A price theory is at the core of economics. The question of 'what is produced, and for whom' is decided, under capitalism, by prices. The signals prices send to producers and consumers coordinate economic activity. You can pick Marx, who argued that producers set prices based on the cost of variable (current labor) and fixed (past labor) costs, you can pick Walras, where a hypothetical auctioneer unites people at agreeable prices for both producers and consumers, or some flavor of one of these basic types, or someone totally different like Sraffa. In all cases your goal is (or should be) not just to fit reality in the sense of fitting a curve to data but fitting the psychology of the people involved in the transaction, and the real process itself. That's why Marx's theory makes much more sense to me- he says that businessmen set prices based on their costs, which are determined by labor, and attempt to capture a greater share of the social surplus via the medium of money prices. Mobility of capital leads to a uniform rate of profit, changes in the organic composition of capital (more machine, less human) lead to a decline in the rate of profit because the social surplus that is available to be appropriated is gradually less and less because only current labor can produce real value, etc. etc. down to the final crisis.
Additionally, Marx is more useful for video games because video games, unlike academic papers or data summaries, are all about the process. If the process of an economy growing and producing things feels weird even if the end results are similar to what we observe in the real world (as with some of the games we're going to talk about below) then the theory has no explanatory power whatsoever, and this is obvious. It is less obvious in a paper or book where a theory that has little relationship to reality appears to have the same conclusion as reality: we sort of gloss over the process and skip to the results, and if those are justified we ignore the 'sausage making' that led to them. In a video game this bad science is put to the test. The results are almost meaningless- we're spending all of our time in the middle of things, in the real process that produces those results. And we spend a lot of time thinking about that process and picking it apart. If it has holes they are ruthlessly exploited. That's how video games can serve as a crucible of economic theory in general and price theory in particular. Let's look at a few video games and how they deal with price, and what the consequences are for the mechanics of the game.
Video games that deal with economics even in a grazing way have to make a similar choice. In the first Victoria there was a fixed list of prices: however many canned goods were being produced, they would all be sold to the computer and tossed into the sea (deleted from memory) at a fixed price which would change over time according to a table. Machine parts in 1836 are absurdly expensive and if you're lucky enough to be in a position to build a machine parts plant (lucky enough to be Britain, basically) your national treasury will go buckwild. But towards the end of the century, the price declines, coincidental with, but not because of, the fact that more countries can produce machine parts. Even if only one machine parts factory were ever built, the price would follow the same path.
Age of Empires 2 has a market where the player can buy scarce resources from the computer- they are added to his stockpile out of thin air. And the price changes over time depending on how the player uses the market. If you need a lot of stone in the Castle age, the price of stone will go up every time you buy some of it (all prices are in gold, by the way, and gold can be mined or gained in trade with other players, by sending carts to them) and will go down every time you sell. Prices are more than 'sticky' - they are completely permanent absent direct intervention by the player. And every player has his own set of prices for stone- there's no opportunity for arbitrage, unless during multiplayer you were to coordinate a sale of stone or something else using the 'tribute' function. But that's not what the tribute function was intended for and I've never seen it used that way, even in competitive AOE2. And of course the labor units in AOE games are slaves and require zero upkeep, so the 'value' of gold or stone can be expressed in the seconds it takes to mine or gather it, because the cost of the laborer's upkeep is nominal (a flat one-time 50 food payment). Additionally, there is a fixed amount of stone, wood and gold on the map, and only food can be regenerated indefinitely (as long as you can get wood to build farms over and over again). So after everything's been mined and chopped, an entirely different economy comes into force, with food as the basic commodity that determines the production of all the others. I've never played a game long enough to investigate this, I think, scientifically, it's impossible to get this far because you would pass out from boredom before you got there.
The second Victoria added global supply and demand to the game. Now, the price of canned goods and machine parts would be determined by the supply for those goods and the demand for them in the game. This requires minute tweaking of 'needs,' however. If the capacity of an individual plant to produce cement and the global demand for cement are out of wack (as they almost always are) then your cement will go unsold and the factory will go bankrupt. It's also interesting to note that there was no way to scale down production except in a drastic fashion: as long as a factory is open it will produce as much as it possibly can, and the factory owner (if it's the state, or an individual capitalist) has no power whatsoever to set their own price. They don't even submit possible prices to a theoretical Walrasian auctioneer: they get the price that the computer figures out a priori.
It's interesting that all of these games, even the ones that have a more 'free-market' veneer, rely on the autonomous and unaccountable actions of a computer. Agents (even simulated ones, like AI) have no effect whatsoever on their ability to run the economy. This actually mirrors the Walrasian auctioneer, who functions like a computer in matching mutual price desires automatically, with no input beyond an initial list from actual producers and consumers. But nobody has tried to give producers and consumers autonomy and the freedom to set their own prices for commodities, like Marx (and the real world) does.
It's also interesting to note that only one of these games, the Victoria series, deals with debt, but in a really bizarre way: debt is automatically public regardless of the economic system in place (laissez-faire capitalism, state capitalism, communism), and is not connected to the factories or RGO (farm, pasture, mine, etc). Capitalists save for factories and plop down their money in big chunks, but never go bankrupt, and never take out loans: they can only run out of money to fulfill their commodity needs and then demote to other types of pops, none of which are able to save towards the construction of factories like capitalists are. Your national bank can make loans to other countries but these (and the loans you take yourself, if your country is not state-capitalist) can only be used to pay for schools, the army, etc. When you build factories in foreign countries you do it out of state coffers, as well, or individual capitalists pool their funds and do it themselves once they've saved up enough.
Victoria's incoherent finance and price theory, and the absurdity of most of the rest of video games' engagements with economics, are a product of not doing it the Right, meaning the Marxian, way. Prices in the real world are not computed ahead of time according to a formula, however complex, they are set by people with their own motivations, chief among which is to stay in business and make a profit. Prices have a real basis, not in the completely abstract and useless raw quantity of something demanded but the price of the things that went into that something. An economy where prices really were determined by supply and demand alone would look something like an economy in Victoria II- very weird, not at all like any sort of capitalism any of us would recognize. And an economy based on the one from the Age of Empires series would be a complete hellscape, like I Have No Mouth And I Must Scream except with the player in the place of the malevolent computer and the last remaining tortured humans as the farmers endlessly stabbing the dirt with tiny shovels, their magically appearing product being turned into ever smaller amounts of gold, ever tinier amounts of wood and stone, until building a single barracks takes five years and the construction of a castle would take millenia. A video game that tries to match the complexity of real situations and the decisions that real people face has a better chance of not turning out this way, and being more fun and more engrossing. And maybe even reflecting on poor science in the real world.