N N Talebs Probability QuestionsDoes the Ketogenic Diet Work for Brain Cancer? Doctors find it hard to do a trial so many make you believe that it does not work rather than simply claiming "it is just untested" (conflating the absence of evidence for evidence of absence).
Does the Ketogenic Diet Work for Brain Cancer? Finding Signal in a Small Sample.N N Talebs Probability Questions2025-03-07 | Does the Ketogenic Diet Work for Brain Cancer? Doctors find it hard to do a trial so many make you believe that it does not work rather than simply claiming "it is just untested" (conflating the absence of evidence for evidence of absence).Alexander Taleb: Why Do Planes Crash?N N Talebs Probability Questions2025-07-22 | Alexander Taleb, aerospace engineer, lectures at the #RWRI summer school 2025 (#RWRI 20) on plane safety. The lecture has close to 2 hours of Q&A.Stephen Wolfram visits RWRI 20 (The Real World Risk Summer School, 2025)N N Talebs Probability Questions2025-07-21 | The great Stephen Wolfram spends 2 hours discussing the ruliads and computational irreducibility at the Real World Risk Institute Summer School.Discussion with Pierre Zalloua on AncestorsN N Talebs Probability Questions2025-04-28 | A discussion with Pierre Zalloua about his book Ancestors on the Genetic & Cultural History of the Levant. It is the first scientific history of the area, which debunks many historical accounts.MINI LECTURE 19 - Conditional Probability & A Fallacy in Cardiology Concerning LDLN N Talebs Probability Questions2025-03-14 | People with normal Lp(a) have much lower risk factors. We tend to overmedicate them. The normal levels of LDL should be revised upwards for them.The IQ debate, a conclusionN N Talebs Probability Questions2025-01-25 | Just to show, in conclusion to my article "IQ is a pseudoscientific swindle", that psychologists arguing with it proved that IQ doesn't explain up to 99.5% of things it is supposed to explain. It's not my data, but theirs that they did not interpret correctly.
medium.com/incerto/iq-is-largely-a-pseudoscientific-swindle-f131c101ba39Universas Bernoulli for Portfolio Simulation: Correcting the Empirical DistributionN N Talebs Probability Questions2024-09-11 | The empirical distribution is not empirical, full of Turkey problems. "Real tails" do not show in past samples because of their property under fat tails. Empirical distributions are, by design, interpolating; we fix by extrapolating and extending the tails. Note the dispute: In French, Bernoulli is pronounced "Ber-noo-yi" not "Ber-noo-li", much like consigliere is pronounced "consiyyere".Scott Pattersons Chaos Kings, Part 4N N Talebs Probability Questions2024-08-16 | Geoengineering -- Mao's Four Pest Problem -- Brexit -- How antagonism between bad politicians can turn into good government -- The "LongTermism" nonsense of SBF -- The Final ChapterScott Pattersons Chaos Kings, Part 3N N Talebs Probability Questions2024-08-01 | + Why Ilmanen-style "simulations" of tail performance are both theoretically and empirically flawed because of 1) need to extrapolate extremes, and 2) The Niederhoffer problem. + The birth of the Non-Naive Precautionary Principle in East Anglia, in a pub discussion with Rupert Read.Scott Pattersons Chaos Kings, Part 2N N Talebs Probability Questions2024-07-26 | +The Turkey Problem + Meeting Mandelbrot & discovering a wealth of processes generating grey swans + Meeting Kahneman + The 3-day meeting with the Harvard geneticist George Church, involving Elon Musk, Anne Wojcicki, Larry Page, Richard Thaler, & others that triggered the non-naive version of the precautionary principle.Scott Pattersons Chaos Kings, Part 1N N Talebs Probability Questions2024-07-21 | Scott Patterson discusses with Nassim N Taleb his book Chaos Kings, chapter by chapter. This is Part 1.STEPHEN WOLFRAM VISITS RWRI 19, SUMMER SCHOOL 2024N N Talebs Probability Questions2024-07-05 | TOPICS: 2nd law, generative AI, computational Irreducibility, climate modeling.
