In 2003, we reviewed “Moneyball,” Michael Lewis’s book about Billy Beane and the Oakland A’s. The book, we noted, had become a sensation, despite focussing on what would seem to be the least exciting aspect of professional sports: upper management. Beane was a failed Major League Baseball player who went into the personnel side of the business and, by applying superior “metrics,” had remarkable success with a financial underdog. We loved the book—and pointed out that, unbeknownst to the author, it was really about behavioral economics, the combination of economics and psychology in which we shared a common interest, and which we had explored together with respect to public policy and law.

Why isn’t the market for baseball players “efficient”? What is the source of the biases that Beane was able to exploit? Some of the answers to these questions, we suggested, might be found by applying the insights of the Israeli psychologists Daniel Kahneman and Amos Tversky, on whose work behavioral economics greatly relies. Lewis read the review, began to take an interest in the whole topic of human rationality, and, improbably, decided to write a book about Kahneman and Tversky. He kindly even gave us credit for setting him down this path.

Though we were pleased that Lewis was taking an interest in our field, we admit to being skeptical when we heard about his book plan. Granted, Lewis has shown many times before—not only with “Moneyball” but also with “The Big Short,” his book about the real-estate market, and “Flash Boys,” which is about high-speed trading—that he can write a riveting book about an arcane subject. And we did not doubt the appeal of the book’s main characters: one of us had written several papers with Kahneman, and the other had known Kahneman and Tversky since 1977 and had collaborated with both men. (Tversky died in 1996, at the age of fifty-nine. Kahneman, now eighty-two, is blessedly still very much with us.) Both of us had been deeply influenced by their joint work on the psychology of judgment and decision-making. Still, how was Lewis going to turn a story about their lives into the kind of page-turner that he’s known for? Kahneman and Tversky were brilliant, but they did most of their work together more than thirty years ago, and they worked primarily by talking to each other, switching between English and Hebrew. Where’s the book?

Our skepticism was misplaced. The book, titled “The Undoing Project: A Friendship That Changed Our Minds,” captivated both of us, even though we thought we knew most of the story—and even though the book is just what Lewis had said it would be, a book about Amos and Danny, two men who changed how people think about how people think. Lewis accomplishes this in his usual way, by telling fascinating stories about intriguing people, and leaving readers to make their own judgments about what lessons should be learned. He provides a basic primer on the research of Kahneman and Tversky, but almost in passing; what is of interest here is the collaboration between two scientists. Having written several articles and one book together, we have firsthand experience in both the joys and struggles of getting two minds to speak with one voice, and the conflicts that can arise when one author is a fast writer and the other likes to linger over each word. And while one gleans a good deal about teamwork from the book, Lewis doesn’t spell the lessons out. Instead, the reader learns through observation, getting as close as anyone could to being in those closed rooms where the two men worked.

In 1968, Tversky and Kahneman were both rising stars in the psychology department at the Hebrew University of Jerusalem. They had little else in common. Tversky was born in Israel and had been a military hero. He had a bit of a quiet swagger (along with, incongruously, a slight lisp). He was an optimist, not only because it suited his personality but also because, as he put it, “when you are a pessimist and the bad thing happens, you live it twice. Once when you worry about it, and the second time when it happens.” A night owl, he would often schedule meetings with his graduate students at midnight, over tea, with no one around to bother them.

Tversky was a font of memorable one-liners, and he found much of life funny. He could also be sharp with critics. After a nasty academic battle with some evolutionary psychologists, he proclaimed, “Listen to evolutionary psychologists long enough, and you’ll stop believing in evolution.” When asked about artificial intelligence, Tversky replied, “We study natural stupidity.” (He did not really think that people were stupid, but the line was too good to pass up.) He also tossed off such wisdom as “The secret to doing good research is always to be a little underemployed. You waste years by not being able to waste hours.” Managers who spend most of their lives in meetings should post that thought on their office walls.

Early in his career, Tversky was a “mathematical psychologist,” which meant that he used formal models to characterize human behavior. He didn’t care for metaphors: “They replace genuine uncertainty about the world with semantic ambiguity. A metaphor is a cover-up.” He was organized and highly disciplined. His office was spotless; there was nothing on his desk except a pad, a mechanical pencil, and an eraser. (Even Tversky made mistakes.)

If there were a pad and pencil in Daniel Kahneman’s office, on the other hand, Kahneman would struggle to find them. Born in Tel Aviv when his mother was visiting family, he spent his childhood in Paris, speaking French as his first language. His father was a chemist in a cosmetics company. In 1940, the German occupation put the family at risk. Hiding in the South of France, they managed to survive (with the exception of his father, who died in 1944, from untreated diabetes). After the war, the rest of the family immigrated to Palestine.

