When choosing a restaurant, should we visit an old favourite where we know we shall get a good meal, or try somewhere new, which may be better or worse? This illustrates the ‘explore/exploit trade-off’, a model developed in psychology and computer science as a study of behaviour in situations of imperfect knowledge.1 One can readily appreciate the dynamic relationship between the two approaches. How many restaurants must we try before being confident of having found the best one to return to? Most academic journals focus on exploration, welcoming descriptions of new disorders, novel pathophysiological mechanisms and trials of new treatments. In contrast, Practical Neurology aims to exploit what we already know—enjoying a good meal on return to a favourite restaurant.

In a clear example of exploiting knowledge, Michael Foster and colleagues review first-line immunosuppression in neuromuscular diseases on page 328, bringing together what is known from clinical …