To predict which distinct polymorphs of a material (e.g. SiO2 or ZnO) are likely to be formed during synthesis is, one might argue, the holy grail of computational solid state chemistry. This would not only allow one to rationalise synthesis routes towards a particular polymorph and highlight promising areas of the periodic table for experimental exploration but also provide insight as to whether a certain desired property (e.g. a large pore-size in the case of nanoporous zeolites) is obtainable for a given chemical composition. In practice this is a challenging problem as besides thermodynamics it requires an accurate calculation of nucleation and growth kinetics - something that is in principle possible, but currently computationally intractable for all but the simplest systems due to the inherent complexity and/or long simulations times required. However, even when one ignores kinetic issues it is still possible to make very relevant predictions about the fate of polymorphs during synthesis. Through a careful calculation of a materials (free) energy landscape one can derive the materials phase diagram, showing which polymorphs are thermodynamically stable for a given pressure and temperature. Moreover, one can then characterise metastable polymorphs in terms of their (free) energy difference with the stable polymorph and the number of other metastable polymorphs close lying in (free) energy. Finally, one can prepare databases of such hypothetical (metastable) polymorphs for a given chemical composition. To date such studies have been mostly limited to simple alkali salts and silica (including siliceous zeolites) but potentially this methodology can be applied to materials formed from the complete periodic table. Proof of the latter is the very recent theoretical work into polymorphism of such systems as AlF3, SiS2 and ZnO.