Game Theory’s Application in Behavioral Ecology
Game theory and physics [1] – https://aapt.scitation.org/doi/abs/10.1119/1.1848514
This research paper, by Christoph Hauert of the University of British Columbia, explains the application of evolutionary game theory to reveal new insights into the discipline of behavioral ecology. In fact, the abstract states “we employ the prisoner’s dilemma to discuss new insights gained in behavioral ecology using methods from physics,” something quite discussed in class [1]. It goes on to describe the evolution of cooperation. Specifically how evolution through working together goes against the principles of Darwinism. This leads to the tie with the prisoner’s dilemma, one of the most famous examples of game theory specifically about cooperation. Game theory has allowed evolutionary scientists to do a similar process. Here, they consider an infinite population of a certain species and assume some random amount is the number of individuals that are willing to cooperate and the others who are not willing to. Then, they calculate the payoffs to determine what the strategies of the subset populations lead to. This helps lead to conclusions about the population dynamics of a certain species. This paper raises an issue about classic game theory. The author goes on to state that this school of thought leads to a prediction of an “undesired outcome of mutual defection
and economic stalemate, where no one receives any benefits
for the sake of reducing costs” [1]. To work around this problem, Hauert states that instead of considering an infinite population, instead, we consider a structured population that is confined to a given zone. This limits interactions among individuals and leads to those individuals that cooperate to form groups and can succeed, even if there is a subset of the population that refuses to cooperate. The article continues to dive into a very detailed analysis of game theory and its relation to condensed matter physics, which is quite interesting yet confounding.
The relation to the work we have done in class is specifically about game theory and networks. We have learned a great deal about game theory including identifying equilibriums and determining dominant strategies. However, something I found interesting was the author’s conclusion that classical game theory, which I believe we work with in class, often leads to undesired outcomes instead of the max benefit for both parties. On further inspection, it does make sense for Hauert to claim this because he is dealing with an infinite number of individuals in the population and splitting them with two opposing strategies. When he changed his analysis to use a more structured population, this led to a conclusion that we are more used to.
Additionally, the author, Hauert, connects to the topic of networks, something we have intensively discussed in this class. In the example of the cooperative and noncooperative species, he discussed the formation of a network among those cooperative individuals while the noncooperative individuals remained isolated and scattered. Hauert also talks about how small-world networks can be easily generated using a simulator. These networks are a “natural combination of high local connectedness and a few long-range connections that result in short average path lengths between any two nodes” [1]. This is quite related to the theory of the six degrees of separation, another topic relevant to the discussion of our class.
Overall, this article provides some great insights on studying game theory using physics and simulations and is quite relevant to the topics we have learned such as Nash Equilibriums and constructing networks.