Finally, the illusions of validity and skill are supported by a powerful professional culture. We know that people can maintain an unshakable faith in any proposition, however absurd, when they are sustained by a community of like-minded believers. Given the professional culture of the financial community, it is not surprising that large numbers of individuals in that world believe themselves to be among the chosen few who can do what they believe others cannot.
The Illusions of Pundits
The idea that the future is unpredictable is undermined every day by the ease with which the past is explained. As Nassim Taleb pointed out in The Black Swan, our tendency to construct and believe coherent narratives of the past makes it difficult for us to accept the limits of our forecasting ability. Everything makes sense in hindsight, a fact that financial pundits exploit every evening as they offer convincing accounts of the day’s events. And we cannot suppress the powerful intuition that what makes sense in hindsight today was predictable yesterday. The illusion that we understand the past fosters overconfidence in our ability to predict the future.
The often-used image of the “march of history” implies order and direction. Marches, unlike strolls or walks, are not random. We think that we should be able to explain the past by focusing on either large social movements and cultural and technological developments or the intentions and abilities of a few g coဆreat men. The idea that large historical events are determined by luck is profoundly shocking, although it is demonstrably true. It is hard to think of the history of the twentieth century, including its large social movements, without bringing in the role of Hitler, Stalin, and Mao Zedong. But there was a moment in time, just before an egg was fertilized, when there was a fifty-fifty chance that the embryo that became Hitler could have been a female. Compounding the three events, there was a probability of one-eighth of a twentieth century without any of the three great villains and it is impossible to argue that history would have been roughly the same in their absence. The fertilization of these three eggs had momentous consequences, and it makes a joke of the idea that long-term developments are predictable.
Yet the illusion of valid prediction remains intact, a fact that is exploited by people whose business is prediction—not only financial experts but pundits in business and politics, too. Television and radio stations and newspapers have their panels of experts whose job it is to comment on the recent past and foretell the future. Viewers and readers have the impression that they are receiving information that is somehow privileged, or at least extremely insightful. And there is no doubt that the pundits and their promoters genuinely believe they are offering such information. Philip Tetlock, a psychologist at the University of Pennsylvania, explained these so-called expert predictions in a landmark twenty-year study, which he published in his 2005 book Expert Political Judgment: How Good Is It? How Can We Know? Tetlock has set the terms for any future discussion of this topic.
Tetlock interviewed 284 people who made their living “commenting or offering advice on political and economic trends.” He asked them to assess the probabilities that certain events would occur in the not too distant future, both in areas of the world in which they specialized and in regions about which they had less knowledge. Would Gorbachev be ousted in a coup? Would the United States go to war in the Persian Gulf? Which country would become the next big emerging market? In all, Tetlock gathered more than 80,000 predictions. He also asked the experts how they reached their conclusions, how they reacted when proved wrong, and how they evaluated evidence that did not support their positions. Respondents were asked to rate the probabilities of three alternative outcomes in every case: the persistence of the status quo, more of something such as political freedom or economic growth, or less of that thing.
The results were devastating. The experts performed worse than they would have if they had simply assigned equal probabilities to each of the three potential outcomes. In other words, people who spend their time, and earn their living, studying a particular topic produce poorer predictions than dart-throwing monkeys who would have distributed their choices evenly over the options. Even in the region they knew best, experts were not significantly better than nonspecialists.
Those who know more forecast very slightly better than those who know less. But those with the most knowledge are often less reliable. The reason is that the person who acquires more knowledge develops an enhanced illusion of her skill and becomes unrealistically overconfident. “We reach the point of diminishing marginal predictive returns for knowledge disconcertingly quickly,” Tetlock writes. “In this age of academic hyperspecialization, there is no reason for supposing that contributors to top journals—distinguished political scientists, area study specialists, economists, and so on—are any better than journalists or attentive readers of The New York Times in ‘reading&#oulဆ8217; emerging situations.” The more famous the forecaster, Tetlock discovered, the more flamboyant the forecasts. “Experts in demand,” he writes, “were more overconfident than their colleagues who eked out existences far from the limelight.”
Tetlock also found that experts resisted admitting that they had been wrong, and when they were compelled to admit error, they had a large collection of excuses: they had been wrong only in their timing, an unforeseeable event had intervened, or they had been wrong but for the right reasons. Experts are just human in the end. They are dazzled by their own brilliance and hate to be wrong. Experts are led astray not by what they believe, but by how they think, says Tetlock. He uses the terminology from Isaiah Berlin’s essay on Tolstoy, “The Hedgehog and the Fox.” Hedgehogs “know one big thing” and have a theory about the world; they account for particular events within a coherent framework, bristle with impatience toward those who don’t see things their way, and are confident in their forecasts. They are also especially reluctant to admit error. For hedgehogs, a failed prediction is almost always “off only on timing” or “very nearly right.” They are opinionated and clear, which is exactly what television producers love to see on programs. Two hedgehogs on different sides of an issue, each attacking the idiotic ideas of the adversary, make for a good show.
