It is said sometimes that physicists constantly simplify the problems they are trying to solve by considering simple models. This is true but not the whole story, or even the important part of it. Simple models must also be relevant, otherwise they are worthless. This is usually expressed by saying that the said simple model contains the important/relevant physics one is trying to address. This, I think, is far more crucial to what physicists do than the simplicity of their models which is a reflection of our modest abilities. It is a reflection of our immodest abilities in finding the correct solution.

The reason I am writing this is that today I saw Richard Muller's lecture webcast of Tue 11/7/2006, which started, due to the occasion, by his icecream model of elections. It is a simple model that argues there is no or little difference between the candidates in a two-party system. However, it does so by assuming a whole load of assumptions that are not explored in Muller's short discussion. For instance, it assumes there is a continuous, static, and spacially coordinated one-dimensional spectrum of political opinions. It also assumes that the candidates can move more or less continuously over this political spectrum without losing any votes on the opposite side of their oppponent. In short, it is a simple model, but one that does not contain the right physics.