-19

What happens when crowd-sourcing fails utterly, that the technically incorrect answer gets up-voted while the technically correct answer is down-voted because is is unpopular. This is possible because crowds are known for group-think and are not known for logical, reasoned thinking. It is quite possible for a provably-correct answer to fail the popularity contest of voting. What happens then? Can we do a correctness review?

marked as duplicate by Shadow The Curly Braced Wizard discussion Mar 12 '16 at 6:52

This question has been asked before and already has an answer. If those answers do not fully address your question, please ask a new question.

  • 6
    This has been asked approximately 712,514 times before. Do a quick search on meta for "wrong answer". – Dan Bron Mar 12 '16 at 0:04
  • Are you serious that I have asked nothing new? I think not. I am asking to prove the technical correctness of an answer. Perhaps you would like to consult academic sources to reverse votes. – Brent Kirkpatrick Mar 12 '16 at 0:05
  • 8
    I am serious, yes. Are you surprised that others have asked this before? And no, I would not like to consult anything. I was suggesting something you could consult. No matter. Having a dupe in your question history is no big deal. – Dan Bron Mar 12 '16 at 0:07
  • Do you know what a proof is? If a question remains open long enough, someone could supply a logical proof. I realize that this suggestion may not be novel, but it is unlikely to have been suggested 7 hundred thousand times. – Brent Kirkpatrick Mar 12 '16 at 0:09
  • 9
    You know what's a cool word? hyperbole. – Dan Bron Mar 12 '16 at 0:10
  • There are levels of accuracy that Stack Exchange could employ. Only one level is crowd-sourcing for popular answers. If you look at the academic method of finding good answers, one uses peer-review (similar to crowd-sourcing) and increasingly rigorous scientific justification. The proof (en.wikipedia.org/wiki/Proof) is the highest mathematical expression of correctness. Why not use levels of accuracy here on Stack Exchange? – Brent Kirkpatrick Mar 12 '16 at 1:04
11

While Dan may be mildly, yet entertainingly, overstating things, he is not too far off the mark [just by a couple of orders of magnitude ;-)]:

Wrong answer search results on SE Meta

OK, some of those may be about deleteing, correcting, or editing wrong answers, but the question has been asked many times before, not just on SE Meta, but also on the individual Meta sites.

Yes, sometimes there is a crowd mentality that prevails and the wrong answer can, even over time, get substantially more votes than the correct one. Having the wrong answer accepted by the OP can also accentuate this "erroneous" voting. I have seen some examples of this, across various SE sites.

However, generally, even if the wrong answer has been accepted, the votes tend to go with the correct answer, over time. Unless, of course, it is an obscure question and/or it is on a site with little traffic, such as a Beta site. In such a case, the two answers may only get, for example, 5 votes between them, in which case, the erroneous answer could look like the correct one, as accepted by the "crowd", albeit a small one.

You have only been on SE for two days. Relax a little, things don't necessarily happen that fast on SE. Even though you have racked up 12 answers in a few hours, other people here may not have the energy, nor the time, to keep the votes coming as quickly as you would like.

  • Please see the notions of evidence (en.wikipedia.org/wiki/Evidence) and proof (en.wikipedia.org/wiki/Proof). Sometimes the non-intuitive or non-popular answer is indeed the correct one. Only sufficient evidence or proof can establish its correctness, not voting. – Brent Kirkpatrick Mar 12 '16 at 1:06
  • 5
    @BrentKirkpatrick: While you're technically correct (the best kind of correct!) unfortunately, humans are flat incapable of actually constructing and recognizing 100% reliable proofs 100% of the time. Incapable. In the absence of flawless perfection, then, what science actually does is rely on ... consensus. Consensus that an argument is generally correct. That's what SE voting attempts to measure. It's far from perfect, certainly, but its fundamental flaws are non-trivial to fix. – Nathan Tuggy Mar 12 '16 at 2:16
  • 3
    meta.stackexchange.com/q/268666/215590 is a proposal to try to keep low voted but accepted answers from staying above more upvoted ones forever. – PolyGeo Mar 12 '16 at 2:39
  • @NathanTuggy Thanks for the comment. I agree. The question was what to do about this. I general I agree that it is difficult to do much except rely upon reasoned consensus over time. We hope that reason and logic eventually sways the consensus. Thanks for the comment! – Brent Kirkpatrick Mar 12 '16 at 3:49
  • 5
    You yourself need to present evidence that you've identified something that's a problem in practice here, @BrentKirkpatrick. – Josh Caswell Mar 12 '16 at 5:26
9

