Archive for September, 2009

Uncertainty and hubris

Sunday, September 27th, 2009

“Before I went to Tech, I couldn’t even spell ‘engineer’, now I are one.” Similarly, I only learned the the word hubris a few years ago. I knew many examples of hubris, I just didn’t have a good fifty-cent word to describe them.

I’ve just finished reading The Black Swan – The Impact of the Highly Improbable, by Nassim Nicholas Taleb. It was, at turns, controversial and obvious, worrisome and strangely comforting, annoying and great fun. NNT, as he is often called, is a clever writer, well-read and well-educated, practical, and self-assured. He has apparently thought about, and worked with, uncertainty, probability, luck, and knowledge for most of his life.

Trying to summarize any book is risky, particularly one with such complex and nuanced subject matter. The term “Black Swan” refers to a significant event that no one has seen before.  NNT says that Black Swans have three attributes – “unpredictability, consequences, and retrospective explainability.” Trying to explain the events leading to a Black Swan is similar to using a puddle of water on the kitchen counter to deduce the shape of the ice cube before it melted (or to prove, indeed, that there ever was an ice cube.)

Black Swans may be positive or they may be negative. They may occur suddenly, such as a sudden financial disaster or world event, or they may be slower in happening, such as the discovery of coherent light and its many uses (also known as the “laser”, and its many children, including optical computer networks, CDs, and DVDs.)

Because the book is full of stories, illustrations and examples, it was fun to read. He includes an appropriate number of well-chosen graphs and pictures, and only a few “math” sections, which he warns about in advance so the “non-technical” reader may skip them. He seasons his arguments with quotations ranging from Yogi Berra to Benoit Mandelbrot.

He does not hesitate to say that people who use precise mathematical models to predict “risk,” such as financial analysts, are foolish. He thinks that the Nobel Prize in Economics is generally given to charlatans. Needless to say, this has not made him popular with a large segment of professional money managers and other people who get paid quite well to make profound predictions, nor with the bogus experts he calls “empty suits.”

I won’t claim to have already figured out the things NNT talks about, but some of the points of the book supported some simple personal observations that have bugged me for some time.

  1. People who base their beliefs on anecdotes instead of statistics. As a recent example, consider people who tell stories about long waiting lines for medical treatment in a country with universal health care (anecdotes), but ignore World Health Organization statistics which show otherwise.
  2. What I call “straight-line projections,” wherein someone bases decisions on a few data points that can be extended in a straight line to demonstrate a wildly successful outcome. As an example for this one, I offer several businesses (including a national furniture chain and a popular doughnut maker) who appeared to have recently borrowed money and expanded their stores without any consideration for possible uncertainties in their calculations.
  3. A lack of understanding of “percentage growth,” which is doomed to decrease as growth increases. When I moved to the county I live in, it was experiencing “double-digit” growth. Now that growth has “slowed” to single digits, even though the number of people entering the county is larger every year. Perhaps that misunderstanding of the difference between growth and percentage growth accounts for some of our budgetary problems.
  4. The smug tone of documents which predict market changes and project the popularity of proposed new service ideas. I don’t know what bugs me worse: the overconfident perspective, or the fact that things never happen as predicted.
  5. The fact that most significant inventions are discovered accidentally, rather than by rational, linear thinking. This one reinforces my own particular research style, which is to just play around with interesting stuff until something useful emerges. (I’ll let you know how that turns out.)

Perhaps the most dramatic thing I read in the book was a footnote in a section bemoaning the concentration of U.S. financial institutions into gigantic, interconnected, bureaucratic banks, a frightening “financial ecology” largely based on convincing but overly simplistic risk measurements. “[T]he government-sponsored institution Fanny Mae, when I look at their risks, seems to be sitting on a barrel of dynamite….” The book was published in 2007.

He also discusses other patterns of human behavior that contribute to our short-sightedness, including

  • the difference between absence of evidence and evidence of absence (just because I’ve never seen something happen doesn’t mean it never has or will)
  • confirmation bias (only noticing events that confirm what you already think)
  • the “narrative fallacy” (we prefer compact stories over raw truth),
  • our inability to resist explaining why things happened as they did, even though the events, players, and interactions are complex beyond anyone’s understanding; this is particularly prevalent in the political arena, and
  • the fallacy of silent evidence (history only records the parts that fit the understanding of the historian.)

One more thought came to me as I read the book. Years ago (and I’ve forgotten where), I read a comment on the study of “non-linear equations.” It turns out that the primary reason we study linear equations is not because there are so many linear systems in the world (there aren’t), but because the mathematics are manageable. The comment was something to the effect that referring to non-linear equations is a little like referring to zoology as the study of “non-elephant animals.” That seems to be the only explanation for why we use overly-simple mathematical models to do projections – because the math is manageable. And because they seem to work. That is, until they fail dramatically.

