overcrowding and underfunding: the lessons of “mini metro”

Last week, several people emailed me to make sure I had seen Mini Metro, a simple but absorbing transit game that’s come out in alpha.   

Mini metro

Between deadlines, I played a little, but in two hours we reached a win-win condition: The program was sure it had beaten me, and I was pretty sure I’d exhausted its possibilities.

The deal here is that you can draw lines but you can’t site stations.  You build your first humming three-station metro, but then new stations just appear, further and further out, emitting graphical screams until you extend lines to serve them.   

Untitled 2

Each line has just one vehicle on it, so as the line is extended its frequency drops, so it visits each station less often, increasing the risk of crowding.  That's accurate in its own way: Under any fixed budget, frequency and extent are inversely related.  

Survive enough time and you’ll get upgrades: an additional line, or a chance to increase a line’s speed and capacity so that it drains its stations  more often. 

But it’s not enough.  Before you know it, stations are getting overcrowded and when one of them finally fails, remarkably, the game calls this “losing” and ends.  You see it as a little clockface loop on a crowded station, in this case the next to last station from the west end of the Blue Line.

Untitled 3

You are not offered options to encourage flexible work hours, run shuttle buses, or hire Japanese train-packing specialists.  Every round I played ended up in the “station failure” condition.  

There's plenty of nonsense here.  Optimizing for capacity is a very different problem than optimizing for customer utility or freedom.  With a single train on each line, you add capacity by doubling up lines at key stations, or sometimes by creating odd loops whose purpose is to visit each station often, not necessarily run straight lines that would be useful to the customer.  In the session above, a looping downtown shuttle seemed useful, for example, though these are usually disappointing in reality.   Still …

There are may ways to construct a transit game, but if you want it simple and sexy, a metro game that focuses on capacity is not a bad way to go.   What’s more, although many of the assumptions are absurd, the game’s final message is surprisingly accurate:  Politicians demand that transit systems spread out but not that they provide enough intensity — whether that means frequency, speed, or in this case capacity.  Transit agencies are always being told to spread themselves thin. 

In Mini Metro, the relentless appearance of new stations (often in awkward and expensive places) mimics this constant pressure on transit agencies to spread out horizontally.  The slow trickle of funding – most of it spent just keeping up with that spread – is never enough to stop overcrowding.  Sooner or later, transit agencies have to invest in core capacity – new lines or bigger platforms  where the failure points are, which is often downtown.  But to too many local leaders, that’s read as “spending too much money downtown instead of in the neighborhoods,” or "they already have a transit line, so why are they getting another when my district/city/ward has none?" 

That's why Los Angeles's crucial new downtown rail link is called the Regional Connector, emphasizing (truly) that the whole region benefits from it.  Toronto is considering renaming its proposed “downtown relief line” for similar reasons.  

Imagine how a campaign to get every city councilor playing a well-designed game might raise consciousness.  

If any developers out there are interested in working with me on this, let me know.  As longtime readers know, I’ve been thinking about this for years.

7 Responses to overcrowding and underfunding: the lessons of “mini metro”

  1. Lior March 8, 2014 at 8:54 am #

    The more I think about this, having my background as a computer programmer, I just think that this type of game is impossible to develop. Games of this nature will always be simplification of reality. If there is something city councilors should not be taught – is to simplify the complex and unpredictable future of city planning.
    I did like though Vancouver’s interactive public outreach ( https://www.humantransit.org/2013/12/vancouver-interactive-public-outreach-on-network-design.html ). This should be the direction, a very local-focused explanation of possibilities.

  2. Fedor Manin March 9, 2014 at 3:46 pm #

    The game in question is extremely fun and addictive (for me, at least) and part of this is the simplicity which makes it have little to do with transit planning, or anything else for that matter. I would much rather play it than an actual transit planning game.

  3. Sean Gillis March 10, 2014 at 9:05 am #

    Jarrett – you forgot the most important point: what was your best score 😉

  4. Aaron Priven March 10, 2014 at 11:54 am #

    You didn’t mention the game’s most significant unreality, which is that it’s free to tear down a line and rebuild it somewhere else.
    It’s a fun game, but its relationship to transit is just skin-deep. That’s not a criticism. If you’re going to oversimplify transit enough to make it a game, better to introduce some things that are obviously not true in real life so as to avoid people basing their thinking on it. Better this than the “well, we thought the real amount of parking necessary was ugly” of SimCity.

  5. Nathanael March 13, 2014 at 7:28 pm #

    I’d make a Simutrans mod if you want to make a well-designed transpo model game. Simutrans has a few advantages:
    (1) it’s completely open source (free software)
    (2) it’s got a very open-ended modular design
    (3) it already has separate bus and rail traffic models
    (4) people are already trying to make it more realistic (in the “experimental” mod). I’ve actually been working on this.
    – The biggest lack is any coherent model of growth, but that really shouldn’t be too hard to put in. I was working on it last year but I got distracted.
    – The second biggest is scheduling, which actually seems to matter less than you’d think, since you *can* implement headway-based operation.
    – The revenue model is dopey but easy to alter in code. If you’re a programmer.
    Most of the existing paks are designed to model intercity transport, but there’s no particular *reason* for that, just people’s tastes.

  6. Patrick Sunter March 18, 2014 at 5:34 pm #

    Hi Jarrett, nice review of Mini Metro.
    Well, not exactly a game, but there is more and more good software, some of it Free/Open Source, being developed to help understand transit networks, and that are becoming closer to being able to support decision-making and analysis.
    I know you’ve already featured OpenTripPlanner and Mapnificent, a nice new one for animating transit networks by Vasile Cotovanu is https://github.com/vasile/transit-map, see “Swiss Trains” demo:- http://simcity.vasile.ch/sbb/
    I am keeping a bit of a list of the open source tools in this space here as part of my PhD work :- http://www.appropedia.org/Open_Source_GIS-T_Public_Transport_Tools_Review

  7. S M Sabri Ismail (sabre23t) September 11, 2015 at 5:01 pm #

    Minimetro [1] is already at Beta 33 version released 9 Sep 2015. Now with Gif animation playback/export, and audio.
    I am wondering about that “every city councilor playing a well-designed game” Jarrett mentioned. Any progress?
    [1] http://dinopoloclub.com/minimetro/