A Preponderance of Perpendicular Planets.

Observing the Rossiter-McLaughlin effect during a planetary transit allows the determination of the angle $\lambda$ between the sky projections of the star's spin axis and the planet's orbital axis. Such observations have revealed a large population of well-aligned systems and a smaller population of misaligned systems, with values of $\lambda$ ranging up to 180$^\circ$. For a subset of 57 systems, we can now go beyond the sky projection and determine the 3-d obliquity $\psi$ by combining the Rossiter-McLaughlin data with constraints on the line-of-sight inclination of the spin axis. Here we show that the misaligned systems do not span the full range of obliquities; they show a preference for nearly-perpendicular orbits ($\psi=80-125^\circ$) that seems unlikely to be a statistical fluke. If confirmed by further observations, this pile-up of polar orbits is a clue about the unknown processes of obliquity excitation and evolution.