Interactive Nutrition Label
The thirteen scored dimensions as a single at-a-glance label, scored live by the same API that powers the home-page address search. Under the overall score the label reports two headline grades — the building (how it is built) and the site (what surrounds it) — because a well-built house on a hard site and a poor one on an easy site average to the same middling number, and the composite alone can’t tell you which you are looking at. Both are national percentiles, so an A on either means the top fifth of US homes. Enter any U.S. address or place name below (or use your current location). Nothing is scored until you do. This home mode auto-fills the real building’s details from public records (USACE structure inventory + Census) and scores it. Edit any detail in Refine building details to correct it and the label updates. Prefer a what-if? Switch to What-if build or Compare builds to see how five hypothetical construction profiles move each dimension at the same location (the location-driven dimensions shift with the address while the construction-driven ones stay comparable). You can also search by the name of a place or business: the typeahead resolves it to its street address. Each dimension also shows where it lands vs US homes as a national percentile. Your scored location is kept in the page link, so you can bookmark or share it, and it’s pre-filled on your next visit. A “cost over a mortgage” strip present-values the dollar difference in energy and expected disaster losses against a typical comparable, and Compare builds mode puts two profiles side by side with a per-dimension delta.
About this data: Every profile here is scored live by the project's HTTP API (the CLI simulator behind HTTP) at the location you enter above (Cooper-Young, Memphis by default). The five construction-driven dimensions (resilience, energy, durability, environmental footprint, and infrastructure burden) are modeled from each build's configuration and stay comparable across locations. Location-driven dimensions are resolved for the scored location: Health (CDC PLACES), Socioeconomic (Census ACS), and Walkability (EPA National Walkability Index) are each scored against the national distribution of US census tracts. These sources are bundled, keyless, and comparable across locations. Where a location's tract can't be resolved they show N/A: excluded from the composite rather than filled with a placeholder, so a strong build isn't unfairly penalized for a missing input.
Climate Projections is a sub-county downscaled climate-hazard score (USGS CMIP6-LOCA2 ensemble mean for heat, precipitation & drought, SSP2-4.5 to SSP5-8.5, mid-century, plus an Argonne ClimRR 12 km Fire Weather Index fire leg) sampled at this location's census tract, identical across profiles. The composite is the mean of the scored dimensions. National grades use the absolute thresholds A ≥ 80, B ≥ 60, C ≥ 40, D ≥ 20, F < 20, and each dimension also shows a national percentile (“vs US homes”).
The “cost over a mortgage” figure is a comparative present value of only the two dollar-defensible flows (modeled energy cost and expected annual disaster loss) discounted in constant (real) dollars at ~4% real (homeowner mortgage opportunity cost; ~2% OMB social rate shown as the band), with no real cost escalation. It deliberately excludes purchase price, property tax, maintenance, and the qualitative dimensions, so it answers “which home is cheaper to run and insure,” not “which is cheaper all-in.” Full methodology and citations: research/lifetime-cost-research.md.
The small confidence dot on each row (● High, ◐ Moderate, ○ Low) and the composite’s “N of 9 scored” line report data quality (source, geographic resolution, completeness) on a channel kept deliberately separate from the grade, so “high confidence” never reads as “good score” (a parcel can be confidently an F). It is a provenance judgement, not a statistical margin; the only true numeric interval drawn is the faint whisker on Climate Projections, spanning the SSP2-4.5 → SSP5-8.5 emissions-scenario range. Full methodology and citations: research/uncertainty-confidence-research.md.