A mathematical mind works in reverse: a matrix is a special case of a tensor, a vector a special case of a matrix.

In probability, those NonRussians who start from special cases make yuuuge mistakes in higher dimensions (like idiotic psychologists/geneticists w/IQ problems).
You can see the problem with:
+ Distributions: economists & statisticians have no consciousness that the Gaussian is a special case & can't handle the more general.
+ "Correlation" is special to linear association & people think it is "evidence based".
+ "Sharpe ratio", factor...
For instance something as basic as correlation (2 dimensional metric) is misinterpreted in the IQ literature & elsewhere.

academia.edu/39797871/Foole…

• • •

Missing some Tweet in this thread? You can try to force a refresh
 

Keep Current with Nassim Nicholas Taleb

Nassim Nicholas Taleb Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @nntaleb

11 Jan
SOCIAL MEDIA & TRUMP

1/ Remember that social media got Trump elected.

Twitter decided to not be a vehicle for violent insurrection against elected government.

There is no possible evidence that can convince conspiracy weirdoes that the election, ANY election, isn't rigged.
2) Now one MUST wonder: what abt those on the left who also called for violent insurrection agst elected govmnt?

Perhaps there is something about right wing-racial supremacist weirdoes that is especially dangerous, or morally repelling compared to rebellion by race victims.
3) So the compromise is that Twitter MUST ethically ban some pple on the other side who violate similar rules threatening the democratic process.

And if the democratic process seems too flawed owing to lobbies working both sides, consider an active targeting of these lobbyists.
Read 6 tweets
31 Dec 20
DEPARTMENT OF BAD MISTAKES W/PROBABILISTIC CONCEPTS

LANCET compares life duration of athletes to the life expectancy AT birth for males.

They shd compare them to the conditional expectancy of someone active in same age group!

Effects shrinks yuugely.

thelancet.com/journals/lance…
2) 2nd error: conditional on being an athlete one must be in good health so the comparison must be ONLY between otherwise healthy non-athletes & these athletes, conditional on being alive at that age.

Note: (unconditional) LE changes have been mostly affected by child mortality.
3) Finally, for the point they are making (whether extremes lengthen or shorten LE) they must use a comparative cohort of mild exercisers, nothing else.
Read 4 tweets
24 Dec 20
COVID REUNIONS
The great John Gray was told by a friend that his cat was vegan. "Do you let your cat out?" "Yes", he answered.

So if pple claim they "don't see anybody", ask them if they have seen their children. Odds are they did.

As in finance, all it takes is one weak link.
AIRLINES & WEAK LINKS
@AirFrance's planes are Hepa-filtered, extremely safe; passengers/crew are forced to wear masks. They spent millions to increase the safety.

But they ferry people at CDG in buses that are not ventilated, w/passengers packed like sardines.

Explain.
I guess @ParisAeroport has some explanations to give us.
Read 4 tweets
21 Dec 20
PROBABILITY QUIZ DU JOUR

In 1784, at a dinner in Paris, Cagliostro claimed to be 3500 y. o., in permanent middle age; he partook of most historical battles; he was immortal.

Condorcet exposed an error: C could not be that old.

Explain C's error & what to do for C to be #Lindy
ANSWER: If you don't age, you must have a constant force of mortality wrt accidents (external/internal vascular).

A tiny pr. of an accident cumulates over time to a constant additional Life Exp, a memoryless distr.

Some animals have 20y LE regardless of age.
Human LE <200y
MORAL: For a human to be immortal he/she must be error-free.
Aging= increasing force of mortality from > fragility.

Debate whether people cease to age at ~100 but become accident prone at a constant high rate, 110 same risk of death as 115!
Lightbulbs are fragile w/constant FM.
Read 5 tweets
18 Dec 20
There must be something about BTC that makes the person statistically clueless.
Separate BTC from these idiots.
3) Hint: If Sweden has a "79% larger" elderly population than Norway (in proportion), there must be a yuuuuge life expectancy diff betw the 2.
You don't need to understand probability/expectation beyond high school level to see the error.
BTC activists have intellectual issues.
Read 4 tweets
15 Dec 20
Something unpopular:
The bulk of the "losses" in #Lebanese #lollar deposits is just interest not made at all (& interst on interst).
Many, many are actually ahead or breaking even.
#Yieldhogs were making >5-7½% extra for years, >60-100% per decade.

Yieldhogging is not risk-free.
2) #Yiledhogging is actually defined and described in Fooled by Randomness (2001).

3) Depositors were clearly not 100% victims, particularly when wealthy & supposed to be sophisticated.
80% of Lebanese depositors have cashed out (since it's a Pareto law: 1% of depositors have >50% of deposits).

FBR (2001)
Read 4 tweets

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!

Follow Us on Twitter!