From progenitor studies, type II-Plateau supernovae (SNe II-P) have been decisively and uniquely determined to arise from isolated red supergiant (RSG) stars, establishing the most homogeneous --- and well understood --- progenitor class of any type of core-collapse supernova. The physical process by which these stars explode, however, remains a mystery. A fundamental clue to the nature of the explosion mechanism is explosion geometry: In short, are supernovae round? Because young supernova atmospheres are electron-scattering dominated, their net linear polarization provides a direct probe of early-time supernova geometry, with higher degrees of polarization generally indicating greater departures from spherical symmetry. In this presentation I will describe the ongoing work being carried out on RSG explosion geometry by the SuperNova SpectroPOLarimetry project (SNSPOL), with a particular focus on SN 2013ej -- an SN II-P that exhibited remarkably high polarization just days after the explosion, and for which twelve epochs of spectropolarimetry trace an intriguing tale about its geometry deep into the nebular phase.