How this works
The estimate starts with what actually happened nearby: the share of homes of the same type that sold above asking in your zip code over the last two periods of Redfin data. It then adjusts that starting point for the things that move a listing’s odds, and turns the total back into a percentage.
The adjustments, each from the model file rather than the code: price per square foot against the local going rate for homes of that size, where the price sits against the median sale price, months of supply (capped at 6 months over normal: past that a glut is a glut), the recent trend in sale-to-list ratios, the month you list, the property type, the 30-day change in the 30-year mortgage rate, and a penalty if the home has already sat unsold with no offers.
Price is the lever, and it works differently in each direction. Overpricing costs about 6 times as much per point as modest underpricing earns, up to a cap. Underpricing does something else as it deepens: the local market’s verdict fades. A base rate describes homes priced normally (Redfin’s series excludes sales far from list), and a home priced well under the going rate is bought on arithmetic rather than mood, because buyers who would pay market will pay less. By 27% under, the local verdict no longer matters and the estimate reaches the ceiling in any market. Studies of asking prices point the same way: a lower asking price draws more visitors and more offers (Han and Strange, Journal of Urban Economics, 2016; Anundsen, Røed Larsen and Sommervoll, 2018), though how far the effect goes in each market is exactly what the fitting step will measure.
Where the numbers come from. Market data: the Redfin Data Center monthly market tracker, imported monthly. Mortgage rates: Freddie Mac’s Primary Mortgage Market Survey, via FRED. Model version 1.0.2-defaults.
Estimates are capped at 3% and 97%. The coefficients are published at /data/odds-model.json.
Questions
What are the chances of getting multiple offers?
It depends on the local market and the price. This tool estimates it from the share of nearby homes that recently sold above asking, then adjusts for your price per square foot, months of supply, the season and mortgage rates.
How accurate is the estimate?
It is a structured estimate from market data, not a prediction for a specific home. Condition, photos, marketing and timing all matter and none of them are in the data. The methodology section says exactly what goes into it.
What price would give the best chance of multiple offers?
The tool solves for the list price that reaches the target estimate and shows it next to your price, so you can see what the trade-off costs.
Where does the data come from?
Redfin’s public Data Center for market data, updated monthly, and Freddie Mac’s weekly mortgage rate survey via FRED.