The Ghost’s Story

Everybody knows John Nash now, thanks to the film “A Beautiful Mind,” even if they are not entirely certain exactly what it was that John Nash did. The story of the Nobel laureate’s schizophrenia is to be told by Mike Wallace on Sunday evening (March 17) on “60 Minutes,” and again on “The American Experience” on public television on April 28. And of course the Academy Awards will be given Monday night. It may be a good time, therefore, to recall the career of another man, whom relatively few remember.

John McDonald was the business journalist who brought the advent of game theory to the attention of the readers of Fortune Magazine and the book-reading public over the course of thirty years. In retrospect, I suppose that it also can be said that he contributed to the eclipse of Nash, since he conveyed the impression that the field had sprung full-blown from the brow of the far better-known thinker John Von Neumann. To John McDonald, the really interesting Nash was Charles, the automotive pioneer.

Does it matter that the market for reputation wasn’t efficient? I think not. Yet the story of McDonald’s remarkable career sheds some light on how we get our news of intellectual developments — through a combination of the windshield and the rear-view mirror.

McDonald’s interest had its beginnings in a series of articles he set out in 1945 to write for Fortune on various games. The first one chosen was poker, but the draft (he said) turned out to be “nostalgic without drive.” Fortune in those days was an impossibly grand enterprise, headquartered in the Empire State Building and home to a remarkable collection of writers. McDonald strolled own the hall to the office of his colleague, the young economist John Kenneth Galbraith, who called his attention to an eminent if somewhat obscure authority on games in Princeton by the name of John von Neumann.

McDonald engaged Von Neunann and his sidekick Oskar Morgenstern in a series of conversations. They had just published “Theory of Games and Economic Behavior.” What emerged were two articles — “Poker, An American Game” in 1948 and “A Theory of Strategy” in 1949 — followed in 1950 by a little book in 1950, “Strategy in Poker, Business and War,” complete with whimsical illustrations by Robert Osborn. All ambitions of the new science are set out there in lucid language, beginning with the straightforward acknowledgement that the key conceptions all had been discovered through mathematics, that “rigorous medium of logic.” So solid was McDonald’s reporting that the book was still selling reliably 25 years later.

The project that occupied the writer’s middle years had to do with other things, at least on its surface. In 1954, he began a collaboration with Alfred Sloan, the marketing genius who had built General Motors into the most formidable corporation in the post-war world. Ten years later Sloan’s book, “My Years With General Motors,” ghost-written by McDonald, appeared and became a classic _ but only after an epic behind-the-scenes battle in which GM lawyers sought to suppress it for fear it would become evidence in an antitrust suit. McDonald’s final book, “A Ghost’s Memoir,” describes this story in beguiling detail, mostly at the expense of business guru Peter Drucker, whose 1990 introduction to a new edition of Sloan’s book offered a thoroughly potted account of the origins of the project. McDonald’s book will appear in May, three years after the author’s death at the age of 92 — a poignant reminder of a more graceful epoch in business journalism.

In 1975, McDonald returned to the topic of the underlying importance of game theory with “The Game of Business.” As before, he sought out experts — William Lucas, Lloyd Shapley and Martin Shubik — for advice. His goal, he wrote, was to illuminate “The rules, choices and strategies in important business events.”

Much had changed in game theory by then. Throughout the 1950s, the work had been done mainly in Princeton and at the Air Force-supported RAND Corporation in Santa Monica — and then almost entirely by mathematicians, for Von Neumann’s theory was too mathematical to interest most economists at first. Moreover, much of this work had the character of breaking away from the initial vision. When a distinguished group gathered at the Institute for Advanced Study in Princeton last month to pay homage to the original pioneers, mathematician Harold Kuhn surveyed the early breakthroughs — Nash’s theorems, Shapley’s statistical approach, Al Tucker’s delineation of the famous prisoner’s dilemma as the quintessential description of the interplay between competition and cooperation.

