MIN-LESSON 9b: 2 min technical comments for Mini Lesson 9.

[@Sprajt]

I know about the central limit theorem, but I never thought about it in this way before. Very interesting.

[@ASingh21112]

Mr.Taleb, I was a critic until I read the black swan. Now i have your books on my shelf next to Popper, Hayek, and Schumpeter. Thanks for opening my eyes and my mind!!!

[@miraclemaxicl]

Mathematica examples are my favorite. Thanks.

[@MrCan94]

Hi N N Taleb, love your content but would you mind uploading better quality (at least 720p) videos? thanks!

[@nntalebproba]

Sorry, NO.

[@sillyfarmerbilly8872]

So simple, I get it! I'm going to try this myself and play around with the idea, thanks for showing me how!!

[@EvgenyChepil]

Can you advise what software to try this?

[@sillyfarmerbilly8872]

@@EvgenyChepil He is using Mathematica (also known as Wolfram), which seems like it's the best but it's not free. I've been using Python. It's a very general use language but they are modules and packages that you can load to give you about the same functionality as mathematica (or so I'm told). Those packages are numpy, pandas, matplotlib, pyplot, etc. If you pair that with a notebook tool called Jupyter notebook, it even looks very similar to mathematica. You can learn how to use all this with KZbin, just start with the word python and the names of those packages, or search "python for data science" or something similar. freecodecamp.org is a great channel for that. If you get stuck on something just Google it and they will always be some nerd somewhere who posted a solution for you in a Reddit or GitHub post (kind of like using Linux). ... But if you would rather pay for the right tool and you don't want to spend the time to learn a programming language that has broader applications, just pay Wolfram to use mathematica.

[@lfsa86]

I think he also explained this point with an example of portfolio management theory in an episode back.

[@KarimTraderFutures]

Merci pour le partage.

[@KarimTraderFutures]

Bonjour Monsieur Taleb, je vous souhaite un joyeux anniversaire ainsi qu'un bonheur plénier, au moins aussi équivalent au bonheur que m'a apporté la lecture de vos ouvrages.

[@HubertHeller]

Example: The more times you roll a dice (uniform distribution) the more likely the shape of the distribution of the means tends to look like a normal distribution graph. Is my understanding correct? (I am not a statistics pro.)

[@nntalebproba]

Oui.

[@ibraheemmoosa]

The flip side of Central Limit Theorem. If I understand correctly, this means you cannot recover the original distribution from aggregate.

[@maxutaibi5986]

I wish someone would call me friend the way Nassim calls us friends

[@DanielJanzon]

Suppose you evaluate two individuals from two distinct populations based on their merits. If the populations have different means, necessarily the uniform distribution of the individuals must be different too. E.g. shifted like [a,b] and [a+c,b+c]. If you use a Bayesian update approach to correct for the "base rate fallacy" would you use the uniform distribution as prior and end up with the same prior for both individuals or would you use the population distribution and end up with different priors? If the latter, the population "leaks" over to a correct evaluation of the individuals. So if they have equal merits, you would end up picking the one with best "base rate" = your best guess before being presented with any evidence particular to the individual.

[@nicolaswalter1523]

How do I work at Universa LP

[@jamesjackson5955]

Love this simple demister. This assumes the individuals are uncorrelated i.e. you can't for from the group to the individual if the individuals are uncorrelated. Please correct me if I'm wrong

[@nntalebproba]

You can't do Central Limit (at least not in this form) if there is correlation/dependence.

[@daudkaun3807]

Give us a library tour of yours

[@twistedbydsign99]

Nassim is this simpsons paradox?

[@nntalebproba]

Not quite

[@000FURQAN]

can anyone tell me which software Nassim use for visualization?

[@nntalebproba]

Mathematica

[@samirelzein1978]

King!

[@sillyfarmerbilly8872]

If anyone wants to try this but doesn't want to commit to subscribing to mathematica (proprietary programming language he's using)... If you buy a raspberry pi it comes with mathematica free.