Tuesday, August 26, 2014

Welcome to Game Theory - Fall 2014

Welcome to Game Theory. If you are open to the possibility, this course can fundamentally change the way you look at the world--or at least that is my hope. Why do I think this? The main reason is that game theory fosters what I call "outside thinking" or, if you like, empathy. Game Theory is ultimately about seeing the world through others' eyes--walking a mile in their mocassins--and the adjusting your choices in light of theirs.

Sometimes, situations can be adversarial and the insights gains from empathy are used to frustrate the others' plans, to pre-empt or forestall their best options, or to cause confusion so that they make the wrong choices. When we think of games, such as chess, checkers, backgammon, poker, and so on, we often think of "zero sum" situations where, for one side to win, the other must lose. Empathy in such situations is often called "outsmarting" the other player.

Yet, in many business situations, the "game" may be adversarial, but need not be. Consider two firms competing in price for extremely price sensitive consumers. At first blush, the game seems purely adversarial---for one firm to win the business, the other must lose. And in the short run this is indeed the case. But taking a longer view, one sees possibilities for cooperation. It is in neither firms' interest to enter into a price war, Both desire to avoid such an outcome and are willing to sacrifice some to achieve this, providing the sacrifices are perceived as fair and that they are not too costly compared to pursuing purely selfish ends. Gae Theory is about these situations as well---how can firms bridge their differences and live in harmony.

Apart from empathy, the key learning from the first class was:

Look forward, reason back
This principle underpins much of Game Theory. Roughly, it means that, in making a plan, one needs to look ahead to the likely outcomes, accounting for the reaction of the other players in the game. In chess, it is not enough to simply spot a winning position several moves hence. One also needs to account for the likely reaction of the other player. If she can block the intended advance, generate a more potent threat, or simply capture the piece required for complete the position, then the original plan becomes worthless. Only plans that anticipate reactions are worthwhile.

But how can one anticipate future reactions? The course is Game Theory and not Mind Reading. This is all true, but here we benefit from the fundamental principle of choice imposed by economics:

Individuals will choose whatever action they perceive to be in their best interests.
In a way, this hardly needs to be enunciated as a principle for it seems self-evident. Yet it is important to dispose of some myths about this premise. First, the principle does not say that individuals will pursue whatever action nets them the most money. If such an action is immoral, unethical, or perhaps harms others, an individual may well elect not to pursue it, yet still abide by the tenet above. The tenet takes no position as to the factors that might reasonably constitute a person's "interest." Money might be part of this, but so too might be status, conformity to some set of behaviors deemed to be moral by the individual in question. It might account for the interests of others---family, friends, pets, or even strangers.

Look forward reason back, together with the behavioral tenet suggest that if we can understand what others consider "their interests," we can anticipate their choices and thereby conform our own choices to meet a particular end.  We need not be mind readers, but we do need to be empathetic to anticipate others' choices.

Our hiring game highlighted some of the differences between inside and outside thinking. Inside thinking suggests that, if we see a potential employee with a dazzling resume but a middling interview, we interpret these past successes as signals of quality and therefore ignore our own perceptions. Or, we might go the other way and simply discount past successes altogether, as Somit apparently did during our game.

Outside thinking, however, enables us to interpret the signal content of past choices, to interpret how much weight to give to the resume versus the interview. Early on in the process, hiring choices indeed represent a signal about the interview. With a stream of successes, however, hiring decisions convey no information, since the same choice, hire, is made regardless of the interview result. Awareness of this possibility, what is termed an information cascade, can only be had through empathy--understanding why others made the choices they did.

The rest of the course explores these ideas in two ways. First, given the rules of the game, how can we learn to empathize and thereby predict the actions of others. Second, how can we change the rules of the game to produce a better outcome for ourselves and, possibly, others we care about.

No comments: