> The same is true of a black hole. Yes, as I wrote in the second paragraph: >> An object can take a hyperbolic orbit arbitrarily close to a black hole withou

> The same is true of a black hole. Yes, as I wrote in the second paragraph: >> An object can take a hyperbolic orbit arbitrarily close to a black hole without falling in. It occurs to me that in your various comments here you are thinking of the point mass (or divergence of the Kretschmann scalar or whatever) as the black hole. Conventionally, and for good practical reasons, practically everyone working with astrophysical and theoretical black holes define the horizon as the black hole. It's frequently tempting to think of the point-mass in Schwarzschild as the generator of the event horizon. After all, it's usually described as being a surface at r = 2GM/c^2, with "M" doing the heavy lifting, if you'll pardon the expression. However, Schwarzschild is an eternal black hole, rather than one that forms by gravitational collapse. For the case where there is some matter and no black hole -> some matter + a black hole, it is the early configuration of the "some matter" that generates the event horizon. If one, following Lemaître-Tolman <https://en.wikipedia.org/wiki/Lema%C3%AEtre%E2%80%93Tolman_metric https://en.wikipedia.org/wiki/Lema%C3%AEtre%E2%80%93Tolman_m...>, takes a spherical shell of radiation with a total momentum-energy comparable to a galaxy all moving radially inwards, and starts the spherical shell at billions of light-years from the shell's centre, then anything already at the centre (even a small interplanetary civilization) is already inside the event horizon before the civilization's home planet formed. Barring faster-than-light travel, nothing within (or produced by) the "victim" solar system will be able to cross outside a surface near the trailling edge of the inrushing radiation: at early times all possible low-speed trajectories from the victim solar system ultimately recurve back to a (set of) point(s) within it and at late times all possible high-speed trajectories recurve inwards. Near the latest time in the collapse, everyone outside the victim solar system can conclude the victims are inside a black hole, even though for hours to days (and much longer, if we make the total mass of the shell extremely large) the victims inside will still be going about their business wholly unaware (because "c") of their fate. Horizon = yes. Singularity = no ("not yet", perhaps). Finally, > 90% of the matter orbiting a black hole will never fall into it. is probably wrong, especially if one takes "never" literally. The dynamical spacetime in the near horizon region is on its own probably enough for orbital decay of anything in close orbit (a few tens of R_{crit} ~ R_{Schwarz.}) and there are plenty of forcing functions on bodies in elliptical orbits at greater remove, particularly in galaxy cenrtres and globular clusters.