Foxes, by contrast, are complex thinkers. They don’t believe that one big thing drives the march of history (for example, they are unlikely to accept the view that Ronald Reagan single-handedly ended the cold war by standing tall against the Soviet Union). Instead the foxes recognize that reality emerges from the interactions of many different agents and forces, including blind luck, often producing large and unpredictable outcomes. It was the foxes who scored best in Tetlock’s study, although their performance was still very poor. They are less likely than hedgehogs to be invited to participate in television debates.
It is Not the Experts’ Fault—The World is Difficult
The main point of this chapter is not that people who attempt to predict the future make many errors; that goes without saying. The first lesson is that errors of prediction are inevitable because the world is unpredictable. The second is that high subjective confidence is not to be trusted as an indicator of accuracy (low confidence could be more informative).
Short-term trends can be forecast, and behavior and achievements can be predicted with fair accuracy from previous behaviors and achievements. But we should not expect performance in officer training and in combat to be predictable from behavior on an obstacle field—behavior both on the test and in the real world is determined by many factors that are specific to the particular situation. Remove one highly assertive member from a group of eight candidates and everyone else’s personalities will appear to change. Let a sniper’s bullet move by a few centimeters and the performance of an officer will be transformed. I do not deny the validity of all tests—if a test predicts an important outcome with a validity of .20 or .30, the test should be used. But you should not expect more. You should expect little or nothing from Wall Street stock pickers who hope to be more accurate than the market in predicting the future of prices. And you should not exp
ect much from pundits making long-term forecasts—although they may have valuable insights into the near future. The line that separates the possibly predictable future from the unpredictable distant future is inဆ yet to be drawn.
Speaking of Illusory Skill
“He knows that the record indicates that the development of this illness is mostly unpredictable. How can he be so confident in this case? Sounds like an illusion of validity.”
“She has a coherent story that explains all she knows, and the coherence makes her feel good.”
“What makes him believe that he is smarter than the market? Is this an illusion of skill?”
“She is a hedgehog. She has a theory that explains everything, and it gives her the illusion that she understands the world.”
“The question is not whether these experts are well trained. It is whether their world is predictable.”
Intuitions vs. Formulas
Paul Meehl was a strange and wonderful character, and one of the most versatile psychologists of the twentieth century. Among the departments in which he had faculty appointments at the University of Minnesota were psychology, law, psychiatry, neurology, and philosophy. He also wrote on religion, political science, and learning in rats. A statistically sophisticated researcher and a fierce critic of empty claims in clinical psychology, Meehl was also a practicing psychoanalyst. He wrote thoughtful essays on the philosophical foundations of psychological research that I almost memorized while I was a graduate student. I never met Meehl, but he was one of my heroes from the time I read his Clinical vs. Statistical Prediction: A Theoretical Analysis and a Review of the Evidence.
In the slim volume that he later called “my disturbing little book,” Meehl reviewed the results of 20 studies that had analyzed whether clinical predictions based on the subjective impressions of trained professionals were more accurate than statistical predictions made by combining a few scores or ratings according to a rule. In a typical study, trained counselors predicted the grades of freshmen at the end of the school year. The counselors interviewed each student for forty-five minutes. They also had access to high school grades, several aptitude tests, and a four-page personal statement. The statistical algorithm used only a fraction of this information: high school grades and one aptitude test. Nevertheless, the formula was more accurate than 11 of the 14 counselors. Meehl reported generally similar results across a variety of other forecast outcomes, including violations of parole, success in pilot training, and criminal recidivism.
Not surprisingly, Meehl’s book provoked shock and disbelief among clinical psychologists, and the controversy it started has engendered a stream of research that is still flowing today, more than fifty yephyဆЉ diars after its publication. The number of studies reporting comparisons of clinical and statistical predictions has increased to roughly two hundred, but the score in the contest between algorithms and humans has not changed. About 60% of the studies have shown significantly better accuracy for the algorithms. The other comparisons scored a draw in accuracy, but a tie is tantamount to a win for the statistical rules, which are normally much less expensive to use than expert judgment. No exception has been convincingly documented.
The range of predicted outcomes has expanded to cover medical variables such as the longevity of cancer patients, the length of hospital stays, the diagnosis of cardiac disease, and the susceptibility of babies to sudden infant death syndrome; economic measures such as the prospects of success for new businesses, the evaluation of credit risks by banks, and the future career satisfaction of workers; questions of interest to government agencies, including assessments of the suitability of foster parents, the odds of recidivism among juvenile offenders, and the likelihood of other forms of violent behavior; and miscellaneous outcomes such as the evaluation of scientific presentations, the winners of football games, and the future prices of Bordeaux wine. Each of these domains entails a significant degree of uncertainty and unpredictability. We describe them as “low-validity environments.” In every case, the accuracy of experts was matched or exceeded by a simple algorithm.
