Notes on SimCity at 30

Car oriented development looks a lot more viable when you hide all the parking!

Yes, the first attempt at a comprehensive city planning game, Sim City, is 30 years old.  Jessica Roy in the Los Angeles Times has a good piece on how the game helped turn people onto city planning …

Along the way, the games have introduced millions of players to the joys and frustrations of zoning, street grids and infrastructure funding — and influenced a generation of people who plan cities for a living. For many urban and transit planners, architects, government officials and activists, “SimCity” was their first taste of running a city. It was the first time they realized that neighborhoods, towns and cities were things that were planned, and that it was someone’s job to decide where streets, schools, bus stops and stores were supposed to go.

… while also reinforcing some bad 20th century ideologies.  Sim City …

  • conceals the impacts of parking, thereby making car-dependent development look more functional and attractive than it is.
  • requires single-use zoning.  You can’t live above your shop, or have a grocery store in your office building.
  • requires car access to every building.  Pedestrianized urban cores are impossible, no matter the density.
  • treats transit very superficially, not allowing the user to specify routes and frequencies, and giving the misleading impression that any kind of transit, anywhere, produces some vague benefit.  Thus there is nothing to stop you from common mistakes like building high density in culdesacs, where efficient transit could never get to it.

Recently, I did a quick look at available iPad city planning games.  I tried Megapolis, Designer City, Pocket City, and Sim City: BuildIt.   They’re all built on the same four fallacies, and their handling of transit ranges from comical to nonexistent.  (Sim City BuildIt actually starts with a greenfield freeway interchange, leaving no doubt what kind of city they expect you to build.)

My past articles on SimCity are here, here, here, and here.   Sim City gets credit as a pioneer, but it’s run its course.  I hope we see more planning games that try to get transportation right, and games that try to do transit in particular.  If you’re working on one, let’s talk!

  (Photo: BLDGBLOG)


15 Responses to Notes on SimCity at 30

  1. Sean Gillis March 7, 2019 at 9:36 am #

    City Skylines is much better (in most ways) at transit than Sim City. The player is responsible for placing bus stops and connecting them via routes. Same thing for rail and subways – you build the stops and the rail and plan the routes. Players can also specify how many buses (or trains) run on each route, in response to demand. If you want more trips, it costs a lot more, and you also need to ensure your depot and fleet are big enough to support the service.

    The transit downsides to City Skylines are a weird scale/ simulation problem, where even small city transit systems get overwhelmed by demand. I had a city of about 30,000 residents with hundreds of people waiting at bus stops. This seems to have something to do with how trips are modeled, in that every ‘resident’ needs to make their commute trip to/ from work every day, but the simulation wants to show individual buses running in game time. Essentially everyone’s trip seemed to happen at once, instead of being spread through a simulated day. The other obvious scale problem was the cost and capacity of subways, which started making sense (i.e. were affordable to build and operate) in cities under 50,000 residents. Two lines, many stops – 50,000 residents!

    City Skylines is a visual treat, but still lacks a realistic transportation simulation. I think this is partly a question of computing power, as anything above a mid-sized city starts to grind computers to a halt. It also lacks mixed-use zoning, and has way too little parking, and it seems way too easy to get people onto transit.

    One thing to point out, for serious players of Sim City 4 and City Skylines, they can be heavily modified. I’m clueless as to how to do this, but it includes custom buildings and site layouts (could fix the parking) and the ability to change how the simulation works (maybe address some of the transit/ transportation problems). Sim City 4 had modes which allowed pedestrianized streets, with building access, but I’m not sure about Skylines.

    • Morgan Wick March 7, 2019 at 10:04 am #

      This reminds me of the problems I had with Traffic Giant, which you can see in the comments of one of the posts Jarrett links. That game didn’t have everyone trying to get to work at once, but it did attempt to simulate individual people attempting to commute in real time even as the month counter advanced far faster than the apparent movement of the buses should have allowed. The effect was to neuter the effect of frequency and leave you wondering whether to get more buses or bigger buses – the trade offs weren’t always obvious, especially given the size of the maps that could leave the total number of stops on a route in the teens.

      It sounds like what Skylines should have done was make transit service levels and usefulness part of the demand model, with more frequent routes having more capacity and more likely to have residents choose them over driving, without attempting to play out the actual, real-time movement of buses on the landscape. That worked okay for Traffic Giant which was just a transit sim, but it sounds like it clashes with Skylines’ aspirations to a full-fledged city sim.

    • Evan Derickson March 7, 2019 at 11:37 am #

      C:S also has the same problem of requiring car access to every building. There are some mods to make pedestrianized-looking areas, but the fundamental assumption of car use still underlies everything in the game. I find it difficult to make anything in that game that I would actually want to walk around. Interestingly, the studio’s previous two games (Cities in Motion and CiM 2) were transit management simulators which did have the player specify routes and frequencies, even allowing different levels of service at different times of day. I’d love to see Jarrett’s thoughts on those someday.

    • Kenny Easwaran March 8, 2019 at 12:52 pm #

      On one forum I saw Cities: Skylines described as “a city painter, not a city simulator”. It seems right – it creates all the visual imagery of a city, but if trucks or cars or bus riders can’t get where they’re trying to go, they just disappear after a certain period of time to avoid some of the most hilarious failure modes of SimCity 4 (where you could easily get 36 hour travel times on the stretch of highway into your city, which people kept going for!)

