The author does not look at counter values but rather at how the values changes. That reveals dynamics.
replyMy comment still seems relevant? Do frequent commits to correct mistakes imply more "value" than infrequent, but well tested, commits, or what? I don't think it is a reliable signal.
reply