Why are we not finding as many multi-planet systems as we should? For a long time this question puzzled scientists, but a change to a small assumption used in modeling planetary systems now makes sense of the data we have received from the Kepler Space Telescope. Ironically, the assumption that works involves systems being more similar to our own.

The vast wealth of data produced by the Kepler Space Telescope has given us a chance to build statistical answers to the question of how like our Solar System other systems are. Early efforts concluded we're not finding as many multi-planet systems as we should. To explain this, a theory known as the Kepler dichotomy was produced, which assumed there are two sorts of planetary systems – one rich in planets and the other with only one or two. Even those who proposed this idea couldn't explain why there wouldn't be plenty of systems with intermediary numbers.

However, in a paper accepted by the Monthly Notices of the Royal Astronomical Society (preprint on arXiv), Australian National University PhD student Tim Bovaird and his supervisor Dr Charley Lineweaver claim the dichotomy is false. Instead, the results Kepler is giving us can be explained by rethinking the relationships between planetary orbits. Oddly enough, Bovaird and Lineweaver argue the problem has come from an assumption that most star systems differ from our own.

A common simplification when describing the Solar System, particularly in artistic renderings, has all the orbits in exactly the same plane. Although the planetary orbits are actually very close to co-planar (unlike many comets, for example), the alignment is not perfect. If it was, we would see transits of Venus every year or two, instead of a couple of times a century.

Planetary orbits can differ from being perfectly co-planar in two ways. Bovaird and Lineweaver refer to these as “flat” and “flared”, based on whether the distribution of orbital inclinations depends on a planet's distance from the Sun or not.

“Simulations with flared planet systems were slightly easier to perform and that is what researchers had assumed,” Bovaird said in a statement. When Bovaird and Lineweaver repeated the simulations using a flat model, they found it would be slightly harder for Kepler to detect multi-planetary systems than we thought.

In other words, it's not that the galaxy has an excess of stars with a small number of planets, it's just that we're often looking at systems with many planets and only picking up one, because the orbits of the others don't cause them to pass across their star's face, as seen from our location.

Our own Solar System is very flat until the orbit of Jupiter, and then flares significantly for the outer planets. Lineweaver pointed out to IFLScience that Kepler's observations have been all about the inner reaches of other systems – mostly finding planets closer in than Venus. Consequently, systems that most resemble our own look flat to Kepler.

Lineweaver admitted to IFLScience that he is “drawing a long bow”, but he thinks flat systems may be more conducive to life. Flared systems might see volatile molecules, including water, incorporated into planets at different abundances from flat ones. We know that at least one flat system delivered a planet ripe for life, while it is possible planets in flared systems might have less favorable compositions.

^{A flat planetary system would see all planets on a horizontal line in this graph, as the planets (marked by their first letters) do out to Jupiter. Flared systems curve upwards, as seen with the dotted lines. Objects detected by Kepler (with numbers shown on the right hand scale) are overwhelmingly concentrated closer to their star than Mercury. Boviard and Lineweaver/Monthly Notices of the Royal Astronomical Society}