I this talk I will describe a picture which has emerged over the past few years regarding the statistical interpretation of semiclassical gravity and how this relates to wormholes, averaging and the so-called factorization puzzle, the information paradox, and a combinatorial description of 3d gravity.