The existence of chaos among the system of Jovian planets (Jupiter, Saturn, Uranus, and Neptune) is not yet firmly established. Many investigators (Sussman, Wisdom, Holman, Murray, among many others) have consistently measured a Lyapunov time of between 5 and 12 million years. Murray and Holman demonstrated that the chaos arises from the overlap of three-body resonances, and Guzzo has corroborated their theory across a wide range of system parameters. Conversely, other investigators (Laskar, Newman, Grazier, and Varadi, among several others) have compelling evidence against chaos. Namely, Newman et al. have convincingly demonstrated that a sympletic integration using the Wisdom + Holman symplectic mapping with a 400-day timestep reproduces the chaos seen by others, but that the chaos disappears and the orbit converges to being regular as the timestep decreases.

Using high-precision integrations and convergence testing, I demonstrate that the resolution of the apparent paradox is simple. The orbital positions of the Jovian planets is known only to a few parts in $10^7$. It turns out that, within the observational error ellipsoid, there exist both chaotic and regular solutions. Thus, some investigators legitimately find chaos, while others legitimately find no chaos. The question of whether the Outer Solar System is really chaotic cannot be answered using current observations.