Archive for September, 2006
Book review: Expert Political Judgment: How Good Is It? How Can We Know? by Philip E. Tetlock
This book is a rather dry description of good research into the forecasting abilities of people who are regarded as political experts. It is unusually fair and unbiased.
His most important finding about what distinguishes the worst from the not-so-bad is that those on the hedgehog end of Isaiah Berlin’s spectrum (who derive predictions from a single grand vision) are wrong more often than those near the fox end (who use many different ideas). He convinced me that that finding is approximately right, but leaves me with questions.
Does the correlation persist at the fox end of the spectrum, or do the most fox-like subjects show some diminished accuracy?
How do we reconcile his evidence that humans with more complex thinking do better than simplistic humans, but simple autoregressive models beat all humans? That seems to suggest there’s something imperfect in using the hedgehog-fox spectrum. Maybe a better spectrum would use evidence on how much data influences their worldviews?
Another interesting finding is that optimists tend to be more accurate than pessimists. I’d like to know how broad a set of domains this applies to. It certainly doesn’t apply to predicting software shipment dates. Does it apply mainly to domains where experts depend on media attention?
To what extent can different ways of selecting experts change the results? Tetlock probably chose subjects that resemble those who most people regard as experts, but there must be ways of selecting experts which produce better forecasts. It seems unlikely they can match prediction markets, but there are situations where we probably can’t avoid relying on experts.
He doesn’t document his results as thoroughly as I would like (even though he’s thorough enough to be tedious in places):
I can’t find his definition of extremists. Is it those who predict the most change from the status quo? Or the farthest from the average forecast?
His description of how he measured the hedgehog-fox spectrum has a good deal of quantitative evidence, but not quite enough for me check where I would be on that spectrum.
How does he produce a numerical timeseries for his autoregressive models? It’s not hard to guess for inflation, but for the end of apartheid I’m rather uncertain.
Here’s one quote that says a lot about his results:
Beyond a stark minimum, subject matter expertise in world politics translates less into forecasting accuracy than it does into overconfidence
Book review: Evolution’s Rainbow: Diversity, Gender, and Sexuality in Nature and People by Joan Roughgarden
This book provides some good descriptions of sexual and gender diversity in nature and in a variety of human cultures, and makes a number of valid criticisms of biases against diversity in the scientific community and in society at large.
Many of her attempts to criticize sexual selection theory are plausible criticisms of beliefs that don’t have much connection to sexual selection theory (e.g. the belief that all sexually reproducing organisms fall into one of two gender stereotypes).
Her more direct attacks on the theory amount to claiming that “almost all diversity is good” and ignoring the arguments of sexual selection theorists who describe traits that appear to indicate reduced evolutionary fitness (see Geoffrey Miller’s book The Mating Mind). She practically defines genetic defects out of existence. She tries to imply that biologists agree on her criteria for a “genetic defect”, but her criteria require that a “trait be deleterious under all conditions” (I suspect most biologists would say “average” instead of “all”), and that it reduce fitness by at least 5 percent.
Her “alternative” theory, social selection, may have some value as a supplement to sexual selection theory, but I see no sign that it explains enough to replace sexual selection theory.
She sometimes talks as if she were trying to explain the evolution of homosexuality, but when doing so she is referring to bisexuality, and doesn’t attempt to explain why an animal would be exclusively homosexual.
Her obsession with discrediting sexual selection comes from an exaggerated fear that the theory implies that most diversity is bad. This misrepresents sexual selection theory (which only says that some diversity represents a mix of traits with different fitnesses). It’s also a symptom of her desire to treat natural as almost a synonym for good (she seems willing to hate diversity if it’s created via genetic engineering).
She tries to imply that a number of traits (e.g. transsexualism) are more common than would be the case if they significantly reduced reproductive fitness, but her reasoning seems to depend on the assumption that those traits can only be caused by one possible mutation. But if there are multiple places in the genome where a mutation could produce the same trait, there’s no obvious limit to how common a low-fitness trait could be.
Her policy recommendations are of very mixed quality. She wants the FDA to regulate surgical and behavioral therapies the way it regulates drugs, and claims that would stop doctors from “curing” nondiseases such as gender dysphoria. But she doesn’t explain why she expects the FDA to be more tolerant of diversity than doctors. Instead, why not let the patient decide as much as possible whether to consider something a disease?
