Why don't we find more polymorphs?

Acta Cryst. (2013). B69, 313-328 (http://doi.org/nsk)

Few molecules have a uniquely favorable way of packing with themselves. The approximate symmetry in the intermolecular interactions of caffeine results in many almost equi-energetic structures, giving rise to complex disordered phases.

We have computationally generated thermodynamically plausible crystal structures for 200 organic molecular systems and compared these crystal energy landscapes with the results of experimental solid-form screening studies. There are usually far more structures within the energy range of possible polymorphism than known polymorphs. Seven possible explanations, ranging from the approximations in the calculation of crystal energies to experimental limitations in finding and structurally characterizing polymorphs, are given. This highlights the need to understand crystallization better before we can develop programs to predict polymorphs.