Methodology

In plain English

A report-card style grade for this block — how safe it is and how much investment and activity it sees, compared to similar Boston blocks over the past year.

• Safety — crime incidents, fire calls, and police encounter reports in the area
• Opportunity — building permits, 311 service requests, county job market, metro CPI, and tract-level Census income context
• Census ACS — population, poverty, and income at tract/place grain for interpreting signals (reference only)
• Both sides matter — a block can feel safe but quiet, or busy but facing more incidents

Compared to other Boston blocks like this one, using public city data from the last 365 days.

Letter grades work like a school report card: A is strong overall, C is about average for similar blocks, F is well below. The number out of 100 is the same idea on a finer scale.

This is not a home value, insurance rating, or guarantee about future crime. It is a snapshot to help you compare places.

What it measures (technical)

Production block scores use model version v1.1-block. The score fuses safety (crime, fire, FIO, EMS when available, climate resilience) with opportunity (permits, valuation, 311, inspections, business licenses, county unemployment, wages, tract income) on H3-9 blocks. Each block is compared to jurisdiction peers over a rolling 365-day window.

Safety sub-score (0–100)

Weights adapt when EMS coverage or weather data is available for the jurisdiction.

• Crime, fire, and FIO — inverted peer percentiles (lower incidents → higher score)
• EMS — inverted peer percentile when ≥30% of jurisdiction blocks have EMS data
• Climate — inverted climate-stress percentile when weather is bridged to the block

Typical blend (no EMS/weather): 50% crime · 30% fire · 20% FIO. With EMS + weather: 44% · 26% · 16% · 7% climate · 7% EMS.

Opportunity sub-score (0–100)

• Permit volume, declared valuation (log-scaled), 311 activity — peer percentiles
• Inspections and business licenses — peer percentiles when present
• County unemployment — inverted (BLS LAUS, county grain)
• County average weekly wage — normalized (BLS QCEW, county grain)
• Tract median household income — peer percentile when ACS is available

Typical blend with income: 20% permits · 16% valuation · 16% 311 · 15% unemployment · 13% income · 9% wage · 6% inspections · 5% licenses. County labor metrics are context — not block-level precision.

Composite & letter grade

Composite = √(safety × opportunity). Letter grades: A ≥ 85, B ≥ 70, C ≥ 55, D ≥ 40, F < 40.

Scores are reproducible from stored place_genome features and raw counts — no hidden state. Blocks with no incident activity in the window may receive a context-only grade ranked against peers using county and tract context; confidence is capped accordingly.

Place Index (trajectory score, Pro)

A separate decomposable index (0–100) from OSS safety/opportunity, 6-month crime and permit momentum, lifecycle state, crime volatility, climate resilience, and displacement pressure. Headline index equals the weighted sum of five sub-indices — no hidden state.

• 30% safety · 25% opportunity · 20% momentum · 10% volatility · 15% displacement pressure
• Momentum blends lifecycle stage with permit growth minus positive crime growth
• Pressure sub-index = displacement risk score × 100 (0–100 scale)

Higher volatility and pressure raise the trajectory score — they measure civic dynamism and change intensity, not desirability alone. Not a credit, insurance, or lending score.

Limitations

• Not a real-estate valuation or insurance rating.
• Sensitive domains (FIO) carry caveats; see our Trust & Data page.
• Missing data for a component reduces confidence — we show sub-scores separately.
• Scores are calibrated per city. Active metros today include Boston, Cambridge, Chicago, Atlanta, Cincinnati, Seattle, and Austin; Atlanta-metro candidate cities (Decatur, Sandy Springs, Marietta, and others) use block grid + county context with FBI UCR crime benchmarks where no open incident feed exists.
• County-level labor metrics (BLS LAUS/QCEW) and FBI agency benchmarks are context — not block-level precision.

Trust & Data →

Displacement early-warning (Analyst Solo)

A separate pressure score (stored 0–1, displayed 0–100; not part of the letter grade) that rises when permit activity, permit valuation, 311 volume, and assessed values accelerate together on a block. Intended as a 12–18 month neighborhood-level early warning — not a prediction about any person or household.

• v1 weights: permits 30% · valuation 25% · 311 15% · assessments 30%
• Levels: watch ≥ 55, elevated ≥ 70, high ≥ 82 (on 0–100 scale)
• Do not use for lending, enforcement, or individual targeting.
• Confidence drops when assessment history is missing for a window.