I comment on how chaos might be defined. A sample of dynamical systems that have quasi-periodic forcing functions is then considered. The normal approach found in the literature is to start with an ordinary differential equation, change to a difference equation, and then plot a graph. The question of how to detect a strange non-chaotic attractor without the underlying ordinary differential equation is posed and some pointers are given as to a possible method of solution using statistical analysis. © Austral. Mathematical Soc. 2010.