Back in the old days, sometimes two good TV shows came on at the same time. Some people would just choose one show to watch, but I would bounce between them. I would change the channel as soon as the commercials started for one show, then do the same at the commercial break on the other show, and keep following that pattern. (It turns out it wasn’t a great strategy, as I usually missed the best parts of each show.)
Just like me bouncing between TV shows, planets can bounce between stars. In most cases a planet just orbits one star at a time. There are a few exceptions, like “Tatooine binaries”, in which a planet orbits around a pair of close binary stars. And as we know, free-floating planets don’t orbit a star at all.
A bouncing planet will orbit one star for a while, then bounce off to orbit a different one. This bouncing can take place in binary star systems in which two stars share a common, relatively wide orbit (about 100-1000 au in the examples, where 1 au or “astronomical unit”, is the Earth-Sun distance). Bouncing can only happen when a planet has the right orbital energy. Too little energy and the planet is stuck orbiting one star. Too much energy and the planet flies off and is free from both stars’ gravity. In between is where things get interesting.
This diagram shows four examples of planet orbits in a binary star system. The number C represents a measure of the orbital energy (more precisely, the Jacobi Constant) of a planet’s orbit. The planet cannot enter the gray regions, so for C=5 the planet is restricted to orbiting the red star. For C=3.5 or 3.1, the planet bounces between stars but, after a wandering path, ends up escaping entirely. In between, when C=3.7, the planet bounces between the two stars but cannot escape.

It’s pretty awesome to imagine a planet bouncing back and forth between two stars like in the setup with C=3.7. But how would that situation ever arise?
A planet cannot form in a bouncing configuration, but it can be kicked into one. The orbital energy of a planet can change. One way this can happen is when it is kicked by the gravity of other planets in the same system, as in the case of planet-planet scattering (which I’ve discussed a lot in the past, such as in When good Jupiters go bad):
Imagine this. A system of Jupiter-like gas giants forms around one member of a binary system. The gas giants’ orbits start to cross and the planets gravitationally scatter each other. So far, this story looks a lot like the animation above. Around a single star, the most common outcome of scattering is that one planet is ejected to interstellar space, leaving one or more survivors on stretched-out (“eccentric”) orbits around the star. But around a binary star, a scattered planet often bounces around to the other star before being ejected.
The following animation shows the evolution of two planets scattering off of each other in a binary system consisting of two Sun-like stars separated by 250 au. First the planets are trapped around the red star, but after being gravitationally scattered, one planet bounces to the other planet many times before finally being ejected from the system entirely.
You might wonder whether a bouncing planet could be captured by the other (blue) star. In the simple setup like the one shown above, capture is impossible. The bouncing planet needs to slow down and lose orbital energy to stay in orbit around the other star, and there is no way for that to happen.
But capture is possible in real systems. For instance, if the action happens early enough that the other star has a disk of leftover gas, then the scattered planet’s orbit can lose energy by plunging through the disk and be captured. The same thing can happen if there is a disk of leftover rocky or icy bodies. If the other star has its own planets, then the incoming scattered planet can even gravitationally disrupt the native planetary system and end up surviving in orbit. Another less-likely outcome is that the scattered planet could collide with one of the planets — this would have the effect of moving all of the planet’s mass from one star’s orbit to the other’s, but the collision would dramatically shake up the planet (and wipe out any chances for survival).
A planet can also be transferred between binary stars during stellar evolution off of the main sequence (see this blog post for more discussion of stellar evolution). This is because the disallowed orbits (the shaded regions in the image and animation above) depend on the stars’ masses also. As a star becomes a red giant and loses mass, a “window” can open that allows the planet to bounce to its companion star. That window then closes after more mass is lost from the star (details in this paper).
Planet bouncing and transfer is nice and all, but does it ever happen in real life? Almost certainly, yes! Binary stars are a dime a dozen. There is even an astronomical saying: “three out of every two stars are in a binary system.” Gas giants form around ~10-ish% of stars, so there have been billions of opportunities in the history of our Galaxy for planets to bounce between binary stars and to be captured.
I think the concept of planet bouncing can lead to some pretty cool science fiction scenarios. Let me explore a couple.
Setup 1. A system of planets forms around a star with a binary companion. The system contains three gas giants and three rocky planets, including one — let’s call it Earth 2.0 — with the right conditions for liquid water and life. A few hundred million years after the planets form, the giant planets go unstable and scatter each other all over the place. Earth 2.0, which by now hosts a civilization, is caught in the crossfire and kicked outward. Its orbit bounces back and forth between the two stars. In the meantime, one of the gas giants is kicked outward and also starts bouncing. Earth 2.0 passes close to the gas giant while in orbit around the second star — the encounter removes enough orbital energy from Earth 2.0 to strand it on a stable orbit around the binary star. The civilization on Earth 2.0 must adapt to its new system and its much colder, stretched-out orbit, while some of its writings still recall the “good old days” before the instability…
This story could also be told from the point of view of the inhabitants of a planet around the second star, who see something strange happen in the planetary system orbiting their binary companion, and then a hundred or more years later are confronted with an intruding planet…

Setup 2. Life develops on a system of five large moons around a gas giant in a binary star system. The gas giants undergo a dynamical instability and scatter. The three outer moons are stripped from the gas giant. One is captured by another gas giant which, along with the other two moons, are scattered out and start bouncing between the two binary stars. One of the scattered moons is kicked onto a stable orbits around the second star, becoming a planet rather than a moon (also called a “ploonet“). The scattered gas giant and second scattered moon are both eventually ejected into interstellar space as free-floating planets. Luckily, both former moons have enough thermal blanketing to maintain liquid water. The scattered moon can retain liquid water for another billion years under a thick layer of ice, and the captured moon is heated by tidal dissipation from interactions with its gas giant captor. Meanwhile, the engineers of the ploonet spend their time developing propulsion technology to push a spacecraft hard enough to be able to bounce back to their star of origin…
(Side note: moons really are destabilized and sometimes captured when giant planets go unstable and gravitationally scatter off of each other, as you can see in this animation I made [from this paper]):
There you have it: planets bouncing between binary stars. Boom!
Additional Resources
- Nick Moeckel and Dimitri Veras‘ paper: Exoplanets bouncing between binary stars
- Kaitlin Kratter and Hagai Perets‘ paper: Star-hoppers: planet instability and capture in evolving binary systems
- Binary Systems: friends or foes? (blog post)
- How planets die: When good Jupiters go bad (blog post)
- How planets die: Roasted, Toasted and Swallowed by their evolving stars (blog post)
- Exo-moons: Innocent bystanders during giant planet instabilities (blog post)
Very cool. I wonder if “juggling” stars are possible. A planet in a hyperbolic trajectory goes from one star in a binary system to the other with the stars’ orbit being timed to sling it back. Probably not stable or metastable. Maybe there is a way to make it stable. What if there were two planets being “juggled” and doing momentum transfer..?
Many thanks Sean!
Hmm — so you mean the planet would bounce back and forth every orbit? Hmm, I wonder whether that is possible…. It would be pretty cool if that setup was possible. I imagine it could be fine-tuned but I’m honestly not 100% sure how to figure that out. I will think about it.
Yeah, every “orbit” for the planet, but not necessarily for the binary. Let me know! 🙂
I think the planet would have to be in a retrograde orbit in the simplest case.
this may or may not be a part of a story i am writing…maybe…okay most likely a couple stories.
Awesome — I’d love to see what you come up with!
Yeah, it would make a good disaster story. I’m imagining something like Nightfall. 🙂