Methodology

A score should narrow research, not hide it.

Five measured components create one comparable metro screen. Eligibility, direction, weights and missing-data rules stay fixed across the current release.

Last updated August 2, 2026
Score components
5
Jobs, rent, affordability, supply and climate
Required coverage
All five
Incomplete metros remain unscored
Published range
0–100
Relative percentile screen, not a forecast

What the score answers

The composite score is a relative screen across US metro and micropolitan markets. It asks how a market ranks on five measured conditions in the current release; it does not estimate a property's return or predict what the market will do next.

Each component becomes a national percentile from 0 to 100. The weighted contributions are added and rounded to the published whole-number score. A higher score means a stronger combination under these definitions, not a guarantee that the market or a particular deal is suitable.

Five components, one published weight set

ComponentWeightMeasured inputBetter rank
Job growth30%Employment change over the trailing 12 monthsHigher
Rent trend25%12-month change in measured market rentHigher
Affordability15%Annual rent as a share of median household incomeLower burden
Supply pressure15%Permits per 1,000 residents and months of inventoryLower pressure
Climate risk15%Expected annual building loss relative to building valueLower loss ratio

Gross yield, migration, investor share, taxes, wages and other page evidence remain decision inputs, but they are not silently added to this score. Keeping the boundary fixed makes two published scores comparable.

Incomplete evidence receives no score

A metro is scored only when all five buildable components exist. Missing rent, labor, affordability, supply or climate evidence produces an unscored page with the available raw figures and a reason. It does not receive a neutral component or a score based on a smaller denominator.

Why not reweight a partial score?

A two- or three-component score can look more extreme simply because fewer signals pull it toward the middle. Publishing it beside a five-component score would imply a comparability the evidence does not support.

Percentiles, source pools and outliers

Percentiles are calculated against the eligible national observation pool. Job growth is ranked within its own BLS instrument: CES measures payroll jobs located in selected metros, while LAUS measures employed residents and covers a broader geography. The two series are labelled and are not treated as the same instrument.

Rent growth, job growth and rent burden are winsorized at the 1st and 99th percentile before ranking. This limits the influence of a single small-sample or revised print without deleting the market or replacing its displayed source value.

Derived figures keep their arithmetic visible

A derived number is published only when its inputs exist. Core relationships include:

  • gross yield = monthly rent × 12 ÷ home value
  • rent burden = annual rent ÷ median household income
  • price-to-income = home value ÷ median household income
  • score = sum(component percentile × component weight)

The release checks that contributions add to the score, weights add to the full scale, chart totals reconcile and captions can be recomputed from their series. A failed relationship stops the release rather than appearing as a public warning badge.

How written analysis is bounded

Location analysis is generated from the verified evidence packet for that location. It may explain a trade-off or ask a next due-diligence question, but it is not allowed to widen the factual record. Numeric claims must match a supplied figure, recompute from supplied figures or be removed before publication.

Writing quality does not override evidence quality. If the facts are incomplete, the page keeps the measured modules and missing-value labels instead of substituting a confident generic article.

A current-build example

In the current build, Crawfordsville, IN has a composite score of 88. Its stored component contributions are:

  • job growth: 25.20
  • rent trend: 23.25
  • affordability: 14.55
  • supply pressure: 11.85
  • climate risk: 13.65

The example is read from the same current database as the location pages; it is not a hand-maintained illustration. Open the market page to inspect the inputs and source periods.

What the method does not know

The score does not know a building's condition, achievable unit rent, insurance quote, tax assessment, financing, renovation plan, management quality or neighborhood block. It also does not forecast appreciation or guarantee that a percentile remains stable after the next source revision.

Use the score to prioritize research. Use the page evidence to challenge the score. Use current property records, local inspection and professional advice to decide on a deal.