History Doesn’t Repeat. But Mistakes Do.
This week’s book is Nassim Nicholas Taleb’s first salvo against modernity, the 2004 classic Fooled By Randomness. I’ve read this book several times and it is still a difficult thing to parse into simple articles and broad reviews. Mostly because everything in this book continues to carry the sort of insight that stretches my thinking in very painful ways.
I’m not alone, either. This book was a phenomenon when it first arrived. Tom Peters called it “life-altering”. Maggie Mahar beautifully said that it was “The book that rolled down Wall Street like a hand grenade.” I call it the book everyone wants to have read.
It is also a book that started a fundamental change in my thinking. To this day, when combined with his next two books, I can’t think of anyone who has done more to help me gain clarity on the nature of all that happens around us. Not because of his writing, per se. His writing is great. But Taleb did more than just share his thoughts. He transparently shared the underpinnings and gave readers the introduction to a much broader library. It is through Taleb that I initially learned about Systems Thinking, Stoicism, Falsifiability, Complexity, and much more.
One such concept that he shares is the Lucas critique. So to begin this week’s review, consider the following from our author:
We are human and act according to our knowledge, which integrates past data. [Therefore] If rational traders detect a pattern of stocks rising on Mondays, then, immediately such a pattern becomes detectable, it would be ironed out by people buying on Friday in anticipation of such an effect.
There’s a lot in this statement and it’s hard to toss it out without much introduction. So let’s add a little more from the author:
We take past history as a single homogeneous sample and believe that we have considerably increased our knowledge of the future from the observation of the sample of the past.
Which is to say that we frequently look at large-scale events and treat their history as a strong marker of what will occur again. “I’ve seen this before,” we say. “It’s going to be just like last time.”
And yet, in complex (i.e., interesting) situations, the very ability to identify a potential outcome based on past events is what makes the new situation completely different from those that came before. A fine example of this comes from the initial (incorrect) read that I took from Ray Dalio’s increasing warnings about a pending downturn.
Notice the emphasis on the aesthetic: “the economy looks like 1937”. What, exactly, did 1937 look like? How can I, Dalio, or the Business Insider reporter possibly convey that information in a single article?
And does this mean that there are no counterfactuals? Are there no other years that were very successful despite also “looking” like 1937?
I don’t know. All I know is that the article sounds very interesting, was fun to read, and therefore got me thinking that history is repeating. Just as Dalio said.
But history, as we think of it, does not repeat. Not on the scale they discuss in that article. And certainly not on the scale that the Lucas Critique attacks. More on that in a moment. There are things that repeat but broad history ain’t one of them.
For now, this means that an interesting article is highly likely to be an inaccurate one. To know why is to dive into the strange comparatives that make Taleb such a fun read.
Rock Stars versus Dentists
Taleb was the first to popularize (for me, anyway) the beauty of statistical thought. And one of the illuminating aspects of this book is the manner in which it helps me distinguish success from the interplay of skill, luck, and scale. There are averages here that must be respected.
Consider the expected success of a rock musician and that of the dentist. We can easily understand that the rock musician’s profession has a very winner-take-all quality. I can’t tell you who the most popular rock musician is today but I can confidently say that there are only a few. Rock music, and all other music, is highly scalable. It can be shared with the entire world and so the world clusters around a discrete few. Outside that precious few, there are hundreds and thousands of other rock musicians that toil in obscurity. This is the stuff of the attention economy and the nature of popular music.
How many rock-and-roll musicians pay their bills with their music? 400?
How many rappers and EDM artists? 2,000?
I’m just guessing.
Meanwhile, the American Dentist Association states that there are 195,722 dentists in the United States. These folks enjoy a rock-steady middle- to upper-middle class lifestyle and would be considered by many to be very successful. They have to drill a whole lot of teeth to keep their success intact but they can. And they do. Because dentistry doesn’t scale. There isn’t a handful of dentists out there that everyone clamors to visit. They can’t do the work remotely. This is the stuff of the service economy, particularly where services don’t scale.
So scale clearly has an effect here, right? Rock music scales quickly with attention and popularity and has a winner-take-all quality. Dental services do not. I’ll never go to a dentist in any of the millions of locations they occupy outside my immediate 50-mile vicinity.
