How peer matches are made

1. Organization taxonomy & compatibility

Maryland delivers fire/rescue through countywide systems, independent municipal departments, and volunteer corporations inside county systems. Each organization is typed, and only organizations in a compatible comparison class are ever compared head-to-head. County-level systems and independent municipal departments compare together as “system-level” organizations. A countywide system is never compared with a single volunteer corporation or a special-purpose (airport/military) department.

2. Similarity factors

Similarity is a weighted blend of scale-aware closeness on these factors (weights are documented and unit-tested; they sum to 1.0):

• Population served26%
• Annual incident volume24%
• Number of stations14%
• Career personnel12%
• Operating budget10%
• Area protected (sq mi)6%
• Population density4%
• EMS share of calls4%

Size-like quantities (population, incidents, budget…) use a ratio closeness robust across orders of magnitude; bounded shares (EMS %) use a linear closeness. The adjacent smaller/larger baselines are the nearest organizations by a composite size index, guarded so a vastly larger or smaller organization is never presented as “adjacent.”

3. Four independent scores

Similarity — how alike two organizations are (0–100). Never a verdict on quality. Resource capacity and Operational scale are descriptive percentiles of an organization’s size and workload among Maryland systems — bigger is not better. Data confidence reflects how much verified public evidence backs the profile. These are kept separate on purpose so scale is never mistaken for performance.

4. Missing data

Similarity is renormalized over only the factors both organizations report, so an unknown value never rewards or penalizes a match — it only lowers the evidence backing it (shown as “matched on N shared fields” and in Data Confidence). Values are never invented or estimated; unknown stays unknown. Calculated values (population density, funding per resident) are labeled and name their inputs.

5. Source transparency

Population and land area are U.S. Census primary for every organization. Department operational fields (stations, personnel, incidents, budget) come from official department pages and adopted county/city budgets where available; where only a secondary source (e.g. Wikipedia citing primary data, or news coverage) was locatable, it is used and flagged. Every displayed number links to its source. Determinism: every ranking breaks ties by slug, so results are stable.

Scope & limits

This is a Maryland-only pilot covering system-level organizations. Volunteer corporations within county systems are intentionally not compared against countywide systems. Several smaller systems have limited public data (budgets and incident totals are often unpublished) — their matches are directional and flagged with a “Limited” data-confidence rating. Not a substitute for a formal accreditation peer set (e.g. CPSE/CFAI).