The Friendly Face Fallacy

Mike is a friendly, good looking middle-aged man who lives in Manhattan. He is having drinks with his friend Brent, a sharp thinking interlocutor.

Mike: Yesterday, a woman approached me asking for directions, which I happily provided. This is not the first time this has happened to me. People ask me for directions all the time! I must have one of those faces.

Brent: Mike, you fool. Don’t you see your illogical reasoning?

Mike: No. Please, explain.

Brent: How many lost people have looked at your face and decided not to ask you for directions?

Mike: How could I possibly know the answer to that question?

Brent: You cannot, which is my point.

Mike: I don’t follow.

Brent: True, people have asked you for directions. But it’s nearly impossible to know if  you “have one of those friendly faces” without first knowing how many people have looked at your face and turned the other way. Your judgment only considers success stories. It’s like trying to measure how good David Ortiz is at hitting home runs when your evidence is a highlight reel of every single one of his home runs. You need to account for how many times he didn’t hit a home run — all the strikeouts, ground outs, fly outs, singles, double, triple, etc.

Mike: I think I see what you are saying. But I still think I am at least above average.

Brent: Possibly, but unlikely. For that you would have to know what the average is in the first place. And I assume you don’t have data on how many times the average Manhattanite is approached for directions. Your intuition is not a good guide to determine that number.

Mike: So I am just an typical Manhattanite who is approached for directions as much as everyone else?

Brent: Most likely – but you’re an above average friend.

Mike: Thanks, Brent.

Brent: You’re welcome.


We Don’t See Non-Events

A wrecked car idles on the side of the road; four bottles of wine and two handles of whisky are in the back seat—the driver, an adult male, is unconscious in the front seat. The police arrive and call the paramedics. The situation looks bleak.

What do you make of this story? It’s difficult to avoid jumping to the intuitive conclusion without first considering the number of sober drivers who drive with alcohol. Every year, many cars that contain alcohol get into accidents. How many are caused by the driver consuming that alcohol? How many involve sober drivers who just happened to have alcohol?

The mind processes information fast, often generating a clean casual chain with just a few variables—Car crash. Alcohol. Unconscious driver. Drunk driving. But if we want to understand what happens in the world we must consider information that is not immediately apparent. This, however, is difficult.

In 1967, two psychologists presented participants with the following numbers: 147, 724, 947, 421, 843, 394. What do they have in common? Four is in each of them. Now consider another list: 239, 639, 865, 795, 261, 756. What do these numbers have in common? Four is in none of them. “What can we learn from this?” asks Rolf Dobelli in his book The Art of Thinking Clearly. “Absence is much harder to detect than presence.”

An analogous problem occurs when we attempt to understand business performance. If, for example, the price of a stock decreases while the CEO was on vacation, it’s intuitive to connect the two and conclude that the stock decreased because the CEO was on vacation. But this is the stuff of headlines. We must remember all those vacationing CEOs whose companies managed just well without them, even if such news does not make the front page.

On Listening

Epictetus - Enchiridion CXLII

Nature has given to men one tongue, but two ears, that we may hear from others twice as much as we speak.

La Rochfoucauld - Collected Maxims V:142

As great minds have the ability to say much in few words, so, conversely, small minds have the gift of talking much and not saying anything.

Attributed to Plato, but origins likely unknown

Wise men talk because they have something to say; fools, because they have to say something.

Plutarch - On Listening

Anyone who has acquired the ability to listen in a self-controlled and respectful fashion is receptive to and retentive of any remarks that are useful, while any that are useless or false are quite transparent to him and easily detectable, because he is–as is obvious–aiming at the truth rather than at winning an argument, and does not pitch in head first for a fight.

Why You Should Almost Never Take DayQuil

Regression to the mean: if the first measurement is extreme, the second is likely closer to the average. If the second measurement is extreme, the first is likely closer to the average. In the long run stocks, athletes, academics, the time it takes to fly from New York to Chicago, your immune system and your intelligence will hover around an average.

Consider the so-called SI Jinx. Cover-warranting performances are usually extraordinary, which means the athlete will regress back to his normal level of performance. Jordan was on the cover of SI so many times because his normal level was extraordinary. A coach yells at his team after a poor performance. The team plays well the next match. It appears that the coaches’ antics caused performance to improve, but the team simply regressed back to their normal level of play. You have a shitty week, go to your therapist, and the next week is normal. You would most likely have returned to normal without a visit to the therapist.

People say “I was sick, took a pill/went to the doctor, and got better” to suggest the pill/doctor visit caused improvement. This is almost always an incorrect inference, implying a causal relationship where no such thing exists. Minus bad injuries or extreme illness, the body heals itself perfectly fine without intervention. Since your immune system will normally regress back to its default state, do yourself a favor and do nothing if you experience a non-fatal illness. (no DayQuil).

Inaction is an optimal strategy much more than you think. Yet humans are plagued by a bias to do something to correct mistakes/errors/hardship when such things are commonplace events that your body and mind are perfectly prepared to handle.

Why were bleeding, vomiting and purging practiced for some 2,350 years? Doctors who performed these heinous operations were trying achieve something the body was also trying achieve: health. When bleeding, vomiting, and purging did not kill a patient they appeared to save the patient. This is a straightforward example of confusing a correlation with a cause. Ironically, the same problem still pervades medicine today: we intervene when intervention is unnecessary, and if the body self-heals we believe (mistakenly) that the intervention (and not the body’s normal recovery processes) helped. To make matters worse, we recommend the intervention to others–doctors prescribe it and big pharma mass produces it.

Three errors from Wootton’s “Bad Medicine”

Three errors outlined by David Wootton in his very excellent book Bad Medicine:

1. We ignore non-events:

What we need in cases such as these is a history, not of progress, but of delay; not of events, but of non-events; not of inflexible logic but of a sloppy logic, not of overdetermination, but of underdetermination.

Because we have not listened out for the screams, we never hear the eerie silence that fell over operating tables in the 1850s.

Since this book argues that real medicine begins with germ theory, at its heart there is a most puzzling historical non-event: the long delay that took place between the discovery of germs and the triumph of germ theory.

2. Knowledge and therapy are together in hindsight, but separate in reality:

We tend to assume that where there is progress in knowledge there is progress in therapy… but before 1865 progress in knowledge rarely led to improvements in therapy.

Between the sixteenth and the nineteenth centuries, ideas about the body changed fundamentally, but therapies changed very little. Bloodletting was the main medical therapy in 1500, 1800, and 1850. The discovery of the circulation of the blood (1628), of oxygen (1775), of the role of haemoglobin (1862) made no difference; the discoveries were adapted to the therapy rather than vice versa.

Harvey announced that the heart pumped blood through the arteries in 1628; yet the use of the tourniquet in in amputations, which one would have thought was an absolutely elementary application of Harvey’s theory, was first pioneered by Jean Louis Petit (1674-1750), roughly a century later. Leeuwenhoek saw what we would now loosely call germs, or more accurately bacteria, through his microscope in 1677; yet in 1820 microscopes had no place in medical research, and in 1881 the conflict between germ theorists and their opponents was only just entering its final phase. Penicillin was first discovered not in 1941 but in 1872. And so on.

3. Status quo and confirmation errors plagued the progress of medicine:

To think about progress, you must first understand what stands in the way of progress–in this case, the surgeon’s pride in his work, his professional training, his expertise, his sense of who he is… the obstacle was the surgeon’s own image of themselves.