Protection methodology

Protection is a proprietary 1 to 10 address-level signal for fire station response access. Lower is stronger. The active FPS-v2 method uses routed road distance, estimated drive time, and station type.

Inputs

The system identifies a nearby HIFLD fire station, routes from the submitted US street address, and records road miles and estimated minutes. When either routed distance or time is missing, the calculation returns no Protection number.

Distance has two remote gates. More than 5 road miles produces a base 9, and more than 7 road miles produces a base 10. At 5 miles or less, drive-time buckets produce a base value from 1 through 8. A station whose name signals volunteer service adds one point, capped at 10. Other or unknown types retain the base value.

Bands and polarity

• 1 to 3: Strong
• 4 to 6: Adequate
• 7 to 8: Limited
• 9 to 10: Remote

These labels describe routed station access. They do not predict an emergency outcome.

Station type limits

The volunteer signal comes from words in the public station name, such as “volunteer” or “VFD.” It is a name heuristic rather than a certified staffing roster. A missing signal cannot establish career staffing.

Inputs we exclude

Protection excludes hydrant distance or density, water supply and flow, building construction, alarms, sprinklers, staffing detail beyond the name-based station type, apparatus, mutual aid agreements, real-time dispatch, and traffic incidents. It does not produce an official ISO, PPC, or FSRS result and has no affiliation with those systems.

Common misreads

A city inventory does not equal an address score. Straight-line map distance does not equal road miles. A low number does not describe wildfire likelihood. A station name without a volunteer keyword does not prove paid staffing.

Versioning and updates

FPS-v1 used distance and time. FPS-v2 adds the station-type adjustment described above. Method changes are labeled so readers can interpret results against the active rules. HIFLD inventory snapshot dates appear on each context hub.

Complete base bucket table

Exactly 5.0 miles uses time. Exactly 7.0 miles receives base 9. FPS-v2 adds one for a volunteer name signal after the base and caps the final at 10. Strong covers 1 through 3, Adequate covers 4 through 6, Limited covers 7 through 8, and Remote covers 9 through 10. Lower is stronger.

Frozen illustrative arithmetic

These examples teach fixed rules and do not report live properties.

Version and missing data

FPS-v1 used distance and time. FPS-v2 retains that base and adds the volunteer name adjustment. Both road miles and drive time are required. Missing either input returns no Protection number. A straight line, city average, ZIP centroid, inventory count, or guessed default cannot fill the route.

Source and exclusion controls

HIFLD provides station candidates and context rows, not departments. CITY does not establish jurisdiction. Protection excludes hydrants, water supply and flow, building construction, sprinklers, alarms, apparatus, detailed staffing, mutual aid, live dispatch, current traffic, and incident conditions. It cannot predict response or loss. A station name signal is a heuristic and does not certify staffing. Protection does not calculate ISO, PPC, or FSRS classifications, and FireProtectionScore has no affiliation with those organizations or programs.

Read the examples, sources, and score-reading guide, then check a US street address.

Calculation order and reproducibility

A reproducible reading follows the same order every time. First confirm that both route inputs exist. Second test whether road miles are above 7, then whether they are above 5. Only when the route is at or below 5 miles should time select a base. Third apply the volunteer name signal and cap the result at 10. Fourth map the final value to its band.

This order matters. Applying the name adjustment before the distance gate, rounding a boundary into a different comparison, or allowing minutes to override a remote gate would produce different arithmetic. The published table is the controlling content reference for readers.

Selection and routing uncertainty

The address must resolve to a usable property point, a station candidate must be selected, and the routing network must return miles and minutes. Geocoding changes, station-source updates, road edits, access restrictions, and routing-engine changes can alter those inputs. A result describes the modeled route available to the active calculation rather than a permanent physical guarantee.

A nearby station pin cannot establish a route. Straight-line distance can cross water, slopes, restricted land, divided roads, and disconnected streets. Station inventory count also cannot establish route quality because it omits the property start point and network connection.

Responsible comparison

Compare results only after checking version, miles, minutes, selected station name, and whether the volunteer heuristic applied. A band-only comparison can hide a large input change that stayed inside one band or a small adjustment that crossed a band edge. Keep the underlying facts with the final label.

Do not average property results into a ZIP or city value. The method is address-native, and aggregation would introduce a new formula, weighting choice, and geographic claim that this methodology does not define. Context hubs remain source inventories for exactly that reason.

Worked boundary checks

Consider three illustrative routes with no volunteer signal. At exactly 5.0 miles and 7 minutes, time gives base and final 5 Adequate. At 5.01 miles, the first distance gate gives base and final 9 Remote, and time no longer governs. At exactly 7.0 miles, base remains 9 Remote; at 7.01 miles, the upper gate gives 10 Remote. These comparisons show why the words above and at or below matter.

Now hold a base 6 route fixed. With no volunteer name signal, final 6 is Adequate. With the signal, final 7 is Limited. The route did not change, but the disclosed FPS-v2 adjustment crossed a band boundary. At base 10, the same adjustment has no numeric effect because the result is capped.

Method governance

A future change to inputs, bucket edges, gate comparisons, station selection, type adjustment, missing-data behavior, bands, or polarity would require a new version and updated examples. Editorial wording cannot change formula semantics. Snapshot refreshes and routing-provider updates should remain identified as data changes rather than method versions.

The examples table acts as a regression reference for public explanations. New copy should reproduce E1 through E6 exactly, preserve E7 as an address-geometry narrative without a ZIP number, and keep E8 as a separation between Protection and wildfire context. Any conflicting arithmetic should be corrected before publication.