In a way this is sort of like "should I down vote ..." (which forms one solution to the issue) but is about specious answers (i.e. right sounding answers which upon closer examination are wrong) and how to deal with them.

I see 4 possible actions: 1) Put in a parallel answer, 2) Put in a parallel answer but refer to the other answers pointing out the flaws, 3) Comment in the original answer simply stating that things are wrong, 4) Comment in the original answer with all the details pointing out the flaws.

In a recent item I responded with 3), and poised the question in a way that if the "expert" knew what he was taking about they'd say "oops, ou're right that was silly of me let me fix that" and I also down-voted the response. The response I got was a snarky answer, and a retributional down-vote. Kinda funny actually.

The reason I choice the #3 approach is that like it or not, people look at the higher ranked answers as the correct answers, popularity seems to replace accuracy. So I'd rather have the original author give the best/nuanced answer rather than a wikipedia style answer. And I'd rather have the right answer than get points.

Another aspect is that it also allows one to figure out if the original answerer actually knew what they were talking about. In this case the specific domain knowledge was very weak.

If I'd answered like #1 - parallel answer then it probably won't see the light of day. a #2 Answer - rebutting the original answer in a another answer is messy and rude. and a #4 answer - cleaning up details in their answer also seems rude as well.

But it seems a downvote is also rude.

And this leads to a bigger question, should things that correct the "known" answer from the past, say items that are 2 years old, be done? I see people talking about a repository of knowledge, but the fact is someone searching for an answer only will look at the top rated one and not the new correction to that.

Maybe that is the nature of these types of systems, that they can tend towards a cargo cult type of body of knowledge.

So what should I have done?

  • 3
    \$\begingroup\$ Downvoting is not rude if you truly believe the answer is wrong. In that case I think you should really leave a comment explaining what you think is wrong. After all, you could be wrong or not think of something. If the original author fixes the question, you can remove the downvote if you like. \$\endgroup\$ Oct 9, 2012 at 17:26
  • \$\begingroup\$ Sounds like the basis for an answer @OlinLathrop. I agree, the core reason to vote is for the most useful answer to be at top and least at bottom, something that is wrong guarantees a downvote for me, but I would suggest basing your comment on if you think it will lead to productive result. \$\endgroup\$
    – Kortuk
    Oct 9, 2012 at 23:28
  • \$\begingroup\$ Thanks for the feedback. And yes, it is easy to be wrong or to have a off day, but conversely, when there is domain specific knowledge which isn't commonly known the answered may not even know that they are wrong or rather off base. However, the real answer to this may very well be that there really is no single answer to this. \$\endgroup\$ Oct 10, 2012 at 1:31

2 Answers 2


The best thing to do is to downvote incorrect answers. Downvoting is not rude if you truly believe the answer is wrong. That's actually exactly what it is there for. Upvotes for good answers and downvotes for bad answers makes the good answers bubble to the top, which is the goal of this site.

While you are not obligated to provide a reason for a downvote, or even mention you did it at all, I think it is common courtesy and constructive to do so. One problem with this system is that while answers are essentially peer reviewed (that's a good thing), those reviews are not. If you think something is wrong and downvote it, it's only fair to let everyone see the reason and the author therefore a chance to rebut or fix it. You could be the one that is wrong or overlooked something. It is only thru a open exchanged of learned dialog that we can eventually get to the right answer, and just as importantly, everyone can see that. If it turns out you eventually agree your initial assesment was wrong or the author fixes the answer, you can undo your downvote. Of course it helps if the authors pings you in a comment to let you know he fixed things, else you may never know.

I strongly object to your options of commenting in the answer. That is not the place for a dialog, and I think is extremely rude. If you have issues, downvote and leave a comment below the answer. Editing someone else's answer should be limited to fixing obvious typos, spelling, and sometimes grammar and word usage. I forget exactly the site policy on this, but I personally think changing the meaning of someone else's answer is just plain wrong. Comment, downvote, and/or write your own answer instead.

  • \$\begingroup\$ +1 Official (non-meta) site policy: stackoverflow.com/faq#editing \$\endgroup\$ Oct 10, 2012 at 17:53
  • \$\begingroup\$ I must have misread it, or maybe there was a miscommunication, I read, "Commenting in the answer" as writing a comment, but I could be wrong. \$\endgroup\$
    – Kortuk
    Oct 10, 2012 at 21:35
  • \$\begingroup\$ @Kortuk: Since he specifically said commenting in the answer, I can only take that to mean editing the answer with in-line comments. That is not the way to do things here. Comments should be added after the answer using the mechanism specifically provided for that purpose. \$\endgroup\$ Oct 11, 2012 at 21:43
  • \$\begingroup\$ @OlinLathrop Yes, i agree they should be added the other way, I think that he misscommunicated. He has never done that, but he has left comments. \$\endgroup\$
    – Kortuk
    Oct 12, 2012 at 5:08

I agree that #2 is wrong, but I don't have a problem with #4. #1 is fine, but you should still correct the incorrect answer (through comments). I don't see the difference between #3 and #4.

Option #5: If the answer is salvageable, and you fear retribution from options 1-4, you can edit it yourself. Not only will the original high-voted answer (and your edits) remain at the top of the page, but it's a friendly way of correcting mistakes without coming across as rude.

Site policy encourages editing: stackoverflow.com/faq#editing

  • \$\begingroup\$ That's an interesting idea. I suppose it would be up to the individual answerer to be accepting. If they had a firmly held belief that they were right that could rapidly degrade into an editing war vs. a comment war. \$\endgroup\$ Oct 10, 2012 at 1:32
  • \$\begingroup\$ @rawbrawb True, but it's worth a shot if you are pretty certain you are correct. \$\endgroup\$ Oct 10, 2012 at 1:34
  • 1
    \$\begingroup\$ I think editing someone else's answer to change the meaning is always wrong. \$\endgroup\$ Oct 10, 2012 at 17:30
  • \$\begingroup\$ @OlinLathrop: It's certainly a personal choice, and I don't do it often. The few times I have done it, however, the original authors have been appreciative. \$\endgroup\$ Oct 10, 2012 at 17:44
  • \$\begingroup\$ @OlinLathrop: Also, my stance only applies to answers that are almost right (salvageable). I agree that editing an answer to completely change the meaning is wrong. \$\endgroup\$ Oct 10, 2012 at 17:52
  • \$\begingroup\$ @OlinLathrop Changing what you user meant I disagree with also, you should post a different answer, but correcting an error that is probably just a mistake, minor math error leading to it all being wrong and correcting and flagging the poster. \$\endgroup\$
    – Kortuk
    Oct 10, 2012 at 21:37

You must log in to answer this question.

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