I'm somewhat frustrated at the seemingly bandwagon mentality at Stack Overflow with respect to voting sometimes. I've seen a great number of questions voted closed as off-topic, duplicates, etc. without the voters giving them adequate attention, as evidenced by the fact that they aren't actually off-topic, duplicates, etc. if you take the time to read through them. It's as though some users with voting privileges see that others have voted a certain way, and just blindly vote along-side them sometimes.

I'd like to propose that SO stop displaying these votes to you until after you actually vote on it (assuming you ever do). I feel like not displaying this information would force users into giving the question more attention and voting according to their own opinion, rather than promoting bandwagon voting.

Although I disagree with the conclusiveness of the "evidence" cited in it, this post touches on the same concept.

Edit: One such example, as requested: Why does bcmul return a number with a scale different than the one I specified? (Disclaimer: I was the OP).

  • 1
    Any examples of questions that were incorrectly closed?
    – yannis
    Dec 26, 2013 at 16:58
  • All of his questions... duuuuh @Yannis
    – qwertynl
    Dec 26, 2013 at 16:59
  • Hmm... stackoverflow.com/questions/20765060/… There have been many debates on the subject of "why does program language X do Y?" questions. I believe the consensus is that these questions are not on topic here as they generally cannot be answered but by the design team, or they can be answered by simply reading the specs for the language. It's a sore spot that usually ends up with the question being closed as off topic/opinion based or being closed as dupe of a similar question (your case).
    – user1228
    Dec 26, 2013 at 17:00
  • Exactly @Won't :-P
    – qwertynl
    Dec 26, 2013 at 17:01
  • 1
    I know you're joking @qwertynl, but the OP is a 6K user with some excellent answers. Don't think this is yet another pointless "Y U CLOSE MY PRECIOUS QUESTIONS" rant (although, without examples we won't know for sure)...
    – yannis
    Dec 26, 2013 at 17:02
  • @Yannis See Wont's comment.
    – qwertynl
    Dec 26, 2013 at 17:02
  • @qwertynl No, I will not click on a link Won't shared. I learned that lesson the hard way. ;P
    – yannis
    Dec 26, 2013 at 17:03
  • @Yannis Hmmm ok. I dunno why but ok. Here is my version of the link: stackoverflow.com/q/20765060/1022697 :-)
    – qwertynl
    Dec 26, 2013 at 17:04
  • I added an example (same one @Won't cited). To clarify, this virtually never happens my own questions (this is the only one I can remember in the past couple years that I completely disagreed with in retrospect), so I'm not one of "those people." I do see it in effect quite often for other users' questions, though.
    – FtDRbwLXw6
    Dec 26, 2013 at 17:15
  • Ummmm Your question works both ways.... by putting your post here you got it reopened (and I agree that it was a duplicate)
    – qwertynl
    Dec 26, 2013 at 17:31
  • @qwertynl I happen to agree that it was a duplicate question, but only incidentally. If you read the answers on the duplicate, the OP accepted one that didn't actually answer the question formally. And that's something I've struggled with before - what do you do in that case? Technically it's a duplicate question, but the duplicate has an accepted answer that doesn't actually answer the question (the OP was probably only interested in "making the code work" rather than getting the answer to the question). I opted for a re-post, hoping to get a real answer.
    – FtDRbwLXw6
    Dec 26, 2013 at 17:54
  • 1
    Awarding a bounty to the answer that you think answers it might be something you could so @drrcknlsn
    – qwertynl
    Dec 26, 2013 at 18:05

3 Answers 3


Your proposal presumes that the reason people cast votes is because other people cast votes, but perhaps people cast votes for legitimate reasons most of the time.

It's worth noting that questions which attract certain kinds of votes are placed in review queues, where community members can add their vote to the tally. So, in a sense, we encourage this sort of behavior. Without it, removing poor questions from the site would be difficult.

There are reasons why this is not a problem. One of the biggest reasons is that the close reasons were refined and clarified, so that they can be used in a more instructive and unambiguous way. This more or less prevents the abuse that might otherwise occur with vague, broadly-specified close reasons that could apply to a wide range of scenarios.

Nowadays, unfair closures are exceedingly rare. They occur when the OP puts too much rant in their question, or their question itself is an edge case. The remedy is to post on Meta, as you just did, and get enough consensus for a reopen, which in fact just occurred.

