25

I always search SO hard before asking a question. And let's be honest, the site isn't necessarily optimized for search (answers tend to reuse question terminology, which often isn't the broadest relevant terminology), so sometimes finding what I'm looking for is laborious.

My question is, are we punishing searchers? I don't ask a lot of questions (to date, zero), quite frankly, because I can always find the answer. Sometimes I find it elsewhere and forget to re-ask on SO, which admittedly is my own fault - but I think the point holds.

However, I routinely see very-near-but-not-exactly-duplicate questions voted up and answered with no recourse. And I don't see that as a mostly solvable problem.

Maybe I'm just venting because I feel like I am punished for not asking questions without searching first. What does Meta think?

| improve this question | |
  • 15
    Related: Give an incentive for finding duplicate questions and any number of questions discussing how much searching sucks, which I can't find right now because searching sucks! – Ether May 18 '10 at 1:06
  • Can you please give examples of near-duplicate questions? – Andrew Grimm May 18 '10 at 7:21
  • @Andrew_Grimm well, I don't have a bunch of time to dig, but today I ran across 2853677 which is very similar (as asked) to 62503. The question led to some interesting discussion around IntPtr and marshaling, but as asked it was IMO a near dup of multiple questions around keyword type aliases in C# which are answered quite plainly. – hemp May 18 '10 at 9:17
  • @Ether: I would point out that I'm primarily targeting "near dup" questions here because I believe the system for addressing obvious duplicates can be successful. It's the questions which are really close to duplicate, but just different enough that no one is likely to vote to close them - those are the sticky wickets here. Often I can deduce the answer to my specific question by reading answers to a very similar but slightly different question. When that happens, I don't feel it's appropriate to ask the very similar question. – hemp May 18 '10 at 9:25
  • Related: Should I ask a question I know the answer to? meta.stackexchange.com/questions/17463/… because sometimes your questions are answered via google-search already. – rlb.usa Jul 20 '10 at 22:31
11

I think three things could help this.

  1. The tags system is polluted, and rarely useful. Integrating the search with tags, eliminating the cruft tags, and perhaps suggesting tags based on question content could help. Additionally mods could get a list of questions with similar content and tags (upon request) so they can quickly evaluate new questions as a dupe.

  2. Building on that last part, a system for finding & reporting duplicate questions at multiple levels could be useful. When a user with low rep asks a question, maybe they have to review a few similar questions to be sure the new question isn't a dupe before it's actually posted. And a mechanism for reporting dupes, other than flagging it for a mod, or posting it in the comments would be nice for normal users. And again, a mechanism for mods to instantly search existing questions for similar content & tags would be useful. A "punishment" mechanism could also be part of this, but I doubt it would work as intended, as new users are too easily scared away by such negativity.

  3. An improved search system has been talked about many times before. I think pulling the tags taxonomy into the equation would help, but again you get into the tags problems. It's really a issue of where to start, and having a plan of where you're going with it all at the same time.

| improve this answer | |
  • 2
    the "pollution" of the tags system is why I support meta.stackexchange.com/questions/48417/… – Jeff Atwood May 18 '10 at 5:06
  • 2
    @Jeff Atwood Why not also support a tag blacklist/whitelist curated by high rep users? This was proposed by Diago a while back, yet nothing's been done about it. – alex May 18 '10 at 8:05
  • I'm upvoting because I like part 2. Proposing likely duplicate questions during the process of asking a question may go a long way toward reducing the near-duplicate issue. I've seen this work fairly well on UserVoice. I understand that Ask currently searches for title matches, but obviously that's a very narrow set. – hemp May 18 '10 at 9:41
  • @Jeff, I definitely agree, cleaning up the Tags system would allow it to be a useful building block for searches and automated dupe-detection systems. – Chris S May 18 '10 at 14:12
  • 1
    @Hemp, I think you're also hitting on part of the problem with current searches, different people word the titles of the same question very differently. – Chris S May 18 '10 at 14:16
  • +1 for "And a mechanism for reporting dupes". Would then enable rewarding reporters of duplicates. With rep rather than flags, that is. – Raedwald Oct 21 '11 at 0:09
6

If you have a question you want an answer to, isn't getting a correct answer quickly the ultimate reward? You're likely to get the answer quicker through a successful search than through carefully crafting a question and waiting for an answer, even though answers on SO do usually come remarkably quickly.

Surely that should be much more important than the (reasonably small) amount of reputation you get for asking a question.

| improve this answer | |
  • How did you get 31.4k rep on Meta...? – uɐɯsO uɐɥʇɐN May 18 '10 at 7:19
  • 5
    @George: Um, by answering questions... – Jon Skeet May 18 '10 at 8:15
  • The ultimate reward .. asks the guy with more rep on SO than Chuck Norris. ;) <br/> It's true, I agree that with regard to the things that should matter in life, getting the answer quickly is more valuable. But for the more trivial things (like building up SO rep to reach the ill-defined "participates in the community" qualifier by prospective employers) rep matters more than answers. – hemp May 18 '10 at 8:57
  • 3
    @hemp: If you want to earn rep, that's a lot easier to do with answers than with questions. Take the time you've saved by finding the answer to your question via a search, and spend it answering other questions :) – Jon Skeet May 18 '10 at 9:34
  • @Jon_Skeet: That's probably good advice overall. I found this to be an interesting problem w/ regard to SO and wanted to pose the question to the community. Plus it earned me a bunch of rep. Ironically, most of my rep on meta is from asking this question ;) And before you go all C.N. on me, yes I realize meta awards reputation differently than SO - never let facts get in the way of an apparent irony. For great justice! – hemp May 18 '10 at 9:57
  • 1
    @hemp: indeed; I've built up vast swaths of reputation on meta purely by being the first to exploit a new meme in its infancy, and getting out of the game before it jumps the shark. That, and shooting down Jeff's comments, seem to be winning formulas. (I'm saying this tongue in cheek, but many here will see partial truth in what I say.) – Ether May 18 '10 at 16:22
3

SO is really designed to reward answerers more than questioners. The reward for questioners and searchers is a good answer.

| improve this answer | |
  • 1
    The OP is stating what has been a problem - the lack of good search in SO. It is sometimes easier to just ask a new question than search. – tim May 18 '10 at 1:51
  • With badges like the Electorate badge, I disagree meta.stackexchange.com/questions/38222/… – Christian Payne May 18 '10 at 4:07
2

Just a thought: In terms of SEO, duplicate questions (and answers to them) would result in "better" search results on Google, etc. because it makes The Triology more relevant for the keywords contained.

| improve this answer | |
  • To that end, I would rather see one question with lots of clarifying answers and comments which will do even more to improve SEO. Perhaps it would be better to encourage people to further refine already answered questions. – hemp May 18 '10 at 9:59
  • @hemp I'm not an SEO expert but I was thinking along the line of more (duplicate) posts resulting in more relevant search results attracting more visitors. Of course content itself would gain more from refining existing posts. – Filburt May 19 '10 at 6:14
1

I've noticed that I frequently find answers, or at least highly relevant threads, in the Related questions shown on the right hand side after my question has been saved. Curiously, they are not the same set of Relateds shown on the initial composition page. I don't know if this a fact or a mental illusion, but what is certainly true is that even though I always search before posting a question, the Relateds are usually new to me.

| improve this answer | |

You must log in to answer this question.

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