A core idea of WIKI is to take imagination as simulated belief revision (ISBR): imagining that P is the case and wondering what follows in the imagined scenario, is a bit like learning that P is the case and revising one’s beliefs accordingly.  This makes the issue of the epistemic credentials of imagination (more) tractable, by assimilating it to the better-understood process of reasoned belief revision upon learning. This paper digs into the idea, qualifies the similarity between real belief revision and its simulation in imagination, clarifies what sort of ‘reasoning’ is involved in both. The paper then expands on conditional beliefs as the characteristic output of ISBR, showing how it makes sense to assess them as justified or unjustified. It proposes Four Constraints on the workings of ISBR and shows that they are represented in both AGM belief revision theory and Bayesian conditionalization.