When a star falls sufficiently close to a supermassive black hole, it is disrupted by the enormous tidal forces and then the debris feeds a sudden burst of accretion. Such tidal disruption events (TDEs) provide an ideal laboratory to study the otherwise dormant black hole population, accretion disk physics, and many general relativistic effects. However, due to numerical challenges, simulations have not been able to capture the full range of timescales and physical processes needed to compare with observations. Here, I will present a new hybrid method that circumvents the challenges and allows global simulations of realistic TDEs. Our model provides a unification of the diverse behaviors of TDEs found in the X-ray and optical bands. On-going surveys like ZTF and eROSITA promise to increase the current sample size from a few dozen to many hundreds. I will summarize the open questions that may be solved in the near future.