This.
Also, the unacknowledged inconsistency of his arguments is ridiculous: ~"Unpredictable events happen more frequently than people expect. I'm 100% right about this prediction of the future and you are an idiot if you disagree".
Also, the unacknowledged inconsistency of his arguments is ridiculous: ~"Unpredictable events happen more frequently than people expect. I'm 100% right about this prediction of the future and you are an idiot if you disagree".
Taleb's Black Swan concept is that unpredictable events happen more frequently than people expect, and have an outsized impact on the outcomes of a model. Events that are not predictable are not included in the predictive model. Take the frequency of event and exclusion from the model, add in asymmetric results (small changes to input parameters can lead to huge changes in outcome), and that's his theory.
Is that accurate? What are the issues with that?
- it's hard (impossible) to predict population statistics from a sample with size 1, ceteris paribus. Over the short term, you have sample size 1 for rare events, over the long term ceteris paribus doesn't hold.
- people "coerce" everything into a normal distribution in order to get a handle on things which is of course wrong if the thing you look at is not normally distributed. Finance comes to mind.