But it's not clear to me what the player should be doing because buildings are placed by the NPCs and then they make their own decisions according to self interest. What is the player doing except watching? Setting laws maybe, but it doesn't feel very fun to play yet.
Certainly the player can just be one more player in the economy against/among the AI. This is basically the premise of Rise of Industry. And more actively, less realistically, Offworld Trading Company.
Running one business is sort of the premise of the tycoon genre.
Generally, trouble is fun (Losing is Fun), so you may want to put pressure on the player to make the thing work. Maybe it naturally runs to ruin or is fragile to balance properly. (In Dwarf Fortress, danger follows growth and success.) There could be adversaries. Or the world could be dangerous.
They could be in charge of the macro economy or a town or a business. There could be pickles that the player has to work their way out of. Recessions and Depressions. Think about how towns collapse, businesses fail, people lose jobs. Businesses also cheat, wreck the commons, avoid taxes, break the law, exert monopoly power, form cartels, fight for dominance. There is a lot of trouble here for the player to cause, or to struggle against.
Sometimes I like to watch sim games run to see what will happen 'naturally', or setup AI battles in strategy games. The best toys allow players to enjoy them however they want.
Thus, the primary thing you need to introduce in a sufficiently complex simulation is those points of interaction, to influence the simulation. Essentially god games and dwarf fortress are the points of inspiration — and then you need a reason to optimize the economy, which is either internal (need to eventually produce some expensive thing to win, like factorio’s rocket) or external (e.g. need to continuously improve the defensive infrastructure against ever-stronger foreign attacks).
Another way of looking at it is on levels of operational complexity — whether the player is acting at the level of tactics (a single fight; Starcraft), operation (a single campaign/theater; total war), or strategy (resource control and long-term politics; grand strategies).
From the sounds of it, if your AI manages the cities themselves on their own, then they key interaction point is on the resource dependencies - land acquisition, trade routes and manipulation, etc. The player’s goal is to learn to select resources the AI can best make use of, and acquire them efficiently, to ever-expand the city itself (or perhaps multiple).
Also note there must always be some negative feedback loop, e.g. pollution, to contend with.
Also note that IMO random events like Godzilla attacks and disasters are a bit of a cop-out — it’s just shit that happens, but you can’t do anything about. It’s more a nuisance than anything else. Jurassic Park tycoon was slightly more interesting — dinos always had some chance of escaping, becoming more common as you had a bigger park (negative feedback), and disasters like typhoons just exploded that risk. The typhoon damage wasn’t of note — it was the rampaging t-rex’s it unleashed because of your fragile placement of barriers.
That last part is important, I think, because you can still interact with the disasters such that they weren’t plainly random events. SimCity on the other hand is just a nuisance trigger every now and then that you can’t plan for, mitigate, or really do anything about.
Add some sort of story or goals. Encourage business starting, innovation trying to get more futuristic or get to space or something? But to Achieve these goals, the player has to pass the right laws, manage their economy, ensure sufficient funding & taxes, etc. The core game loop being passing laws and seeing how it impacts behavior, then repeat to achieve a goal
Could be an interesting colony sim or something
A few different ways a player can effect their nation's economy include: unlocking technologies, conquering territory for resources, expanding a nations sphere of influence for resources and markets, electing parties with different economic policies (planned economy vs laissez faire), playing with taxes/tariffs, encouraging certain industries, etc. The economic simulation isn't super deep, but it is deep enough to be engaging and fulfilling to a player who likes these sorts of spreadsheet simulator games.
A friend of mine who’s a strong advocate of this philosophy said that the reason why Keynes is so popular in the real world (fiat currency, national banks, micro/macro economics) is that it gives politicians something to do.
It’s worth thinking about what economic philosophy your game is following or is biased towards (I always felt that the older civilization games were biased toward “research is good” and against democracy as the senate made it too difficult to go to war. Last one could be argued to pro democracy if you try to bring the lesson out of the game and into the real world. )
But zoning, setting taxes, building infrastructure, police and military are things that no one has really managed to leave to the market forces.
