What is the best method for researching anything? History, biology, physics, politics? The answer is, from my perspective: just make a quantitative model. Or at least a semi-quantitative model.

Only once one starts to account of the flow of sufficiently immutable things, such as mass, energy, people, time, atoms, monetary value, does it emerge where the gaps in one’s reasoning are. Good rhetorics hides these gaps, and a story might feel convincing. However, if you account for the immutable things, and keep in mind the sufficiently invariant things, which could be the total number of people, or atoms, or grain, or energy in the system, then suddenly the question emerges where someone was for a few years in the 14th century? Who else was there, since three people cannot possibly do this much in such amount of time? How exactly did they get the amount of revenue they needed? Which chemical reactions are existing besides the well-known biochemical pathway? Where does this little bit of energy go?

No doubt this is surprisingly laborious and seems maybe even unnecessary. The reward is the discovery of stories that do not go in the expected way, that there is more to a question. Perhaps even some intellectual humility, if thanks to the model one knows the gaps with some certainty, and perhaps can even say the precise conditions under which story A must be the case, and the conditions for story B to be true. We may never know how the unknown conditions are, or not for a long time. But we know with certainty what we don’t know. If deciding between the stories is important, then we will hence know exactly what to research next, which is a most valuable result.