Should you vote based on how others voted?

Question

Sometimes I come across questions or answers that I appreciate and believe deserve an upvote. The question or answer, though, already has ten or more votes. The post, I believe, is good, but is it good enough for fifteen upvotes, for twenty? Has the post already received its fair share of attention and votes?

Is voting a simple operation of asking yourself whether a post has merit, and voting accordingly? Or is voting an attempt to, as part of a crowd, give the question a proper rating?

The latter seems simpler, but this means that number of votes is no indication of quality.

More than a tenth of my total reputation comes from my answer to one question. My answer wasn't particularly good, but the question was, for whatever reason, well trafficked. I suppose I am appreciative of the added rep, but I can't at all confidently say that answer deserved six times the recognition of my next best answer—and probably it is inferior to many of my one vote answers.

Answer 1

It's totally for you to decide whether or not you vote (either way or not at all). On the one hand, you may decide not to vote on a post with a high score because you think its score is high enough. On the other hand, you could up vote because the post is useful to you.

Answer 2

I just upvoted your answer and the question and a couple of other answers because I wanted to signal my joy at literate Qs and As.

Actually, I think the second answer was better than yours, but it came in several hours later. The early bird does get more worms, deserved or not. There is much craziness in the voting here: my best question got only four upvotes, and the answer, which required some admirable research, got only four or five votes. Asinine drivel can get many, many votes.

So, my answer to your question is to express your opinion, and not pay attention to the opinions of other people. And....if you really like an answer, and it already has 20 votes in one day, postpone your upvote to the next day (or even the day after that) so your vote will count.

Answer 3

People are people, that is, they're not simple algorithms, devoid of context. People are irrational or rather they are rational within their own virtually continuous set of prior data that is nonetheless bounded.

When an SE user votes, they're doing it in the context of reading the words, but also questions they've recently read, the current set of votes, how other things have been voted on in the past, and on and on.

So yes, psychologically, a voter may like an answer but think 'Hmm...I don't really need to vote, it has enough already', or oppositely 'Lots of people have voted up, I think I will too' or contrariwise 'This is a terrible answer but someone down voted once, and that's enough of a signal' or.... all sorts of other, let's say, not playing by exactly the nominal rules.

Oh, one of your answers was super upvoted, beyond what you think it's worth? That may not be such a psychological thing as that it got on the Hot Network Questions queue that every site sees, and so it attracted a lot more traffic than normal.

So, the rational thing to expect is that the quality of questions and answers is judged in an unbiased, context-free manner. And that's just not the case, here or in other reality.

Answer 4

Might I suggest that you completely ignore any votes on any question or answers and just do what you think best (not what you think the community might think is best)?

What does it matter that the vote count is already high if it's a really good answer? Good answers deserve to be upvoted. Upvote the way you feel and think. Upvote by your own criteria.

Does an answer need to be useful to me to be upvoted? Not at all. I can think of no particular usefulness for one of the top voted answers but it was so well researched and entertainingly written that relativism has no place in voting. If I came across another well researched and well written answer not nearly as entertaining, I'd still upvote it. It doesn't matter what other people do.

I could see if you were new here wanting to understand how voting works, but you're not. What does take more time, though, is realizing how little reputation matters. So don't overthink it. It's not like you're parceling out food, clothing, housing, or career opportunities. It's more like you're handing out tiny bits of approval. If you approve for any reason, give it an upvote.

Answer 5

Well, I have to address the phenomenon, aka the elephant-in-the-answer box.

But first things, first. Voting on the worthiness of a question is not the same as voting on whether an answer to a question is good.

For answers, I upvote an answer based solely on my understanding of the question and deciding on which answer best fits the bill. Sometimes, two answers may be good or cover more ground together than separately. So I upvote two different answers.

The phenomenon that drives me round the e-bend is: upvotes on answers that are completely off track when a good answer is present but has gone unrecognized by all the upvoters.

I call this: the null-set or shiny-pretty-thing bias. That is a polite way of saying the upvoters simply don't have the knowledge to know either way whether an answer is actually really good. They read an answer, don't know the subject matter, and upvote because everything "sounds good". They vote for the shiny, pretty thing. The incredible part of this is that often this occurs in a context where a really good answer is present and overlooked. When I really do not know the subject matter, I do not vote. Period.

The other phenomenon is that the right answer is completely different from what most readers (voters) are able to recognize, and, not only do they upvote what everyone else has upvoted, they downvote what is actually a good answer but one they simply do not understand.

Unfortunately, there is a lemming effect, and people do seem to sometimes just follow the crowd. What other explanation can there be to explain a huge number of unmerited upvotes for certain answers? And also, the converse: a large number of unmerited downvotes on certain answers?

Please note: I only am commenting here on answers, and not on questions.

Re this question: I think it is useful, so I am upvoting it. Re the answers: I think they all say something useful, so I am upvoting all of them; they each add some interesting insight on how people vote.

Answer 6

What does it mean when a post (question or answer) is up- or down-voted? Does it mean it's a good post? It's a high quality post? It's a funny post? It mentions cats?

I believe everyone has their own reasons for voting and that there is no consensus on what a vote means or why you should vote or should refrain from voting. Voting is highly subjective.

I think I've only ever cast two downvotes, so I won't discuss those. I upvote questions if I have (almost) the same question, or the question is really thoughtful, i.e. now that I've read the question, I'm curious too and would like to know the answer.

I upvote answers when they fully answer the question and are not repeating previous information, i.e. the earliest answer that satisfies my curiosity.

I don't upvote simply because others have upvoted, but because I believe it deserves to be upvoted, however, if there are a lot of answers, I don't always read all of them, leading to the top answers being more likely to receive an upvote from me, simply because I don't see any potentially good answers further down. Sometimes I sort the answers by age to try and mitigate this.

However, feel free to upvote just because the post mentions cats.

Answer 7

When a post already has a lot of votes that agree with yours, your voting signal is diluted. It's maybe a bit less helpful to future visitors. But I really only make that calculation when I'm running out of votes for the day. At that point, I might consider reserving my remaining votes to signal usefulness of posts that have received less attention. Otherwise, I just go ahead and vote'em as I sees'em.

Answer 8

It is not necessarily the best answers that get the most votes.

People vote on many different factors. Each to their own. One thing to bear in mind is that some people vote on which answers were useful to them, rather than focusing only on the quality of the answers or the research (and expertise) required to answer them.

More recent answers may give more insight than older ones but these were not a candidate for earlier voters or to be selected as the "acccepted" correct answer. Similarly votes given long after the question is given will be a result of search queries for problems they encounter or questions they want answered as well.

They may not even have the expertise themselves to judge whether an answer is reliable. I think this works well in programming and Math SE sites where it is possible to evaluate whether the solution works yourself. Whether this model applies well to Linguistics and Language is a matter for a longer debate elsewhere. There are cases where it works well and other matters (such as best word choice or translation) that are arguably too subjective to be put to a vote. The Japanese language site in particular has encountered issues with this and discussed it more on their Meta.