This is going to sound like sour grapes: it isn't, as I can hopefully demonstrate. I think it makes a useful case study to raise an interesting question.

One of the first questions on the site that I attempted to answer was Was Neil Gaiman's Stardust influenced by C.S. Lewis? because I knew a fair amount about the genesis of the book and I thought I could show that the questioners thesis was, at the very least, a minor aspect in the creation of the work.

The reason is that by co-incidence, I'd just finished reading a novel called Lud-in-the-Mist. Habitually upon finished a novel I like to google around to see if there are any interesting literary allusions connected with it. In this case, I found a large number with Stardust. Try it yourself: google for "Lud-in-the-Mist Stardust" and see what comes back.

I tried to distill these oft-remarked upon similarities into an answer and show that there was a strong connection between the two books and that, in turn, these strong connections suggested an influence from Lewis was unlikely.

Another answer did the same, but they got the wrong book. Their suggested influence, a writer called Lucy Clifford, was named by Gaiman in the same article I used as a reference but only in passing. He named (and praised) Lud-in-the-Mist more frequently. Google "Lucy Clifford Stardust" and you'll see that, aside from the original interview, there's almost nothing.

I would hope that demonstrates that what I'm suggesting is a more factually accurate answer than the other. Yet it's sinking. I presume that's because I went out on a limb and made a statement about Lewis belonging within the great tradition of Tolkien, which Gaiman was trying to escape from in writing the novel.

It's fair that I was probably on shaky ground make that allusion: it's interesting to consider but I presented what was a weak thesis fairly strongly and that's going to get people's backs up. I should probably go and edit it, and will do so once this has run its course on meta: but the damage is already done.

So: we have a situation where a badly worded but accurate answer has been outvoted by a clear but incorrect one. This worries me, because anyone looking at that question and the highest voted answer alone is going to walk away with the wrong answer. It seems to me this is a big potential danger for this site, or indeed any SE site, where there's no scientifically agreed 'correct' answer.

Hence the question: on what basis should we be "voting" for things, especially on an SE site where there are dubious criteria for what qualifies as the "right" answer?

  • @yannis Sometimes: it depends how much visibility things get. Anyway, as I said, I will edit it in time: to do so right now would either render this question incomprehensible or demand substantial edits. And while I may not be entirely clear about what I'm asking here, it seems to me there is an important issue at the bottom if it, somewhere.
    – Matt Thrower Mod
    Jan 30, 2017 at 13:13
  • 1
    For the record, I (the OP) have upvoted your answer but not the other one. I haven't accepted an answer yet, because it's still possible that something even better might come in. We're still in private beta and our questions haven't had many eyeballs yet.
    – Rand al'Thor Mod
    Jan 30, 2017 at 19:42
  • 2
    IMO, don't ask how to vote (people should vote how they see fit, not to match what the community sees fit), rather ask what you can do to improve the quality of the answer.
    – Möoz
    Jan 30, 2017 at 20:55

2 Answers 2


Stack Exchange gives us a wide range of tools, for those cases where upvote/downvote isn't expressive enough.

  • A valuable answer written in a way that obscures its value, should get edit-suggestions that fix it, and/or comments that suggest what might be corrected.
  • An answer that is popular, but incorrect, should likewise be commented on. If the poster made an error, point it out. They may edit and correct it. Or, they may disagree with you, but at least the disagreement will be visible - and if you're correct and persuasive, then (a) votes may reflect that, and (b) your comments will be preserved along with the answer.

None of this guarantees that the best answer will always rise to the top (and also, as we all know from our own experience, nothing guarantees that your answer is the best one, even when you're sure it is :P ). But in general, it does a pretty good job.

  • Yes, very good points. As a veteran of several other SE sites, I really should have thought of this myself: suggested edits especially are a good approach here.
    – Matt Thrower Mod
    Jan 30, 2017 at 20:17
  • 1
    Good answer, rather than determining how people should vote, we should be setting out guidelines to encourage users to actively improve the quality of posts.
    – Möoz
    Jan 30, 2017 at 20:57

We should be voting down things that seem to be low quality, up voting things that are high quality, and close voting things that are off-topic/etc.

Now, there is significant wiggle room in those directions because what one person perceives as low quality might be meh or good quality to another. These votes are inherently subjective opinion.

AS per your answer, I'm not understanding the need for this meta. You've got 3 up votes and 1 down vote (as it stands as I write this). So you've 1 up vote less than the top answer and 1 more down vote, making the total score of 2 with the other answer a score of 4. Also, neither answer has been accepted by the asker yet, so either Rand hasn't read either answer yet, thinks they're both on par with each other, or feels that neither properly answered the question.

But again, this meta was stemmed from one down vote. Down votes happen, that's just the way the internet works. It's only when the down votes pile up compared to the up votes when you really need to worry about the quality of something on the site.

You must log in to answer this question.

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