2 days ago (see 2) I posted something into the !programmer_humor@programming.dev community which got almost 1000 upvotes (see 1) while the user numbers for the last week are at 131 (see 3), this doesn’t add up. How can 131 users upvote almost 1000 times?

  • Kalcifer
    link
    fedilink
    arrow-up
    34
    arrow-down
    2
    ·
    edit-2
    8 months ago

    That’s because, currently, the community stats that you see in the sidebar are only from your instance – community stats are currently not federated. Afaik, federated community stats are going to be implemented in 0.19.

  • Ategon@programming.dev
    link
    fedilink
    arrow-up
    27
    ·
    edit-2
    8 months ago

    users/day and users/week is from all instances. Only subscribers are from your own instance (so the other two comments are wrong)

    The reason is cause that only counts users who have either posted or commented and not those who have upvoted only

  • rglullis@communick.news
    link
    fedilink
    English
    arrow-up
    9
    ·
    7 months ago

    Do voters count as active users? I am somewhat confident that the user count is just related to people who posted or commented on the community.

  • ShaunaTheDead@kbin.social
    link
    fedilink
    arrow-up
    8
    ·
    7 months ago

    It’s definitely from other instances voting on it. For example. I’m on the @kbin.social instance but I’m commenting on your post from @jeemy.jeena.net

    Also, this doesn’t really have to do with your point but it’s interesting none-the-less, it’s really hard to keep view count numbers accurate when you’ve got multiple servers involved. The view count is accurate and every view is being counted, but any time you look at it, it’s not exactly correct. If you think about how each server might get a view count update from the server at different times, and then add the views that only it sees to the view count, and then when it goes to add it’s view counts to the central count and get a new updated number there may have been dozens of other server that have updated the view count at various times between then and now.

    It all gets very complicated but I think it’s kind of neat that a view count number can be accurately kept track of and yet every time you look at it, it’s not exactly correct.

  • zeppo@lemmy.world
    link
    fedilink
    English
    arrow-up
    7
    ·
    8 months ago

    The weekly & subscribed user count is only users on that specific instance. Also someone doesn’t have to view a community to upvote a post.

    • Jeena@jemmy.jeena.netOP
      link
      fedilink
      arrow-up
      1
      arrow-down
      1
      ·
      8 months ago

      But I’m on a one person instance, so that would only show one user but doesn’t.

      • Nawor3565@lemmy.blahaj.zone
        link
        fedilink
        English
        arrow-up
        6
        ·
        8 months ago

        The number is displayed on the instance the post is on, so it doesn’t matter which instance you view it from (or if you’re not logged in at all), the number is calculated from the active users who are members of the instance you’re viewing.

      • Optional@lemmy.world
        link
        fedilink
        arrow-up
        1
        arrow-down
        1
        ·
        8 months ago

        So you’re running your own instance, and that instance had 131 people/week login? Or did you login 131 times?

  • kamenLady.@lemmy.world
    link
    fedilink
    arrow-up
    9
    arrow-down
    2
    ·
    8 months ago

    The upvote logic is so complicated on Lemmy, that idgaf anymore. It’s not like it really was important in the reddit days, just funny somehow…

    The first versions of Sync for Lemmy’s still had the added upvote count on my profile page - a Sync for Reddit leftover, but vanished quickly.

    This post obfuscated the Lemmy upvote system to a almost opaque level for me lol

  • aion@lemmy.world
    link
    fedilink
    arrow-up
    4
    arrow-down
    1
    ·
    8 months ago

    Disclaimer: I have not looked at how Lemmy calculates those numbers.

    I suspect this is a combination of multiple factors:

    • The stats could be averages over the last several days/weeks.
    • The stats could be estimates, not exact numbers.
    • The stats probably only account for people who look at the community directly.
    • People do not need to look at the community to upvote a post, people can upvote from the site-wide feed.