Figure 7.

Competition with a highly risk-averse opponent in Experiment 4. (A) Risk-sensitivity value is plotted for the baseline and for each block of the competitive task. A positive value indicates a risk-seeking strategy. *denotes p < 0.05 from the risk-neutral value (i.e., 0). The sub-optimal risk-seeking strategy at the baseline improved during the competition. The shaded gray area denotes the last 50 trials of the competitive task. (B) Change of aim point (mean endpoint) when competing against the highly risk-averse opponent. The opponent aim point was set further away from the boundary line. We replicated the results in the risk-averse group (Fig. 3B). The subjects’ aim point decreased from the baseline and retained during the competition. Data is averaged across the subjects, and the shaded area denotes the standard error of the mean. (C) Aim point modulation as a non-linear function of the opponents’ relative aim point. The subjects’ relative aim point (mean endpoint in the competitive task A_c - mean endpoint in the individual task A_i) is plotted against the opponents’ relative aim point (mean endpoint of the opponent A_o − A_i). Sixty-six data points (11[2nd–12th] blocks × 6 subjects) are plotted.