A constant worrier, Kahneman is an early riser who often wakes up alarmed about something. He is prone to pessimism—claiming that, by expecting the worst, he is never disappointed. This pessimism extends to the expectations he has for his own research, which he likes to question: “I get a sense of movement and discovery whenever I find a flaw in my thinking.” In our own collaborations with Kahneman, we saw this close up, as he would proclaim, at what seemed to be the final stages of some joint work, that he had just discovered a fatal problem with our whole approach and that we would have to give up or start all over again. He was usually wrong about that—but sometimes he was right, and the constant worry made the work much better.

Tversky liked to say, “People are not so complicated. Relationships between people are complicated.” But then he would pause and add, “Except for Danny.” So, yes, they were different, but those who saw them together, spending endless hours just talking, knew that something special happened when they applied their two very different minds to a problem. Lewis both captures and sharpens the contrast between them, showing us why their collaboration was impossibly incongruous and yet perfectly complementary.

The names Kahneman and Tversky are now well known among social scientists, but even experts in the field will not know the story of how their collaboration began. At the beginning of their careers, they worked in different branches of psychology: Kahneman studied vision, while Tversky studied decision-making. Like much of psychology, these topics can be studied only indirectly; one can’t directly monitor what people see or think (yet). In those days, mathematical psychologists like Tversky conceived of thinking in much the same way as economists: choices were thought to be made more or less “correctly,” as people incorporate new information and make good choices for themselves. By contrast, those studying vision made much use of common mistakes such as visual illusions. (What does the fact that we see what seems to be water on a desert highway tell us about the vision system?) As Kahneman put it, “How do you understand memory? You don’t study memory. You study forgetting.”

In the spring of 1969, Kahneman invited Tversky to speak at his seminar. Tversky chose to outline some cutting-edge experiments about how people learn from new information. The experiments seemed to demonstrate that ordinary people were close to being rational; they thought like “intuitive statisticians.” Though the presentation was impressive, Kahneman thought that the experiments were, as Lewis writes, “just incredibly stupid,” and that they demonstrated no such thing. Insisting that judgments are more like sensory perceptions (and similarly prone to error), he went after Tversky hard, as people do in the best academic environments. Tversky almost never lost an argument, but he lost this one.

Very much in character, Tversky reacted to this loss by coming back for more. His friend Avishai Margalit, the distinguished Israeli philosopher, calls the session “Kahneman and Tversky’s Big Bang.” He recalls meeting an agitated Tversky, who “started by dragging me into a room. He said, ‘You won’t believe what happened to me.’ He tells me that he had given this talk and Danny had said, ‘Brilliant talk, but I don’t believe a word of it.’ ”

Before long, Kahneman and Tversky were in constant conversation. They worked intensely in a small seminar room or a coffee shop, or while taking a long walk. The sessions were private; no one else was invited to join. As they began to produce work together, each sentence would be written, rewritten, and rewritten again, with Kahneman manning the typewriter. (Tversky never did master the art of the keyboard.) On a good day, they would write a paragraph or two. Everything was produced jointly; they did not really know where one’s thought ended and the other’s began. Graduate students “now wondered how two so radically different personalities could find common ground, much less become soul mates,” Lewis writes. One reason was that “Danny was always sure he was wrong. Amos was always sure he was right.”

That actually did help. While Tversky was “the most terrifying mind most people had ever encountered,” he was uncharacteristically receptive to Kahneman’s ideas. Kahneman, for his part, found Tversky’s arrogance surprisingly liberating: “It was extremely rewarding to feel like Amos, smarter than almost everyone.” And they laughed together—a lot. As Kahneman said, “Amos was always very funny, and in his presence I became funny as well, so we spent hours of solid work in continuous amusement.”

What followed was a period of extraordinary creativity—the best and most original work that either of them had done, or would do, at any stage in his career. In the period between 1971 and 1979, they published the work that would eventually win Kahneman the Nobel Prize in Economics. (The prize would certainly have been shared with Tversky had he still been alive. Nobel Prizes are not awarded posthumously.) There were two distinct themes: judgment and decision-making. Judgment is about estimating (or guessing) magnitudes and probabilities. How likely is it that a billionaire businessman from New York with no experience in government gets elected President? Decision-making is about how we choose, especially when there is uncertainty (meaning almost all the time). What should we do now?

Kahneman and Tversky showed that, in both of these domains, human beings hardly behave as if they were trained or intuitive statisticians. Rather, their judgments and decisions deviate in identifiable ways from idealized economic models. Most of the importance of Kahneman and Tversky’s work lies in the claim that departures from perfect rationality can be anticipated and specified. In other words, errors are not only common but also predictable.

