Loading models...
sweep
-
N 0
P(capture) -
Presets
Parameters
Live readouts
Live graphs
Domain of danger of each animal, as a multiple of the average risk 1/N (sorted). Interior animals have zero exposure; the edge carries it all.
Current hunt: distance predator to target (m, teal), predator speed (m/s, orange); vertical ticks mark target switches (confusion).
Experiments
A. Group-size sweep. Repeated simulated hunts for N = 1 ... 64 under three conditions: dilution only (no confusion, no vigilance), plus confusion, plus vigilance (many eyes). Per-capita risk = P(capture per hunt) / N. With dilution alone P does not depend on N, so the log-log slope is -1.
B. Does a selfish animal lower its own risk? Take a herd, pick its most exposed animal, let only that animal follow a movement rule for 20 s while the rest drift, and compare its exposure with the start. The "none" row is the control (selecting the most exposed animal alone makes the ratio fall).
C. Checks against exact results.