Modern optical fiber networks transmit data at astonishingly high rates—at least a hundred billion ($10^{11}$) bits per second in a single optical fiber. The error rates caused by physical phenomena are amazingly low—in the range of one error per trillion ($10^{12}$) bits or lower—but still need to be detected and corrected. How can one simulate such a system effectively and efficiently to understand and deal with the transmission errors? Not by a straightforward Monte Carlo simulation; the number of sample points would be prohibitively high. This is an example of the type of problem addressed in this issue's SIGEST paper, “A Method to Compute Statistics of Large, Noise-Induced Perturbations of Nonlinear Schrödinger Solitons” by R. O. Moore, G. Biondini, and W. L. Kath, originally published in 2007 in the SIAM Journal on Applied Mathematics. As the paper's title suggests, the type of communication network that it considers is a soliton-based transmission system. (Solitons are wave pulses that maintain their shapes while traveling at constant speed.) The mathematical model that is used is the nonlinear Schrödinger equation, which governs the propagation of optical pulses under simplified conditions. The authors' approach is not limited to this simplified model, however, as they have shown in subsequent work. The key technique that is used by the authors to conduct effective and efficient simulations is importance sampling. This technique does what the words connote: concentrate the sample points in regions that are most consequential. Of course, this is much easier said than done. In this instance, the authors accomplish this by using approximate analytical knowledge about the behavior of the transmission system that comes from soliton perturbation theory to bias the sampling to regions where events of interest occur. Importance sampling is an important technique in general, and one of the bonus aspects of this paper is that it provides readers with a nice general introduction to importance sampling as well as a specific application. A consequence of this research is the ability to effectively simulate soliton-based optical transmission networks governed by the nonlinear Schrödinger equation several orders of magnitude more quickly than by standard Monte Carlo simulations. The numerical simulations towards the end of the paper illustrate this. The overall result is a very nice demonstration of the use of sophisticated and deep mathematics to aid in the understanding of an important practical problem, written in a very accessible manner that should interest a broad range of SIAM readers.