Paul Meehl

Paul Meehl is presented as one of the most versatile psychologists of the twentieth century and as a central figure in the book’s argument for disciplined prediction. His work confronted the prestige of expert intuition with an awkward empirical possibility: when a judgment depends on uncertain, complex information, a simple rule may outperform a highly trained professional. Meehl’s importance lies not in dismissing expertise altogether, but in showing that confidence, sophistication, and personal contact are not reliable substitutes for tested predictive validity.

Identity and Intellectual Range

Meehl held faculty appointments in psychology, law, psychiatry, neurology, and philosophy at the University of Minnesota. His interests also extended to religion, political science, and learning in rats, an unusually broad range that made him difficult to place within a single academic specialty. He was both a statistically sophisticated researcher and a fierce critic of empty claims in clinical psychology, yet he also practiced psychoanalysis. The book’s narrator regarded him as a hero after reading his work, although the two never met.

Meehl’s intellectual character is defined by this combination of breadth and skepticism. He could engage philosophical questions about psychological research while insisting that claims about prediction be tested against outcomes. His career therefore becomes an implicit counterpoint to the image of the intuitive expert: wide knowledge may be valuable, but it does not by itself demonstrate that a judgment method works.

Clinical Judgment versus Statistical Rules

In Clinical vs. Statistical Prediction: A Theoretical Analysis and a Review of the Evidence, Meehl reviewed 20 studies comparing subjective predictions by trained professionals with statistical predictions produced by combining a few scores or ratings according to a rule. One typical study asked counselors to predict the grades freshmen would receive at the end of the school year. The counselors interviewed each student for forty-five minutes and had access to high school grades, several aptitude tests, and a four-page personal statement. The statistical formula used only high school grades and one aptitude test, yet it was more accurate than 11 of the 14 counselors.

The same pattern appeared in other domains, including violations of parole, success in pilot training, and criminal recidivism. Meehl’s finding was disturbing because the counselors possessed more information and had the opportunity to interpret it personally, while the formula operated on a narrow selection of variables. The comparison suggested that additional information can become a liability when human judgment combines it inconsistently or gives undue weight to vivid but weak evidence.

Meehl’s book provoked shock and disbelief among clinical psychologists, and the controversy continued for more than fifty years. By the time the book’s account was written, roughly 200 studies had compared clinical and statistical prediction. About 60 percent found significantly greater accuracy for algorithms; the remaining studies were ties. Because statistical rules are normally much less expensive to use than expert judgment, a tie effectively favored the formulas. The evidence covered medical outcomes such as cancer longevity, hospital stays, cardiac-disease diagnosis, and susceptibility to sudden infant death syndrome; economic outcomes such as business success, credit risk, and career satisfaction; governmental decisions involving foster parents and juvenile recidivism; and miscellaneous predictions including football results and Bordeaux wine prices. These were described as low-validity environments, settings in which uncertainty and unpredictability limit the reliability of intuition.

Why Algorithms Often Prevail

Meehl suspected that experts often try to be clever. They think outside the box and search for complex combinations of features, even though complexity may improve a prediction only in unusual cases and more often reduces validity. Simple combinations can be more dependable because they apply the same weights to the same information. Studies even found that human decision makers could remain inferior to a formula after being shown the formula’s score: they believed they had additional case-specific information that justified overriding it, but they were wrong more often than not.

The principle was not absolute. Meehl’s famous broken-leg rule allowed a formula to be disregarded when decisive new information appeared, such as learning that a person who was expected to go to the movies had broken a leg that day. A broken leg is both rare and powerful evidence. The rule’s force comes from its exception: judgment should not replace a formula merely because a person feels more perceptive, but genuinely decisive information can justify intervention.

Meehl also emphasized inconsistency. Professionals frequently reach different conclusions when evaluating the same information twice. Experienced radiologists classifying chest X-rays as normal or abnormal contradicted themselves 20 percent of the time on separate occasions. A study of 101 independent auditors found a similar degree of inconsistency, and a review of 41 studies suggested that such variation was typical among auditors, pathologists, psychologists, organizational managers, and other professionals, sometimes even when reevaluation occurred only minutes later. For Meehl’s argument, this was not a minor defect of style: an unreliable judgment cannot be a valid predictor of anything.

The book connects this inconsistency to the context dependence of System 1. Unnoticed environmental stimuli can alter thought and action; a cool breeze on a hot day may make an evaluator slightly more positive, while the timing of food breaks may affect parole judgments. A formula does not fluctuate with such incidental circumstances. Given the same input, it returns the same result, making it more stable even when human judgment feels richer.

Place in the Book’s Larger Argument

Meehl’s research gives empirical structure to the book’s broader account of heuristics and biases. Intuitive thinking is often fast and economical, but it can answer a simpler substitute question rather than the intended one, neglect base rates, exaggerate the diagnostic value of vivid evidence, and construct a coherent story from whatever information happens to be available. Bayesian reasoning requires prior beliefs to be combined with the diagnosticity of evidence; disciplined prediction requires comparable attention to calibration, consistency, and outcomes rather than to the subjective plausibility of an explanation.

His work also illuminates the contrast between experts and formulas in later examples. Political and economic forecasters in Philip Tetlock’s study performed worse than equal weighting across three possible outcomes, while financial officers’ forecasts of the following year’s Standard and Poor’s index had a correlation with actual returns that was slightly below zero. Meehl’s pattern helps explain why knowledge can increase an illusion of skill without producing dependable prediction. The problem is not simply ignorance. It is the human tendency to overinterpret limited evidence, add unstable complexity, and overlook the noise introduced by changing contexts.

Meehl therefore occupies a crucial methodological role. The book’s research program does not ask whether an expert sounds persuasive or possesses impressive credentials; it asks whether the method is externally validated, properly calibrated, and reproducible. In that framework, Meehl’s legacy is a demand for humility before uncertainty. The strongest prediction may come not from the person who tells the most compelling story, but from the modest rule that applies a few relevant facts in the same way every time.