This book was brought to my attention by good friend, and gaming colleague Wayne and I immediately rushed out, bought the book and read it in a couple of days. I was both pleased and disappointed with the book.
Richards J. Heuer Jr.’s Psychology of Intelligence Analysis is an excellent examination of how analysts interpret information, form judgements and frequently mislead themselves without realising they have done so. Although written primarily for the American intelligence community, its arguments apply equally well to military planning, historical research, risk assessment, wargame design and almost any activity requiring decisions to be made from incomplete or uncertain evidence.
The book brings together articles originally written by Heuer between 1978 and 1986 for use within the CIA Directorate of Intelligence. They were revised and published as a collected volume in 1999.¹ This origin explains both the book’s strengths and some of its limitations. It is practical, accessible and firmly connected to the problems encountered by working analysts. At the same time, it occasionally feels like a series of related instructional papers rather than a single argument developed steadily from beginning to end.

Heuer begins with the uncomfortable proposition that analysts do not simply receive information and interpret it objectively. The mind organises what it sees through existing expectations, experience and assumptions. These mental structures are necessary because the amount of information available would otherwise be overwhelming, but they also determine which details are noticed, which are ignored and how ambiguous evidence is understood. Two analysts can therefore examine the same material and reach different conclusions without either being dishonest or incompetent.
One of the familiar examples is the ambiguous drawing first published by William Ely Hill in 1915, which can be seen either as a young woman looking away or as the face of an elderly woman.² Once one interpretation has been recognised, it can become surprisingly difficult to see the other. The point is simple, memorable and hardly new, but it remains effective. Analysts do not merely respond to what is before them. They respond to what their minds have already decided is likely to be there.
This leads to one of the book’s strongest arguments. More information does not automatically produce better analysis. Once an analyst has adopted an explanation, additional evidence is often interpreted in a way that supports the existing position. Evidence that fits is welcomed, while contradictory information is discounted as unreliable, exceptional or irrelevant. Expertise provides valuable knowledge, but it can also produce confidence in an established model long after that model should have been reconsidered.
Heuer’s discussion of memory is equally useful. Memory is not a perfectly organised archive from which information can be recovered unchanged. What people remember is reconstructed, influenced by later experience and organised according to what now appears important. Analysts may therefore remember evidence as being more consistent with their final judgement than it actually was. This is particularly dangerous when reviewing previous assessments because the outcome can make earlier uncertainty appear less substantial than it really was.
The middle chapters move from diagnosing these problems to considering methods for controlling them. Heuer examines the value of externalising analytical thought through written assumptions, diagrams, matrices and clearly stated alternative explanations. The most important of these methods is the Analysis of Competing Hypotheses. Instead of selecting the explanation that initially appears most convincing and searching for supporting evidence, the analyst identifies several plausible explanations and compares how well each survives the available evidence.³
This is a valuable reversal of the usual process. Evidence that agrees with a hypothesis is often less useful than it appears because it may agree equally well with several alternatives. The more revealing evidence is that which is difficult to reconcile with one explanation but consistent with another. The method does not remove judgement from analysis, nor does Heuer pretend that it can. Its value lies in making assumptions visible and forcing the analyst to confront alternatives that might otherwise be dismissed too quickly.
The James L. Adams nine dot puzzle serves much the same purpose. The reader is asked to pass four connected straight lines through nine dots without lifting the pencil. Most people unconsciously treat the outer dots as the boundary of a square, although no such boundary has actually been imposed. The solution requires the lines to extend beyond this imagined enclosure.⁴ It is another familiar exercise, but it neatly demonstrates how easily people accept restrictions that exist only in their own understanding of the problem.
The later chapters examine the more familiar collection of cognitive biases. These include the tendency to give excessive weight to information that is vivid, recent or easily remembered, the search for patterns in events that may be largely random, anchoring upon an initial estimate, overconfidence in judgement and the tendency to interpret outcomes as having been more predictable than they actually were. Heuer is particularly effective when explaining how these biases distort estimates of probability.
This is where the book overlaps most clearly with the literature on risk perception. Anyone with experience in that field will recognise most of the usual mechanisms. People overestimate dramatic risks, underestimate familiar ones, anchor upon the first figure presented, search for causal explanations and become increasingly confident as they receive more information, even when the additional material contributes little to accuracy. The work of Daniel Kahneman, Amos Tversky, Paul Slovic and other researchers is clearly present throughout Heuer’s discussion.⁵
For that reason, I found relatively little in the book that was completely new. The ambiguous images, altered playing cards, probability exercises, nine dot problem, competing explanations and warnings about confirmation bias belong to the familiar equipment of courses on judgement, decision making and risk. Readers who have worked extensively in these areas will probably recognise nearly all the tricks before Heuer explains them.
That is not a serious criticism of the book. Its achievement is not the discovery of an entirely new psychology. It is the translation of established psychological research into the daily practice of intelligence analysis. Heuer shows why these apparently simple demonstrations matter when an analyst is attempting to understand military preparations, political intentions, deception, economic change or the behaviour of an adversary. A harmless mistake in a classroom puzzle can become extremely serious when the same mental habit is applied to a national security assessment.
There are, however, important limitations. Heuer sometimes treats the findings of psychological experiments as more settled and universally applicable than they really are. Many of the studies discussed in the book involved relatively small groups completing artificial tasks under controlled conditions. Intelligence analysis is a considerably more complicated activity involving trained professionals, extended investigation, institutional review and adversaries deliberately attempting to deceive them. The movement from a laboratory experiment to professional intelligence practice is therefore not always as straightforward as the book occasionally suggests.