As Meehl pointed out with justified pride thirty years after the publication of his book, “There is no controversy in social science which shows such a large body of qualitatively diverse studies coming out so uniformly in the same direction as this one.”
The Princeton economist and wine lover Orley Ashenfelter has offered a compelling demonstration of the power of simple statistics to outdo world-renowned experts. Ashenfelter wanted to predict the future value of fine Bordeaux wines from information available in the year they are made. The question is important because fine wines take years to reach their peak quality, and the prices of mature wines from the same vineyard vary dramatically across different vintages; bottles filled only twelve months apart can differ in value by a factor of 10 or more. An ability to forecast future prices is of substantial value, because investors buy wine, like art, in the anticipation that its value will appreciate.
It is generally agreed that the effect of vintage can be due only to variations in the weather during the grape-growing season. The best wines are produced when the summer is warm and dry, which makes the Bordeaux wine industry a likely beneficiary of global warming. The industry is also helped by wet springs, which increase quantity without much effect on quality. Ashenfelter converted that conventional knowledge into a statistical formula that predicts the price of a wine—for a particular property and at a particular age—by three features of the weather: the average temperature over the summer growing season, the amount of rain at harvest-time, and the total rainfall during the previous winter. His formula provides accurate price forecasts years and even decades into the future. Indeed, his formula forecasts future prices much more accurately than the current prices of young wines do. This new example of a “Meehl pattern” challenges the abilities of the experts whose opinions help shape the early price. It also challenges economic theory, according to which prices should reflect all the available information, including the weather. Ashenfelter’s formula is extremely accurate—the correlation between his predictions and actual prices is above .90.
Why are experts e yinferior to algorithms? One reason, which Meehl suspected, is that experts try to be clever, think outside the box, and consider complex combinations of features in making their predictions. Complexity may work in the odd case, but more often than not it reduces validity. Simple combinations of features are better. Several studies have shown that human decision makers are inferior to a prediction formula even when they are given the score suggested by the formula! They feel that they can overrule the formula because they have additional information about the case, but they are wrong more often than not. According to Meehl, there are few circumstances under which it is a good idea to substitute judgment for a formula. In a famous thought experiment, he described a formula that predicts whether a particular person will go to the movies tonight and noted that it is proper to disregard the formula if information is received that the individual broke a leg today. The name “broken-leg rule” has stuck. The point, of course, is that broken legs are very rare—as well as decisive.
Another reason for the inferiority of expert judgment is that humans are incorrigibly inconsistent in making summary judgments of complex information. When asked to evaluate the same information twice, they frequently give different answers. The extent of the inconsistency is often a matter of real concern. Experienced radiologists who evaluate chest X-rays as “normal” or “abnormal” contradict themselves 20% of the time when they see the same picture on separate occasions. A study of 101 independent auditors who were asked to evaluate the reliability of internal corporate audits revealed a similar degree of inconsistency. A review of 41 separate studies of the reliability of judgments made by auditors, pathologists, psychologists, organizational managers, and other professionals suggests that this level of inconsistency is typical, even when a case is reevaluated within a few minutes. Unreliable judgments cannot be valid predi
ctors of anything.
The widespread inconsistency is probably due to the extreme context dependency of System 1. We know from studies of priming that unnoticed stimuli in our environment have a substantial influence on our thoughts and actions. These influences fluctuate from moment to moment. The brief pleasure of a cool breeze on a hot day may make you slightly more positive and optimistic about whatever you are evaluating at the time. The prospects of a convict being granted parole may change significantly during the time that elapses between successive food breaks in the parole judges’ schedule. Because you have little direct knowledge of what goes on in your mind, you will never know that you might have made a different judgment or reached a different decision under very slightly different circumstances. Formulas do not suffer from such problems. Given the same input, they always return the same answer. When predictability is poor—which it is in most of the studies reviewed by Meehl and his followers—inconsistency is destructive of any predictive validity.
The research suggests a surprising conclusion: to maximize predictive accuracy, final decisions should be left to formulas, especially in low-validity environments. In admission decisions for medical schools, for example, the final determination is often made by the faculty members who interview the candidate. The evidence is fragmentary, but there are solid grounds for a conjecture: conducting an interview is likely to diminish the accuracy of a selection procedure, if the interviewers also make the final admission decisions. Because interviewers are overconfident in their intuitions, they will assign too much weight to their personal impressions and too little weight to other sources of information, lowering validity. Similarly, the experts who evaluate the quas plity of immature wine to predict its future have a source of information that almost certainly makes things worse rather than better: they can taste the wine. In addition, of course, even if they have a good understanding of the effects of the weather on wine quality, they will not be able to maintain the consistency of a formula.
Thinking, Fast and Slow Page 26