A transit planning simulation

Four thousand people work in that neighborhood.Nobody has ever run a line to it.

Lines
8
Stations
185
Trains
61
Clock
05:30

The clock runs a minute a second. By the morning peak you’ll know if that neighborhood ever gets a train.

    • Installs and plays offlineBoth editions install as apps and run with no connection; the city packs ship inside. Signing in is the only thing that needs one.
    • One optional account, nothing hiddenSign in with Google or an e-mail and password — for cloud saves, a career, friends, or to buy here. It holds your e-mail, a username and what you sync; the site counts page views, sign-ups and purchases with no name or address attached, and the app asks once before sending anything about crashes or how a run went.
    • The free edition is the whole modelThis week’s metro under this week’s mandate, 2025, five years — with every system in it. Not a crippled demo.
    • 79 metros, US and internationalFifty-four US Census packs plus twenty-five in Europe, Asia and South America. The maps are not interchangeable: US jobs are LODES; international jobs are derived from land use.
    • Open source, AI disclosedMIT licence. Built with AI assistance, stated up front and in the README.

    Sold already?

    All aboard.

    Download the free or full edition and build the network you keep sketching on the back of a ticket. Keep scrolling to learn how the game works.

    Free: one real metro and one mandate a week, five years on the clock. Full edition: all 79 metros, every decade, every mandate — $25 once. Ten percent of every sale goes to pro-urban organisations; buy Mac or Windows direct and an extra 30% (Valve's Steam cut) goes with it. MIT licensed.

    Launching late November. Join the mailing list for the exact date, first access before general release, and the full edition at $10 instead of $25.

    AWhat it is

    You are the planner. Not the driver, and not the mayor.

    You inherit a real metro — its real neighborhoods, its real transit, its real traffic — and decide where to spend money you answer for. A travel-demand model of the kind an MPO actually runs tells you who rode, and why.

    Today
    5 km

    902,500 926,900 weekday journeys +24,400

    Ashland Avenue Light rail, Irving Park Road to 79th Street: 20 stations over 22.5 km, every 8 minutes, 3-car trains.

    17,100 of them out of cars, 7,400 off foot and bike; the line itself boards 24,500 a day — nested-logit mode choice

    With one line
    ASHLAND
    Every figure here comes from the engine, run on the real Chicago of 2025. Today is the system's own reported boardings; the answer is that plus the weekday journeys the model finds once the line is added. The existing 'L' is in the model too, so these increments are, if anything, conservative.

    Read a real city before you draw on it

    Hover any neighborhood for its real population, density, jobs by sector and median income — with its GEOID, so you can check it against the Census. Where the Census suppresses a figure, the pack says unknown rather than inventing one.

    Census 2020 PL94-171 (P1/P3/H1) · ACS 2018–2022 B19013 · LEHD LODES8 2022

    Commit money you answer for

    Every alignment has a price, and tunnel runs four to five times surface. Cost and schedule risk are drawn once, when you commit, from the run's own seeded stream. You cannot reload to escape a bad draw.

    Log-normal overrun, calibrated on Flyvbjerg’s 258-project sample

    Run a service not just a map

    Track is not service. Headways across six periods including evening, night and weekend; consist lengths; skip-stop and express patterns; fleet ageing and procurement lead times. Three operating dials the board votes on, and a workforce that notices what you do with them.

    Service level, wage ratio and maintenance funding, each voted on

    Find out if you were right and why

    Four-step demand — generation, distribution, nested-logit mode choice, assignment — solved to equilibrium over crowding and congestion. When ridership moves, the panel says what moved it, and shows the residual it could not explain.

    Attribution, not decomposition

    What it is not: a route-painter where ridership is a function of how pretty the line looks, and not a city-builder where you zone the demand into existence. The demand is already there, measured, and largely indifferent to you.

    BThe decisions

    They arrive whether you were ready or not.

