This is not an argument against self-citations, it's an argument about how self-citations should be counted. There's also a fun formula for expressing estimated linear regression coefficients as weighted averages of local slopes.

from Regression and other storiesSection 8.3: Least-squares slope as a weighted average of pairwise slopes: I was reminded of this after an email discussion with Shravan Vasishth, which began with the following note from him: I just saw this article on the h-exponent for games: https://arxiv.org/abs/2402.04607 Isn't it odd that Google Scholar doesn't have a […]

Source