Abstract: Living systems are highly complex and detailed, requiring a convergence of many disciplines, with different experimental, computational, and theoretical approaches. Part of what makes them so complex is that they involve dynamics at many interacting scales: synaptic plasticity, for example, involves short- and long-term interactions, affecting everything from pairs of neurons to cortical organization, and influencing an organism’s interactions with its environment. Similar problems have been addressed in statistical physics (my background area), where, for example, macroscopic properties of gases (like pressure and temperature) have been related to microscopic interactions through fundamental forces. While not as complex as living systems, one of the keys to the successes in physics is the strategic simplification of physical systems. I will show how such a simplified model of neural networks led to a testable hypothesis and how this is now being used to understand neurological disorders. To finish, I will describe a well-established systematic method from theoretical physics used to determine which details can be ignored, which must be included, and the obstacles to applying this technique to problems in neuroscience.