A lot of statistic analysis is based on parametric statistics. One of the most crucial assumptions is the bell shape distribution or normal distribution of the sample or population under central limit theorem. However, the normal distribution could be regarded as a typical distribution. There is a good deal of atypical distributions.
Earthquake, sea contour, river bank shape, stone shape, wind speed, thunder, fern leaf shape, stock price, liquidity, star size, ... the list of atypical distributions is very long. Using normal distribution to represent the distribution of a lot of things may compromise the quality of the analysis because of oversimplification and over-generalization. We should watch out for this to avoid calamity to the world such as the earthquake and tsunami in Japan in recent years which caused the biggest nuclear event in human history.
The world also pays more attention to this knowledge development. The popularity of Chaos Theory, The Butterfly Effect, Fractal Theory, and Black Swan Event exhibit this trend.