  • 1
    I tried to use wording in my post that was non-absolute. I recognize that voting in a lot of (most?) cases is legitimate, but I feel like people glaze over the demonstrable fact that it often-times isn't. As far as I can tell, there is no practical purpose for displaying other users' votes prior to you actually voting, and it's a hard sell to convince me that it never influences other voters, so the question is then - why show them if it serves no good, and can potentially be bad?
    – FtDRbwLXw6
    Dec 26, 2013 at 17:12
  • 1
    I personally like the fact that I can see how other people are voting before I vote. It validates what I was thinking all along: there's something good (or bad) about this post. I have no qualms about disagreeing with the other voters, however, and I'm pretty sure that many other community members have the same streak of independence that I do.
    – user102937
    Dec 26, 2013 at 17:18
  • Fair enough, but wouldn't that same validation still exist if you got to see the other votes after-the-fact? It just wouldn't be able to influence your vote...
    – FtDRbwLXw6
    Dec 26, 2013 at 17:26
  • Perhaps it is because I'm a moderator and my vote is binding, but I prefer to see if some other community members agree with my thinking first before casting my vote, especially on edge cases. I don't worry too much about the clear-cut cases.
    – user102937
    Dec 26, 2013 at 17:27
  • 1
    Community moderation works best when each reviewer judges posts by independently applying his discretion and wise judgement. So I disagree with the part of this answer that says horde voting is not a problem. At least mindless horde voting, should be avoided ...
    – Dilaton
    Dec 26, 2013 at 22:49

If there is pile-on and mindless close-voting without reading (a point I don't actually concede) it would be caused more by the review queue than by the existence of the close (2) link at the bottom of the question.

However, the cure for this is to comment: "this isn't actually a dupe because that is about Fortran and this is about APL" or whatever. And vote to reopen if you like. And edit it to clarify and highlight the different aspects of this question.

Closing doesn't have to be permanent, so if something is closed and then reopened again there is no harm done.

  • I agree completely with this, and it's exactly what I do. Though, I feel like having recourse when things go wrong is only a small comfort, and doesn't actually do anything to prevent them from going wrong in the first place, which is what I'm hoping we can do.
    – FtDRbwLXw6
    Dec 26, 2013 at 17:20
  • @drrcknlsn with 7,000+ new questions every day, you need a fast and efficient closing process. In such a process wrongful closings will happen. It's the recourse process that needs improvement IMO.
    – Pekka
    Dec 26, 2013 at 17:35
  • @Pëkka I have no illusions that we'll ever be able to prevent 100% of wrongful closings, but I feel like hiding votes would prevent a lot more than we do now. I'd also support bettering the recourse process instead of or in addition to hiding votes, though.
    – FtDRbwLXw6
    Dec 26, 2013 at 17:57
  • 1
    @drrcknlsn I'm pretty sure Kate is right and it's the review queues that cause the piling on, not individual users seeing the question and noticing the (n) next to the close button. If that's correct, hiding the vote counts wouldn't help
    – Pekka
    Dec 26, 2013 at 18:02
  • Closing is not permanent, unless you have in addition to horde closing leave closed -ing by the same people who did the first, too ... That is why for example on physics you get hardly any wrongly closed or improved question reopend ...
    – Dilaton
    Dec 26, 2013 at 22:53

The core assumption here overlooks a very important aspect of closing. The close vote review. Although the close vote review queue is large, the first thing a reviewer will see is the most recently asked question with a close vote according to their filter.

This is a contributing factor to closure speed and should not be overlooked. Most questions which are closed do not deserve to be open. And those which do deserve it often are either re-opened or discussed on meta.

Moreover, in the review queue when filtering on a type of closure, users are more likely to choose that same closure reason. While there may be some validity to that being "bandwagoning", people tend to find what they are looking for, and it does not have anything to do with showing the current amount of votes on a question.

When the votes are shown in the close vote review queue, they are not always the same as users tend to choose what they feel fits the question most accurately. Is it a perfect process? No. But I do not believe that showing the counts affects the process negatively.

All that being said, when a close vote specifically for duplicate is showing in a comment, other users will generally be more inclined to agree than to fully analyze. I believe this stems from a reluctance to post duplicate answers, although it will still happen.

As a result of that inclination, when someone points out that your question is a duplicate, and you believe it is not, you must be attentive to that and explicitly define how your question is unique.

In the exact case you link, it is true that you show in your question that you are not interested in the "number_format" (which is the answer in the linked duplicate - although this was reopened, the linked duplicate (confused by PHP's bcmul() scale) was what started this post here on meta), you must make sure to specifically say that your question is not a duplicate of the other post for that exact reason if you want to be fully transparent about the issue.

In conclusion, I do not believe that the vote count leads to bandwagoning - whether it is shown in the review or on the post. I do believe that being attentive will allow your question to remain open and that being attentive is a key part of communication.

P.S. - Closing as duplicate can take some inspection at times, even to familiar users. Perhaps voting to close as duplicate could require at least a score of 5 in one of the tags shared by both the parent and eligible child duplicate.

  • "I do not believe that showing the counts affects the process negatively." - Can you think of any way to test this belief empirically? Is there any data one could analyze that would support or refute it? (Maybe looking at the range of spread of close votes for close votes that come through the review question when filtered by close reason, vs those that don't?)
    – D.W.
    Jan 1, 2014 at 22:11

You must log in to answer this question.

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