    Depth is not a list of features — it is whether the systems bite. Every decision below is one the model computes both sides of, and the constant underneath it is named so you can go and disagree with the constant. A choice with only an upside is not a choice, so none of these has one. And none of them finishes when you make it: what you decide here changes which crises come looking for you later.

    B Platform 1 Next: go underground
    B Raise the fare The trap · 1 of 6
    Upside Revenue, every time

    |e| < 1, so the arithmetic never turns against you.

    What it costs Riders, approval, subsidy

    It empties the system, and the subsidy that follows riders leaves with them.

    Measured elasticity −0.52 — steeper than the −0.33 planners quote, because that rule was fitted to cities where riders had no car. Measured by sweeping the fare, not assumed

    The dice are real. You load them. Two different kinds of trouble come down the line, and the difference between them is the whole game.

    Some of it is genuinely chance. A severe winter storm closes the region and every street-running line stops. You could not have prevented it, you are not being punished, and the model does not pretend otherwise.

    The rest you commissioned, years earlier. Every event is drawn against a weight the model reads off the city you actually built — so a derailment is a random event whose odds you set a decade ago. Track held at 90 per cent condition draws one about an eighth as often as track let go to 40. Let the maintenance backlog grow and the press investigation gets likelier too, because the backlog is one of the figures it reads. Take an unpopular agency into a contract year and the operators are likelier to walk.

    And some of it is not random at all. An election is not an event. It is a date, and it is on the calendar from the day you take office.
    cite: Derailment odds read worst track condition; the press investigation reads the maintenance backlog; the contract reads approval and council support

    CThe city answers back

    Point at a district. It will tell you what it is — and the figures it read to decide.

    Nothing here was assigned by a designer. Every tract works out its own character each year from its own census and employment numbers — so a neighbourhood can become a secondary centre while you play, and the reason is sitting in its figures.

    ResidentialResidentialCampusCampusResidentialMixedMixedResidentialDowntownMixedMixedResidentialDowntownDowntownRetail corridorResidentialDowntownMixedMixedRetail corridorRetail corridorIndustrialOffice parkIndustrialIndustrialOffice parkOffice park

    Seven characters

    Hover or tap any parcel. These are the seven things a census tract can be in this model, and which one it is gets recomputed from its own numbers every year of the run.

    Ordered by evidence: a campus signal is unmistakable and rare, so it is tested before the generic jobs-versus-residents split that everything else falls out of.

    A diagram of the categories, not a real metro. In a run these are your actual tracts, and the office park that appears on your ring road in year six appeared because the jobs really moved there.

    How growth follows a station — the method

    How a station acts: regional growth is allocated each year by where development already is, tilted toward tracts within walking distance of an open station and fading to nothing at the edge of that range; jobs are tilted harder than homes; no tract grows faster than a fixed share of itself in a year. The character is then re-read from the new numbers. Slow on purpose — over thirty years it is the difference between a corridor and a car park. cite: Growth allocated yearly by existing development and station access, jobs tilted harder than homes

    Why this matters more than it sounds. The office park is named in the code as “the hardest thing in transit to serve well”. If your growth keeps producing them on the ring road, that is the city telling you what your network did to it.
    cite: Derived annually, never assigned

    DSteps to victory

    Six ways to win, and every one of them is set by your city.

    No score to beat, no target somebody picked. Each condition is computed from the metro you inherited, so winning in Houston is a different job from winning in Boston. Pick a metro and read its six bars.

    2025 · 2,637,686 residents · the real system carried 53,699,248 boardings in 2025 (NTD, fixed-route modes)

    1. Gauge: regional transit mode share. Today 1.2%, needed 1.2%. 1.2%

      regional mode share, held for a year

      Transit Capital

      today 1.2%needed 1.2%

      What the real system carries today, on a network of your own.

    2. Gauge: peak road congestion relief against the no-network counterfactual. Today 0%, needed 10%. 10%

      peak congestion relief

      Roads Unclogged

      today 0%needed 10%

      Against the Minneapolis–Saint Paul that never got your network — the counterfactual, not the day you started.

