Warmachine. Which is interesting, because I called my high school project The War Machine. It earnt me a perfect grade at the time; I still take pride in that project of mine.
Written in Pascal, The War Machine (MY War Machine) was a computer turn based strategy game. Its first [independent] module was a map designer: users were able to design a rectangular map of any sensible size they can think of and populate it with different types of terrain. They were then able to place army units from two sides on that map.
The second module was the game itself: the single player picked a previously created map, chose a side, and then started issuing commands to their army unit (like move, attack, retreat or deploy). There was some complexity included into the engine, such as managing supplies and logistics. The main event, however, was the artificial intelligence that allowed the computer to assume the opponent's side and fight the player. Pretty effectively so, if I might add.
Looking back at The War Machine, I find it amazing how little research I had done for the project. Not for lack of trying; it's just that it was very hard to acquire information during those pre-internet times. Things we take for granted nowadays, like online searches or Wikipedia researches, simply weren't there. I can see the differences these make with my pre-teen son's school projects. Seriously, I do more research in order to decide what to have for dinner than I did for that school project of mine!
That research relied pretty much on three sources. The first was Dungeons & Dragons rules for determining the results of army to army fights. You may recall D&D is a game focusing on individual characters; those rules I'm referring to were there to answer the need for determining what would happen if armies collided during those individual characters' campaigns.
That D&D set of rules was called The War Machine. I politely borrowed the name for my project. It was those rules that pretty much governed the AI (Artificial Intelligence) of the computer side.
That AI was built around ideas borrowed from another book I purchased years earlier, called Artificial Intelligence and the Dragon Computer. [For those of you who don't know, probably most of you, the Dragon was a personal computer from that very first round of personal computers for the home.] Essentially, the computer player calculated The War Machine rules' scores of each option available to its army units and went with the choices that got it the maximum score (read: effectiveness).
My third source of inspiration was a strategy board game called Ogre that my uncle bought me* a few years earlier. It was a two player turn based game where one player controlled a single very powerful but slow tank, the Ogre, as it fought to cross the battlefield. Opposing that powerful Ogre was the second player, armed with numerous faster but not as capable army units. At the time, Ogre was analysed to the death in various forums; books were written about it, and I read some of those. Essentially, the experts concluded that the Ogre should almost always win (barring severe lack of luck with the dice) with one single exception: if the opponent deployed only hovercraft a to fight the Ogre and used them in a way that saw them attacking and then retreating away each round (so as to avoid the counterattack).
Ogre acted as a test case for my game. I designed a map that was, essentially, a copy of that board game's map, and deployed units across it as per the original game. I then played it out against the computer and exchanged roles between Ogre and opponent. Lo and behold, my simulation behaved exactly as that Ogre published analysis said it would!
I was a proud boy.
Image copyright © 1984 Keith and Steven Brain, used under the assumption of fair use