Deliberation Measure

The deliberation measure is designed to determine how much time an agent spends `deliberating.' It works by separating the actions into two partitions, active and passive. Passive actions are taken as being those that wouldn't take any thinking. The `do-nothing' action is the most obvious example of a passive action.

This measure determines the average, median and quartile lengths of a string of passive actions for each agent, then averages these values. Many theories of emotion cite speedy processing as a function of emotion, and that applies here. The idea of the measure is that the time spent performing passive acts is interpreted as time spent in deliberation. If the agents evolve to spend less time thinking before performing actions when necessary, this is analogous to at least some of the functions of emotion.

For example, say an agent performed a string of actions PAAPPPAPA, where P stands for a passive action and A stands for an active one. The agent has two strings of passive actions of length one and one string of length three. Its deliberation value would be $\frac{1\times{}2+3\times{}1}{3}=2.66$.