I am not so naive to think that this wasn't asked/discussed, but I have not found a close enough fit to my question (yet). Also, I will appreciate any improvements to my question (also as per language).

So: I guess that naively, as an outsider, one would think that the most upvotes next to an answer should mean that the answer is the most correct one, so "correctness of the answer" could be the answer to my question. However, after some experience with the site, one can argue that "usefulness" is a much better concept, both intuitively and given the broad/fuzzy nature of questions asked (sometimes the answer have rather the character of a discussion, or even an opinion on some SE sites). As far as I understand, this is the reason behind the current paradigm: "This answer is useful." appears when hovering above the upvote button. (I wonder, though, what the prehistory of this was.)


  1. as per correctness - not all readers have the knowledge to judge the correctness of an answer (which can be evaluated quite objectively in case of e.g mathematics or computer code, which is even more straighforward), partly because they didn't know the answer in the first place and that is why they landed there, which touches the author of the question, too.

  2. as per usefulness - it is not so clear whether usefulness is meant with respect to the ability to solve/understand the solution of the question as asked by author, or to solve a related question of a user who was directed to this question because of a similar, but maybe not identical, problem.

Perhaps a more correct interpretation (on some SE sites) of the upvotes is something like "plausibility" of the answer, how correct/useful it SEEMS to people, as far as they can judge.

Also, as most upvoted answers tend to be most read and hence most upvoted, a positive feedback loop kicks is. How this would be taken into consideration in the interpretation of the upvote button?

So our hypotheses are (please feel free to add):

  • most useful (currently accepted paradigm)
  • most correct (not always a measurable criteria)
  • most plausible (appearing to be correct)
  • most relevant?
  • best explained (could still be incorrect, though)
  • best by some other metric ? ...

I realize this question is far from being perfect/clear/complete, so feel free to through some ideas. Maybe "usefulness" is just fine and all this is superfluous.

Related: How does accepting an answer work?


3 Answers 3


I think the answer is simple: there is no "correct meaning" for that. Each person is free to vote as they see fit.

That's the base idea behind the voting system.

What we do have are guidelines and tips, mostly the tooltip that you mentioned:

This answer is useful

This should guide the one who considers how to vote and make them think if the answer might be useful for people, not only for the one who asked the question. That's not a rule, it's just a tooltip. If there were strict rules for voting, it would diminish and nullify the whole meaning of the voting system.

There's even whole for this: When should I vote?


Most users, who have enough reputation to upvote, have seen the answer and thought it deserved an upvote.

We currently do not have the technology to read the users' mind and determine why they thought a given post deserved an upvote.


Stack Exchange operates on the Wisdom of the crowd model.

It is similar to a jury trial in that we hope that the majority are smarter, hopefully less susceptible to bias or corruption and therefore more likely to reach the correct result than relying on a single individual expert - the judge in the case of a jury trial.

Wrong answers are usually not useful, although in the absence of any positively scoring answers, downvoted answers may at least tell you some directions you shouldn't take.

Answers that are plausible but incorrect has certainly been magnified as an issue by LLMs such as ChatGPT, which can confidently give you the wrong answer in some cases. When ChatGPT first launched, people simply pasted both correct and incorrect answers into Stack Exchange en-masse and caused chaos for a while.

In general all we can hope for is that people use their best judgement when they vote. They may be experts, they may not but have the same problem and find the answer helps them or they may have just lost their keys. We expect the number of people who have lost their keys to be a pretty small group and therefore that their votes will not change the final outcome very much.

  • 1
    We should not forget that on Stack Exchange sites the voting system is set in place even on posts where the "Wisdom of the crowd" "should not be used". This is because there is only one voting system, i.e. there is no way to set a different voting system on per-site meta, or per mandatory tags (discussion, feature-request, support, bug)
    – Rubén
    Commented Jan 25 at 1:36

You must log in to answer this question.

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