AevoraeXAevoraeXSign inStart free trial

Every score opens into its anatomy.

See what shaped the score, what changed it and how much confidence to place in it. Every ranking comes with components, adjustments and caveats you can inspect.

Explore the componentsHow scoring works

The published score is the computation.

The engine blends each component as weight times sub-score, applies gates and adjustments in order, then shrinks the result toward 50 according to confidence. The anatomy you open in the product is that same breakdown, and the engine test requires it to recompose to the score.

  1. Weighted components

    Each lens has components, such as price-cut pressure or all-in cap rate. A component has a sub-score and a weight, and the blend is weight times sub-score, summed.

  2. Ordered adjustments

    Gates, multipliers, caps, bonuses and penalties are applied in a fixed order. An adjustment that did not apply is still listed, marked as not applied.

  3. Confidence shrink

    The last step pulls uncertain results toward 50 according to the confidence of the inputs.

  4. City percentile

    Percentiles are computed against the complete city pool for that lens, not against the rows visible in a browser. Top is percentile 95 or higher; Strong is 80 or higher.

  5. Caveats and rationale

    Caveats travel with the score, so missing data is visible instead of silently treated as evidence. A plain-English rationale says why the property ranks where it does.

Colour is rank, not the raw score.

Top 5% · Top tierTop 20% · Strong tier

Green is reserved for the top 5% of a city for a lens; amber is the top 20%. The raw score never chooses the colour, and public pages show rank and tier only.

The engine rejects a breakdown when recomposition differs from its score by more than 0.01. Version, city pool and confidence travel with every anatomy.

Five component catalogues.

These are the signal families each lens reads from the listing. Labels may be refined when the engine breakdown documents are published.

Motivated seller

The motivated seller lens

Labels Labels may be refined when the engine breakdown documents are published. Each score shows its own weights.

Motivated seller component families
ComponentDrivers
Distress flags from listing textForeclosure, REO, short sale, auction · Estate wording · As-is, vacant, tenant-occupied
Price-cut pressureCut velocity · Cut depth · Cut recency
Listing historyRelists · Withdrawn history
Time on marketDays on market · City median
Price versus derived valueList price vs derived value · Implied equity · Years since last sale
Description keywordsMatched phrases

Fix & flip

The fix & flip lens
Fix & flip component families
ComponentDrivers
MAO gapMaximum allowable offer · List price
Levered returnProjected ROI · Rehab range
Seller pressurePrice cuts · Days on market
Insurance and exit penaltiesClimate exposure · Exit conditions
Valuation confidenceSold comps · Model confidence

Buy & hold

The buy & hold lens
Buy & hold component families
ComponentDrivers
All-in cap rateNOI · All-in basis
Yield versus the cityEstimated yield · City yield
AppreciationArea price trend
Schools and amenitiesSchools · Walk score
HOA and debt gatesMonthly HOA · Debt coverage
Florida condo riskBuilding age · Assessment exposure

Mid-term rental (30+ days)

The mid-term rental (30+ days) lens
Mid-term rental (30+ days) component families
ComponentDrivers
Location fitHospitals and employers · Mobility
Seasonal economicsMonthly rent estimate · Seasonal adjustment
Suitability for 30+ daysFurnished · Pool · Lease minimum

Airbnb (nightly)

The Airbnb (nightly) lens
Airbnb (nightly) component families
ComponentDrivers
Location and points of interestNearby attractions · Mobility
Proxy economics (labelled proxy)Rent-based proxy · Seasonality proxy
Property fitGuest-ready features · AI photo summary
Restriction gatesHOA restriction · Known listing restriction

What we do not have.

Listings come from Redfin (MLS data as distributed via Redfin). These fields are not in the engine today and never appear as current evidence in the product.

Owner data, deeds or liens

Not available

Tax delinquency

Not available

Permits or code violations

Not available

Court or probate records

Not available

Municipal short-term-rental rules

Not available

Nightly-rate (ADR) or occupancy data

Not available

Model versions.

Each anatomy names the model version that produced it. A public version history will be published here.

Open the anatomy on your own city.

7-day free trial, no card required.

7-day free trial, no card required.See the demo

Questions about the method.

Does the score use a market baseline?

No. It is a weighted blend of component sub-scores, followed by ordered gates, multipliers, caps, bonuses or penalties, and a confidence shrink toward 50.

What makes a listing worth a call?

Its percentile inside the city pool. Top means percentile 95 or higher; Strong means 80 or higher. The raw score does not drive the colour.

Can two people reproduce the score?

Yes. With the same breakdown version and inputs, the engine's recompose test must return the published score within 0.01.

What data is missing today?

Owner data, deeds, tax delinquency, permits, liens, code violations, court and probate records, municipal short-term-rental rules, and nightly-rate and occupancy data are not available today.

Keep reading.