In order to discuss the fundamental aspects of calculating nucleation rates via atomistic simulations, we have to remove as many layers of complexity as possible from both the system under investigation and the conditions in which we are working on it. We start by focusing on supercooling or supersaturation regimes that lie well away from the spinodal limit where the free energy barrier associated with the formation of a critical nucleus is vanishingly small. Conversely, we have to avoid, by necessity, very mild supercooling or supersaturation as well; in that case, the size of the critical nucleus is usually too large to be even taken into account by means of atomistic simulations. Then, we shall consider almost exclusively homogeneous nucleation, in principle, the simplest scenario available to us. However, homogeneous nucleation is rarely observed in reality, with heterogeneous nucleation being much more common. Indeed, it is often very challenging to experimentally measure crystal nucleation rates without any influence in terms of impurities, which almost always manage to facilitate the crystallization process.3,4 Note, however, that much of what is discussed in the following is equally relevant to studies of heterogeneous nucleation. In addition, we will largely avoid the emergence of two- or even multi-step nucleation processes (biomineralization being a prominent example) as well as confinement effects. Excellent reviews on these topics and, broadly speaking, on the subject of crystal nucleation as a whole can be found in, e.g., Refs. 2, 3, 13, 17, and 18.