How many planets are there?

A planet crossing in front of its star dims it by a tiny, repeating amount. That is the transit method, and it requires the orbit to be lined up almost edge-on from here, which most are not.

A planet also pulls its star around a small orbit of its own, shifting the star’s spectrum back and forth. That is the radial velocity method, and it favours heavy planets close in.

Both are biased in known directions, which is what makes correcting for them possible.

Take the detections, divide by the probability that each kind of planet would have been detectable, and the answer becomes a rate rather than a count. Analyses of the Kepler survey’s Sun-like stars put the occurrence of small planets in modest orbits at a substantial fraction of stars.

Different teams working from the same data have reached different numbers, because the correction rests on judgements the data do not settle: how completely faint signals were recovered, how many of them are real, and how large their stars are.

This is the number everyone wants and the one the data support least. It requires extrapolating from planets on short orbits, which are easy to detect, to planets on orbits of a year or more, which are not, and for many stars like the Sun the habitable zone lies beyond the longest orbit Kepler could detect.

Published estimates for how common such planets are differ by more than a factor of ten, and the spread is honest rather than sloppy: most of it comes from where the habitable zone is drawn and how far the data are stretched.

The boundary with brown dwarfs is a convention rather than a discovery, and objects drifting through the galaxy with no star at all may outnumber the stars. The survey that counted them calls them planets; the IAU’s working definition does not.

Definitions are not facts about nature, and arguing about them is only useful when the argument changes what gets measured.

The most common kind of planet found so far has no analogue in this solar system: larger than Earth, smaller than Neptune. The reading with most support is rocky cores wrapped in hydrogen, and some researchers argue for worlds rich in water. It is not settled, and the answer decides what the commonest planets in the galaxy are made of.

Their atmospheres are now measurable for a handful of favourable cases. A handful is not a census.

Thousands are confirmed. The corrected rates imply that planets outnumber the more than a hundred billion stars in this galaxy, and NASA puts the total past a trillion. The uncertainty is a factor rather than a percentage.

That is a real result and it is not a precise one, and those two things are allowed to be true at once.