Simple rules for simple models

In a world dominated by large-scale numerical simulations and applications of machine learning to biology, medicine and neuroscience (and complex systems, in general), each driven by the availability of 'big data' and decreasing costs of computation, the art form of constructing simple models is no longer in the focus of scientific attention. Yet, the most successful of these 'minimal models' have brought about a revolution of our understanding of complex systems, in particular in the life sciences, and are still shaping our view of the world today. It is hard to look at a power-law distribution without thinking of the Bak-Tang-Wiesenfeld sandpile model. And it is near impossible to reflect on synchronization without imagining coupled phase oscillators. Here we have put together a few ideas on how to formulate and analyze simple models and to draw conclusions from them.