Game Theory and Python: An educational tutorial to

game theory and repeated games using Python

Nikoleta E. Glynatsi1,2 and Vincent A. Knight1
1 Department of Mathematics, Cardiff University, Senghennydd Rd, Cardiff CF24 4AG 2 Max
Planck Institute for Evolutionary Biology, August-Thienemann-Straße 2, 24306 Plön

DOI: 10.21105/jose.00078
Software
• Review
Summary
• Repository
• Archive
These materials are an open-source educational tutorial aimed at introducing participants
Submitted: 05 December 2019 to game theory and more specifically to repeated games. The tutorial uses the open-
Published: 24 May 2021 source library called Axelrod-Python and Jupyter Notebooks making this an open-source,
License reproducible and interactive tutorial.
Authors of papers retain copy-
right and release the work under
a Creative Commons Attribution Statement of Need
4.0 International License (CC BY
4.0).
We are mathematicians and Python programmers with interests in game theory and peda-
gogy. This tutorial was created as a resource for computer workshops and for introducing
mathematicians to programming. As a result, this tutorial is aimed at two groups of
individuals:

• those familiar with Python (programmers) who want to start to learn game theory,
• mathematicians with little or no programming knowledge as a pathway to program-
ming through the interesting subject that is game theory

Game theory is a field of applied mathematics interested in strategic interactions. Game

theory itself has a number of subfields and the one considered in this tutorial is the area of
repeated games. Repeated games have been the subject of research (Axelrod & Hamilton,
1981; Glynatsi & Knight, 2021; Nowak, 2006) but have also been used extensively as an
entry to the subject for students at undergraduate level courses (Brokaw & Merz, 2004;
Knight, 2015). This tutorial focuses on the Iterated Prisoner’s Dilemma. The advantage
of studying the Iterated Prisoner’s Dilemma is that it models situations in which self-
interest clashes with collective interest, thus it provides a framework for illustrating the
usage of mathematics in real-life decision making.

Content

These materials consist of four tutorials that focus on the Iterated Prisoner’s Dilemma,
the notion of strategies and computer tournaments whilst using Axelrod-Python (The
Axelrod project developers, 2016).
In 1980, a political scientist called Robert Axelrod ran a computer tournament of the
Iterated Prisoner’s Dilemma where strategies written in computer code would repeatedly

Glynatsi et al., (2021). Game Theory and Python: An educational tutorial to game theory and repeated games using Python. Journal of Open 1
Source Education, 4(39), 78. https://doi.org/10.21105/jose.00078
 choose between self and collective interest. The strategies would decide on their next
action using the history of previous interactions and the winner was decided based on
the average score. The open-source package Axelrod-Python was created in order to
reproduce Axelrod’s original work but to also serve as an educational and research tool.
Following the tutorial, participants will be able to reproduce Axelrod’s tournament but
also create their own unique tournaments from a selection of more than 200 strategies.
Furthermore, participants will be able to progress their comprehension of the topic by pro-
ducing strategies and implement them in Python. The new strategies can then be placed
in different tournaments in order to access their success and weakness. This progression
of the tutorial fits within a constructive framework of learning (Jordan et al., 2008) and,
by the nature of participants actively writing the required software, an active learning
setting (Freeman et al., 2014). Participants with knowledge of Python will benefit from
an understanding of game theory and participants with no knowledge of Python will also
be introduced to basic programming concepts through the tangible ideas of the Prisoner’s
Dilemma.

Usage

The tutorial Game Theory and Python can be used in a workshop environment or through
independent learning.

