A more piecemeal simulation released by Maxis back in 1991. Possibly some inspirations (red ants vs black ants, food represented in green circles/spheres):
Simant lost a lot of its appeal once I worked out I could enter a new zone and queen-rush the enemy with a 50/50 chance of winning, and just send in a new queen if I failed.
For me all the various SimSomething games were more about figuring the rules, limits and possible abuses of the simulation design than the intended actual game.
That "revenge is mine" feeling is one of my favorite parts of RPGs.
"What's the matter, horde of orcs? You beat me twelve times when I was level four, why are you scared of me now? Could it be fireball? Oh I see it could. And what is this little stick I'm carrying? Still a nonmagical quarterstaff? Oh, wait, dropped that when I picked up the Wand of Lightning Bolt."
Pick up Thomas Schelling's Micro Motives and Macro Behavior; he originated many of these models (the most mind-blowing to me was that a small segregation preference among individuals balloons into way more segregation than was desired just because of physical constraints inherent in the geometry of neighborhoods).
I did a course on modelling complex systems for my honours studies; this was one of the recommended readings. I'm glad I bought it.
Most books about systems modelling refer to one of the commercial modelling programs. Or they are quite generic and hand-wavy. The nice thing about TTATJ is that the software (NetLogo) is downloadable[1]. The book is on the Kindle too, which is convenient.
This reminds me of Rich Hickey's Clojure Concurrency talk[1], where he uses a simulation of an ant colony to explain Clojure's concurrency features. Hickey's version[2] is a lot simpler, but remains an interesting starting point if you want to hack on something similar.
I have a really beautiful simulation of ant foraging I coded up in Processing a while back. I'm currently porting all of this kind of work I have done to Python (via Pyglet) and will have to remember to post a link to it here when it's done. (See my profile info for linke to a very small sample.)
* Edit: "A very small sample" of the kind of thing I've been doing. No samples of the ant simulation to share just yet...
http://en.wikipedia.org/wiki/SimAnt