    3. Gauge: net tonnes of CO2 avoided a year. Today 0 t, needed 52,754 t. 52,754t

      net CO₂ avoided a year

      Carbon Cut

      today 0tneeded 52,754t

      0.02 t a resident, net of what the network itself burns.

    4. Gauge: on-time performance, needed 80%, in a surplus year with rider satisfaction above 65%, held for a year. 80%

      on-time, in a surplus year

      Held the Line

      today set at the start of the runneeded 80%

      Surplus, satisfaction above 65% and punctuality — all three at once, for a year.

    5. Gauge: share of the metro within walking distance of a stop. Needed at least 10%. 10%

      of the metro within a walk of a stop

      Within Reach

      today set at the start of the runneeded 10%

      Or 60% more than you inherit, whichever is higher.

    6. Gauge: cut in the average regional commute. Today 0%, needed 10%. 10%

      off the average commute

      Ten Minutes Back

      today 0%needed 10%

      Measured on every resident, not only on the ones who ride.

    Where these figures come from — the method

    Figures are the 2025 bars; an earlier decade scales the inherited share up — sixteen-fold in 1920, when a third of urban trips were by streetcar. Population is the pack’s Census 2020 tract sum; boardings are the agency’s latest National Transit Database year across fixed-route modes. cite: Each bar set from the metro you inherited, with absolute floors and caps

    None of these is a checkbox. They pull against each other on purpose: coverage costs farebox recovery, and the carbon target punishes exactly the empty-bus coverage the access target rewards. Winning all six means finding the trade-off.
    cite: Relative targets, with absolute floors

    EThe metros

    79 of them, built by a script from open data.

    183,301,439 residents across 35,752 zones — 31,521 US census tracts and 4,231 OSM land-use zones. No API key, no account, no scraping. Each pack is a file: every zone in the metro with its real population and jobs, the observed or derived commute matrix, the water, the roads, and whatever transit the agency currently publishes a GTFS feed for — drawn in that agency's own colours, with real opening years, so the 1960 map shows the 1960 network. Forty of the 79 start with real rail on the map; the other 39 start with nothing, which is also true of them.

    Free this week One of the thirteen metros marked Free is this week's free assignment, with one mandate, in 2025, for five years. It changes every Monday at 00:00 UTC.

