SELLOUT is a real-time 2 player game, designed around the concept and dynamics of the Prisoner's Dilemma. Whether the players will cooperate with each other or sell the other one out is the core mechanic behind each round.
In an iterated fashion the game is played multiple rounds, allowing participants to speculate on how their competitor is thinking. It intends to bring out strategies surrounding decision making in non-cooperative, non-zero-sum games.
Taking Robert Axelrod and his Iterated Prisoner's Dilemma tournaments from the 1980s as a reference; the installation aims to provoke questions on how self interest, human cooperation and trust are intertwined.
ExhibitionsOct 2019 Kunstuniversität BestOff 19 (BestOFF 100%) splace am Hauptplatz, Linz AT, October 8-22Sep 2019 Ars Electronica Festival 2019 (IC @ Ars Electronica) Post City, Linz AT, September 5-9Background
Invented in 1950 by Merrill M. Flood and Melvin Dresher, Prisoner's Dilemma (PD) is an example formalized and studied in game theory, analyzing strategies surrounding individual decision making in a non-cooperative, non-zero-sum game. It has been fundamental to theories of human cooperation and trust.
In the example, 2 prisoners who were caught by authorities are interrogated in separate rooms with no communication between each other. They are both given 2 options: each prisoner can either stay silent to cooperate with the other one, or they can sellout to betray them.
As written down by mathematician Albert W. Tucker, the offer is:
- If both betray each other, they will both serve the designated sertence.
- If A betrays B, but B stays silent; A will be set free, while B gets a heavier sentence (and vice versa).
- If both of them remain silent, hence cooperate, they will both get a lesser sentence than usual.
In SELLOUT the iterated variant of PD is implemented, also referred to as the "Peace-War Game". The game is played multiple times in a row, in order to allow the participants to experience different psychological factors coming into play based on previous rounds. Many academics have been interested in observing outcomes from the Iterated Prisoner's Dilemma (IPD) over the years, inculding political scientist Robert Axelrod (The Evolution of Cooperation, 1984) who also held an IPD tournament in 1980 for computer strategies to compete against each other .
The installation will consist of 2 chairs placed back-to-back, with noise-cancelling headphones (or ear protectors) next to them. The participants will be asked to wear the headphones for isolation from the space and the other participant meanwhile sitting very close to them, only facing backwards.
The game will start once both participants access the local network provided by a Raspberry Pi, then through a website connect to the game server on a handheld device. If they don't have one, a smartphone may be provided to them, depending on the availability and possibility of supervision.
Participants can either cooperate (C) or defect (D) in a round, and the results will be shared with both at the end of every round. After the last round of the game, the final results will be calculated and revealed.
Other than playing against each other, the participants are asked to beat a required minimum score in order for the game to be a "successful" one. Therefore the game is balanced in a way that the participants cannot simply choose beforehand to just cooperate all the time, because the number of rounds do not add up to the minimum score only with the outcome from the cooperation scenario. In the end someone will have to sellout in order to win, and this changes the whole structure for trust and cooperation.
 Axelrod, R., 1987. The evolution of strategies in the iterated prisoner's dilemma. The dynamics of norms, pp.1-16.