Crime

Crime in Microlandia is a spatial feedback loop, not a flat city-wide rate. Each tile carries a dynamic attractiveness score that decays a little each day, leaks a fraction to its four neighbors, and gets a small boost whenever a crime actually happens there.

One petty theft makes the next petty theft on the same block slightly more likely, and broken windows spread outward. Left alone, a hotspot stays sticky for about two weeks before it fades. The numbers driving this are taken from the Short and colleagues model of hotspot crime, the standard reference in spatial criminology.

Where criminals come from

Criminals are not parachuted in. They come from your own population. Once a month, the city scans two groups:

• Adults (eighteen and older) who are not in school or university, and whose bank balance has dipped below the basic-needs threshold. Whether they have a job or not does not matter. A working citizen whose pay cannot cover food and rent is just as eligible as an unemployed one. If they roll, they walk off the job and pick up crime.
• Unschooled teens between thirteen and eighteen who are out of school and without work.

Each candidate rolls against a monthly chance derived from the annual rate on this page. Unemployment, homelessness, dropout rates, and household balance all feed the size of this candidate pool before the dice are even thrown.

Police and response

Every police station you build with staff hires officers, and those officers split their time between patrol and dispatch. Patrol units (two officers per unit) walk a fixed distance each shift, taken from this page. When a crime fires, an officer is dispatched to the scene from any staffed station.

If the dispatcher catches the criminal, the city pays the incarceration bill out of the police budget. When local prison capacity is available, that bill accrues monthly across the sentence at the local per-inmate housing rate. When every local cell is full, the criminal is exported to an outside facility, which bills the city up front for the full sentence at the (higher) exported rate, plus a one-time transport fee. Either way, a petty theft conviction barely registers, while a homicide books decades of housing.

The clearance rate (the share of crimes that end in an arrest) varies by crime type, from very low for property crime to very high for homicide. Those numbers match the real-world averages from FBI data.

This page also lists the monthly police officer salary. That bill lands whether your stations are catching criminals or sitting idle.

Building a prison

By default, a caught criminal is exported to an outside facility. That facility charges the city up front for the full sentence at a marked-up annual rate, plus a one-time transport fee. The whole bill lands in a single day, so one homicide arrest can dump close to two decades of housing into your police budget in one line.

A city-owned prison flips this. Inmates housed locally accrue at the lower local annual rate, billed monthly across the sentence, with no transport fee. Capacity is fixed; once it fills, new catches overflow back to the exported rate.

The tradeoff is up-front cash and zoning. Building a prison is a major capital decision, and it carries a sizable NIMBY footprint, so site it well away from residential neighborhoods or watch nearby approval and rent both sag. The per-inmate-year savings alone will not pay back the construction cost in any realistic play session. The real reason to build is to smooth the cash-flow shock from long sentences (homicide is nearly two decades on its own), which is exactly where the exported-upfront model hurts the most.

Why crime moves families

After every crime, the victim household rolls a chance of moving out. The chance climbs steeply with the severity of the crime: petty theft barely budges anyone, burglary moves some, and a homicide on the block is a near-certain trigger. So a single unsolved spree can quietly drain a neighborhood weeks after the headlines fade. And because each crime nudges the local attractiveness field upward for about two weeks, the same block stays slightly more likely to be hit again.

Your levers

• Build more police stations and hire enough officers to actually cover the map.
• Cut the offender pipeline upstream. Jobs, schools, and shelters reduce the candidates the monthly roll has to draw from.
• Accept that prevention is cheaper than clearance. The full incarceration bill plus the rent loss from a fleeing renter tend to dwarf the cost of avoiding the crime in the first place.

Parameters

Vandalism fine amount

$300

Fine charged when police catch a vandal. Matches typical municipal fine for vandalism.

⚠️ Source pending

Vandalism repair cost

$300

Cost to repair public infrastructure damaged by vandalism.

⚠️ Source pending

Clearance rates by crime

Clearance rates by crime type (percentage of reported crimes cleared by arrest or other exceptional means).

Source: FBI Crime in the Nation (2023)

Move out probability by crime renter

Probability that a renter household moves out after experiencing a crime, by crime type.