    All 79 metros, as a ledger 183,301,439 residents · 35,752 zones · 40 with rail on the map
    Metro Region Residents Zones Difficulty Existing rail lines Page
    HartfordNortheastConnecticut899,498235DemandingnoneCity page →
    BostonFreeNortheastMassachusetts3,155,919745Forgiving21City page →
    Newark & North JerseyNortheastNew Jersey3,119,659717ForgivingnoneCity page →
    BuffaloNortheastNew York954,236260DemandingnoneCity page →
    New York CityFreeNortheastNew York8,804,1902,324Forgiving29City page →
    PhiladelphiaFreeNortheastPennsylvania3,683,718926ForgivingnoneCity page →
    PittsburghNortheastPennsylvania1,250,578394DemandingnoneCity page →
    ProvidenceNortheastRhode Island660,741145DemandingnoneCity page →
    BirminghamMid-Atlantic & SouthAlabama674,721189PunishingnoneCity page →
    Washington DCFreeMid-Atlantic & SouthDistrict of Columbia689,545206ForgivingnoneCity page →
    JacksonvilleMid-Atlantic & SouthFlorida995,567218PunishingnoneCity page →
    MiamiMid-Atlantic & SouthFlorida4,646,1421,122Demanding4City page →
    OrlandoMid-Atlantic & SouthFlorida2,289,420414PunishingnoneCity page →
    Tampa & St. PetersburgMid-Atlantic & SouthFlorida2,418,869606PunishingnoneCity page →
    AtlantaFreeMid-Atlantic & SouthGeorgia3,554,303936Punishing6City page →
    LouisvilleMid-Atlantic & SouthKentucky782,969216PunishingnoneCity page →
    New OrleansMid-Atlantic & SouthLouisiana824,778309DemandingnoneCity page →
    BaltimoreMid-Atlantic & SouthMaryland1,440,243418Demanding5City page →
    CharlotteMid-Atlantic & SouthNorth Carolina1,115,482305PunishingnoneCity page →
    Raleigh & DurhamMid-Atlantic & SouthNorth Carolina1,454,243298PunishingnoneCity page →
    MemphisMid-Atlantic & SouthTennessee929,744249PunishingnoneCity page →
    NashvilleMid-Atlantic & SouthTennessee1,305,096284PunishingnoneCity page →
    RichmondMid-Atlantic & SouthVirginia925,547235DemandingnoneCity page →
    ChicagoFreeMidwestIllinois6,208,4181,550Forgiving8City page →
    IndianapolisMidwestIndiana1,324,670310PunishingnoneCity page →
    DetroitMidwestMichigan3,949,1731,218PunishingnoneCity page →
    Minneapolis–Saint PaulFreeMidwestMinnesota2,637,686668Demanding3City page →
    Kansas CityMidwestMissouri1,077,257309PunishingnoneCity page →
    St. LouisMidwestMissouri603,156208Demanding6City page →
    OmahaMidwestNebraska775,130210PunishingnoneCity page →
    CincinnatiMidwestOhio830,639226DemandingnoneCity page →
    ClevelandMidwestOhio1,264,817427DemandingnoneCity page →
    ColumbusMidwestOhio1,323,807328PunishingnoneCity page →
    MilwaukeeMidwestWisconsin1,346,467399DemandingnoneCity page →
    Oklahoma CityTexas & PlainsOklahoma1,091,820337PunishingnoneCity page →
    AustinTexas & PlainsTexas1,899,205425PunishingnoneCity page →
    Dallas & Fort WorthTexas & PlainsTexas6,695,0661,507Punishing8City page →
    HoustonFreeTexas & PlainsTexas5,553,9241,248PunishingnoneCity page →
    San AntonioTexas & PlainsTexas2,009,324375PunishingnoneCity page →
    PhoenixMountain WestArizona4,420,5681,009PunishingnoneCity page →
    TucsonMountain WestArizona1,043,433270PunishingnoneCity page →
    DenverFreeMountain WestColorado2,473,074592Demanding10City page →
    BoiseMountain WestIdaho726,072183PunishingnoneCity page →
    Las VegasMountain WestNevada2,265,461535PunishingnoneCity page →
    AlbuquerqueMountain WestNew Mexico676,444176PunishingnoneCity page →
    Salt Lake CityMountain WestUtah2,207,316473Demanding6City page →
    Los AngelesFreePacificCalifornia10,014,0092,495Punishing6City page →
    SacramentoPacificCalifornia2,206,197507PunishingnoneCity page →
    San DiegoPacificCalifornia3,298,634736Punishing6City page →
    San Francisco BayFreePacificCalifornia4,486,6871,034Forgiving12City page →
    San JosePacificCalifornia1,936,259408PunishingnoneCity page →
    HonoluluPacificHawaii1,016,508329DemandingnoneCity page →
    PortlandFreePacificOregon1,837,201418Forgiving10City page →
    SeattleFreePacificWashington4,018,762860Demanding2City page →
    BerlinEuropeBerlin3,149,817211Demanding93City page →
    BarcelonaEuropeCatalonia1,228,82293Forgiving20City page →
    MadridEuropeCommunity of Madrid2,030,103156Demanding12City page →
    LondonEuropeEngland4,885,364228Forgiving14City page →
    ParisEuropeÎle-de-France2,423,794192Forgiving57City page →
    RomeEuropeLazio1,291,712190Demanding5City page →
    LisbonEuropeLisbon983,712150Demanding11City page →
    WarsawEuropeMasovia1,521,650224Demanding35City page →
    AmsterdamEuropeNorth Holland1,038,104155Forgiving59City page →
    PragueEuropePrague1,169,775204Demanding84City page →
    StockholmEuropeStockholm1,325,520187Demanding98City page →
    ViennaEuropeVienna715,219161Forgiving304City page →
    BangkokAsiaBangkok2,553,855222Punishing112City page →
    Hong KongAsiaHong Kong1,660,484117Forgiving117City page →
    OsakaAsiaKansai605,446152ForgivingnoneCity page →
    TokyoAsiaKantō3,090,463218Forgiving108City page →
    MumbaiAsiaMaharashtra4,490,480125Punishing21City page →
    SeoulAsiaSeoul2,108,959191Forgiving36City page →
    SingaporeAsiaSingapore1,436,591162Forgiving14City page →
    TaipeiAsiaTaipei570,808111Demanding93City page →
    Buenos AiresSouth AmericaBuenos Aires3,333,821112Demanding24City page →
    BogotáSouth AmericaCundinamarca2,415,240138Punishing1City page →
    Rio de JaneiroSouth AmericaRio de Janeiro2,960,775107Demanding28City page →
    SantiagoSouth AmericaSantiago3,669,069202Demanding9City page →
    São PauloSouth AmericaSão Paulo4,849,464223Punishing16City page →
    What counts as an existing line — the method