• Workshop: The material has been designed for a 2 hour workshop. A suggested
timetable is:
– 0:00 - 0:15 Installation (guidelines are given in the README)
– 0:15 - 0:30 An introduction to game theory and the Iterated Prisoner’s Dilemma
– 0:30 - 0:55 Creating matches and tournaments using Axelrod-Python
– 0:55 - 1:20 Writing strategies and contributing to Axelrod-Python
– 1:20 - 1:50 Playing against strategies of the Iterated Prisoner’s Dilemma
– 1:50 - 2:00 Closing remarks and wrapping up
In a workshop environment we suggest that the instructor has familiarized them-
selves with the written parts of the tutorial beforehand. The written parts of
each notebook include an introduction to the topic being covered. It is advised
that the instructor discusses the introduction to the topic followed by them typing
out/running the material while the participants follow using their own machines.
The instructor should encourage the participants to try the exercises of each note-
book alone or with other participants. Before moving to the next notebook the
instructor should encourage a discussion amongst everyone regarding the results of
the exercises each had and their interpretation.
• Independent Learning: An independent learner should aim to spend the same
amount of time on the material, approximately 2 hours. A suggested timetable is:
– 0:00 - 0:15 Installation (guidelines are given in the README)
– 0:15 - 0:30 An introduction to game theory and the Iterated Prisoner’s Dilemma
– 0:30 - 1:00 Creating matches and tournaments using Axelrod-Python
– 1:00 - 1:30 Writing strategies and contributing to Axelrod-Python
– 1:30 - 2:00 Playing against strategies of the Iterated Prisoner’s Dilemma
If the tutorial is being followed by an individual learner, we suggest that the learner
reads the written parts of each notebook followed by running the tutorial and com-
pleting the exercises. The individual should take some time to reflect on the results
of each notebook and their interpretation.

Glynatsi et al., (2021). Game Theory and Python: An educational tutorial to game theory and repeated games using Python. Journal of Open 2
Source Education, 4(39), 78. https://doi.org/10.21105/jose.00078
 These instructions can also be found in the README. Though we recommend that
the material can be covered in 2 hours, in both a workshop environment and through
independent learning, some audience/learners might need more time to complete all the
exercises. We recommend any facilitator/learner to adjust as necessary.

Recent Uses

This tutorial was originally formulated to run a game theory workshop at PyCon Namibia
2017. It has attracted the attention of the community ever since and has been used in
other events.

Acknowledgements

The tutorial has been made possible due to the existence of the open-source package
Axelrod-Python. We would like to express our appreciation to the maintainers of the
package as well as anyone that has ever contributed to it.

References

Axelrod, R., & Hamilton, W. D. (1981). The evolution of cooperation. Science, 211(4489),
1390–1396. https://doi.org/10.1126/science.7466396
Brokaw, A. J., & Merz, T. E. (2004). Active learning with monty hall in a game theory
class. The Journal of Economic Education, 35(3), 259–268. https://doi.org/10.3200/
JECE.35.3.259-268
Freeman, S., Eddy, S. L., McDonough, M., Smith, M. K., Okoroafor, N., Jordt, H., &
Wenderoth, M. P. (2014). Active learning increases student performance in science,
engineering, and mathematics. Proceedings of the National Academy of Sciences,
111(23), 8410–8415. https://doi.org/10.1073/pnas.1319030111
Glynatsi, N. E., & Knight, V. A. (2021). A bibliometric study of research topics, col-
laboration, and centrality in the iterated prisoner’s dilemma. Humanities and Social
Sciences Communications, 8(1), 1–12. https://doi.org/10.1057/s41599-021-00718-9
Jordan, A., Carlile, O., & Stack, A. (2008). Approaches to learning: A guide for teachers:
A guide for educators. McGraw-Hill Education (UK).
Knight, V. (2015). Playing games: A case study in active learning applied to game theory.
MSOR Connections, 14(1), 28–38. https://doi.org/10.21100/msor.v14i1.254
Nowak, M. A. (2006). Five rules for the evolution of cooperation. Science, 314(5805),
1560–1563. https://doi.org/10.1126/science.1133755
The Axelrod project developers. (2016). Axelrod: Version 4.3.0. https://doi.org/10.
5281/zenodo.1405868

Glynatsi et al., (2021). Game Theory and Python: An educational tutorial to game theory and repeated games using Python. Journal of Open 3
Source Education, 4(39), 78. https://doi.org/10.21105/jose.00078