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.

I think we all know about the fastest gun problem. My question relates to something similar, but quite different - why do we seemingly reward the fastest answer with votes?

If you look at some of the simple (but valid) questions that get asked, chances are you'll see at least one answer with a fairly large number of upvotes, more than I'd say is appropriate for a simple answers that you'd get for questions of this sort. Examples:

The obvious explanation is that simple questions like these attract lots of people to answer them, and once the first answer gets posted everybody stop and vote on that answer instead. It's certainly not wrong - we are suppose to vote for the correct one after all, but I've always got the feeling (esp. when it's my answer) that somehow I/they don't deserve these much votes for simple answers like this. Of course this partly have to do with the fundamental problem that questions and answers that have more views also gets proportionally higher votes, but it's in these cases that the issue is most obvious.

So, should is it appropriate to give these simple answers our votes? How about the "+1, you beat me to it" mentality? Or am I too stingy with my upvote?

Related to: Fastest Gun in the West Problem

share|improve this question

7 Answers 7

up vote 13 down vote accepted

Before StackOverflow came along, it was not uncommon to wait a day or two to get an answer on another forum. If I ask a good question on StackOverflow, I am surprised if I do not get an answer within 30 minutes. This is why fast answers are encouraged.

share|improve this answer
11  
That's a good point, 美. –  ring0 Aug 30 '10 at 15:04
3  
@ring0, +1 for fanciness –  Pops Aug 30 '10 at 15:36
    
faster,first and then ... –  robinmau Oct 11 '10 at 7:32
    
And? Anything wrong with that? If the question is so specific that you can't find any answer on the whole Internet, then you surely can wait one day or two. And meanwhile do a bit more of your own research, and maybe answer the question yourself, for the benefit of everybody else. I'm not advocating this extreme, though. But a better balance could be necessary. –  Mario Rossi Aug 13 '13 at 6:28
    
And, on the contrary, allowing too quick responses we are probably promoting lazyness. Have others read the manual. Have others do the research. To the point that we need rules to govern questions. –  Mario Rossi Aug 13 '13 at 6:32

I happily vote for multiple correct answers when they make substantially different points, or if they start with a near-duplicate explanation of what's causing the problem but then go on to provide different (correct) solutions.

I vote only for the first one of multiple near-duplicate answers. This doesn't mean I think the people who posted the second, third, fourth (etc.) near-duplicate answers don't deserve a reward for their efforts if such answers are posted within a minute or two of each other and thus it is obvious all parties were busy typing at the same time, however in the long run I think it is better for SO if such duplicates are deleted because it reduces the clutter for people who read the question in the future. Therefore I pick one duplicate to vote for and with no better way to choose I take the first. (As for people who post duplicate answers hours later, I think they deserve downvotes for deliberately cluttering things up.)

So, consistent with my previous paragraph, if I'm one of the people who posted a near-duplicate answer (and I haven't provided some extra information not in the other answers) I will delete my own answer even after receiving votes unless I was first. (And yes, if you go back through my answer history you'll no doubt find examples where I didn't delete my duplicate answers, but I guess I used to care about reputation points more than I do now.)

"How about the "+1, you beat me to it" mentality?"

Sometimes I vote with a "+1, you beat me to it" type comment, which certainly could be taken literally as "I'm rewarding you for speed", but what I actually mean is "This is a good answer that deserves upvotes, and I don't see any point in posting my own near-identical answer even though I'd almost finished typing it when this answer went up."

share|improve this answer

There is a danger that rewarding quick and dirty answers promotes, too much, quick and dirty questions of the sort that are best not asked in the first place.

That is, like I-Forgot-To-RTFM-And-I-Would-Not-Know-How-To-Anyway questions.

I don't think necessarily a poster of such question thinks "There will always be some suckers wasting their time checking the docs and trying out things for me, why should I do the work myself?", rather it may be much in the same way as we don't associate the joy of having sex with the unjoy of changing diapers, we just go for the sex -- and likewise, when there is an immediate-reward system in place, people just ask the questions.

share|improve this answer

To raise a contrary opinion: previously to SO I have spent much time in news:comp.lang.c and other usenet groups. There was no rapid-response, but there was a very strong peer review process: if an answer was dangerously wrong (such as advocating the buffer-overrun inducing gets or getting arrays and pointers confused), it would be commented on. Threads could last for days, with interesting discussion going back-and-forth. You could be sure that, by the end of the discussion, you'd know all the caveats of every suggested answer. The discussion wouldn't stop because two people had +10 answers; each new followup got equal screen time.