Stephen Wolfram's discussion with participants at the Real World Risk Institute Summer School 2024, with Nassim Nicholas Taleb, Robert Frey, Raphael Douady, and others.How to bluff when you dont know what you are talking about.N N Talebs Probability Questions2024-06-24 | Humphrey Bogard and his friend infiltrated a Nazi cell in New York, impersonating two Nazi engineers. They had to bluff their way by answering questions when they really didn't know anything about the subject matter.MINI LECTURE 18: How to build a positive definite correlation matrix for Monte Carlo simulationsN N Talebs Probability Questions2024-03-17 | How to build by tinkering a positive definite correlation matrix for Monte Carlo simulations for a given dimensionality.MINI LECTURE 17: Maximum Ignorance Probability (a bit more technical)N N Talebs Probability Questions2024-03-15 | What do you do when you must estimate a probability and there is ZERO information? A surgeon did 60 operations with 0 failure. What's her or his failure rate? I didn't realize that the trick wasn't in the literature until John Hughes wrote a paper in Statistics and Probability Letters based on my blog post presenting the idea.MINI LECTURE 16: On the Absurdity of Standard Genetics Reports.N N Talebs Probability Questions2024-03-13 | Why reports from genetic firms saying you are 30% Greek, 15% Sicilian, 5% Martian are complete misrepresentations.Stephen Wolfram visits RWRI 18 (Summer Workshop)N N Talebs Probability Questions2023-08-20 | ...MINI LECTURE 15 - Conditional vs. unconditional correlation: twin studies overestimate heredity.N N Talebs Probability Questions2023-02-25 | Description The genetics of twin studies have a bias showing more heredity than in reality, owing to a statistical artifact. The twin studies for heredity is based on comparing the correlation between 2 identical twins minus that between 2 fraternal ones (assumed to be sharing half their genes). The use of fraternal twins as control is assumed to extract the "environmental" factors.
Problem: Correlation is conditional and psychologists think it is unconditional . We show how the math is entirely different.
The core error is that genes and environment are not separable and additive.MINI TUTORIAL: How to look at risk ratios & deal with disinformation: Saturated Fats & Nina TeicholzN N Talebs Probability Questions2022-12-07 | Ms. Nina Ticholz backed up her saturated fat propaganda with a major paper showing that saturated fats don't cause cardiovascular disease that was "ignored by mainstream media". Turns out the paper shows the opposite of what she claims. I show how to read risk ratios, confidence intervals, etc. Note on Terminology: by "odds" I am using the betting/gambling, not the epidemiological terminology, here the CDF/Survival Function.A medical paper that does not replicate. Why?N N Talebs Probability Questions2022-08-04 | Case study in non-replicability of a medical paper. I discuss it with Dr. Fouad Fayad from U.S.J. in Lebanon. Details are in Retraction Watch: apparently the only reply to his case by the lead author is of the fact-free data-free sort "he doesn't like me" which does not appear to have a statistical explanation (nor does it explain why the paper does not replicate).
Expert reports finding evidence of manipulation and confirming Dr Fayad's claim are at fooledbyrandomness.com/Fayad.pdfNo, Covid is not an old persons problem.N N Talebs Probability Questions2022-06-01 | Covid increases mortality rates proportionally. And if one considers such diseases dynamically they can reduce life expectancy even more, hitting the young disproportionally. Telling a 40 y.o. "you have a 10% higher probability of death" surprises them. Many "young" people fail to think dynamically, and realize they will be older some day. Think of a 30 y.o. saving for the future, but not considering the effect of the pandemic of 2050.QUANTITATIVE FINANCE 2: We dont use Black-Scholes, the simpler derivation vindicating Bachelier.N N Talebs Probability Questions2022-04-30 | How instead of using the limit of dt (dynamic hedging) one uses the limit of dK (static hedging, K is the strike) and one can use Bachelier's formula.QUANTITATIVE FINANCE 1: Deriving Black-Sholes via Itôs lemma (the dynamic hedging approach)N N Talebs Probability Questions2022-04-24 | ...Disinformation and Fooled by RandomnessN N Talebs Probability Questions2022-03-25 | We are not naturally good at dealing with information. Disinformation artists confuse you by focusing on noise over signal by playing on saliency, the same effect as the one discussed in Fooled by Randomness. We mistake the particular for the general, details for the ensemble, and noise for signal --all from the same mental bias.QUANT