Just shooting a question out to the HN world.<p>It seems to me that the symptoms of most of the problems in today's economy play out via unpredictable network effects.<p>Obviously the majority of the causes of the problems, I would think, are a combination of political retardation (literal and figurative, with apologies), as well as faulty (if not fraudulent) accounting.<p>Has anyone ever attempted to devise a stats program where our literal economic relationships (individuals connections to banks, companies, salaries, etc) are treated as constants (e.g. run within code), and the "problems" (politics, bad math, etc) can be adjusted variably to see their effects?<p>Would this even work? Are there too many "known unknowns"? Am I stupider for having said that? I'm just a regular web developer asking big comp sci questions, and feel kind of lost.