Christian Hilbe

For further information on the project please refer to or contact the project's co-supervisor Nikoleta E. Glynatsi:

Modeling the coevolution of reciprocity and network structure

Repeated games have been the centerpiece when studying the evolution of human cooperation. They describe situations where two individuals interact repeatedly, and have the ability to condition their actions depending on the outcomes of previous interactions. Individuals can then reciprocate good and bad actions.Theoretical and empirical work has shown that direct reciprocity can lead to high levels of cooperation.

Another form of reciprocity is network reciprocity. In network reciprocity an individual's interactions are limited. An individual can only interact with others that they have ties with.

The aim of this project is to consider both direct and network reciprocity, and thus explore the emergence of cooperation on networks with direct reciprocity. More specifically, the aim is to develop models where individuals are allowed to choose both their reactions to bad and good actions, and can choose with which other individuals they have ties with. These models will be explored using a combination of mathematical approaches (from graph theory, stochastic processes, and linear algebra), and computer simulations.

To accompany the theoretical results on networks with direct reciprocity, the project will explore cooperation using real world data. The aim is to use an open source package which was developed for [1] in order to collect articles' meta data and assert the collaborative behaviour of authors.

Candidates should have a genuine interest in human behaviour, evolution and software development, a good mathematical background, and ideally computing skills.

[1] Glynatsi, N. E.; Knight, V. A.: A bibliometric study of research topics, collaboration and centrality in the Iterated Prisoner's Dilemma. Palgrave Communications

Go to Editor View