In (2014), Mark Sainsbury has claimed, first, that there is a variety of fiction operators, in particular the "according to f"- operator and the "in f"- operator, whose semantic contribution to the complex sentences they contribute to generate is different, and second, that no worlds-based semantical treatment works for any of them. In this paper, I want to hold that, when suitably reinterpreted, Sainsbury is utterly right as to his first claim, yet just partially right as to his second claim.