This is something I wanted to bring up for a while now, so here we go.

If you use Reddit, you are probably familiar with the “karma score” that it displays on user profiles, based on the points from all of the user’s posts and comments. Lemmy also has this for now, but I think it is bad and should be removed.

Having a global score like that makes sense from the perspective of a company like Reddit, because it encourages users to post more, which increases “engagement”, giving them more money from ads and investors.

But from the community perspective, such a score has a lot of negative effects, like users (or bots) posting low quality content with the only goal of increasing their karma score.

Mastodon has given a good example for how to do it differently. For the most part, numbers are hidden (like boost or fav count), and only visible for a single post at a time. I saw a lot of comments that this helped to create a healthier discussion culture because people are encouraged to look at the actual content, and not on some numbers.

So what do you think about this? Any thoughts or suggestions?

  • worldCritic@lemmy.ml
    0·
    4 years ago

    Copy Netflix

    Every post should have a different score per reader, not per author, prediction-driven, based on how a reader has voted in the past and weighing in votes from others based on the extent by which the reader’s past votes align with other voters.

    Complex? yes, but if you ignore complexity this is what best serves the community. I want posts I’m most likely to find insightful to rise to the top of my own view. Posts I’m unlikely to draw value from should be folded (click to expand).

    • dead10ck@lemmy.ml
      1·
      4 years ago

      Oh goodness, please no. This is not a media site where a recommendation engine is appropriate. Only showing people what they want to see is what makes Facebook such a toxic echo chamber. Machine learning is not the answer to everything.