Résumé
This paper uses a CGT TU game modified into a coordination experiment to explore the causal effect of context on players’ behaviour. An analytical framework focusing on four attributes representative of the game’s context is proposed and an experimental protocol based on this framework allows testing hypotheses regarding the influence of context on players’ choices. Results show that attributes such as Repetition and Communication seem to have a higher influence than Illustration on players’ behaviour. The peculiar nature of the experimental results in the control group, showing the emergence of a focal point other than the outcome prescribed by the theory, allows discussing the expected “noise” observed in the treatments from a new perspective.