SO doesn't encourage long discussions, nor does it reward latecomers because people who commented earlier don't return to a question. As an example, here's a thread where I took part in early discussion but I lost track of it because people started talking amongst themselves instead of directly addressing @Philip. Similarly I stopped looking at new answers on the thread (like this one). This might be because I trawl the new questions feed rather than the hot questions; but rereading a hot question for new answers is not easy; you want to show people immediately the new stuff.

So what I really want is this: a combination of SO's rapid response with Usenet's potential for long discussion and QA. I think the new tag-specific "new answers" and "hot answers" will go some way to achieving this by highlighting latecoming answers. I look forward to using them more frequently.

share|improve this answer
1  
Cruise the C++ questions and see how fast the comments come for void main() and other undefined behavior. We may not explore the subject in detail, but we do have peer review. (Less so now that Neil has gone, unfortunately.) –  David Thornley Aug 30 '10 at 17:36
    
You may be interested in this (declined) feature-request: Actively dismiss comments/replies (“mark as read”). –  Tobias Kienzler Aug 31 '10 at 7:28
    
@Tobias: that would help with following my own questions, but it doesn't help when I want to follow new answers and comments on someone else's question, because afaik there's no way for me to even get pinged for that. –  Philip Potter Aug 31 '10 at 9:39
1  
@Philip: You can click on the star next to the question to mark it as favourite. Which is a misleading label, since it actually means exactly what you wanted - "keep me notified of updates". Also, each question has its own RSS feed, e.g. meta.stackoverflow.com/feeds/question/62791, firefox also shows an RSS-symbol on the questions URL. –  Tobias Kienzler Aug 31 '10 at 10:22
    
@Tobias: that's good but not ideal. I have to look in two separate tabs: "summary" and "favourites", when it should just be one tab "stuff I want to see". –  Philip Potter Aug 31 '10 at 10:47
    
@Tobias: and furthermore, favourites don't update me on comments, only answers. Nor do the RSS feeds. Sometimes I want to follow the comments on a particular answer, because interesting discussion happens there. –  Philip Potter Aug 31 '10 at 11:49
    
@Philip: I can see comments added when I click on the "favourites" tab –  Tobias Kienzler Aug 31 '10 at 12:41

It's fairly simple actually. The vast majority of pageviews come in the first 5-10 minutes. And each of those users must

  1. Click on link
    • Page loads, scores and answers "locked"
  2. Read post.
  3. Read answers.
  4. Vote.

Once they've read the original post, they'll most likely read the first answer, and if it is good, they'll upvote it. If 10 people do this at approximately the same time, it means that each user is voting up a correct answer with only one or two votes.

While you are supposed to upvote or downvote in isolation (regardless of other votes) people simply do not think that way. But the way that quick and dirty answers get a lot of upvotes is that everyone can agree with them and it doesn't take much thought.

share|improve this answer
    
IMHO upvotes and downvotes are the least of the problem. The real problem is 10x10 minutes = 100 minutes absolutely wasted in producing duplicate information. OK. Not all of it is duplicate, but at least 60 minutes are. I'm sure we could give a better use to that time. –  Mario Rossi Aug 13 '13 at 6:37

Being able answer a question quick, simple and concise is very nice.

Sometimes it can be nice to give a detailed explanation about something, but other times, people don't really want to read through a wall of text, but just want to see the solution.
That's probably also part of the reason. Not all upvoters want to read a novel, but instead quickly see a short correct answer (first one will usually be at the top anyways) and acknowledges it by upvoting it.

share|improve this answer

While FGIW is a real problem, rewarding fast answers isn't a bad thing. One of the things that makes SO so great is that you can get a good answer to your question in minutes.

share|improve this answer

You must log in to answer this question.

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