How do you explain the orbits of Uranus and Neptune without invoking a 'Planet X'?

Computer simulations of the orbits of these three planets seem to suggest that they are in a 'chaotic' state, particularly Neptune and Pluto.

What this means is that, in addition to Neptune and Pluto being in a roughly 2:3 orbital resonance pattern, (two Pluto 'years' equals about three Neptune years) the orbits themselves are far more complex over time than a simple calculation from Newtonian Physics would suggest, over the span of many orbits. They are not perfect ellipses, and the planets trajectories do not rejoin themselves to make a closed curve.

This doesn't solve all of the mystery of the previously reported deviations in their orbits, but it suggests that the need for a Planet X to account for them is not as great as it was once thought to be. In other words, the deviations fit the model of a chaotic system better than they do the need for a more distant perturbing body in an orbit very highly inclined to the plane of the solar system.