THE LANDLORD ALGORITHM
One recommendation. Every landlord. Same morning.
You manage one portfolio.
You run 16 apartment buildings in Larkspur, a fictional city. Your owners have set a two-year goal: collect in rent ◇ scenario.
Like five competing firms in town, you subscribe to Ledgerline, a fictional rent-pricing service. Every month it sends a recommended rent for each of your buildings, computed from pooled lease data from every subscriber, including your competitors, who mostly accept it.
Each month you choose: accept the recommendation, price below it, or opt out and set rents yourself from public listings.
How to play
- The map is a rent heat-map: blue is cheaper, red and purple are more expensive. Lit windows are occupied homes; dark windows are empty.
- Your buildings have a gold outline and star. Navy badges mark competitors on the same software. Unmarked buildings are independent landlords.
- Tap or click any building to see its numbers.
- Watch both ledgers: your rent collected, and the city's empty homes and average rent.
- 24 months, one decision each. About 5–8 minutes.
Every number in the game is a scenario assumption for a fictional city, listed under “Scenario assumptions.” Ledgerline and every firm in Larkspur are invented and model no real product or company.
The simulation's assumptions
These are the rules of the fictional city you just ran. They are scenario choices, not measurements of any real market, product or company.
Your results
Your actual numbers from this run (scenario units).
Your ledger
The city's ledger
Month by month
The documented evidence
Only these real-world facts are used. Everything else in the game is scenario.
- In August 2024 the U.S. Department of Justice sued RealPage. On August 23, 2024, the DOJ and eight states filed a civil antitrust suit alleging that RealPage's rent-pricing software let competing landlords share nonpublic, competitively sensitive data about rents and lease terms and use it to align their rents.
Sources: U.S. Department of Justice, Justice Department Sues RealPage for Algorithmic Pricing Scheme… (Aug. 23, 2024); Minnesota Attorney General, Attorney General Ellison sues RealPage (Aug. 23, 2024). - Several large landlords later settled. In January 2025 the DOJ amended its complaint to name six large landlords. Cortland agreed to a proposed consent decree that month; Greystar (August 2025), LivCor and Willow Bridge followed with proposed settlements, and RealPage itself reached a proposed settlement in November 2025.
Sources: U.S. Department of Justice, Justice Department Sues Six Large Landlords… (Jan. 2025); Federal Register, Greystar proposed final judgment (Sept. 5, 2025); U.S. Department of Justice, Proposed Consent Decree with LivCor; Multifamily Dive, Willow Bridge settlement; U.S. Department of Justice, Justice Department Requires RealPage to End the Sharing of Competitively Sensitive Information… (Nov. 2025). - These are allegations and negotiated settlements, not findings of guilt. RealPage said its settlement was not an admission of liability, and Greystar admitted no wrongdoing. RealPage has said it had already changed its software to remove nonpublic data.
Sources: HousingWire, RealPage DOJ settlement (2025); USA Today via AOL, Nation's largest landlord, Greystar, settles price-fixing suit with feds (2025); ProPublica, Greystar settlement (2025).
What the game does not claim: Ledgerline is invented. The game does not reproduce RealPage's software, its formulas or its effects on any real city's rents or vacancies.
The conclusion the author draws
This is Corey Spangler's reading, not a finding of the simulation.
No landlord in this game decided to fix prices. Each one, including you, made the individually sensible choice: take the recommendation. The coordination lived in the shared data, not in anyone's intent. And once enough of a market runs on one model fed by everyone's private numbers, that model's best move can be to leave homes empty, because a vacancy protects everyone's comps better than a cheaper lease does. The question isn't whether landlords are good people. It's whether competitors should be allowed to pool private pricing data through one recommender in a market people can't easily walk away from, like housing.
Disagree? Good. Play the two years again: opt out, price below, try to hit the goal without the dark buildings, and tell us which assumption you think is wrong and why.