6
$\begingroup$

Am I mistaken to think that logically the number of upvotes to a question should be greater or equal to the number of favorites?

The small difference between these two numbers in this post: Examples of common false beliefs in mathematics is amazing.

$\endgroup$
2
  • 4
    $\begingroup$ Somewhat related discussion on Mathematics Meta: Why favorite but not upvote? $\endgroup$
    – Martin Sleziak
    Commented Aug 12, 2020 at 5:44
  • $\begingroup$ Oops, sorry. The typo was bothering me in the HNQ list, and I forgot that this one might not already be on the front page. Sorry for the bump! $\endgroup$
    – LSpice
    Commented Aug 21, 2020 at 22:10

1 Answer 1

Reset to default
11
$\begingroup$

No, not entirely. An average question indeed has more upvotes than bookmarks (the new name of favorites). See below for some statistics.

Both vote types (bookmarks are stored as votes in the database) are used in different ways:

  • You are expected to upvote a post when its content is useful and clear; most people use bookmarks to refer to the question later on. That includes situations where you think a post is actually bad and you want to check later if it is improved. (Though you can now follow the post, which is a better feature for this.)
  • Upvoting requires 15 reputation, bookmarking just requires having an account.
  • You can't bookmark answers; perhaps people didn't think it was a good question itself but did appreciate one or more answers.

A minor nitpick, but the question you linked to doesn't have 832 upvotes, it has a score of 832, composed of 841 upvotes and 9 downvotes. You can see that (if you have 1000 reputation) by clicking on the score.

enter image description here

Here is a question: Video lectures of mathematics courses available online for free with 588 bookmarks but only a score of 358; for another popular question: Thinking and Explaining the amounts (score: 331 vs. number of bookmarks: 330) are even closer than in your example.

It's possible to do some data analysis on the distribution of score vs. number of bookmarks with the Stack Exchange Data Explorer, but I'm not sure what to try. Suggestions are welcome in the comments. Here is some proof that the average question has a higher score than the number of bookmarks.

Here is a (log-log) plot of the number of times a certain score (yellow) or number of bookmarks (blue) appears, generated by this SEDE query:

enter image description here

So there are more questions with 1 bookmark than that there are questions with a score of 1, but for other 'counts' it's the reverse.

$\endgroup$
3
  • 5
    $\begingroup$ For very highly-upvoted questions, people might not upvote a question if they like it, but feel it has been upvoted "enough" and its score accurately reflects its quality. That may be a partial explanation for the question Ali linked. $\endgroup$
    – Will Sawin
    Commented Aug 19, 2020 at 20:45
  • $\begingroup$ @WillSawin I agree that this is probably a real phenomenon. I'm conflicted as to whether it's a good, bad, or indifferent phenomenon. As I think about it, there are lots of factors at play. But maybe it should suffice to say the following: when deciding whether to upvote a question / answer, one's primary consideration should simply be the straight-up question "does this question/answer have merit?" -- the question "does this question/answer already have 'enough' votes?" should carry much less weight (if any) as a rule. I wonder if there has been discussion of this on meta before? $\endgroup$
    – Tim Campion
    Commented Aug 22, 2020 at 15:13
  • $\begingroup$ In case anybody else is curious, I did just find a SE meta post about this aspect of voting. $\endgroup$
    – Tim Campion
    Commented Aug 31, 2020 at 19:22

You must log in to answer this question.

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