In the end, I was left feeling like the Christian who decided he would turn every one of his life-decisions over to God. He sat on the edge of his bed for four hours one morning waiting for divine guidance, until he finally realized he was going to have to pick out which shirt to wear all by himself. (There’s a book I read long ago called Decision-Making and the Will of God that, as I recall, addresses this poor fellow’s dilemma.)

While NNT has some modest suggestions for how we should then live, he claims that the main benefit of this analysis is a better understanding of how things work, a respect for the possibility of errors and worst-case scenarios, and an awareness of the potential of the impact of Black Swans on our lives.

Good luck!

Letter to the Editor

Friday, September 18th, 2009

I recently received an e-mail from a good friend of mine in which he made a mildly disparaging comment about the current administration’s attempts at health care reform. Well, my response was swift, and probably too harsh. Little did he know the nerve he had touched. But I’m getting ahead of myself. The story starts before my Letter to the Editor, in fact, at least 7 years before. My family and I have been supportive of improvements to the nation’s healthcare situation for at least that long. At one candlelight vigil for the uninsured around that time, I heard the then-CEO of Grady Hospital give as clear an explanation of the need as I have ever heard, backed up by several other quite reputable speakers. The need seemed so clear at the time.

Fast forward to this year, when real change actually looks possible. I have wanted to do more than just donate, so I was looking for the right opportunity. It came when The Gwinnett “Daily” Post (GDP) published a “Viewpoints” column full of misleading and incorrect statements. I decided I would write a Letter to the Editor (LTE). Not a whiny collection of my personal opinions, but some Facts about our healthcare system that have bugged me ever since I learned them. So I did it. I sat down and wrote a carefully worded, even-tempered response, and submitted it to the GDP. I sent it early in the same afternoon the Viewpoints were published so it would arrive in time to be seen, and recognized as relevant, incisive, and to-the-point.

That night I woke up thinking about my Letter, and even worried if one word was literally accurate. I got up early the next morning, verified that my Facts were indeed true, then slowly made my way through the morning paper, in my usual fashion. When I reached the editorial page, I casually glanced over it, and was disappointed to find that there was nary a single LTE there. I was even more disappointed to find that they request that LTEs be limited to 200 words. I quickly checked my submission, and it was more than twice that.

So I hastily edited out a big chunk, but left the Facts intact, and it came out to 198 words. I quickly re-submitted it. But the next morning, my precise prose again failed to appear on the editorial page, although two other letters did, including one that had to be at least 600 words.

I went through several stages: regret at wasting the time submitting my thoughts, discouragement at the GDP’s editorial staff overlooking my gems of wisdom, fear that the web submission process doesn’t work, but one emotion dominated the others. I was suffering from “lack of information publication.” I had some information that I desperately wanted to share with someone. I didn’t necessarily expect to change the world, but I did want to give people the chance to judge it on its merits.

I began thinking that I would post the information. I write a couple of occasional blogs in a couple of different contexts, and I decided that if only two people read my facts, that would be two more than had seen them on the GDP editorial page. Two over zero is, like, infinity. It would be an infinite percentage more. Who can argue with a number like that?

So I had been mulling over all of this for several days when my friend sent his ill-fated e-mail. I was ready for him. I took my original LTE, bolstered by some additional thoughts, edited it to fit his comments, and fired back a 10-paragraph missive that probably sounded a lot like a rant. It was only the next day that I realized that most of the emotion behind my letter to him didn’t come from his casual comment, but from years of longing for real healthcare reform, and paragraphs of editorials that I considered misleading and even incorrect, and days of wondering what to do with my Facts. It was like the bursting of a dam.

Without further ado, and in an attempt to keep any more dams from bursting, here are the Facts I wanted the GDP to publish. It’s up to you to decide what to do with them. But you can rest assured that they are factual. (I’ll be glad to share the sources with you.)

1) The United States is the only industrialized country on the planet that doesn’t provide health care for all of its citizens. This would be fine if we had the best health. But….

2) In world health rankings, we are NOT EVEN IN THE TOP 20! This includes factors such as longevity and infant mortality. On average, the poorest residents in England live longer than the wealthiest Americans. If health care were football, we wouldn’t even get invited to a bowl game. But there is one factor in which we do lead…

3) Cost. We pay more per person for health care than any other country, about 50% more than Sweden, and more than double everyone else. So we’re paying more and getting less.

As a patriotic American, I believe we can do better than this.

In closing, I think people with vested interests in the current system are determined to distract us from the actual facts. But we Americans are clever people. I believe we can figure out a workable direction for health care reform, and I believe we need to do it now.

And a final warning. I submitted another LTE this morning. If they don’t print this one, keep an eye on this space. I’m liable to post something else!

Thanks for reading.

UPDATE: In response to several requests, I have published a longer, more partisan page (where I am the partisan) that contains some of my additional thoughts, link, and comments about health care reform. The page is at