The number of variables would almost certainly be less than, say, our simulations of the weather. Admittedly, we are far from being able to make reliable long term climatic predictions, but as our mathematical sophistication grows, so also does the accuracy of the predictions.
The stock market knows all about this. Futures traders lie awake at night dreaming of mathematical models (mostly charts) which reveal the underlying trends of gold and grain trading.
Arguably, the social behavior of genuinely stupid and evil people would be far easier to model than independently and free minded people
Said differently, in a political chess game it should be much easier to defeat a Bush/Republican (or dem "centrist"), than a free minded liberal or independent. The reasons for this are twofold.
The first is the self evident fact that G. Duhbya is the tip of an iceberg of genuinely stupid people. If the neocons, for example, were half as smart as they claim to be, then why is
Ergo, using Jesus' criteria of "Ye shall judge the by their fruits", the neocon lobby is demonstrably stupid, however much they may decorate their unAmerican agendas with obfuscating rhetoric.
Of course, to comment on the IQ of religious fanatics is like shooting fish in a barrel. Anyone who actually thinks the Earth (or is it the local group?) is going up in a cloud of Armageddon smoke any day now, necessarily has a primate-like cerebral cortex.
The second reason the pugs should be easy to defeat in a political chess contest is their highly predictable utter selfishness and greed (evil?); since the more predictable the variables, the more predictable the model.
Thus, a fairly reliable simulation of the herd behavior of stupid and evil people (i.e., Bush/Republicans & neocons) should be well within reach of a pc with lots of megs and a select few mathematical programs, e.g., Mathematica and/or Maple, plus some statistical SAS, or even Minitab. Obviously, this is a minute sample of the quantitative programs out there, but the point is that even these few would probably be adequate to assist in the construction of a testable model.
Ironically, the chief vices of the Bush/Republican Administration (i.e., greed and stupidity) are precisely their chief vulnerabilities, since a predictable foe is more easily mathematically modeled (and defeated!) than a creatively intelligent one.