FINANCE 1 - Why We Never Use the Black Scholes Equation, 1N N Talebs Probability Questions2022-02-10 | The first part explaining the Bachelier equation and how options were priced traditionally.Conversation with Yaneer on Covid with mild disagreements.N N Talebs Probability Questions2022-01-27 | Yaneer Bar Yam and Nassim Nicholas Taleb discuss the Covid situation, with some disagreements.MINI LECTURE 14 A First Course on Fragility, Convexity, and Antifragility (Nontechnical).N N Talebs Probability Questions2022-01-21 | A first, very introductory presentation of fragility as linked to both nonlinearity and dislike of variations. Antifragility is almost the opposite, limited to a specific range of variations. Explains: + Why everything fragile must be concave. + The medical S curve. + Why harm to the climate is necessarily nonlinear in dose response. + How hospitals can be overcrowded unless there are redundancies. Futher discussions will be more technical.MINI LECTURE 13b - Technical Appendix. How to fix the problem of power laws with compact support.N N Talebs Probability Questions2022-01-11 | Technical Appendix to the paper on violence: What do you do when the data looks like it is powerlaw distributed over a broad range, but cannot be technically a power law? We use a dual distribution and transport parameters between one and another.MINI LECTURE 13 - Claims that Violence Has Dropped Are Not StatisticalN N Talebs Probability Questions2022-01-10 | Violence is from Extremistan, hence requires some more sophisticated tools since LLN works slowly. We see how Pinker's thesis is bogus. We look at ways to integrate the factual unreliability of historical accounts. We look at transformations to analyze violence using power law tools since the worst case is bounded at contemporary population level. Two Papers with Pasquale Cirillo arxiv.org/pdf/1505.04722.pdf [TECHNICAL, PHYSICA A] papers.ssrn.com/sol3/papers.cfm?abstract_id=2876315 [NONTECHNICAL, NOBEL FOUNDATION]MINI LECTURE 12: How to Look at the Risks of Covid VaccinesN N Talebs Probability Questions2022-01-02 | How to look at the risks of Covid vaccines, why they much lower than you think. We never had a larger monitored sample size in history and it allows events that on average show up later to manifest themselves very early on. Rationale: It takes a long time in a casino for someone to win 8 times in a row. But if 8 billion people played at the same time you would certainly witness a minimum of such events every day.MINI LESSON 11b: Drawdowns and LogsN N Talebs Probability Questions2021-12-24 | Quick presentation of drawdowns and the necessity to use logarithms for returns.MINI LESSON 11: Path dependence in 2 ½ minutesN N Talebs Probability Questions2021-12-20 | Explaining path dependence in sequences and why mean and variance are insufficient metrics. (Corrected video, moved max drawdown to 11b)MINI LESSON 10: Simpsons Paradox & Its Exploitation by Covid Sociopaths in 3 ½ min.N N Talebs Probability Questions2021-11-26 | In every age bracket, the vaccinated live longer than the unvaccinated (using all cause mortality). However as a group the unvaccinated appear to have a longer life expectancy. This is because the vaccinated tend to be older (hence more likely to die). I explain Simpson's Paradox in general.
Note: I used the vaccinated/unvaccinated ratio for 50-60 vs 10-20 of Oct 2021, so don't bug me if it rose since; no effect on the point so long as there is an inequality.Supply Chain: The Convexity in 2 ½ minutes.N N Talebs Probability Questions2021-11-22 | An antifragility tutorial, with application to supply chains.MINI LESSON 9c: Detecting Quackery in a Psychology PaperN N Talebs Probability Questions2021-09-29 | A simple case study where model error shows the claims in a paper published in a "prestigious" psychology journal to be false. Why worry about low R^2.MIN-LESSON 9b: 2 min technical comments for Mini Lesson 9.N N Talebs Probability Questions2021-07-27 | Why the statistical properties of a group will NOT particularize to any individual member of the group. Explained in 2 minutes. The example uses central limit.MINI-LESSON 9: Evidence Based Science & Mistakes in Particularizing the General (Simplified)N N Talebs Probability Questions2021-07-22 | 1) The law of large numbers (properties of aggregates) works in one direction. Why you can generalize from particulars, never particularize from generals. 2) The difference betweern clinical, statistical, and risk management approached. Why they don't scale. 3) Never compare Mediocristan to Extremistan (Covid to car accidents).Comments on the Bitcoin Black Paper (Why Bitcoin is Worth Exactly 0)N N Talebs Probability Questions2021-07-02 | The main paper and the supplementary equations are at arxiv.org/pdf/2106.14204.pdf and fooledbyrandomness.com/blog/index.php/2021/06/24/the-bitcoin-currencies-and-bubbles-paper
