Show HN: UpvoteRate – Towards a fairer ranking formula for Hacker News

Hello HN! About 1.5 years ago we published an idea on how to improve the Hacker News ranking algorithm (https://felx.me/2021/08/29/improving-the-hacker-news-ranking-algorithm.html https://felx.me/2021/08/29/improving-the-hacker-news-ranking...). The main idea was to counteract the rank-upvotes feedback loop by using attention as negative feedback. We received very valuable input from the HN community (https://news.ycombinator.com/item?id=28391659 https://news.ycombinator.com/item?id=28391659), and have since come up with something that we think could be a real contribution: the upvoteRate metric. upvoteRate is an estimate of how much more or less likely users are to upvote a story compared to the average story. We think this provides more stable and meaningful comparisons between stories. We would propose it as an alternative to the simple upvote count (which is distorted by feedback loops) in the ranking formula. A Hacker News client called "Quality News" shows this metric for live Hacker News data (https://news.social-protocols.org/ https://news.social-protocols.org/). And you can read about how we calculate this metric here: https://github.com/social-protocols/news#readme https://github.com/social-protocols/news#readme We hope for interest and support from the Hacker News community to encourage official ranking experiments on the Hacker News frontpage.