    Existing rail lines are the metro, light rail, commuter rail, streetcar, BRT and ferry services in the agency's published GTFS feed, bus routes excluded. Difficulty is the game's own three-step scale. Europe, Asia and South America are on their own tabs. Every metro, with its real figures →

    FThe decades

    An era is not a vehicle catalogue. It moves nine levers at once.

    How many people there are, where they live, how many cars they own, what building costs in real terms, how hard land is to take, what the fare is, who pays the operating deficit, how willing Washington is to help, and what technology exists. In 1920 transit was a business that had to turn a profit. By 2025 seven of every ten operating dollars are somebody else's political decision.

    1920 → 1950

    The Streetcar City

    Transit is how the city works, and it turns a profit. Almost nobody owns a car, land is cheap and easy to take, and there is no such thing as federal transit money.

    Fare norm
    $0.68
    Cars per household
    0.32
    You may build
    Streetcar, commuter rail, metro, ferry, bus
    The eight decades, side by side every lever that moves
    1920

    The Streetcar City

    Transit is how the city works, and it turns a profit. Almost nobody owns a car and there is no such thing as federal transit money. You are running a business, not an agency.

    Cars / household
    0.32
    Operating subsidy
    0%
    Federal match
    0%
    Real build cost
    0.42
    1950

    Postwar Boom

    The suburbs are filling and the cars are arriving. Ridership has peaked and begun its long slide. Construction has never been cheaper in real terms, and there is still no subsidy if you lose money.

    Cars / household
    1.00
    Operating subsidy
    0%
    Federal match
    0%
    Real build cost
    0.38
    1960

    The Freeway Era

    The streetcars have been torn out and the Interstates are being driven through the city. Transit is in freefall, Washington funds highways and nothing else, and the political climate is openly hostile. The hardest era to win.

    Cars / household
    1.22
    Operating subsidy
    5%
    Federal match
    0%
    Real build cost
    0.44
    1975

    Fiscal Crisis

    The city is broke, the fleet is falling apart, riders feel unsafe. But the oil shock has made driving expensive and Washington has finally started paying — eighty cents on the dollar for capital, if you can get the application in.

    Cars / household
    1.52
    Operating subsidy
    45%
    Federal match
    80%
    Real build cost
    0.62
    1992

    Rail Renaissance

    Flexible federal funding has arrived and every city wants a light rail line. Land is harder to take and environmental review is real, but for the first time in forty years the money is on your side.