Source: Effect of Criminal Victimization on a Household’s Moving Decision - Laura Dugan (1999)

Move out probability by crime owner

Probability that a homeowner household moves out after experiencing a crime, by crime type.

⚠️ Source pending

Distance patroled by police officers per shift

5000

The distance a police unit (2 officers) can effectively patrol during a single shift (8 hours). This affects the area that is covered daily by each police station in the city.

Source: ICMA Center for Public Safety Management

Incidents per police office shift

1

Number of crime incidents that a single officer can be dispatched to and handle during a single shift (8 hours).

Source: NYPD

Criminality rate

0.015

Annual probability that a working-age, low-cash adult (not in school) becomes a chronic offender. Calibrated to first-arrest base rates among poor US adults (~1-2% annually). The criminal flag is sticky — only cleared by accumulating $1M or reaching retirement age — so this is a stock-recruitment rate, not a per-incident or reoffense rate.

Source: Mitchell, J. et al. (2022). Homelessness and predictors of criminal reoffending: A retrospective cohort study

Attractiveness decay rate

0.05

Per-tick decay rate (ω) for the dynamic crime attractiveness field. B decays ~5% per day, giving a half-life of ~14 days. From Short et al. (2008) hotspot model.

Source: Short, M.B. et al. (2008). A statistical model of criminal behavior

Attractiveness diffusion

0.1

Fraction (η) of attractiveness that spreads to cardinal neighbors each tick. Creates the near-repeat / broken-windows effect. From Short et al. (2008).

Source: Short, M.B. et al. (2008). A statistical model of criminal behavior

Attractiveness crime boost

0.5

Boost (θ) added to the dynamic attractiveness field at a tile when a crime occurs there. Creates positive feedback: crime → higher attractiveness → more crime. From Short et al. (2008).

Source: Short, M.B. et al. (2008). A statistical model of criminal behavior

Inmate export transport cost

$4,000

The logistical cost of transporting a criminal to an out-of-city facility when no local jail is available. Covers private extradition contractor fees, or the overtime, mileage, and processing fees for two sworn deputies to perform a long-distance transport.

Source: Police1 - The Hidden Costs of Inmate Transport

Inmate housing cost per inmate

$47,000

Annual incarceration cost per inmate when housed OUT OF JURISDICTION (no local Prison, or local Prison full). The full bill is charged upfront when a criminal is caught, equal to this rate multiplied by the sentence length in SENTENCE_YEARS_BY_CRIME. Reflects the marked-up contract rate other counties charge to take in inmates, on top of housing, personnel, food, and healthcare.

Source: Vera Institute

Local inmate housing cost per inmate

$32,000

Annual incarceration cost per inmate when housed in a city-owned Prison with free capacity. Lower than INMATE_HOUSING_COST_PER_INMATE because the city pays its own operating cost rather than another jurisdiction’s contract rate, and no inmate transport is needed. Vera’s nationwide local-jail operating cost averages ~$33k/inmate/year.

Source: Vera Institute — The Price of Jails

Sentence years by crime

Average prison sentence actually served before initial release, by crime type. Multiplied by INMATE_HOUSING_COST_PER_INMATE to produce the upfront incarceration bill when a criminal is caught.

Source: Bureau of Justice Statistics — Time Served in State Prison, 2018

Police officer monthly

$9,500

The monthly salary paid to each police officer, including benefits overhead. NYPD FY25 Personnel Services budget of $5.38B across ~48,844 employees = ~$110k/year per employee. We use $9,500/month ($114k/year) to approximate total compensation including benefits.

Source: NYC Council - NYPD Fiscal 2025 Executive Plan

Prison construction cost

$75,000,000

Upfront construction cost of a 200-bed city Prison. Anchored to Monroe County, Indiana’s planned 400-bed facility ($80.9M hard construction, ~$202k/bed) — scaling down to 200 beds gives ~$40M hard, ~$48M with soft costs. We round up to $75M to cover the higher-end of contemporary US per-bed estimates ($286k–$550k/bed for state projects) without inflating to the police-headquarters tier.

Source: Monroe County (Indiana) Council jail budget reporting