>>I do wanna know how the heck you write a computer model to simulate the behavior of a deluded killer.
The same way you write a climate change model: make stuff up, create assumptions, write the rules of the model to make the facts follow your assumptions, tweak as necessary to renew your government grants.
In what you describe, there is no problem with the computer modeling. It's the making up of data, the use of unproven and unprovable assumptions, and modifications after the fact to achieve a particular outcome that betray climate changists as charlatans.
Scientific computer modeling works and works well in many applications.