More on Strategy and Game Theory
| Peter Klein |
Recent posts on strategy and game theory (here and here) generated quite a lot of discussion here and on social media. Avinash Dixit offers more grist for the mill in his December 2014 Journal of Economic Literature essay on Lawrence Freedman’s Strategy: A History (Oxford, 2013). (An ungated version is here.) Dixit’s essay contrasts the economist’s and the historian’s view of strategy — “strategy” being game theory for the former, a broader, interpretive, interdisciplinary exercise for the latter — but the discussion is highly relevant for strategic management. The management literature has traditionally taken a wide, flexible view of “strategy,” closer to the historian’s sense than the economist’s, though that is rapidly changing as game theory becomes more widespread in strategic management research and teaching.
Here’s an excerpt from Dixit’s opening, which gives you the flavor:
[Freedman] heads the preface with a memorable quote from Mike Tyson: “Everyone has a plan till they get punched in the mouth.” Later he quotes another fighter, the legendary German Field Marshal Helmuth Karl Bernhard Graf von Moltke: “no plan survived contact with the enemy” (p. 104). The game theorist will respond: “Those plans are not strategies. They are incomplete. They fail to specify any action at the node of the game tree where you get punched in the mouth or meet the enemy army, or in the ensuing subgame.” It would be extreme stupidity, or arrogance tantamount to stupidity, for a boxer not to recognize the possibility of getting punched in the mouth. And although avoiding battle may be an important aspect of military strategy in many situations (see pp. 47–9), every plan should include a provision for action if or when battle commences. Tyson, or Freedman, will probably counter that even if the boxer starts with a complete plan that specifies the action for this contingency, the punch will make him forget the plan and react hot-headedly. Modern game theorists exposed to behavioral ideas will admit some truth in this, and agree that the boxer’s System 2 calculations are likely to fly out of the ring when the punch lands and System 1 instincts will take over. But they will add that that makes it all the more important for the boxer to strategize better in advance—to take actions before getting punched, either to reduce the risk, or to arrange matters in such a way that the anger and instinct (or the prospects of such reactions) are put to more effective use, as with the strategy of brinkmanship. More generally, “the art of creating power” often entails strategic moves like commitments, threats and promises that game theorists have analyzed following Thomas Schelling (1960). And Freedman’s picture of “strategy as a System 2 process engaged in a tussle with System 1 thinking” (p. 605) looks remarkably like Schelling’s (1984, ch. 3) “intimate contest for self-command.”
I have a twofold purpose in constructing the above exchange. One is to highlight the difference between the perspectives of economists and historians in thinking about the same situation. The second is to argue that each has something to learn from the other, and a fuller understanding can result from their dialog. The two perspectives share a lot of middle ground, and have useful complementarities.
The thoughtful essay is well worth reading in its entirety.