Modeling Genocide at the System and Agent Levels
require a more advanced model in order to produce a meaningful validation .
Experiment 3 — Uniform Global Threshold to Act
In this experiment , all settings of the Base Model are identical to those listed in Experiment 1 , except for Threshold to Act . Rather than every agent having this attribute randomly set to a floating-point number between 0 and 1 according to a random uniform distribution , it is given a fixed level that is the same for all agents from both groups . The Threshold to Act remains static throughout the simulation .
This experiment steps through Threshold to Act levels from . 10 to . 75 in increments of . 05 , running five replicates for each setting . To avoid simulations running indefinitely , the code is adjusted to stop the simulation after 7,000 ticks , as this is generally the timeframe beyond which the model never produces genocide . Between Threshold to Act levels . 45 and . 50 , genocide does not occur beyond a cutoff .
Figure 5 below shows the general pattern of simulation run length as a function of Threshold to Act .
Figure 5 . Simulation length versus threshold-level .
42