Game Theory and Nuclear War
Game theory, and especially the prisoner’s dilemma, seems to manifest in the current state of relations between North Korea and the United States. What I found particularly interesting about the articles related to this situation is that there is much analysis of game theory methods and how effective (or ineffective) they really are at modeling the situation. It seems clear that the typical numerical loss/gain that is commonly used to introduce the prisoner’s dilemma would not be ideal to describe the complex web of potential consequences that any one action might have.
The basic prisoner’s dilemma representation would describe the result of actions by North Korea and the U.S. as a numerical tuple. But perhaps a better model for this result could be a network – a graph where countries are represented by nodes, and the edges are the relationships between them. This way, we need not argue whether North Korea going on the offensive is positive or negative for them in the short term – we can look at how we expect that move to affect their relationships going forward.
As an example, assume that we are examining the outcome of the U.S. invading North Korea before they can respond. In the classic prisoner’s dilemma representation we might describe this as a net gain for the U.S. and a net loss for North Korea. But is that really an accurate model of the real world? Our graph representation of the situation, though not flawless by any means, might be a better model to look at. China has publicly declared that if the U.S. strikes first, they will intervene. So perhaps we can note this factor in our representation by making the value on the edge between China and the U.S. extremely poor.
By using this model we can grasp a more holistic view of the consequences of any action. If we are prioritizing relations with China, we may not want to make the previously described move. On the other hand, if we are much more concerned with the potential threat from North Korea, it could be a feasible option. There is a lot of room in this approach to add additional statistics on the edges to describe more intricacies in the relationships between countries, and also to develop an optimization function that could be used to calculate the best course of action, depending on what relationships we are most interested in changing.
Sources:
https://www.washingtonpost.com/news/wonk/wp/2017/08/16/what-game-theory-tells-us-about-nuclear-war-with-north-korea/
https://www.forbes.com/sites/kotlikoff/2017/08/11/using-game-theory-to-handle-north-korea/#45712453603e