Software effort estimation is mostly fake research

Article note: I consider most of "software engineering" to be a fake/failed field, but that's more egregious than usual.