Very interesting! Identifying great new content is a big unsolved problem for HN IMHO. Unfortunately, scores are not a good metric to predict, because they are

Very interesting! Identifying great new content is a big unsolved problem for HN IMHO. Unfortunately, scores are not a good metric to predict, because they are not comparable (see https://felx.me/2021/08/29/improving-the-hacker-news-ranking-algorithm.html https://felx.me/2021/08/29/improving-the-hacker-news-ranking...). A better metric might be "upvoterate", defined as how much more or less likely users are to upvote a story compared to the average story. More about that here: https://github.com/social-protocols/quality-news?tab=readme-ov-file https://github.com/social-protocols/quality-news?tab=readme-...