"Being an economist is like being a weather forecaster in a world where there are gods that routinely mess with the weather."
I like this analogy because it works on several levels. Most people appreciate how complicated and hard to predict the weather is. Even though the basic science is well known, there are a lot of confounding factors. Tiny, unobservable fluctuations in the system can get magnified into large effects that seemingly come from nowhere. Economics is also like that, making predictions very difficult.
But we have added complications. The weather forecast never depends on what people think or how they react. Meteorologists never have to worry about psychological factors like expectations or incentives or the political process. But there are lots of 'gods' that can whip up economic storms at will. Maybe they do it as a response to somebody's prayer, and maybe they are just following selfish goals. A meteorologist in this world would have to anticipate these actions in order to make good predictions.
And then there is the added complication that the economists can change the actions of the 'gods'. Economists have to figure out how to do the economic equivalent of telling Poseidon how to manage his storms so as to best further his goals, with the added complication that Zeus and Hera and all the rest are also messing around with the weather at the same time. We have to predict what people will want to do and how their actions will interact in a system that is already full of chaos and instability.