The interesting bit of simulating an economy for a game is not the tradeoff of realism vs computer resources, but realism vs fun. A game where AI actors are ma

The interesting bit of simulating an economy for a game is not the tradeoff of realism vs computer resources, but realism vs fun. A game where AI actors are making prices come very close to costs, as they make the most profitable trades, and demand makes actors switch to alternative goods whenever their preference is priced too high, will turn into a spreadsheet. Not unlike in real life, the work will be just identifying where you have left actual inefficiencies, and end up with an algorithm that exploits it. Not so much fun. Most games that involve trading leave giant holes in the economic model on purpose: The parts of the economy that they copy are the parts that make the game more engaging. It's not that in, say, Railroad Tycoon 2 you couldn't build a more realistic economic model: It's just that the more realistic model would be less fun to play with. Same as in boardgames. Many people love playing Power Grid, and like its market mechanic, but it's 100% attached to a very unrealistic turn order for purchases, which attempts to balance out who is receiving more income. It builds a tradeoff between being the leader or falling behind, which would in no way occur in a more realistic model. Then the winners would just keep winning no matter what, and some people might as well accept defeat in the first turn of the game. This is not just speculation: People in the game industry have playtested games with a stronger economic model all the time, and the more realistic the model, the harder it is to make the game fun. So the real question is, what parts of an economic model are necessary to make the game realistic enough to be engaging, and which parts harm your game?