I'm going to try to express a very big idea. I hope I can do it justice.

Nassim Taleb coined the term "ludic fallacy". Ludic is the latin word for games. Probability theory and game theory are sufficiently advanced to predict economic returns for gambling casinos. However, the economy as a whole is a different animal.

Probability theory and game theory are fashionable on Wal-Street. Quants (analysts) force their data to fit the Gaussian models, and express their uncertainty in standard deviations. Based on these models many people thought they understood how much risk they were taking. We all know outcome when they are wrong. The real world, and subsets like the financial markets, are not like games. The rules are not well defined, and rare events can turn the whole systems upside down.

We don't have any logically sound way of making decisions under uncertainty. Gaussian models are helpful most of the time, but sometimes they are woefully wrong. There are other mathematical models, but they are not logically sound either. We need to acknowledge deep uncertainty. In the words of Donald Rumsfeld, "there are unknown unknowns".




This idea also informs my view of regulation going forward. Many people are calling for the government to regulate the procedures used by credit rating agencies. This is a horrible idea because the regulators have no logically sound procedure. If the government did oversee credit rating procedures, this would lead to an even more dangerous level of misplaced trust in flawed models.