Randomized Benchmarking is an experimental technique for assessing the quality of quantum gates in experimental quantum computing devices. It was proposed less than a decade ago and is now arguably the `industry standard’ when it comes to testing quantum devices. It is also mathematically non-trivial and many aspects of its functioning did not became clear until recently or are not yet clarified. In this overview talk I will review the randomized benchmarking technique, focusing on its mathematically interesting aspects. Then I will discuss several questions about the protocol that have been answered in a more or less satisfactory way by the community (using my own contributions as somewhat of an Ariadne’s thread) and finally I will discuss several issues that are still unresolved and try to convince you that these issues might be of interest to the more mathematically inclined.