  2. Wai Yip Tung March 7, 2019 at 11:55 am #

    I had fun with SimCity. But I understand how it fall short of your expectation.

    On the other end of the spectrum, we have some pretty sophisticated simulation system like BEAM.

  3. Ruediger Herold March 7, 2019 at 12:28 pm #

    Over the years I’ve developed kind of a single player game simulating a growing rail (or street or whatever) network. You’d need:
    1. a piece of (grid) arithmetic paper, it needn’t be large to allow for complex surprises later on in the game.
    2. dice (the more the better but limiting the number is also fine)
    3. something to draw with (the more colors the nicer, but a single ball-pen would also do)
    The dice decide where a place (population: +1) starts to exist – or grows.
    The player then decides where to build a segment of railway to connect places. The results will be quite different according to who you prefer e.g. neighbors, bigger cities, ports etc. (see coverage vs. ridership)
    I could e-mail you a “powerpoint film” how it could look like, but although the rules of this game are very simple, that one would be quite “heavy”.

    • Eric March 14, 2019 at 2:29 am #

      This sounds like a perfect game to be automated as a cell phone app. It wouldn’t even be particular hard to do.

  4. Georgist economist March 7, 2019 at 12:34 pm #

    Not city builder but transit-oriented games: OpenTTD and Simutrans. They have their own issues. (OTTD: passenger trips have no fixed destination, i.e. people will happily pay you to take them anywhere. Simutrans: passengers have a very clear idea of exactly where they want to go, which is all over the map, thus passenger transportation is not economically viable until late in the game. Neither game allows fare policy changes.) But the core of both games is route planning, with the player allocating vehicles to routes. Headways also directly affect how many passengers turn up at stations (but so does the top speed of the vehicles and whether you paid the city to build a statue of you). Roads are not congested with private cars. Cities do grow when served, but they do so in all directions, not just toward the service.

    • Rollmaterial March 8, 2019 at 12:46 am #

      Simutrans-Extended is an attempt at a realistic simulation. Two years ago even a fare policy system was introduced. I highly recommend checking it out!

    • MinchinWeb March 9, 2019 at 2:55 pm #

      Cargo distribution (aka “Cargodist”) ( was added to OpenTTD to address this exact concern (passengers not having a fixed destination) back in June of 2013.

      • Georgist economist March 11, 2019 at 10:09 am #

        No, Cargodist only locks in the destination(s) that other passengers had previously traveled to; but those “original” passengers still behave in exactly the same manner. The choice of what service pattern to create is entirely up to the player, Cargodist only makes it difficult to change the pattern that the player had created earlier.

        (I’m only criticizing the destination—and route!—determination mechanism. Balancing passenger, mail, valuables, etc. creation is a much more coherently done feature. Route determination, on the other hand, is IMO even more stupid: passengers won’t take a direct connection between A and B because they insist on going via C.)

        I was contrasting the above to Simutrans because in the latter, even the very first passenger showing up to a newly built station has a definite idea of where he wants to go.

  5. Joel Williamson March 7, 2019 at 8:29 pm #

    MiniMetro is a much simpler game than most mentioned here. It focuses on building (as the name suggests) a metro system for a city. It shows the advantage of mixed use areas for transit planning, and how transfer stations can get overloaded when served by lines with different capacities.

  6. JJJ March 8, 2019 at 10:00 am #

    Sim City 4 has been heavily, heavily modified over the course of a decade to address most of those points. The only point not addressed is parking, because nobody wants to build a city with sprawling lots.

    Unfortunately, the game engine itself is not optimized, so you can run it on a state-of-the-art desktop today and get the same laggy performance you got in 2005.

  7. Rick Robinson March 19, 2019 at 7:59 pm #

    Raising my hand a bit belatedly, I have been playing with what I call a transit game. At this point it is not

    much of a game, just a set of very generalized design rules. I assign a city so many square miles of residential fabric at various binned density levels, and derive a potential transit ridership based on how many prospective riders each square mile produces. The same idea applied to an arterial corridor gives an estimated ridership for individual lines.

    Mainly I have been playing with the history of a couple of fictional California cities, geographically unconstrained Midvale and more constrained (on a coastal strip) Santa Teresa, both in the current size range of 300-500K for the city proper and on order of a million for the metro area. Like most California cities that aren’t San Francisco, most of their growth is post 1890, when Sprague streetcar techology made it easy for cities to expand beyond walking limits.

    In these conditions, I suspect that sprawl was and still is a default condition. So long as cheap land is available within a decent commute of where people need or want to go, greenfield development with single story frame construction is probably the cheapest way to provide lots of housing for a fast growing city. Perhaps the ideology of single family homes and leafy burbs is more effect than cause.

    You get densification when you run out of cheap land, or the land is so far out that people will increasingly pay extra to live closer in. Hence the modern California situation of people making enormous commutes to cities they can’t afford to live in.

    For both gameplay and instructional reasons you’d want to provide policy levers that can succeed or fail in providing a desired result, but I am still exploring the environment that policy works in.

    @Ruediger Herold, @Eric – any game of this type could be automated pretty easily, I think. So long as you’re content with a somewhat schematic map, not pretty graphics of buildings and elevated lines.

  8. R. R. Miguel March 22, 2019 at 11:31 am #

    Have you played Mini Metro? It isn’t equivalent to Sim City, but a wonderful game (I should emphasize that unlike simulation games like Sim City, it doesn’t attempt/promise to replicate the actual experience of managing a train network – my recommendation is based moreso on entertainment value!)