I believe sim city added disasters because they felt players didn’t really have anything to do...
That’s a projection of the Austrian school’s overtly anti-empirical approach onto Keynesianism.
You can dabble in what happens by enacting and funding laws policies.
For an economics simulator, score the simulated economy on say... a dozen things. Make that into a radar plot and the overall score is the area of the plot.
Now the player can do things by enacting rules that cause changes in how the AI plays the game in an attempt to maximize their score.
It reminds me of what you're looking at [0][1](github[2]).
The project broadly covers what you're doing with agents also starting businesses and going out of business with simple supply chains. There's even some stuff to do with health.
-[0]: https://www.trytobegood.com/projects/hosny
-[1]: https://rhizome.org/editorial/2018/apr/03/humans-of-simulate...
It will definitely make things more complicated, but a morale mechanism will speed up visual feedback on other things.
I like your idea, and it is very enjoyable in games like Crusader Kings/Europa Universal is where player can technically put down buildings, but almost all interactions with the world are via policy making and event based interactions, and most of the actual world moves on via actions of NPCs.
A couple prototype screenshots in case you are curious as to what I'm talking about:
https://twitter.com/_DaveSullivan/status/1377196666663677957
https://twitter.com/_DaveSullivan/status/1375063612281217024
Now something that I must say, but know that it may not be universally shared.
I see you are making a 3D game. But there is a very particular reason why games like Crusader Kings/Europa Universal is/Civilization/Rimworld/Prison Tycoon/Factorio are 2D. They are inherently numbers games, and appeal to specific crowds. These are ultimately tabletops or rather very complex spreadsheets and ultimately as a player you re being statistician. There many successful franchiseare s on the contrary, like Tropico, StarCraft, SimCity but after playing through all of them, I've realised that fancy 3D worlds, while very pretty, take the toll on cognitive processing. Dwarf Fortress, while 3D with multiple layers, forces player to look at, deal with and think about only one at a time, with very distinct..uh..visuals and they do the job better than anything else.
Ultimately you want to convey the information to player as fast as possible, make it as instantly identifiable, and intractable as possible. Graphics, especially faux 3D like Ck3 or Civ4/5/6 are pretty nice and help with that ultimate goal, not hinder.
Just my 2c, and something to think about. Be assured that I'll play your game regardless (I spent many an hours on Tropico series, along with every other game listed above. No regrets :))
From that, there countless trade-off themes emerge – industrialization vs. sustainability, regulation vs. laissez-faire, bilateral negotiations vs. multinational institutions, and so on and so forth
Even Civilization games are essentially computerized board games, there is nothing of fidelity in its historical topics or in its simulation aspects.
Perhaps setting the tax rate and the money supply? Choosing which businesses to nationalize, and which services the government will provide from tax dollars?
You could have challenges, migrate to renewable energy within 10 years, flood your enemies by causing oceans to rise, migrate from a car based economy to public transit, out compete other nations while eliminating tariffs, fix the US economy, win over your war focussed enemies with better science/schools, cause your enemy countries economies to fail, etc.
I've built a couple of different models all of which have failed to be quite fun =)...
I've talked about it before here[0], the newest one I'm working on is much simpler, and focused around player interactivity and giving them knobs to tweak at the outset.
Let me know if you'd like to do a call and chat about this, it would be fun to see what kinds of things we've been attempting =)...
I can imagine it being a fun game for spreadsheet players to try and tease apart how to recover a death-spiralling community by applying new ideas in the right order at the right time to the right people. Ideas can be unleashed into key parts of the sim and only live for so long before they become 'proven discredited' and ignored, or whatever.
Good luck! Simulations are fun.
Maybe allow players to build a scenario, then switch it on and watch it run.
Traffic is one real-life operation done from a position of somewhat centralized planning, so maybe allow players to build bus routes or subways.
Or just sell it to city planners. They are used to joylessly watching everything go awry.
Set objectives for the people, a few non intrusive tools (moving cats around, blowing wind, changing light traffic), and as the angel, the player must influence the world so that objectives are reached.
Not doing what I want ? Eat this volcano !