14
  • $\begingroup$ old catholic school joke, couldn't help it xD $\endgroup$
    – moonbutt74
    Commented Aug 21, 2015 at 5:20
  • $\begingroup$ Very well put. You have my +1 $\endgroup$
    – Richard
    Commented Aug 21, 2015 at 5:52
-3
$\begingroup$

I'm going to go out on a limb here and say that if you're trying to find a way to discourage and delete highly upvoted questions that have attracted lots of upvoted answers from the community, that you have

Lost Your Way.

If anything, you should be doing your best to attract those sorts of questions, especially given the interest in those from outside the community via the HNQ list.

Actively discouraging people from asking popular questions when that's what stopping the site from graduating seems wildly counter-intuitive to me. It seems to me that a better question should be:

Can we accept broadening the scope or would we like to have the best failed site on the launchpad?

$\endgroup$
16
  • $\begingroup$ For the record, I find the suggestion that the answers to the runway question "lack effort" is pretty offensive. I see custom graphics and even an original animation in the top 3 $\endgroup$
    – Richard
    Commented Aug 2, 2015 at 9:33
  • 9
    $\begingroup$ Is "because they're popular" really good enough of a reason to keep them around? Popularity isn't the goal, it's the creation high-quality questions and answers. In other words, making the internet better. Our site fails if we fail to create meaningful content, not if we fail to attract large quantities of users. (Popular questions are also deleted all the time on other Stack Exchange sites, fyi.) $\endgroup$
    – user20
    Commented Aug 2, 2015 at 10:11
  • 1
    $\begingroup$ @Emrakul - If the site never exits Beta (which it isn't looking like doing any time soon, given the decline in question numbers) then you seriously need to consider a change of policy. That doesn't mean deleting your most popular questions though. $\endgroup$
    – Richard
    Commented Aug 2, 2015 at 10:13
  • $\begingroup$ Other stacks (and Yahoo Answers) have the luxury of ignoring stats because they're already graduated. $\endgroup$
    – Richard
    Commented Aug 2, 2015 at 10:15
  • 7
    $\begingroup$ The decline of questions is primarily due to the fact that it's summer, actually. That happens pretty much everywhere on Stack Exchange -- not kidding. Also graduation isn't an end goal. Graduated sites can have enormous swings in traffic, and massive declines. I think you may be putting site growth before site quality, while in reality, growth is far less important in the long run. $\endgroup$
    – user20
    Commented Aug 2, 2015 at 10:27
  • $\begingroup$ @emrakul - Yes, a decline is expected but I'm seeing a consistent downward trend line. $\endgroup$
    – Richard
    Commented Aug 2, 2015 at 11:31
  • 3
    $\begingroup$ What you say about graduation is now outdated. Graduation is no longer the general goal of small SE's. $\endgroup$
    – xnor
    Commented Aug 2, 2015 at 21:36
  • 1
    $\begingroup$ @xnor - Indeed. And while "Success and graduation are not the same thing", the opposite is not true, remaining in Beta indefinitely (or even declining in popularity) should not be viewed as a success of any sort. $\endgroup$
    – Richard
    Commented Aug 2, 2015 at 22:22
  • 3
    $\begingroup$ Maybe so, but our goal is not to graduate. Our goal is to become a collection of high-quality puzzles and answers, which a lot of these open-ended questions aren't. If once we've succeeded in that goal, people are no longer interested in the site, then the main motive behind the site simply wasn't strong enough to begin with -- but we aren't worried about that, for now. $\endgroup$
    – Bailey M
    Commented Aug 3, 2015 at 14:13
  • $\begingroup$ @baileym - As I said, while the goal may not be to graduate (immediately), the goal certainly isn't to not graduate $\endgroup$
    – Richard
    Commented Aug 3, 2015 at 14:33
  • 2
    $\begingroup$ As an aside, the main reason I've heard people across SE tell me they can't take Puzzling Stack Exchange seriously is because of exactly these kinds of questions and answers. Anecdotal, but it's not insignificant, particularly since even Shog9 said the same thing (contextually unclear, but it's in reference to these general types of questions) not too long ago. $\endgroup$
    – user20
    Commented Aug 3, 2015 at 17:58
  • 2
    $\begingroup$ @Emrakul - I've heard the opposite, that people have stopped coming because you're systematically making the site less fun by killing these sorts of questions. $\endgroup$
    – Richard
    Commented Aug 3, 2015 at 17:59
  • $\begingroup$ I'd be happy to discuss it further in chat; we might be overextending the usefulness of comments here. $\endgroup$
    – user20
    Commented Aug 3, 2015 at 18:01
  • 2
    $\begingroup$ @Emrakul - To quote; "Puzzling is supposed to be fun". I'm not seeing the site as fun any more. $\endgroup$
    – Richard
    Commented Aug 3, 2015 at 18:05
  • 1
    $\begingroup$ @Emrakul - "Popularity isn't the goal, it's the creation high-quality questions and answers. In other words, making the internet better" - actually, the goal is to provide content that people want. THAT is what makes the internet better. The problem, here, is that your definition of "high-quality" is at odds with that of other users. You seem to want questions/answers that provide information on how to solve puzzles, whereas other users are wanting questions/answers that challenge them. You have a choice - be a site full of bland information, or be a site full of creativity. $\endgroup$
    – Glen O
    Commented Aug 9, 2015 at 11:17

You must log in to answer this question.