For a two week period during the Joint European Torus 2012 experimental campaign, the same high confinement plasma was repeated 151 times. The dataset was analysed to produce a probability density function (pdf) for the waiting times between edge-localized plasma instabilities (ELMs). The result was entirely unexpected. Instead of a smooth single peaked pdf, a succession of 4–5 sharp maxima and minima uniformly separated by 7–8 ms intervals was found. Here we explore the causes of this newly observed phenomenon, and conclude that it is either due to a naturally occurring self-organized plasma phenomenon or an interaction between the plasma and a real-time control system. If the maxima are a result of 'resonant' frequencies at which ELMs can be triggered more easily, then future ELM control techniques can, and probably will, use them. Either way, these results demand a deeper understanding of the ELM process.