Every AevoraeX score opens into the same anatomy: weighted components, expandable drivers, ordered adjustments, a confidence shrink, city percentile, tier, caveats and a plain-English rationale.
Components and drivers
Each component has a weight and a sub-score. Expand it to see the engine fields behind it, such as days on market against the city median, price-cut velocity, valuation spread, estimated rent, mobility or property fit.
Adjustments and confidence
Gates, multipliers, caps, bonuses and penalties run after the weighted blend, in the order shown. Confidence then shrinks an uncertain result toward 50. The engine tests that recomposing this anatomy reproduces the displayed score within 0.01.
Percentile and tier
Public pages show rank and tier rather than the raw model output. The colour answers one question: where the property ranks inside its city, for example #1 of 2,567 in Orlando, top tier. Top tier means percentile 95 or higher, strong means 80 or higher, and the rest stay grey.
Why it matters when you investigate
The rank prioritises the property. The drivers explain why: for example, long days on market against the city median and recent price cuts. Open the Redfin source and verify those facts before outreach.
What the score is not
It is not a guarantee of a deal. Scores prioritise research and never use protected characteristics. The engine does not currently have owner data, deeds, tax delinquency, permits, liens, code violations, court or probate records, municipal STR rules, or ADR and occupancy data. The full catalogue is on the methodology page.