It's only a game
The Economist
15 June 1996

Economist Banner




"Managers have much to learn from game theory - provided they use it to clarify their thinking, not as a substitute for business experience" 

FOR old-fashioned managers, business was a branch of warfare - a way of 'capturing markets' and 'making a killing'. Today, however, the language is all about working with suppliers, building alliances, and thriving on trust and loyalty. Management theorists like to point out that there is such a thing as 'win-win', and that business feuds can end up hurting both parties. 

But this can be taken too far. Microsoft's success has helped Intel, but it has been hell for Apple Computer. Instead, business needs a new way of thinking that makes room for collaboration as well as competition, for mutual benefits as well as trade-offs. Enter game theory. 

Stripped to its essentials, game theory is a tool for understanding how decisions affect each other. Until the theory came along, economists assumed that firms could ignore the effects of their behaviour on the actions of rivals, which was fine when competition was perfect or a monopolist held sway, but was otherwise misleading. Game theorists argue that firms can learn from game players: no card player plans his strategy without thinking about how other players are planning theirs. 

Economists have long used game theory to illuminate practical problems, such as what to do about global warming or about fetuses with Down's syndrome. Now business people have started to wake up to the theory's possibilities. McKinsey, a consultancy, is setting up a practice in game theory. Firms as diverse as Xerox, an office-equipment maker, Bear Stearns, an investment bank, and PepsiCo, a soft-drinks giant, are all interested. They will no doubt seize on 'Co-opetition' (Doubleday, $ 24.95), because it is written by two of the leading names in the field, Adam Brandenburger, of Harvard Business School, and Barry Nalebuff, of the Yale School of Management. It also helps by using readable case studies rather than complex mathematics. 

The main practical use of game theory, say the authors, is to help a firm decide when to compete and when to co-operate. Broadly speaking, the time to co-operate is when you are increasing the size of the pie, and the time to compete is when you are dividing it up. The authors also argue that, to get a full picture of their business, managers need to think about a new category of firms, 'complementers', which lead your customers to value your products more highly than if they had only your product. Hot-dog makers and Colman's mustard are complementers: buy one and you are more likely to buy the other. So are Intel and Microsoft. 

The most important thing to know about a game is who the players are. A small change in the number of players can have unexpected consequences. The Holland Sweetener Company, a Dutch-Japanese joint venture, discovered this to its cost in the late 1980s when it tried to break NutraSweet's monopoly of the American artificial-sweetener market. NutraSweet managed to keep the predator out, but only after Coca-Cola and Pepsi used the threat of competition to force NutraSweet to lower its prices. 

When competition between two players benefits third parties in this way, there is scope for the beneficiary to split its gains. Holland Sweetener in effect gave up its share of the gains that it had helped Coke and Pepsi to win. BellSouth, a telephone company, was wiser: it insisted on being paid to play. The firm said that it would bid against Craig McCaw for control of LIN Broadcasting Corporation only if LIN paid it $ 54m for entering the fray and a further $ 15m in expenses if it lost the bid. 

One way for a player to do well in a game is to make itself indispensable. Nintendo built its video-games business in the late 1980s by restricting software developers to making five games each, keeping retailers on short rations, and doing much of the development in-house. Nobody else had any bargaining power. By contrast, IBM stored up trouble for itself in personal computers by allowing Microsoft and Intel to establish a lock on the two most valuable bits of the business. 

A second technique is to tempt lots of competing players into the game - for instance by increasing the prize. That is what American Express did in 1994 when it organised a coalition with other big companies to purchase health care. The potential contract was so large that a host of health-care providers got into a bidding war. A third technique is to make intelligent use of a resource which is worth more to your customer than to you. In 1993 TWA lifted itself off the bottom of the airline league by tearing out several rows of seats that were usually empty because the carrier was so unpopular, giving passengers more leg-room - and making the airline popular once more. 

When to stop playing 

Game theory seems a fine way to analyse decisions retrospectively. But is it much help in the heat of battle? The track record of grand ideas imported from other disciplines, notably chaos theory, is not impressive. However, the game theorists have already notched up some significant practical successes. The Federal Communications Commission used the theory to help design its $ 7 billion auction of radio spectrum for mobile phones - and hundreds of mobile-phone companies also used the theory to formulate their bids. 

But, as Peter Scott-Morgan, a consultant with Arthur D. Little, points out, game theorists are worryingly silent about the links between a company's strategy and its internal capabilities. Today's most successful managers craft their strategies on the basis of knowledge of their own companies, and devote at least as much thought to the question of how blueprints will be translated into practice. At their worst, game theorists represent a throwback to the days of such whiz-kids as Robert McNamara, chairman of Ford in 1960 and later defence secretary, who thought that rigorous analytical skills were the key to success. 

Yet nothing can ever substitute for deep first-hand knowledge and experience. So, however sophisticated the games that managers play may be, they will still need to get their hands dirty. 



© 1996 The Economist Newspaper Limited. All rights reserved.