Suggesting weighted geometric mean profile metric

[nkapila6]

Select Topic Area

Product Feedback

Body

I couldn't find the best place to give GitHub product suggestions so thought this would be the best place to do it.

Problem

With rise of AI slop and contributions, stars are no longer a reliable metric to understand the technical ability and capability of a GitHub profile.

My suggestion is to have a star to repo ratio. While this isn't perfect as some actors would slip through. At the very least, it'll filter out folks and shows GitHub cares about quality over volume.

The metric

$$\sigma = \frac{\sum_{i=1}^{n} \text{stars}_i \cdot w_i}{n^{0.5}} \cdot \log_{10}(n + 1)$$

Parameters:

• n = number of repos
• stars_i = stars on repo i
• w_i = recency weight
• α = 0.5 (square root normalization)

Recency Weights:

Activity	Weight (w)
Active in last year	1.0
1–3 years stale	0.7
3+ years stale	0.5

Scenario 1: one viral hit
Dev A
1 repo, 10,000 stars, still active

$$\sigma = \frac{10000 \times 1.0}{1^{0.5}} \cdot \log_{10}(2) = \frac{10000}{1} \times 0.301 = 3010$$

Scenario 2: consistent contributor
Dev B
50 repos, 40 active (avg 500 stars), 10 stale (avg 200 stars)

$$\sigma = \frac{(40 \times 500 \times 1.0) + (10 \times 200 \times 0.7)}{50^{0.5}} \cdot \log_{10}(51)$$ $$= \frac{20000 + 1400}{7.07} \times 1.708 = 3026 \times 1.708 = 5168$$

Dev B beats Dev A

Scenario 3: repo spammer
Dev C
200 repos, 190 have 0 to 2 stars (avg 1), 10 have 100 stars, all stale

$$\sigma = \frac{(190 \times 1 \times 0.5) + (10 \times 100 \times 0.5)}{200^{0.5}} \cdot \log_{10}(201)$$

$$= \frac{95 + 500}{14.14} \times 2.303 = 42.08 \times 2.303 = 96.9$$

Low score creating empty repos doesn't game it.

Scenario 4: someone who abandoned but HQ contribs
Dev D
5 repos, all 3+ years stale, but 8,000 stars each

$$\sigma = \frac{5 \times 8000 \times 0.5}{5^{0.5}} \cdot \log_{10}(6) = \frac{20000}{2.236} \times 0.778 = 8944 \times 0.778 = 6959$$

Still scores well (past impact matters), but less than if they were still active where it would be:

$$\sigma_{\text{active}} = \frac{40000}{2.236} \times 0.778 = 13918$$

Scenario 5: active contrib, modest stars
Dev E
with 30 repos, all active, avg 50 stars each

$$\sigma = \frac{30 \times 50 \times 1.0}{30^{0.5}} \cdot \log_{10}(31) = \frac{1500}{5.477} \times 1.491 = 273.9 \times 1.491 = 408$$

Modest but respectabl reflects real-world utility.

Final Ranking: D > B > A > E > C

Where does it fail?

No metric is perfect. Each comes with its own bias of sorts.. but at least this gets us somewhere?

• should it take into account contributions on widely used open source projects?
• should have a diminishing returns on stars? I think first 1-2k stars indicate project is really useful and everything after that is just popularity
• should formula have a ceiling?

[github-actions[bot]]

💬 Your Product Feedback Has Been Submitted 🎉

Thank you for taking the time to share your insights with us! Your feedback is invaluable as we build a better GitHub experience for all our users.

Here's what you can expect moving forward ⏩

• Your input will be carefully reviewed and cataloged by members of our product teams. 
  • Due to the high volume of submissions, we may not always be able to provide individual responses.
  • Rest assured, your feedback will help chart our course for product improvements.
• Other users may engage with your post, sharing their own perspectives or experiences. 
• GitHub staff may reach out for further clarification or insight. 
  • We may 'Answer' your discussion if there is a current solution, workaround, or roadmap/changelog post related to the feedback.

Where to look to see what's shipping 👀

• Read the Changelog for real-time updates on the latest GitHub features, enhancements, and calls for feedback.
• Explore our Product Roadmap, which details upcoming major releases and initiatives.

What you can do in the meantime 💻

• Upvote and comment on other user feedback Discussions that resonate with you.
• Add more information at any point! Useful details include: use cases, relevant labels, desired outcomes, and any accompanying screenshots.

As a member of the GitHub community, your participation is essential. While we can't promise that every suggestion will be implemented, we want to emphasize that your feedback is instrumental in guiding our decisions and priorities.

Thank you once again for your contribution to making GitHub even better! We're grateful for your ongoing support and collaboration in shaping the future of our platform. ⭐

[itxashancode]

I like the structured thinking here. The recency weighting and square-root normalization directly tackle repo spam and one-hit wonders. Your scenario walkthrough shows it balances volume vs quality better than raw stars.

Where I'd hesitate: computing this for every user in real time. GitHub serves billions of repo views daily; calculating weighted sums across all a user's repos on profile load would be expensive. They'd likely need a precomputed field updated nightly.

Also, "active" detection is messy. Is a repo with only issue activity "active"? What about a dependency update via Dependabot? GitHub's current "pushed_to" event might not capture meaningful contribution.

Your point about contributions to others' repos is the biggest gap. A core maintainer of React with 0 personal repos would score near zero. That's a major flaw.

If you want to push this, I'd suggest:

1. Propose it as an optional profile badge/metric first, not a replacement for stars.
2. Use GitHub's existing "contributed_to" relationship data to include external contributions.
3. Start with a simpler formula: (total stars weighted by recency) / sqrt(repo_count + 1) and test against known quality profiles.