I am currently re-reading “The Black Swan: the impact of the highly improbable” by Nassim Taleb. In chapter 8, he writes about the problem of silent evidence, otherwise known as the survivorship bias. To illustrate this bias, ask yourself: what is the average salary for an actor? Odds are that you will overestimate the figure. Why? Because you are likely thinking of successful actors rather than the entire cohort. There are plenty of aspiring actors working in the Cheesecake factory who you never hear about.

The important feature of the survivorship bias is that we can be led to falsely believe that an entity ranking among the survivors has a special property which is not shared by the unheard of failures.

One possible property that we can be mistakenly led to believe a survivor possesses is that of “antifragility”. This concept was introduced in Taleb’s other book “Antifragile: things that gain from disorder”. Loosely speaking, something is antifragile if it gets better when subjected to shocks. Over time, we should find that such things will win out. When we select survivors in a cohort then, we should find that the antifragile are among them. It would be in error though to reason that being among the survivors implies antifragility: Antifragile entities are more likely to be among the survivors, but being among the survivors is no guarantee of antifragility.

Nassim Taleb, the author of the highly acclaimed book “The Black Swan: the impact of the highly improbable”, gives, in the prologue for his book (I’m using the second edition), three characteristics for an event to be a “Black Swan”: (1) it is an outlier, (2) it has an extreme impact, and (3) it is explainable with hindsight, but not foresight.

At first thought, one could be forgiven for thinking that these characteristics all apply to Covid-19. On closer examination however, the matter is not so clear. Why do I say this? Because there is no such thing as an objective “Black Swan” event!

Taleb himself acknowledges this, also in the prologue, in a footnote where he says that the events of 9/11 were a “Black Swan” to the victims, but were certainly not to the perpetrators. That is to say: although the incident was entirely unforeseeable to the victims, it was fully known to the al Qaeda terrorists who planned it.

So the question really should be “For who was Covid-19 a “Black Swan” event?”. The answer to that, I hope should be clear, is “most of us” (excepting virologists and the like). For governments however, the possibility of a pandemic has long been within the realm of conceivability (so it fails to be a “Black Swan” in the first and third characteristics). For a UK example, an influenza pandemic was ranked as the highest in both probability of occurrence and expected severity in the National Risk Register 2017 (which now has a 2020 edition). Note that the probability of a pandemic from any virus must be at least as large as that for a influenza specific one.

IMPORTANT POINT FOR UK READERS: In the 2017 edition of the National Risk Register, events were given a “relative” probability of occurrence over a 5 year period. In the 2020 edition, numerical values for the probabilities are estimated for an occurrence over a 1 year period (so the report is out of date!). This, in my view, is a major fault in the most recent edition as probabilities will be smaller over a single year period than over a five year period leading to an increased chance of failing to prepare sufficiently for a crisis as it’s deemed “unlikely over the next year”.

I was struggling to decide what to write about, after so long of not blogging (reasons personal), so I decided to take the current leading BBC news article and write my thoughts on it going from top to bottom.

The headline is that the UK Prime Minister (PM) says he has not seen evidence of blackmail against those in the Conservative party who oppose his continued leadership. My initial reaction is “assuming you are telling the truth, you having not seen any evidence does not imply that it is not happening nor that there is no evidence”. The two principles here are that: (1) absence of evidence is not evidence of absence, and (2) not seeing evidence does not imply that it doesn’t exist.

After the headline, the BBC then give a bit of backstory: they say that some Conservative MPs have expressed concerns about blackmail; they say that opposing parties are condemning the PM over the allegations; they provide an example of a Conservative MP who calls the allegations “nonsense”. To the first, we should take any such allegations seriously in a country that cares about the rule of law. To the second, who is surprised at that? To the third, well at least one was bound to say it.

With regard to any allegation, there is always the principle of “innocent until proven guilty” that should be borne in mind. Following that principle, many MPs are reserving judgement until the release of a report by civil servant Sue Grey into allegations of Covid rule-breaking at No 10. I might write on that.

A paragraph in the article particularly caught my attention: “He [MP William Wragg] also said he had received reports of government ministers, advisers and staff at No 10 ‘encouraging the publication of stories in the press seeking to embarrass’ those suspected of lacking confidence in the PM”. What kind of stories could these possibly be?

Following the main news comes the “analysis” by a BBC political correspondent, which – at least in this case – doesn’t actually add to the content of the article. It merely adds drama to the (already dramatic) events.

https://en.wikipedia.org/wiki/Conjunction_fallacy

The following is a famous experiment done by psychologists. Give it a try yourself.

Linda is 31 years old, single, outspoken, and very bright. She majored in philosophy. As a student, she was deeply concerned with issues of discrimination and social justice, and also participated in anti-nuclear demonstrations.

Which is more probable?

1. Linda is a bank teller.
2. Linda is a bank teller and is active in the feminist movement.

---

The researchers found that most participants choose the second option. Did you?

If you did, the researchers will reveal that, so they think, you are wrong.

They will argue that option 2 implies option 1 (correct), and therefore option 1 must have a greater probability.

The argument is correct, but under a condition which they fail to see. By “probable”, they mean the ratio of “favourable” possible events to the total possible events. That is they think the question is about comparing the ratio of the number of bank tellers who satisfy the information about Linda to the ratio of the number of active feminist bank tellers who satisfy the conditions.

This is, unless you know the technical use of the term “probable”, almost certainly not what you understood the question to mean. In daily parlence, the word “probable” is used to mean the same thing as the word “plausible”. Hence, the participants – not aware of an alternative precise definition – correctly choose the most probable (in the mathematical sense of the term now) interpretation of the question. Clearly, you have to understand the question in the sense that it is being asked before you are capable of answering it correctly according to that interpretation.

I imagine the researchers thought the participants dumb.

An influenza pandemic. One which spreads fast, is severe, and harms/kills people of all age groups in high numbers.

This has ranked as the highest civil emergency risk for years. See https://www.gov.uk/government/publications/national-risk-register-of-civil-emergencies-2017-edition.

Estimations of how the chance and expected severity of such a possibility have changed owing to recent events, as well as consideration of how to deal effectively with an eventuality, would be well worth the time and money.

I’ve learned by now to read the original source rather than the media commentary (or at least in conjunction with it). They focus on the juicy bits, regardless of their importance. These are the bits I found most interesting (whether you agree is up to you). See pdf here.

NB: I make no narrative or causal claims based on what is presented here.

DEMOGRAPHICS

CRIME/POLICING/LAW

ECONOMICS

EDUCATION

HEALTH

SOCIAL AND POLITICAL LIFE

---

Caveats:

1. Some of this data is based on the 2011 census, so may not reflect the present situation accurately. Feel free to send me any data which is more recent.
2. Some of this data is for England and Wales, rather than the UK as a whole.
3. Some of this data is based on surveys, so is dependent on response rates, good sampling procedures, and so on.

NB: I make no narrative or causal claims based on what is presented here.

SECTION 1: Crime and Policing

SECTION 2: Demography

SECTION 3: Economics

SECTION 4: Government Finances

SECTION 5: Climate

SECTION 6: Health

SECTION 7: Education

SECTION 8: Social and Political Life

SECTION 9: Government and Legislation

SECTION 10: Belief and Opinion

SECTION 11: Military

---

Sources:

I could not possibly include every country here, so I judiciously picked some. My sources contain more information than given here.

https://ourworldindata.org/

https//thebulletin.org/