8
\$\begingroup\$

While there is a lot of talk about bad questions and how the reduce SNR and what we could do about people not asking bad questions lately, I was wondering about the role of answers in this game.

Specifically the role of answers to bad questions. While in a lot of cases it is understandably borderline and someone might have correctly guessed what its all about, I often see answers that are merely extended comments and guess. Or that answer the question more or less properly, but the question is clearly offtopic. Often already in the process of being closed. Heavily downvoted etc.

I think these answers actually encourage people to ask more bad questions. It seems that in a lot of cases the thought process is "Oh, there are these hardcore people that are yelling at me, but that one nice guy helped me, let me post another question again, maybe he helps again or some other nice guy".

Or does it not?

If it does, what should be the consequence for such answers? Downvoting, even though technically they are a good answer to that question, just as offtopic as the question? Add a delete flag reason for answering offtopic questions?

I find these answers especially annoying because in times where there is a new flood of crappy questions, and you try your best to explain to people why the community does not want this question in that way, someone with much higher rep jumps in and answers it and everyone will wonder "well, when you don't want it, then why answer it?" Also it seems that questions without answers get easier deleted when closed than questions with answer, keeping crappy content available for search.

So in the end -- under the light of building up a database of good ontopic Q&A -- do we want to get rid of offtopic answers as well, or are we more lenient if the answer is offtopic than when a question is offtopic ?

||||||
\$\endgroup\$
  • \$\begingroup\$ Other than downvoting, or flagging as very low quality, we have no mechanism for wiping off topic answers. Perhaps we should, though. \$\endgroup\$ – Scott Seidman Feb 18 '16 at 13:23
  • \$\begingroup\$ As someone who more often tries to put the hours into answers that they deserve, usually to questions that are good but complex, I certainly agree it's annoying to see the lowest quality questions receive answers, sometimes up to 6 or 7 equally limited ones within 5 minutes. Whatever the answerer's reasons for doing so. \$\endgroup\$ – Asmyldof Feb 18 '16 at 13:37
  • 3
    \$\begingroup\$ There actually is a badge for good answers to bad questions. The "Reversal" badge: "Provide an answer of +20 score to a question of -5 score". \$\endgroup\$ – Tut Feb 18 '16 at 14:31
  • 4
    \$\begingroup\$ @Tut: Yeah I know, I think that its quite a bad idea \$\endgroup\$ – PlasmaHH Feb 18 '16 at 14:32
  • 2
    \$\begingroup\$ @PlasmaHH if the answer got >=20 upvotes, how can it be a bad idea? Oviously, even though a -5 question obviously needs some rework, if 20+ people liked the answer enough to upvote it, then it added useful information to the site. \$\endgroup\$ – Robherc KV5ROB Feb 19 '16 at 13:34
  • 2
    \$\begingroup\$ @RobhercKV5ROB: I am saying the badge is a bad idea. It is an encouragement to answer bad questions. \$\endgroup\$ – PlasmaHH Feb 19 '16 at 13:39
  • 1
    \$\begingroup\$ Here is a great example of what you are talking about: electronics.stackexchange.com/q/218546/4512. This homework question didn't show much work, and two people wrote good answers that explained how to go about attacking the problem, but didn't outright solve it. Two more couldn't resist looking smart and gave complete answers. I encourage all here to upvote the good answers and downvote the full solutions to embarassment. \$\endgroup\$ – Olin Lathrop Feb 21 '16 at 21:38
  • \$\begingroup\$ @OlinLathrop I'm surprised this question isn't closed as a duplicate of this one or this one. How many more questions about equivalent resistance should be answered before it's enough? \$\endgroup\$ – Dmitry Grigoryev Feb 23 '16 at 9:08
  • \$\begingroup\$ @DmitryGrigoryev: It isn't easy to find the duplicates, and for some recurring ones we should set up canonical questions some day (I might find time for it in the summer). Also people argue and vote to reopen because it has other values sometiems :/ \$\endgroup\$ – PlasmaHH Feb 23 '16 at 9:10
  • 2
    \$\begingroup\$ It's not that it's hard to find the duplicates, it's just much less rewarding. You can either spend your time searching for a dupe, and in the end you'll look like an ass and get zero rep. Or you can spend the same time writing an answer, look like a nice guy and get a few upvotes. How can we expect NOT to get crappy questions answered with such rules? \$\endgroup\$ – Dmitry Grigoryev Feb 23 '16 at 9:16
  • \$\begingroup\$ @DmitryGrigoryev: That sounds like being the one who found a dup getting rep might be helpful.... or backfire, might be worth starting a meta discussion about that one \$\endgroup\$ – PlasmaHH Feb 23 '16 at 9:17
  • 1
    \$\begingroup\$ @PlasmaHH There's already a feature request about exactly this. It's been 6 years and it's still not implemented nor declined. \$\endgroup\$ – Dmitry Grigoryev Feb 23 '16 at 9:27
  • \$\begingroup\$ @DmitryGrigoryev: Might be worth asking it as a new question, fully aware of it being already there, just to make a point about duplicates ;) \$\endgroup\$ – PlasmaHH Feb 23 '16 at 9:28
  • 1
    \$\begingroup\$ @user1582568: you are utterly and completely missing the mission goal of this site: It is not about answering the question for the user at hand, it is about building a database of good questions and answers that will help future visitors. Having the same question asked and answered all over again leads to lots of crappy content, making it harder for future visitors to find proper answers, making them ask it again and draining resources that would otherwise be useful to answer yet unasked questions. \$\endgroup\$ – PlasmaHH Feb 23 '16 at 19:02
  • 1
    \$\begingroup\$ @PlasmaHH I had missed that point. This database of useful questions and answers will only be useful if it is easy to search, and if it is easy to search I assume that finding duplicate answers should be easy. \$\endgroup\$ – user1582568 Feb 23 '16 at 19:24
8
\$\begingroup\$

