Re:I think the authors missed the real point. (Score 1)

Most people only know about the existence of one distribution: the normal distribution (aka bell curve, Gaussian distribution, many others). The fact is that statisticians routinely use quite a few different distributions, some (called "heavy-tailed") which are used to simulate "rare events."

A few comments on the article:

1. Nothing follows the Gaussian distributions (the "normal distribution" is really a family). This is something humans constructed to understand their world. By the same token, nothing follows this new distribution (yet another page in my handbook, I guess). They just do "well enough" (in addition, the normal distribution is the easiest of all to work with).
2. Why is this a big deal? In fact, who was using the normal distribution to predict earthquakes and airplane crashes, anyway? There are different, more realistic models for this; a lot of studies are being done on "heavy tails" for rare event prediction, etc. etc. Again, just another distribution in the statistician's handbook.
3. For that matter, the use of the Gaussian (and all the models that assume it) need to be re-examined in other fields. E.g. there is evidence that indicates that intelligence follows a bimodal distribution (many people tend to be near one of two "areas" of intelligence, instead of near one grand average like the Gaussian case).

I guess this article shows the general public's lack of understanding of statistics, not that statistics is that easy to understand in the first place.