It seems whenever scientists start to think their models of planetary formation are falling into place, something comes along to challenge them. The latest example is a star with barely enough mass to be a star, orbited by a planet with a mass half that of Jupiter.

Astronomers have noticed that most of the planets we have found around low-mass stars are closer to the Earth or Neptune in size than to Jupiter. This is consistent with theories about the way planets form – fairly simply a gas cloud that only has enough material to create a low-mass star isn't likely to have enough in its outer disk to form large planets.

So the discovery of GJ 3512b was a surprise, and one that attracted so much attention the paper announcing the discovery in Science has 183 authors. Even by the standards of low-mass stars, GJ 3512 is small, one-eighth the Sun's mass. Yet GJ 3512b is a minimum of 0.46 Jupiters, and may be more massive still.

GJ 3512b's orbit lasts 204 days. Considering how faint GJ 3512 is (0.00157 times the brightness of the Sun), less light falls on the planet than Jupiter, but the orbit is very stretched out, creating brief periods of relative warmth. There are signs of a second, much more distant planet, but the signal isn't clear enough to be confident it exists, let alone deduce much about it. Such a planet would, however, explain the long, thin orbit, considered a much more likely outcome where two planets interact than where one orbits solo.

M-type stars, those with mass less than 60 percent of that of the Sun are the most common category in the galaxy, but only 10 percent of the planets we have so-far discovered orbit these sorts of stars. That's partly because we've spent more time looking at stars similar to the Sun, which after all we know are capable of supporting life, than M-types, also known as red dwarfs.

However, it is also because larger planets are easier to spot, and red dwarfs usually don't have anything all that big. Intriguingly, where we have found gas giants orbiting M-type stars they've usually been at large distances, something that also doesn't suit the most commonly used techniques for planet spotting.

The characteristics of the system led the authors to consider an alternative disk instability model for planet formation. This might convert a larger portion of the mass of the proto-planetary material into actual planets, explaining the outcome we see here.

In an accompanying editorial Yale University's Professor Greg Laughlin suggests very low-temperature stars leave their surrounding disks cold enough to collapse efficiently, leading to large planets like this one.