This book is a colorful explanation of why we are less successful at finding happiness than we expect. It shows many similarities between mistakes we make in foreseeing how happy we will be and mistakes we make in perceiving the present or remembering the past. That makes it easy to see that those errors are natural results of shortcuts our minds take to minimize the amount of data that our imagination needs to process (e.g. filling in our imagination with guesses as our mind does with the blind spot in our eye).
One of the most important types of biases is what he calls presentism (a term he borrows from historians and extends to deal with forecasting). When we imagine the past or future, our minds often employ mental mechanisms that were originally adapted to perceive the present, and we retain biases to give more weight to immediate perceptions than to what we imagine. That leads to mistakes such as letting our opinions of how much food we should buy be overly influenced by how hungry we are now, or Wilbur Wright’s claim in 1901 that “Man will not fly for 50 years.”
This is more than just a book about happiness. It gives me a broad understanding of human biases that I hope to apply to other areas (e.g. it has given me some clues about how I might improve my approach to stock market speculation).
But it’s more likely that the book’s style will make you happy than that the knowledge in it will cause you to use the best evidence available (i.e. observations of what makes others happy) when choosing actions to make yourself happy. Instead, you will probably continue to overestimate your ability to predict what will make you happy and overestimate the uniqueness that you think makes the experience of others irrelevant to your own pursuit of happiness.
I highly recommend the book.
His analysis of memetic pressures that cause false beliefs about happiness to propagate is unconvincing. He seems to want a very simple theory, but I doubt the result is powerful enough to explain the extent of the myths. A full explanation would probably require the same kind of detailed analysis of biases that the rest of the book contains.
He leaves the impression that he thinks he’s explained most of the problems with achieving happiness, when he probably hasn’t done that (it’s unlikely any single book could).
He presents lots of experimental results, but he doesn’t present the kind of evidence needed to prove that presentism is a consistent problem across a wide range of domains.
He fails to indicate how well he follows his own advice. For instance, does he have any evidence that writing a book like this makes the author happy?
While browsing through charts of various stocks, I came across a company (Manchester Inc., symbol MNCS) with a chart that’s unusual enough that I had to check around to reassure myself that my primary source for stock market prices wasn’t playing tricks on me.
It has a history of unusually steady increases with few signs of the randomness that I normally see in stock prices. If you had bought at the closing price any day this year and held for ten trading days, it would have closed higher than your purchase price (your average gain would have been over 3 percent), and it was almost as predictable the prior year.
A paragraph in the middle of this Forbes story explains why its market value looks strange.
The only guess I have as to what might cause this is an unusual form of manipulation where the manipulators produce this phenomenon until traders who buy purely on price trends provide enough liquidity for the manipulators to cash out. But even that is pretty implausible – if that’s what’s happening, why wouldn’t they create a bit more day to day randomness to disguise it a bit? And how could they afford to risk as large an investment as I suspect that would take on an approach which seems different enough from anything tried before that it ought to be hard to predict whether it will work?
Book review: The Undercover Economist: Exposing Why the Rich Are Rich, the Poor Are Poor–and Why You Can Never Buy a Decent Used Car! by Tim Harford
This book does an excellent job of describing economics in a way that laymen can understand, although experts won’t find much that is new in it.
Harford’s description of price discrimination is the best I’ve seen, and the first to describe how to tell the extent to which an instance of price discrimination has good effects (the extent to which it expands the number of sales).
His arguments that globalization reduces pollution are impressive for most types of pollution, but for carbon dioxide emissions I’m very disappointed. He hopes that energy use has peaked in the richest countries because he’s failed to imagine what will cause enough increased demand to offset increases in efficiency. For those of modest imagination, I suggest thinking about more realistic virtual reality (I want my Holodeck), personal robots, and increased air conditioning due to people moving to bigger houses in warmer climates. For those with more imagination, add in spacecraft and utility fog.
Some small complaints:
He refers to Howard Schultz as the owner of Starbucks, but he only owns about 2 percent of Starbucks’ stock.
His comment that Amazon stock price dropped below its IPO price fails to adjust for stock splits – a share bought at $18 in 1997 would have become 12 shares worth $8 each in the summer of 2001.
His claim that “Google is the living proof that moving first counts for nothing on the Internet” is a big exaggeration. It’s quite possible that Google success was primarily due to being the first to reach some key threshold of quality, and that many small competitors have matched its quality without taking measurable business away from it.