But skill has an effect, too. A dentist must know how to do dental work in the manner that dentistry has established over a hundred years of practice.
A rock musician doesn’t have to do anything the way it was done before. In fact, it’s better if they don’t. Because their popularity is deeply rooted in their individuality, the novelty of what they offer. So long as there is no one like them, they have a chance. This has nothing to do with a technical skill of any kind. Consider the great Johnny Cash. He was not an accomplished guitarist. Neither was Kurt Cobain. There was nothing “learned” and easily transferable about their work. You can’t make another Cobain or Cash.
They ascended to tremendous heights of popularity and success. Why? Well, we know today that this is largely about right-place-right-time. As Taleb puts it,
Mild success can be explainable by skills and labor. Wild success is attributable to variance.
Variance is a more precise way of saying luck. Dentists don’t need luck. They have skills and labor. Cash and Cobain needed luck. And they got it. Make no mistake: their success is not wholly attributable to luck. But luck played a major part.
Right. Sure. We all get it. Yet, I don’t think we do. Not fully. Far too often, we ignore the durable certainty and modest certainty of a non-scaleable, skill-based endeavor. Dentistry is soooooo boring. Dentistry is soooooo divinely reliable.
Which is to say that if you want to respect reality, randomness, and chance, assume the posture of becoming successful slowly. One tooth at a time. And join a band of fellow dentists to play rock music on the weekends. In other words, create a balance of highly reliable work and highly speculative work. Coincidentally, this is a thumbnail sketch of a larger strategy that we’ll cover in Taleb’s next book, The Black Swan.
Dentists versus Weathermen
Back to the broader topic of history repeating itself. To channel my inner Taleb, let’s start by saying that you should always believe your dentist’s predictions and ignore what your economist says. Why?
Cycle times, observations, and closed versus open systems.
Your dentist has a durable skill built over decades of practice in a finite space of knowable complexity. You teeth and gums are a fascinating and complex system. But again, they are knowable. And your dentist is exposed to many mouths many times a day. To say they have seen it all is kinda accurate. And if they haven’t seen it, their colleagues in the profession most certainly have and actively share it. This is the stuff of a “closed” system to borrow from the work of Gary Klein, our fascinating author of the book Sources of Power (read the review here).
So when your dentist says you are going to get a cavity, his prediction is not based on a whim. There is a deep-seated intuition that is based on a very deep history of cause-and-effect that has played out millions of times in the profession. Dentists have observed the life cycle of a cavity so many times that they can pinpoint it clear as day.
But Norm, meteorologists watch weather patterns every day and can still get it wrong.
True! But that is because the weather is an open system with far more inputs, variables, and relationships. And yet, a seven-day forecast is still accurate 80% of the time thanks to the sheer amount of cycles and observations that have been recorded, regressed, and built into the prediction. When the cycle is relatively short (a 24-hour weather pattern) and easily measurable, it can eventually be patterned to some level of reliability.
Emphasis on the term “some”. An 80% accuracy rate is great for weather but awful for dentistry. No one would want to be that one-in-five patient that sits in that chair.
To Be Continued …
Some may be reading this and wondering … what does any of this have to do with original lead of this article? We started with the Lucas Critique and that general misunderstanding of history. Is there a predictive power to history? I don’t think so. Neither does Taleb and other popular writers like Yuval Noah Harari.
Neither, I think, does Dalio. Not the way it seems. So we’ll continue this exploration with tomorrow’s post. As stated at the beginning, this week’s book is very difficult for me to boil down into clean, standalone articles. But for now, I’ll at least build a few mostly-obvious conclusions from Taleb’s writings on this idea so that people can leave with something:
- Those who succeed at scale enjoy a rare bit of luck (time and place)
- Never wager everything on luck (or getting to scale)
- History is important but, for prediction, models matter more
- Just because something looks familiar doesn’t mean it’s true
- And maybe …
When something of a large and complex scale looks familiar, the prediction that it will continue to look familiar is destined to not be true (i.e. the economy in 2019 versus 1937).
I’m sure that seems obvious. The smartest stuff I read from Taleb and Dalio usually does. And yet it isn’t. So more on that last point tomorrow. As stated several times now, this is heady stuff for me. So as I try to make sense of it, I hope you can do the same.
These are important ideas to share. And difficult for me to share easily.