Ten impact experiments are presented that probe the initiation behavior of an RDX-based explosive. The explosive pellets were impacted by thin aluminum impactors supported by low-impedance polymers, thereby subjecting the pellets to short duration shocks. Impact conditions were used to determine impact tilt, pulse pressure and duration, sample transit time, and whether or not a detonation occurred. Exploding Foil Initiator (EFI) launchers were used as a baseline for evaluating the technique, and it was found that the present technique addresses multiple shortcomings of EFI-based systems. The aluminum foil allowed impacts with pressure durations of 70-80 ns, resulting in input conditions closely matching those of EFI-based systems. The thicker aluminum plate impactors resulted in pulse durations of approximately 600 ns and allowed access to a previously inaccessible region of the initiation curve. Initiation thresholds were bracketed using both impactors, and the data is compared with complementary initiation data from EFI-based systems with good agreement. Based on the results and closely related work, a novel, physically-based criterion is proposed to allow the initiation curve to be predicted based on the explosive's unreacted equation of state and sustained-pulse, shock-to-detonation run distance (the "Pop-plot"), giving a physical basis for previously published observations. Finally, the criterion is applied to the data available to propose a relationship between pressure, pulse duration, and run distance, suggesting that input pulse durations of 50-100 % higher than the critical condition should be applied to approximate the run distance of a sustained pulse.