Skip to content

Impact scoring: Adjust translation weighting to reflect actual contribution volume #670

@dionysous

Description

@dionysous

I am opening this issue to suggest improvements to the Recent Impact scoring system. As shown in the screenshot below, the current implementation functions effectively as an "Activity Score" rather than a real "Impact Score."

Image

Why is that?

The system currently treats a single string translation and a high-volume effort (hundreds of strings) as identical daily events. This creates two distinct issues:

  1. Incentivizes "Gaming": I observe that contributors are incentivized to artificially throttle their work (drip-feeding) to maximize points, rather than contributing when I have the time or capacity.
  2. Devalues Significant Effort: A contributor who spends many hours on a single day working on complex translations is scored the same as someone who spends five minutes on one translation.

I propose the following adjustments to better align the score with actual output while preventing massive point inflation:

Option 1: Fractional Scaling

I propose implementing a diminishing returns model where the first contribution establishes a baseline, and subsequent translations provide fractional growth.

  • Base Logic:
    • The first translation = 1 contribution (for strings > 0).
    • Every additional translation within the same 24-hour window = 0.01 points.
  • Calculation:
    For a total of $n$ translations in a day, the formula would be:
    $$Score = \lfloor 1 + ((n - 1) \times 0.01) \rfloor$$
  • Example: 110 translations results in 2 contribution points ($1 + 1.09 = 2.09$, rounded down to $2$). The points per string could linearly decrease each 100 strings, e.g. to 0.0075 and than to 0.005 etc.

Option 2: Batch/Block Contributions

I propose grouping translations into "contribution blocks" to maintain a logical scale that recognizes volume.

  • Logic:
    • 100 translated strings are grouped into a single "contribution block."
    • Each block counts as 1 contribution (applying the 3x multiplier for high-impact status).
  • Benefit:
    • This rewards scale and high-volume work while keeping the impact scores within a predictable, incremental range that is easy for the community to understand.

Goal

I want the Recent Impact score to reflect community contributions more accurately, ensuring that volume-based work is recognized while maintaining the incentive for regular engagement.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type
    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions