Balancing a game economy before players arrive

You cannot balance a game from player behavior before anyone has played it. You can still give the economy a hard time. In Landlord, that means giving several fake landlords the same starting conditions, then letting them make different sensible and terrible choices for a game year.

The next few decisions need to create the kind of game I meant to make. Can a careful player afford a second property? Does a mortgage get them there faster, with a real chance of trouble? Can someone ignore the game for a while without coming back to an inevitable bankruptcy notice?

Those are balance questions. The dollar amounts are just how the game answers them.

Start with a pace, not a price

Landlord starts a new player with $20,000, a desk, and a market full of buildings. The first economy target was a modest operating margin for a maintained property, about 25 to 40 percent after its recurring costs. The cash-first path should make a second purchase possible after a few game months. A mortgage should accelerate growth but leave less room for a bad month. An absentee owner should lose ground without falling through a trapdoor.

That is enough to judge a number against. A repair bill is too high when the owner who fixes things on time can never get ahead. A down payment is too cheap when buying the next building makes the first hour of the game disappear.

The targets also make disagreements useful. On August 26, 2026, I ran the balance tool at Landlord commit 741cc15. The frugal path produced a 30 percent operating margin. The balanced and mortgage-heavy paths averaged 2.9 properties, with their second purchase at game month 3.1. The mortgage path’s average low point was $821 in cash, and one seed fell to -$256. That is a shape I can discuss. “It felt slow” is a much worse starting point.

Give the game repeatable fake players

That August 26 run sent the same small set of strategies through eight random seeds and twelve game months. One owner buys a duplex in cash and keeps it in good condition. One saves, then uses a mortgage for the next move. Another reaches for mortgages immediately. The last one mostly ignores the work.

They are not meant to predict actual players. They are a rack of weights for the economy. Each strategy puts pressure on a different promise:

• The careful cash owner tests whether ordinary play pays for itself.
• The balanced owner tests the cadence between the first and second property.
• The mortgage owner tests whether faster growth has a visible downside.
• The absentee owner tests whether neglect is painful without becoming a dead save.

The seeds matter because a single lucky year can tell a very flattering lie. A strategy that works once may only have missed a burst pipe. Across the eight current seeds, the cash-first owner averaged a 30 percent operating margin.

Repeatable decisions, a clock, and a report can catch a broken relationship between numbers. The report records cash, net worth, properties, occupancy, and the date of the next purchase. Real players will still surprise me. The scripts make sure the floor is in place before they arrive.

The first pass made incidents the whole game

The first balance run is a historical baseline now. On one maintained duplex, incidents cost about $5,000 over a game year, roughly one per month and $417 per month on average. That was more than the mortgage, tax, upkeep, and ordinary repairs combined. The owner could do the responsible thing and still spend most of the business on boilers, storms, roof leaks, and inspection notices.

That did not create a tense property game. It made the operating margin cling to single digits. A player who happened to draw a few expensive incidents had little reason to care about the rest of the portfolio decisions, because the incident table had already made the decision for them.

The fix was deliberately boring. I lowered the daily incident chance from 4.5 percent to 3.5 percent and cut several of the larger incident costs. A boiler repair dropped from $1,200 to $850; a storm from $700 to $500. The current registry at 741cc15 still uses the 3.5 percent chance, $850 boiler repair, and $500 storm repair. The frequent, cheap problems stayed.

That historical before-and-after run moved the frugal owner from about a 9 percent operating margin to 34 percent. The August 26 run puts the current figure at 30 percent. Incidents still interrupt a plan. They just do not own the whole spreadsheet.

Watch margin and cadence together

Operating margin is the cleanest health number in a management game. For the cash-owned property, it compares rent and recovered arrears with the recurring drains: taxes, upkeep, repairs, and incidents. It leaves out capital moves such as buying and selling, because a good month should not look profitable only because the owner sold the building.

Margin alone is not enough. A game can give a healthy return while making the next meaningful choice too distant. That is why the playtest also records when a strategy buys its second property. In the August 26 run, the balanced and mortgage-heavy paths both averaged 2.9 properties and reached the second purchase at a median 3.1 game months. The mortgage path’s minimum cash averaged $821, and its worst seed reached -$256. Faster progress comes with a thin buffer and, in one of these worlds, an actual shortfall.

The negligent absentee path is harsher. It ended with -$1,257 in average cash, reached a worst-seed minimum of -$7,356, and posted a -84 percent operating margin. It also kept one property and averaged $11,883 in net worth, so the run does not prove that every neglected save is dead. Cash is no longer evidence that the save stays recoverable. This strategy can come back to a liquidity crisis, and recovery needs its own measurement.

Balance lives in the rules players can feel

Landlord keeps its economy values in registries rather than scattering them through game rules. When an incident fires too often or a repair costs too much, the answer is a change to the table and another run through the same worlds. The catch-up system can then keep calculating rent and time away without inventing a different economy for a closed tab. That part has its own explanation.

A spreadsheet cannot replace players. It can give me a better first question. Before people teach me how they want to play, the game should have an answer for owners who repair things, borrow too much, wait too long, and come back after a week away.

Landlord is free to play in a browser. Its persistent portfolio uses the same account as Acres, so a game worth returning to has somewhere to keep the buildings you bought.