• Nalivai@lemmy.world
      link
      fedilink
      English
      arrow-up
      8
      ·
      3 days ago

      The opposite of it actually. It’s harder to hide suspicious spikes in activity this way

    • Alphane Moon@lemmy.world
      link
      fedilink
      English
      arrow-up
      60
      ·
      4 days ago

      This is an issue with Lemmy too. Searching for a community by keyword ranks the results list by subscribers not by MAUs

      • palordrolap@fedia.io
        link
        fedilink
        arrow-up
        4
        ·
        4 days ago

        This kind of already happens there though. Video view counts are visible and often way below a channel’s sub count.

        Sure, there are exceptions (viral video views often far exceeding the sub count), but by and large they’re a good metric for seeing how a channel is actually doing.

        • paraphrand@lemmy.world
          link
          fedilink
          English
          arrow-up
          2
          ·
          4 days ago

          Yeah, you can infer it.

          But no one is making a thank you video because they reached a weekly active users metric.

        • paraphrand@lemmy.world
          link
          fedilink
          English
          arrow-up
          3
          ·
          edit-2
          4 days ago

          That’s news to me. They took away subscriber numbers? YouTubers still talk about them.

          And this is the number behind the plaques they send out to people. I can’t image them stopping that. They won’t send plaques based on weekly average activity numbers.

  • Cid Vicious@sh.itjust.works
    link
    fedilink
    English
    arrow-up
    41
    ·
    4 days ago

    I dunno why they wouldn’t display both. Also cynical me thinks this is a more useful change for advertisers than users.

    That said this would probably be useful for post ranking on the back end, if they’re not already using it. There’s always been a noticeable thing on reddit where posts on large subs with little activity don’t seem to bubble up much on /r/popular. Which makes those subs seem even deader because they don’t get new blood.

    • NihilsineNefas@slrpnk.net
      link
      fedilink
      English
      arrow-up
      8
      ·
      3 days ago

      They don’t want to show both because that would show readers exactly when a couple thousand new accounts all start talking about one polarising topic to get everyone agreeing.

    • icylobster@lemmy.world
      link
      fedilink
      English
      arrow-up
      3
      ·
      3 days ago

      If I’ve learned anything about the corporate world is that they only want metrics that confirm the views they hold or want to push. I think you have a point though. These are metrics that they can reference when trying to get advertising money. If I was an advertiser I’d care more about how many are not bots though.

  • OctopusNemeses@lemmy.world
    link
    fedilink
    English
    arrow-up
    2
    ·
    2 days ago

    Yeah it was a useless metric anyways. What I’ve found is that people use that number to as evidence to some rhetoric. Pretty much every time I point out it’s a useless metric they become very angry.

  • FaceDeer@fedia.io
    link
    fedilink
    arrow-up
    19
    ·
    4 days ago

    One of the rare changes Reddit has made in recent years that seems like a good idea.

    They’re also moving to limit the number of large subreddits that any individual moderator account can moderate, which seems like a good thing. Hopefully they’ll be serious enough about it that they’ll bother to catch the power moderators that simply set up a bunch of different alts for themselves.