Before we realized that supermassive black holes were at the center of large galaxies, how did the models explain galaxies? So, I was watching this show on Netflix about supermassive black holes. I didn't realize that astronomers were surprised to find these black holes at the center of galaxies, because I cannot imagine how models of the centers of galaxies worked without these black holes. In my head, when I imagine the center of a galaxy without a black hole, I picture the stars at the center pulling on each other in a way that would pull their fellow stars rather chaotically out of their nice elliptical/circular orbits. And, when I start to think back to the origins of the galaxy, I can't see this chaotic nuclei spawning these relatively long, stable orbits on the outer edges. So, like the title says, how did these models work? How do multiple stars form stable orbits around nothing, despite each other pulling from their orbits? Another way to ask this would be, how could several stars form stable orbits around nothing?" 
Now, I am sure someone will ask, "Who said they formed their orbits around 'nothing?'" To which my reply is the scientist who was on the "Nuker Team" that discovered the supermassive black hole at the center of the Andromeda Galaxy. Though, to be fair, she did not say they believed there was "nothing" at the center of galaxies before the black holes were found; she said they believed only dust and gas was at the center. I reasoned that if the gravity of the dust and gas was not strong enough to pull itself into planet/star/etc., then it certainly was not exerting enough gravity to keep the inner stars of a galaxy in place; thus, my "nothing." So, if they were wrong and astronomers did NOT believe this, leave a comment saying what they did believe.
 A: The stars in the galaxy don't really orbit the black hole in the center of the galaxy. They all orbit a common center of gravity. Obviously, a lot of the mass is in the black hole, and the center of gravity could very likely lie inside the black hole's event horizon, but it's not required.
Look here for a cool animation of Pluto and Charon orbiting a center of mass that's clearly not in the center of Pluto.
If we have one body that's far more massive than the other bodies, the center of mass will clearly be heavily weighted (literally) towards the center of the massive body. But we could hypothetically have a bunch of similarly-massed stars orbiting a common center of gravity without any particular star being at the center.
Similarly, you could have one system of low-mass stars loosely orbiting each other, then that entire system could be orbiting a single high-mass star. So the low-mass stars would quickly orbit a common center of gravity, then the entire low-mass system and the high-mass star would slowly orbit that common center of gravity.
If the total mass of the low-mass stars was similar to the mass of the high-mass star, the star and the star system would appear to be orbiting empty space, while all the stars in the star system would appear to be orbiting an empty space that's orbiting the above empty space.
A: There is no requirement for a central black hole in a dynamical sense. Many galaxies are not known to have one, or if they do, its mass is relatively small.
The gravitational influence of the SMBH can be quite negligible at distances that are only a tiny fraction of the size of a Galaxy. What I mean by this is that say the BH at the centre of the Milky Way is 4 million solar masses. Well you only have to move a short distance away from the centre before a sphere at that radius contains far more mass in stars, gas and maybe dark matter. The shell theorem tells us that (for a symmetric distribution) an object at this radius just feels a gravitational force due to this matter as if all its mass were at the centre.
Perhaps you are concerned that the centre of mass would be chaotically moving about because of stellar motions? How would that happen? What external force on the system is moving it?
Dynamical models of galaxies would not even include the central black hole as a feature unless they were dealing with the very central core or were specifically interested in the dynamics of stars right near the centre. If stellar densities are unusually high - for example near the core or in star clusters- it may be more appropriate to use n-body simulations which allow for interactions between the stars. Nevertheless, even though the orbits may become chaotic, highly eccentric, or even unbound, stars always orbit the common centre of mass, and the position of this cannot change without the action of external forces. Away from the very centre of the Galaxy, the probability of stellar interactions becomes very small and the stellar motions can be accurately modelled in terms of a smooth gravitational potential that does not vary on the timescale of a stellar orbit.
