How the Virginia Fireground Index is computed
The Virginia Fireground Index aggregates the best available public fire data for each locality — Virginia's 95 counties and 38 independent cities. The Overall score is a transparent normalization of the civilian fire-death rate (Virginia Department of Health vital-statistics deaths per Census population) — the one metric genuinely comparable across every locality. Incident/response counts are shown as sourced context where a locality publishes them, and metrics with no authoritative public source are shown as unavailable. Data Confidence reflects how much verifiable public evidence backs each locality.
The comparable metric (scored)
The one fire metric that is genuinely comparable across every Virginia locality is the civilian fire-death rate: average annual civilian fire deaths per 100,000 residents over 2019-2020. Deaths come from the Virginia Department of Health's Injury & Violence Prevention deaths data (Virginia Vital Statistics — unintentional fire/burn deaths, ICD-10 X00–X09, resident-based, un-suppressed counts); population comes from the U.S. Census Bureau. The rate is a direct, reproducible calculation —(5-year deaths ÷ 5) ÷ population × 100,000— not an estimate.
The Overall score (transparent normalization)
The Overall Fireground Index score normalizes that rate across the 133 county-equivalents: a z-score, inverted so a lower fire-death rate yields a higher score, centered at 50and clamped to 0-100:score = clamp(50 − z × 18, 0, 100). There is no black box. The platform's multi-dimensional scoring engine is reserved for when multiple comparable metrics exist; with one comparable fire metric today, this single transparent normalization is used instead.
Context metrics (shown, not compared)
Where a locality publishes an incident/response figure, it is shown as sourced context with its exact definition and reporting year. These counts use incompatible definitions across localities (total incidents vs. calls for service vs. unit responses, and fiscal- vs. calendar-year), so they are never normalized into a cross-locality comparison — that would be misleading.
Unavailable metrics (named, not hidden)
Civilian injuries, firefighter injuries, and fire dollar loss have no authoritative per-locality public source, so they are shown as “Data not publicly available”with the sources investigated. Missing data reduces Data Confidence; it never blanks a locality.
A note on the death-count basis
Virginia’s fire deaths here are death-certificate (vital-statistics)counts from the Virginia Department of Health, not fire-incident (VFIRS) counts. Every Virginia locality uses the same VDH source, so the index is internally consistent; because different states may use different death-count bases, scores are ranked within Virginia and are not a cross-state comparison.
Data Confidence (a core product)
Data Confidence communicates how much verifiable public evidence backs a locality: the authority of its sources, how many public metrics exist, and how statistically reliable the outcome data is (a locality with 0-2 fire deaths over five years has a real but uncertain rate). A data-rich locality reads differently from a data-sparse one — and that difference is itself information. Confidence is never based on whether every locality reports identically.
See the data sources behind every figure.