For instance: ask people what they think is the ratio of gun homicides to gun suicides in the United States. Most of them will guess that gun homicides are much more common, but the truth is that gun suicides happen about twice as often. The explanation that Kahneman and Tversky offered for this type of judgment error is based on the concept of “availability.” That is, the easier it is for us to recall instances in which something has happened, the more likely we will assume it is. This rule of thumb works pretty well most of the time, but it can lead to big mistakes when frequency and ease of recall diverge. Since gun homicides get more media coverage than gun suicides, people wrongly think they are more likely. The availability heuristic, as Kahneman and Tversky called it, leads people to both excessive fear and unjustified complacency—and it can lead governments astray as well.

The influence of their work has been immense—not only in psychology and economics, where it has become part of the normal conversation, but in every other field of social science, as well as medicine, law, and, increasingly, business and public policy. And this legacy is based on what by current standards would be considered a very small number of papers—eight, to be precise. (They went on to write more papers together in the years that followed, but the foundation was laid with those few from the nineteen-seventies.)

The low rate of output was one of their strengths, and is a direct result of their joint personality traits. Kahneman’s constant worry about how they might be wrong combined perfectly with Tversky’s mantra: “Let’s get it right.” And it takes a long time to write a paper when both authors have to agree on every word, one by one.

The Kahneman and Tversky partnership was extraordinary in terms of its scientific impact—they are the Lennon and McCartney of social science—and even now, when joint work is increasingly common in academia, enduring teams like theirs are extremely rare. In Lewis’s account, the relationship between Kahneman and Tversky was as intense as a marriage. As anyone who has been married knows, marriages can be fraught, and they sometimes dissolve entirely, rarely amicably. Tversky and Kahneman never got divorced, but they did start dating other people, and their relationship became strained.

After the two decided to leave Israel, in 1978, Tversky quickly received offers from Harvard and Stanford (where he ended up). Kahneman, who was looking for jobs jointly with his equally distinguished wife, Anne Treisman, was hired at the University of British Columbia, in Vancouver—a fine university, but lower in status than those that pursued his friend. At a relatively young age, Tversky received honorary degrees from Yale and the University of Chicago.

Although their work had been a true collaboration of equals, Tversky had unofficially been declared the star of the team, which didn’t sit well with Kahneman. Tensions were aggravated in 1984, when Tversky was given a MacArthur “genius” grant, and Kahneman wasn’t. Kahneman was not actually eligible for the award, which is given only to American citizens or residents, but not many people realized this—and, what’s more, when Kahneman moved to Berkeley, two years later, thus becoming eligible, the MacArthur Foundation still did not give him a fellowship. The incident illustrates another one of Kahneman and Tversky’s most famous concepts: loss aversion. When the MacArthur grants are awarded every year, only the most egomaniacal of us read the list and say, “Damn, I lost.” Unless, that is, your best friend wins the prize for work you did entirely together.

The two did not stop being friends, or stop talking nearly every day, or stop working on occasional projects. But once they were separated by distance, and began working with students and other co-authors, their relationship lost its ease. The way Kahneman saw it, “Amos changed. When I gave him an idea he would look for what was good in it. For what was right with it. . . . He stopped doing that.” He noted, “Something happens when you are with a woman you love. You know something happened. You know it’s not good. But you go on.” Tellingly, he added, “I wanted something from him, not from the world.” After one particularly difficult interaction, Kahneman decided, and told Tversky, that they were no longer even friends. “I sort of divorced him.” This is the kind of outburst that Kahneman typically takes back within a few days, as he did at least a dozen times when he declared that he was abandoning for good his book project, which would eventually become the mega-best-seller “Thinking, Fast and Slow.”

In the case of his breakup with Tversky, fate intervened to hasten the inevitable reversal. Only three days later, Tversky called to say that he had just been diagnosed with a malignant melanoma and that he had, at most, six months to live. As Kahneman recalled, “He was saying, ‘We’re friends, whatever you think we are.’ ”

In the remaining six months, Kahneman and Tversky worked on the introduction to an edited collection of papers related to their work, an introduction Kahneman had to finish after Tversky died. Kahneman had (of course) worried about completing this introduction alone, and Tversky had (of course) assured him that he should just trust the mental model that, by now, he surely had of Tversky’s mind. But no one has that model, alas. That is why collaborations are so special: one partner cannot simply replace the mind of the other, even after twenty-five years.

Tversky once said, “It is sometimes easier to make the world a better place than to prove you have made the world a better place.” But it is not hard to prove that Amos and Danny did so—you only have to read those papers that they published in the seventies. Or, for that matter, Lewis’s book.