    Cars / household
    1.78
    Operating subsidy
    60%
    Federal match
    75%
    Real build cost
    0.78
    2010

    Transit-Oriented

    Downtowns are repopulating and development follows the rail line. Federal capital money has grown stingier and slower, and three-quarters of your operating cost is somebody else's decision to keep funding.

    Cars / household
    1.86
    Operating subsidy
    68%
    Federal match
    50%
    Real build cost
    0.92
    2025

    Present Day

    The census data as it actually is. Remote work has hollowed out the peak, costs are the highest they have ever been in real terms, and you are subsidised for seventy cents of every operating dollar — for as long as the politics hold.

    Cars / household
    1.88
    Operating subsidy
    72%
    Federal match
    55%
    Real build cost
    1.06
    2040

    The Carbon Budget

    Driving is priced, density is mandated, and transit is policy rather than charity. Building has never cost more, but nobody is asking whether the network is worth it any more, only whether you can deliver it.

    Cars / household
    1.55
    Operating subsidy
    75%
    Federal match
    65%
    Real build cost
    1.25
    What this diagram is, and is not — the method

    The city above is a diagram of what the eras do to one, not a picture of any real metro: the streetcar lines are torn out and paved over, the freeways cut through, the core empties to parking and fills back in around the stations. Route colours mark transit lines only. Fare norm is the era's typical single fare in the game's figures; cars per household and the modes you may build are from the same table. Population back-casting pivots the real 2020 distribution by the change in the density gradient rather than inventing a distribution from scratch, which keeps the real local structure — where the rivers and the rail yards and the university actually are — while moving the centre of gravity.

    GWhat it costs

    Free to start. Twenty-five dollars to finish — ten before launch. No subscription, ever.

    The free game gives away the depth and limits the breadth, which is the opposite of how demos usually work. You get the whole simulation on this week’s assignment — one real metro under one mandate, in the present day, for five years — the same assignment for every free player in the world, changing every Monday. What is bounded is which city, which brief and how long. Never how real.

    · Free Single ticket
    FreeSteam demo · Mac & Windows
    This week’s assignment One of thirteen metros · one of five mandates This week · changes in under a week, Monday 00:00 UTC
    • ✓
      The complete demand model — nothing simplified
    • ✓
      One real metro and one mandate, drawn each week from thirteen metros and five briefs
    • ✓
      Present day, 2025 · five years in office
    • ✓
      A new assignment every Monday — a game you have started keeps going
    • ×
      This week’s metro only, five years on the clock

    Long enough to open a line, watch it fill, feel the operating cost arrive, face an election and take a fare decision. The whole loop, once, on the same city and brief as everyone else playing free that week. Steam demo, or Interurban Free on Mac and Windows.

    Get the free game
    · Full Single ticket, no expiry
    $25once · not a subscription
    $10 before launch, on the list
    • ✓
      All 79 metros
    • ✓
      Eight decades, 1920 to 2040
    • ✓
      Every mandate and every mode, including the sandbox
    • ✓
      No clock and no weekly rotation — pick anything, play as long as you like
    • ✓
      Plays offline. No connection, no check-in
    • ✓
      Every future expansion, at no further cost

    One tier, one price: $25, once, Steam or desktop. Ten percent of every sale goes to pro-urban organisations; buy Mac or Windows direct and an extra 30% — Valve's cut on Steam — goes with it. The binary is the licence.

    Get the full edition
    10–40%

    Ten to forty percent of Interurban proceeds go to urban advocacy campaigns — 10% of every sale to transit groups, street-safety campaigns and land-use reform organisations, plus an extra 30% (Valve's Steam cut) when you buy Mac or Windows direct. Where the money went will be published here, by recipient, after the first payout.

    Join the community and build and run the transit network you always dreamed of living in.

    This week: one real metro · one mandate, 2025, five years, the whole model. Steam, Mac or Windows.