Personally, I often answer questions that are a bit vague, or possibly would be 'more on-topic' on another SE site, provided that:

  • I feel I have an accurate view of what the OP meant to ask
  • The question is at least relatively close to being 'on topic' (i.e. isn't a blatantly useless question, and has some EE aspect to it)
  • The question gives me the impression that the OP is at least attempting to understand what they're working with & ask an intelligent question.

With that said, I also try to be sure to always edit the question to improve it as much as I can, and leave a comment for the OP, giving them guidance for improving that question, and (hopefully) any future questions they ask.

||||||
\$\endgroup\$
3
\$\begingroup\$

I believe my comment actually answers the question, so I'll post it as such. The truth is, StackOverflow rules encourage people to answer bad questions, and requests to change this situation are ignored. Here are the options users have when they spot a bad question, with outcomes:

  1. Post a quick answer and gain some rep. The risk of losing rep is minimal, because (a) downvoting answers also costs rep and (b) it doesn't feel right to downvote an answer which isn't outright wrong. You'll also look nice and smart.

  2. In case it's a duplicate, take a existing answer and post it (or a slightly modified version) under a new question. This can both gain and cost you some rep, but (a) 1 upvote is still worth 5 downvotes, and (b) you can do this even if you have zero knowledge, so this path is open to many more users.

  3. Skip the question, or downvote and move on.

  4. Vote or flag to close. This requires to spend some time to identify what is wrong (and perhaps leave a comment to the OP), gives you zero rep and makes you look like an ass.

  5. In case of a duplicate, flag or vote to close as such. This gives zero rep, and you'll need to spend time searching for the best dupe target, then spend more time arguing with the OP because his question is about capacitors, not resistors, so it's not a duplicate. Oh, and you'll look like an ass again.

