NetLogo(Wilensky)

I’m not sure what to make of this; it is the NetLogo modelling tool

from NorthWestern University, and I am running the Party example.

I get to set the number of people, the number of groups, and the tolerance of individuals

of each sex for the presence of the other sex in their group. If that tolerance

level is exceeded on any turn, the individuals affected change groups on the next turn.

After a time, same sex groups may form. 

 

One very striking aspect of the exercise is that - for all initial descriptions I tested - the
40% model quickly jammed in a same-sex groups pattern. Because the instructions are to
move when tolerance is exceeded, we know that the simulation had ended.
What if there had been a portion of individuals seeking out the opposite sex at the
same time; that would have been an entirely different story.

This points to an obvious feature of modelling: the computer
is doing tedious calculations in a chiffy, but no more. It is the conditions that rule.

So the first aspect to look at in dealing with a model is being clear on the conditions.
In a complex model, the length of a run can be primordial: who knows when something
minor can kick in. Nothing less than the beginning of the universe is at issue here. Pretty
awesome...