In a recent article (Liang and Lan, (2011)), we showed that the trajectories predicted by general-relativistic
and Newtonian mechanics from the same parameters and initial conditions for a low-speed weakgravity
bouncing ball system will rapidly disagree completely if the trajectories are chaotic. Here, we
determine how accurate the parameters and initial conditions of the system must be known so that
the two different calculated chaotic trajectories are sufficiently accurate for an empirical test.