this post was submitted on 18 Oct 2025
1 points (57.1% liked)

ShareGPT

88 readers
1 users here now

To share LLM text output that others might find interesting.

founded 2 years ago
MODERATORS
 

Right now, big communities dominate the feed. I’m wondering what sort algorithm could level the field so niche or hobbyist communities have a fair chance to get seen.

There’s a good related post: Niche Communities won't be able to reach their true potential until Lemmy adds a sort that takes engagement into account. It puts it well:

“If Lemmy is to truly start having active hobbyist communities instead of being 95% lefty US politics, Shitposts, and some tech stuff, it needs a sort that takes into account the user’s engagement.”

What do you think should be the default sort for a more balanced Lemmy?

top 3 comments
sorted by: hot top controversial new old
[–] PumpkinDrama@reddthat.com 4 points 2 months ago

Use a weighted ranking algorithm that normalizes engagement across community size and highlights statistical outliers.

Goal: Equal visibility potential for small and large communities by emphasizing relative performance over absolute volume.

Outline:

  1. Inputs

    • ( u ): upvotes
    • ( d ): downvotes
    • ( t ): post age (in hours)
    • ( n_c ): number of active users in the community
    • ( μ_c ), ( σ_c ): mean and standard deviation of engagement (e.g., score = upvotes − downvotes) within community ( c )
  2. Compute raw score [ s = u - d ]

  3. Normalize by community size [ s' = \frac{s}{\sqrt{n_c}} ] (Square root dampens the effect of community population size.)

  4. Compute z-score (outlier detection) [ z = \frac{s' - μ_c}{σ_c} ] This measures how exceptional a post is compared to typical posts in its own community.

  5. Apply time decay [ z_t = z \times e^{-λt} ] ( λ ) is a decay constant controlling how fast posts lose prominence.

  6. Sort order

    • Primary key: ( z_t ) descending.
    • Secondary key: recency or total engagement.

Effect:

  • A post that performs far above its community norm (high z-score) rises regardless of community size.
  • Large communities no longer dominate since normalization and outlier weighting make “exceptional” small-community posts equally competitive.

Optional adjustments:

  • Cap visibility to prevent manipulation (e.g., minimum engagement threshold).
  • Use Bayesian shrinkage if communities have low post counts to stabilize ( μ_c ) and ( σ_c ).

This yields a discovery feed surfacing statistical outliers across all communities.

[–] TORFdot0@lemmy.world 2 points 2 months ago (1 children)

Is this not what the scaled sort already does?

[–] PumpkinDrama@reddthat.com 1 points 2 months ago

The scale sort, is simplistic, it gives more weight to posts by communities with low active users, but bot accounts with many posts and no users would always be on top. It doesn't get outliers.