Kaito's algorithm update this time resembles an information sorting problem - pressing the gourd causes the ladle to float. The meaning of "pressing the gourd causes the ladle to float": it is a metaphor for solving one problem only to have another arise, describing a situation where problems arise one after another, causing one to neglect others. Factors affecting sorting: A * weight + B * weight + C * weight + N * weight... Where A, B, C, and N (N factors) are the factors affecting sorting, which may include dozens or even hundreds of dimensions. This is the simplest form of a rule-based algorithm. Now that sorting mechanisms have introduced AI/ML algorithms, it won't be as simple as manually defining it, but most small teams might implement something like this. The biggest issue with Kaito's improvement is that you first need relevance; the weight of the relevance factor is too small, leading to a situation where a portion of the users on the leaderboard (not a small percentage) have almost no relevance to a particular project, meaning you can't find this project in the user's x text. So why would the project party pay? For example, this one, what is starknet? Have I written about it? I don't even know what sapien is. 🤣 Also, OG, I did see him publish a paper half a month ago, and I wrote one too, but that was half a month ago, 30D rank 60? Improving the sorting experience is not an easy task; before improvement, there must be evaluation criteria to assess quality, and quality testing and evaluation must be done before going live. @Punk9277