Which option do you think people prefer?

||||||
\$\endgroup\$
  • \$\begingroup\$ Not getting rep because you point to a dup instead of answering is a really thin argument. You may not get any rep, but you aren't spending much time either. Also keep in mind that most of the time you remember something as a dup because you wrote the similar answer, so pointing to that question as a dup actually gets you rep. This has happened to me many times. I don't view closing as dup as repless at all. Remember to use "user:me" in the search string. \$\endgroup\$ – Olin Lathrop Feb 23 '16 at 11:58
  • \$\begingroup\$ For this to work each user will have to write their own version of "you want a PSU with exacty X Volts and at least Y Amps", "redraw your circuit to see which resistors are in series or parallel", etc. before starting to close questions as dupes. Pointing to someone else's questions won't give me any rep for sure. \$\endgroup\$ – Dmitry Grigoryev Feb 23 '16 at 12:07
  • \$\begingroup\$ Those would rightly get closed as dups. What I meant is that after you've been here long enough, you'll have answered enough questions so that occasionally some of those will be dups to questions you encounter. Except for very rare occasions, you shouldn't go out and create a question and answer just so you can point to dups to it later. I think I've done that once, and that was in response to a meta discussion about how we kept getting that kind of question regularly. That has been a popular dup target, but that sort of thing isn't really what I'm talking about. \$\endgroup\$ – Olin Lathrop Feb 23 '16 at 12:10
  • \$\begingroup\$ Also, I'm not necessarily in favor of giving rep for finding duplicates. For example, reversing any rep changes once the question is closed would also demotivate people to answer bad questions. \$\endgroup\$ – Dmitry Grigoryev Feb 23 '16 at 12:10
  • 1
    \$\begingroup\$ Rep is taken away when a question is deleted. Sometimes though I think we've thrown out some useful answers just because they were attached to poor questions. The value of the answers should be considered when deleting. Search engines can find matches in answers too, and if its good material it should stay. \$\endgroup\$ – Olin Lathrop Feb 23 '16 at 12:13
  • \$\begingroup\$ Related: Six simple tips to get reputation fast on any Stack Exchange site \$\endgroup\$ – Peter Mortensen Feb 23 '16 at 22:33
1
\$\begingroup\$

I personally feel like I have a "higher" tolerance for "bad" questions, in the sense that I am willing to forgive some grammatical / spelling errors when it is very clear that it's someone asking who has English as a second language. I hate to jump down someone's throat just because they're not 100% fluent in the language.

I don't necessarily feel that is encouraging bad questions, but certainly it could be viewed as a tacit approval that poor grammar / etc is tolerated. In those instances though, I generally do not mind also editing the original question to fix those mistakes and make it more clear as to what is being asked.

However, if it's purposefully written like a 14-year old texting or is a completely absurd question asking about silly, dangerous shit then I'd argue that it's probably a bad question.

||||||
\$\endgroup\$
  • 9
    \$\begingroup\$ A bad question is rarely about the grammar and/or spelling. \$\endgroup\$ – pipe Feb 18 '16 at 23:09
1
\$\begingroup\$

How about a message, once there has been a vote to close, when you try to submit an answer, to the effect of "At least one participant has voted to close this question". I'm not sure what the appropriate action would be at that point. Maybe it would force a vote to leave open, or something.

I suppose this leads down a path that gives us three categories of questions -- open, closed, and "jury is out", instead of just the obvious two.

||||||
\$\endgroup\$
  • 4
    \$\begingroup\$ I might write a data explorer query to put some stats next to this tomorrow, but I've seen a significant number of answers posted by users with 3k+ that have voted to close a question, so they seem to think the question is bad enough to be closed but want to get in their few cents worth in and possibly get a few upvotes regardless. \$\endgroup\$ – PeterJ Feb 19 '16 at 14:13
  • \$\begingroup\$ @PeterJ -- what if we started a policy of wiping the answers off closed questions? \$\endgroup\$ – Scott Seidman Feb 19 '16 at 15:10
  • \$\begingroup\$ I wouldn't be opposed to that on closed questions - after all five people have normally determined it's crap and they need a much higher rep than the 15 to upvote. But I think the solution on larger sites is the question (along with answers) can still be deleted by 10k users after a few days (or 20k immediately) with three votes but that doesn't seem to work here because nobody checks delete votes. But I think the reason it's not automatic is very occasionally there's a useful answer to something closed, especially if old and rules might just have changed. \$\endgroup\$ – PeterJ Feb 19 '16 at 15:27
  • \$\begingroup\$ That data explorer query wasn't possible - in the table that contains the close votes the user ID has been removed. \$\endgroup\$ – PeterJ Feb 20 '16 at 2:32

You must log in to answer this question.

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