Galaxy mergers are the foundation of hierarchical structure formation and therefore a key driver of galaxy evolution. Merger rates for massive galaxies are well studied both theoretically and observationally. However, similar studies for low mass dwarf galaxies do not yet exist even though they are the most abundant class of galaxies across all redshifts and will dominate observations with next generation instruments (e.g. JWST, WFIRST). Different merger identification techniques are required for dwarf galaxy pairs due to their morphological differences from massive galaxies. Calibrated observability timescales are necessary to transform pair fractions of low mass galaxies into merger rates, yet these calibrations currently only exist for massive galaxies. Cosmologically motived orbital parameters for dwarf galaxy interactions are needed to calibrate the observability timescale in this low mass regime. Only two local dwarf galaxy pairs have detailed orbital models (the Magellanic Clouds and NGC 4490/4485), and both are consistent with high eccentricities and long orbital timescales. Nevertheless, it is unclear whether dwarf galaxy pairs generically display these orbital characteristics. I will present findings that utilize the Illustris simulation suite to constrain the frequency of dwarf galaxy pairs out to z~4 and show that analogs of the Magellanic Clouds were much more abundant at earlier times. I will also discuss the typical relative separations and velocities of these pairs, placing the Magellanic Clouds in a cosmological context via their dynamics for the first time. This work will enable calibrations of dwarf galaxy observability timescales, which are crucial for interpreting merger rates in the era of JWST.