The age of the supporting research is another limitation. Heuer’s original articles were largely written during the late 1970s and early 1980s, while much of the experimental literature is older still. Later research has refined, challenged or qualified some of the classic findings concerning cognitive bias. This does not make the book obsolete, since its central warnings remain sound, but it does mean that it should now be read as a foundational work rather than as a complete account of contemporary cognitive psychology.
The Cold War setting also shapes the questions being asked. The intelligence examples naturally reflect the institutional concerns of the CIA and the difficulties of interpreting the intentions of the Soviet Union and other state adversaries. The underlying principles remain relevant, but contemporary analysts face a wider range of problems involving fragmented organisations, rapidly changing information environments, social media, automated collection and the deliberate manufacture of enormous quantities of misleading material.
Heuer concentrates heavily upon the individual analyst, which risks making intelligence failure appear primarily a consequence of faulty personal reasoning. In practice, intelligence assessments are produced within organisations shaped by hierarchy, deadlines, restricted access to information, institutional rivalries, political expectations and the preferences of senior decision makers. An analyst may recognise alternative explanations but still work within a system that rewards consensus or discourages conclusions that challenge established policy.
Related to this is the limited treatment of group behaviour. Intelligence products are seldom the work of one completely independent analyst. They pass through meetings, committees, supervisors and several levels of editing. During that process qualifications may be removed, uncertainty may be reduced and an initially cautious judgement may acquire far greater confidence. Groupthink is part of the problem, but so too are status, institutional loyalty and the ability of senior personalities to dominate a discussion.
The Analysis of Competing Hypotheses is valuable, but Heuer may place too much confidence in its ability to discipline judgement. The method depends upon the analyst identifying the important hypotheses in the first place. If the correct explanation is omitted, the matrix can only produce a carefully organised comparison of incomplete alternatives. Assigning evidence as consistent or inconsistent also remains a matter of judgement, while apparently separate pieces of evidence may originate from the same source and therefore be less independent than they appear.
There is also a danger that structured techniques create an appearance of objectivity that they cannot entirely justify. A table, score or matrix looks more rigorous than an ordinary narrative judgement, but its contents still depend upon human decisions concerning which evidence should be included, how reliable it is and what significance it should receive. Structure can expose reasoning and make it easier to challenge, but it cannot convert uncertain evidence into certainty.
Heuer is convincing about the danger of rigid mental models, but gives less attention to the opposite problem. An analyst who constantly generates alternatives and challenges every assumption may become reluctant to reach any conclusion. Intelligence assessment requires intellectual openness, but it also requires decisions to be made within limited time. The practical difficulty is knowing when further reconsideration will improve the assessment and when it will merely postpone judgement.
The book also pays insufficient attention to cultural differences. Most of its psychological research and institutional experience comes from an American setting. Analysts examining another society may misunderstand its behaviour because they impose assumptions drawn from their own political culture. Language, religion, historical memory, social hierarchy and differing concepts of rational behaviour require more sustained consideration than Heuer provides.
Finally, the book is better at identifying analytical weaknesses than demonstrating that its proposed remedies consistently improve the accuracy of intelligence assessments. Structured methods may produce clearer reasoning, expose assumptions and encourage consideration of alternatives, but those benefits do not automatically result in more accurate forecasts. Evaluating analytical methods requires evidence about outcomes, not merely evidence that analysts followed a more disciplined process.
Heuer nevertheless makes clear that simply knowing the names of cognitive biases does not provide immunity from them. An analyst can explain confirmation bias in great detail and still seek evidence that supports a preferred conclusion. The answer lies in disciplined procedures, explicit assumptions, genuine consideration of alternatives and an institutional culture willing to reward disagreement when that disagreement is supported by evidence.
This is ultimately the book’s greatest strength. It does not promise perfect objectivity or suggest that uncertainty can be eliminated. Instead, it offers methods by which analysts can make their reasoning more visible, testable and open to challenge. Its lessons are valuable far beyond professional intelligence work. Historians, military planners, risk managers and wargame designers all construct explanations from incomplete evidence, and all are capable of mistaking a plausible story for a demonstrated conclusion.
Psychology of Intelligence Analysis remains an excellent work. Experienced readers may find that much of its psychological content is already familiar, particularly if they have worked in risk perception or decision analysis. Nevertheless, the clarity with which Heuer brings these ideas together and applies them to practical analytical problems makes the book well worth reading. It is less a collection of startling revelations than a disciplined reminder of how readily intelligent and experienced people can be trapped by their own expectations.
Its limitations also remind us that cognitive discipline is only one part of better intelligence analysis. Psychology matters, but so do organisational culture, source reliability, professional expertise, cultural understanding and the willingness of decision makers to hear conclusions they would rather not receive.
Footnotes
- Richards J. Heuer Jr., Psychology of Intelligence Analysis (Washington, DC: Center for the Study of Intelligence, Central Intelligence Agency, 1999).
- William Ely Hill, “My Wife and My Mother in Law,” Puck, 6 November 1915. Heuer uses ambiguous figures to demonstrate the influence of expectations upon perception.
- Heuer, Psychology of Intelligence Analysis, 95–110.
- James L. Adams, Conceptual Blockbusting: A Guide to Better Ideas, 2nd ed. (New York: W. W. Norton, 1980), 23; Heuer, Psychology of Intelligence Analysis, 67–68.
- Heuer, Psychology of Intelligence Analysis, 111–72. See also Daniel Kahneman, Paul Slovic and Amos Tversky, eds., Judgment under Uncertainty: Heuristics and Biases (Cambridge: Cambridge University Press, 1982).






























