To some extent, I feel your pain. I have had times where an answer that I feel is either subtly incorrect or blatantly wrong has gotten too much attention, while answers that are spot-on get relatively few votes. There are many reasons why this could happen, from the question hitting the Hot Network Questions list to people simply not seeing some flaw in the answer. This can be annoying.

First, consider the following: you could be wrong. This is always my first thought when I disagree with an answer, especially one that's managed to get a lot of upvotes. One avenue of logic holds that if it's you against dozens of people, there might be something you've missed that they've all picked up on. I'm not saying that the majority is always right - you're asking about cases where the majority is wrong - but you should always remember that none of us are right all of the time.

If you've considered all this and concluded that you're still right, your best option is to do one of two things:

  1. Leave a comment on the answer you believe is wrong. Let the answerer know why you think they are in error, and let others know the same thing. Carefully explain your logic in detail, so everyone can understand what you're saying.
  2. Post a new answer that fixes the errors in the other answer. This should not be an outright attack on the other answer - that's neither polite nor what answers are for - but should answer the question in a way that counteracts the incorrect answer. You can mention the other answer, but try not to focus on it.

You can't tell people how to vote, nor should you. Nobody has the power to do that. But what you can do is try to convince people that what you say is correct. Voting can be reversed, once an edit is made to the answer. It's possible, generally, for people to change their minds. Just figure out a way to present your logic such that they'll do so.

  • 1
    +1 for point number 2... well, both points to be fair. – Greenonline Mar 12 '16 at 2:30
5

A vote is seen in the eye of the beholder. Literally.

You should consider that you may be wrong and whether the vote was used correctly or incorrectly is simply an opinion, not a fact. Lots of time, a majority might pick one side and the others pick the opposite side. This doesn't state that the minority are wrong, but they just have a different viewpoint than the majority. There is even a chance that the majority are wrong even.

As the community votes often to maintain the sites and for other reasons, the actually better answer will get more of the votes rather than the incorrect ones. Thus maybe the incorrect answer gets the "most" votes quickly, but the correct answer will catch up over time.

If you seriously believe you should do something, there a few things you can do:

  • Comment on the post - Explain why the answer is incorrect and show ways the answer can be fixed.

  • Create a new answer - This answer should be posted to fix the errors shown in the incorrect answer, not as an attack against it

  • Downvote it - Cast a downvote to show your disapproval about the content about the answer and bring it down

  • Upvote the better answer - This will help the better answer overtake the incorrect answer

  • Do a combo - Maybe do two or three things here, like do the first option with the third option

But realize your actions should not be used to attack the wrong answer or the user that posted the answer.

  • +1 - That is a good list of actions that one could take, to remedy such a situation. – Greenonline Mar 12 '16 at 2:43
  • Wow. Thanks for the discussion. This is great. In general there can be a provably correct answer, regardless of opinion. Indeed, that provably correct answer can be unpopular by votes. Such a scenario is possible. See the notiion of a logical proof (en.wikipedia.org/wiki/Proof). I am really glad that there is an active discussion on this topic! Thanks! – Brent Kirkpatrick Mar 12 '16 at 2:46
  • You can use the accept button to accept the answer you agreed/liked the most – Anthony Pham Mar 12 '16 at 2:47

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