Related Experiment Videos
Game theory and reciprocity in some extensive form experimental games
K A McCabe1, S J Rassenti, V L Smith
1Economic Science Laboratory, University of Arizona, Tucson 85721-0108, USA.
Summary
This study on decision-making in games found significant cooperation, even in single-play scenarios. Contrary to game theory, complete information fostered cooperative behavior, challenging established predictions.
Area of Science:
- Behavioral economics
- Game theory
- Decision science
Background:
- Understanding cooperative and noncooperative behavior is crucial in economic and social interactions.
- Game theory predicts noncooperative outcomes in many scenarios, especially with complete information.
Purpose of the Study:
- To identify replicable principles governing cooperative versus noncooperative behavior in two-person extensive form games.
- To investigate the influence of signaling, reciprocity, and backward induction strategies on decision-making.
Main Methods:
- Utilized nine experimental treatments varying matching protocol, payoffs, and payoff information.
- Examined decision-making processes in two-person extensive form game trees.
- Analyzed the impact of strategy availability (punishing strategies) and interaction probability (repeated play).
Main Results:
- Observed substantial support for cooperation under complete information, even in single-play treatments.
- Findings contrast with traditional game theory predictions which often favor noncooperative outcomes.
- Cooperation levels were influenced by factors like signaling, reciprocity, and the probability of repeated interactions.
Conclusions:
- Complete information can lead to significant cooperation, challenging standard game theory assumptions.
- Behavioral factors like reciprocity and signaling play a key role in cooperative decision-making.
- The study provides insights into the conditions that foster cooperation in strategic interactions.