“It is important to recognize,” said Kuhn, “that the results I have enumerated did not respond to some suggestion of Von Neumann, nor did they follow work that he had outlined proposed; rather they were revolutionary ideas that ran counter to Von Neumann’s preferred vision of the theory.” In his treatise, Von Neumann had stated it was impossible to give a practical all-inclusive formal description (the extensive form) of a game. Kuhn introduced the geometrical game tree in 1950. The 41-year old Von Neumann had invented hierarchical computer architecture in 1944; the 26-year old Nash described parallel processing ten years later.

None of this internal drama was to be found in McDonald’s account. Instead, the author sought to show how present-day game theory could parse some of the century’s most interesting big business stories. Seagrams and Schenley battle over strategic reserves of whiskey. General Motors bests Ford at selling automobiles. J. Paul Getty wins big by defecting in an oil industry game. Howard Hughes dramatizes the strategic advantages of secrecy. Walt Disney stakes all on his brand name. Sears, J.C. Penny and Montgomery Ward learn to cooperate to carve up the general merchandise market. The fundamental distinctions are illustrated — coalition-building vs. individualistic non-cooperative games. Zero-sum, win-or-lose situations vs. those in which in which every player has something to gain. Strictly chancy games vs. those in which skill and strategy play a part. But in the end, McDonald’s magnum opus turned out to be a ponderous and dull book.

Much better expositions of game theory soon would appear. Perhaps the clearest example of how things had changed with the growth of knowledge is “Poker Strategy: Winning with Game Theory” a 1981 book by MIT mathematician Nesmith Ankeny. Where McDonald’s original essay on poker had been philosophical, verbal, open-ended, Ankeny’s apoproach was concise, statistical, closed — a system that could be taken to Las Vegas and played. (“Do not call the opener with any pair less than Kings.”) Ten years later economists Avinash Dixit and Barry Nalebuf offered an up-to-date tour d’horizen in “Thinking Strategically: The Competitive Edge in Business, Politics and Everyday Life,” complete with a succint definition of Nash equilibrium — given what the other players are doing, none wants to change his own game.

A few years after that, Sylvia Nasar reintroduced Nash — newly honored by the Swedish Academy of Sciences and his schizophrenia in abeyance — to his colleagues in the brilliant biography that became the basis for the film. Even then, the movie got the key thing wrong. In a little fable about dating strategies, the I-think-that-he-thinks strategies of the guys assure that the knock-out blonde in a bar doesn’t get a date. That would never happen in a real-life Nash equilibrium; someone would give her a tumble. In a somewhat similar way, McDonald’s 1975 book didn’t have to be as good as his first. It was an unmistakable wake-up call to those who took it to heart that business executives had learned that they could make serious money by thinking deeply about their problems. The mid-’70s were a very good time in which to apply to business school.

The moral is that no single mode of coverage of the social sciences is appropriate. There ought to be a medley at all times. Scientific revolutions occur occasionally. The rest of the time there is normal science. Where once only trespassing mathematicians had dared, today no fewer than eight top economists are working in Princeton on strategy, including Institute professor Eric Maskin. Yet in other fields — in computer science, in biology — the applications of game theory have barely begun.

Indeed, if you went looking for a McDonald-like figure in journalism today, someone reporting the discovery of new intellectual continents, you might find Matt Ridley. The former science editor of London’s Economist is the author of three books on evolutionary biology — “Genome,” “The Evolution of Virtue” and “The Red Queen: Sex and the Evolution of Human Nature” — and of a pretty good account of the 1988 American presidential campaign as well. On the other hand, if you wanted news about economics going forward, filling in the contours of the map that Von Neumann, Nash and the others sketched with high excitement in the late ’40s and early ’50s, you’d look elsewhere. You could do worse, for instance, than to continue to read Economic Principals.

Leave a Reply

Your email address will not be published. Required fields are marked *