Figure 2 | Scientific Reports

Figure 2

From: Humans depart from optimal computational models of interactive decision-making during competition under partial information

Figure 2

Task sequence of the single-agent version of the tiger task (TT). At the beginning of each decision step, the participant saw two doors and made an action choice (CHOICE) from a set of three actions (Listen (L), Open-left (OL), Open-right (OR)). Participants had 5 seconds to make their choice, else a default listen action was chosen for them. A chosen listen action led to a probabilistic observation (PHYSICAL-OBSERVATION, \(70\%\) correct), and the sequence repeated. The probabilistic hints were signified by semi-transparent images of the tiger and the gold-pot on the respective doors. There was always an inter-event-interval (IEI) of \(\sim 1600\) milliseconds. Choosing left/right open action opened the door (OUTCOME), where the participants either received the gold-pot or encountered the tiger (tiger encounter is shown here). Opening the door reset the tiger and the reward to begin the task sequence again. The matrix in the red box shows the potential payouts the participant could expect upon the different action choices (constantly displayed through the TT).

Back to article page