A red supergiant (RSG) is the most common manifestation of a massive star at the end of its life.When collapse of the iron core of the star leads to a successful supernova (SN) explosion, the observational signature is a Type IIP SN, which is the most common type of SN. The successful explosion of a RSG usually leaves behind a neutron star while the rest of the enriched stellar matter is returned to the surrounding environment. The SN mechanism can fail in a reasonable fraction of stellar deaths. In that case, the core unavoidably forms a black hole (BH) that consumes the inner, metal-rich layers of the star and a large fraction of the extended, convective hydrogen envelope remains bound to the BH. This thesis considers the subsequent evolution of the BH and envelope following a “failed” SN including the critical questions of whether there is still an explosion or observable transient and how much mass the BH is ultimately able to accrete. In Chs. 2 and 3, I use three-dimensional simulations to show that the convective envelope of a RSG has too much turbulent angular momentum to accrete and that infall of the turbulent material leads to a ∼1e48 erg explosion of star and associated transient resembling a luminous red nova. Magnetic fields are important in the envelopes of RSGs but we do not know how they will affect the accretion flow following a failed SN. In Ch. 4, I show that the magnetic fields in the envelope are tangled with strengths of the order of hundreds of Gauss, but grow more shallowly during collapse than r^−2 such that angular momentum and magnetic fields became dynamically important at similar radii. In the final chapter, I discuss my work modeling the light curves of failed supernovae. This work will apply to a wider range of scenarios in which mass is stripped from the envelopes of evolved massive stars. An important application of this work is to differentiate between luminous red novae of binary-merger origin and those arising from BH birth in failed SN. Overall, this work is important for contextualizing the upcoming LSST survey on the Rubin Observatory, which will observe 100s–1000 times more luminous red novae than the Zwicky Transient Facility. In addition, these cooler explosions self-enshroud as they form dust; detailed modeling of these events will help leverage the infrared transient sky made accessible by WINTER, DREAMS, JWST, and Roman.