3 ms·
Ah, the cruelty of power law distributions. Problem is they come up everywhere, all the time -- as often, if not more often, than Gaussian distributions. https
by Hasz 8y ago
Ah, the cruelty of power law distributions. Problem is they come up everywhere, all the time -- as often, if not more often, than Gaussian distributions.
https://en.wikipedia.org/wiki/Power_law https://en.wikipedia.org/wiki/Power_law
- throwawaymath 8y agoAs a little aside here, gaussian distributions aren't really that common. Even if we put ignore splitting hairs about "true" Gaussian versus approximate Gaussian (which would just be pedantic), Gaussian distributions seem to show up frequently for two reasons. The first reason is because they're attractive and useful as a model, so many measurements are specifically designed to approximate a Gaussian. These are usually artificial, not naturally occurring. The second reason is because it's easy to abuse the term to refer to anything that looks decently uniform visually, if not rigorously, and hand wave away the details as long as it gets the rhetorical point across. The first phenomenon leads to many natural things being interpreted as Gaussian when that's just an artifact of measurement (as an example, see the BMI scale). The second is very similar to the way you can reliably hear people use the term "exponential" to describe a large absolute increase, even if that increase isn't even quadratic.
- esahione 8y agoWell, the Gaussian shows up a lot because of the Central Limit Theorem. A lot of additive processes will converge to the Gaussian.
- curuinor 8y agothe distribution of distributions that you find in textbooks is power law distributed (or pretty close, iirc). at the top of that power law distribution is, ironically, the gaussian.