I have been working on theoretical and computational studies of the electronic, optical and magnetic properties of semiconductor heterostructures for well over a decade. During this time I have had to follow through various theoretical derivations from either books or research papers and frankly I've struggled time and time again. There never seems to be enough detail and what is commonly a couple of lines in a research paper can literally turn into many pages of precise mathematics. With the advent of computers and their wide application to science, the emphasis of theoretical work has changed. Years ago, theoreticians aimed to produce short neat relationships, which summarised physical effects. The concentration dependency of the metal-insulator transition of Mott is an excellent example of a complex process summarised in a compact equation, which could then be evaluated by hand. The modern approach to the same problem would be to take a microscopic model of a semiconductor, calculate the bandstructure, and then vary the impurity concentration, thus deducing an empirical relationship from the numerics. The material system would then be changed and the process repeated. It would therefore appear that the old way was preferable, but in these times of materials modelling and device design, quantitative results are what we're paid for.