In development of new explosives, it is often necessary to balance a number of attributes in performance while certain formulation constraints exist. Statistical design of experiments (DOE) is a valuable tool for rapid formulation optimization and minimization of costly and hazardous testing. During the development of metal-loaded explosives designed for enhanced blast, it was discovered that upon proper formulation, aluminum additives obtained full reaction by 7 volume expansions, which resulted in extremely high Gurney energies equivalent to LX-14 and PBXN-5 but with lower loading of nitramines. The early aluminum oxidation can be described by Eigenvalue type detonations, where the fully reacted Hugoniot of the condensed phase aluminum oxide and explosive products lies below the unreacted aluminum Hugoniot. Such an analysis describes fully the agreement of aluminum consumption by 7 volume expansions from 1-inch copper cylinder expansion tests and an analytic cylinder model, as well as detonation calorimetry. With the early reaction of aluminum also comes a shift in the gaseous reaction products to higher enthalpy species such as CO and H2, leading to further augmentation of blast. Thus, both the mechanical energy (for fragmentation or “metal pushing”) and blast (for structural targets) are available in a single explosive fill. This provides capability for combined metal pushing and blast in a single explosive that was not previously possible. Development of such explosives and the importance of modern statistical design of experiments will be shared.