| The Art of Strategy |
| Avinash Dixit and Barry Nalebuff, Norton, September 2008, 472 pages, 16.99. |
| Unlike most of the ranks of management advice books which pad out bookshop business sections, here is one which is rigorous, fun and extremely useful all at the same time. This is an almost entirely new version of the authors’ 1991 book Thinking Strategically. It is a ‘how to’ text about game theory, ideal for all of us whose knowledge of game theory isn’t as thorough as we might like. The many applications in this book illustrate the concepts beautifully clearly. |
| My favourite examples are in fact the game shows one of the co-authors designed for American television, such as the diet show on ABC which photographed overweight contestants – whose previous diets had always failed – in tiny bikinis before they dieted. The pictures of all their flab would be shown on network television if they failed to lose 15 pounds in two months. It proved a fantastically effective commitment device; none of them failed to shed the required amount. A great example – and of course it’s also good to know of another gainful employment opportunity for economists. We’ve all heard about auction design, but now a new career avenue in game-show design beckons. |
| On a more serious note, the book demonstrates the advantages that can be gained from strategic thinking, in environments where most people think tactically, if they think that much at all. Just a couple of insights alone make it well worth the price. One is in the context of mixed strategies, where both players will be randomly alternating between two choices in proportions which depend on their payoff matrices. A change in one person’s payoffs as between the two choices can leave their own strategy unaffected but alter the other player’s mix. It takes a little thought to work out, but highlights the importance of interdependence in strategic thinking. |
| The other is the downside of piecemeal decision-making, as a series of best choices by a group can easily leave everyone in a worse position than they were to begin with. The reason is that sequential choices ignore the intensity of preferences – in small steps, lots of people gain a little and one (who is always outvoted) can lose a lot. So everyone in a group can end up with nine small gains and one very large loss. An example application is trade talks. A better result will emerge from considering a whole package rather than taking the discussions bit by bit. |
| The book contains many questions that readers can work out for themselves, with answers at the back, and examples to work through in spreadsheets, but lazy readers like myself can benefit enormously without having to do so. Since reading this, I’ve been seeing everything in terms of game theory, and it feels like having put on a pair of x-ray goggles to view the world. |
| Diane Coyle |
| Enlightenment Economics |
|