Erratum: the equation min 3:4 must include the discounting inside the summation (1/1+r)^iMIN-LESSON 8b: (Power Laws) Q& AN N Talebs Probability Questions2021-06-04 | Q&A on MIN-LESSON 8 (Power Laws). Why violence did not drop (the Pinker Problem) -- Pandemics --What an infinite mean means -- What causes power laws.MINI-LESSON 8: Power Laws (maximally simplified)N N Talebs Probability Questions2021-05-26 | Power laws, extremely simplified.MINI LESSON 7: P-Values and P-Value Hacking: a simplified lecture.N N Talebs Probability Questions2021-05-23 | We saw that 1) many metrics are stochastic, 2) what is stochastic can be hacked. This is the simplification of my work showing that "p-values are not p-values", i.e. highly sample dependent, with a skewed distribution. For instance for a "true" P value of .11, 53% of observations will show less than .05. This allows for hacking: in a few trials a researcher can get a fake p-value of .01. Paper is here and in Chapter 19 of SCOFT (Statistical Conseq of Fat Tails):
arxiv.org/pdf/1603.07532.pdfMINI-LESSON 6: Fooled by Metrics (Correlation)N N Talebs Probability Questions2021-05-14 | A maximally simplified presentation of how metrics are random variables, and how they can be gamed. Uncorrelated variables will produce a correlation in samples. For a more advanced version see the P Values paper: arxiv.org/pdf/1603.07532.pdfThe Gates Foundation is Repeating the Errors of MaoN N Talebs Probability Questions2021-05-08 | The Gates Foundation is financing genetically modified mosquitoes to control natural mosquitoes. This quick video: 1) Discusses the Four Pest Campaign by Mao that helped kill 15-50 Million people around 1960. 2) Explains why genetic modification has nothing to do with natural selection or animal breeding. Scientism is not science.MINI-LESSON 5: Correlation, the intuition. Doesnt mean what people usually think it means.N N Talebs Probability Questions2021-05-04 | A maximally intuitive presentation on what correlation is not, with maximally simplified concepts. [Note that I improvise 100% and I don't prepare] There is a more technical lecture to come. The technical paper is here: https://www.academia.edu/39797871/Fooled_by_Correlation_Common_Misinterpretations_in_Social_Science_MINI-LESSON 4: CLT, The Central Limit Theorem, a nontechnical presentation.N N Talebs Probability Questions2021-05-03 | The CLT allows anyone (including ignorant economists and psychologists) to do statistics by using prepackaged recipes coming from the Gaussian. What are the foundations? How does it work? Where does it not work?MINI-LESSON 3: The Law of Large Numbers. A very intuitive introduction.N N Talebs Probability Questions2021-05-02 | Everything in empirical science is based on the law of large numbers. Remember that it fails under fat tails.MINI-LESSON 2: Fat Tails, a Very, Very Introductory Presentation.N N Talebs Probability Questions2021-04-24 | What are Fat Tails? This is very introductory.
See the whole book (gets technical beyond Chapter 5) https://researchers.one/articles/20.01.00018MINI-LESSON 1: Breaking down intuitively the concept of standard deviation. Why pple dont get it.N N Talebs Probability Questions2021-04-23 | SIMPLIFIED TUTORIAL, 1 (in the series where we break down concepts intuitively): Before we talk about correlation, let's discuss standard deviation, its analog in dimension 1. People don't get while using it as a metric for deviation! This is simplified of course (n, not n-1); sorry nitpickers.. More technical: my book at https://researchers.one/articles/20.01.00018 Also we wrote a paper (Goldstein & Taleb, 2007) showing PhD students in stat-math-finance and fund managers got the point wrong (mismatch between intuition and computation). Yet it is elementary.How to Price an Election: A Martingale Approach- Discussion with Dhruv MadekaN N Talebs Probability Questions2020-10-09 | Discussion with Dhruv of the paper: Election Pricing: a Martingale Approach and other related works.
arxiv.org/pdf/1703.06351.pdfFake Regression by psychologistsN N Talebs Probability Questions2020-10-02 | How to detect fake regressions. Never pay attention to numbers before looking at the graph. First I show how to spot fakess in a sinister paper promoted by Nicolas Baumard, whose work seems to focus on linking people's physical traits to (un)trusworthiness. Second, more technically, I derive the distribution of the slope between two random variables and show how one can game it. Note: In the graphs of regressions I showed a paper promoted by Baumard. The actual regressions in Baumard's own paper are even worse.