How listing climate is scored
People ask when a house is likely to come on the market. These files cannot answer that for a single house. They can show which streets have had more sales than the rest of their town, and which patterns in the public record go with a higher rate. The score is for a street. It counts priced sales that were recorded, and it looks ahead twelve months.
What we can count
The Westchester County GIS last-sale fields stop in 2019. NYS SalesWeb drops connections from our system, so we could not use it. The assessor's file in the county's property records system (PROS) is already used for parcel IDs. It keeps the most recent sale date and price for each parcel, and those dates run through 2026 for Scarsdale, Rye and the Town of Mamaroneck. Larchmont and the Village of Mamaroneck use the Town of Mamaroneck file.
PROS stores only the last priced sale for each parcel. A house can be on the market and never sell, and a house can sell without ever being listed. So the rate on these pages is based on closed sales: last year's recorded sales divided by the number of residential parcels. No listing dates are used.
| Town | Parcels | Transfers 12m | Transfers 24m | Rate used | How |
|---|---|---|---|---|---|
| Scarsdale | 5,288 | 186 | 383 | 3.52% | last 12 months |
| Larchmont | 1,509 | 46 | 114 | 3.05% | last 12 months |
| Rye | 4,085 | 3 | 177 | 2.17% | half of last 24 months |
| Mamaroneck | 3,285 | 104 | 217 | 3.17% | last 12 months |
| Mamaroneck Village | 1,953 | 64 | 117 | 3.28% | last 12 months |
What the model leaves out
The model does not use owner names. Age is not on the roll, and we do not guess it from a name. We do not read Enhanced STAR or other senior or veterans exemption codes, because they describe a person and this page does not forecast what a person will do. We do not use obituaries, divorce dockets or estate wording. There is no ranked list of households.
The public data file for each town lists streets only. The build works out a rate for each parcel in memory so it can add them up to an expected count for the street. Those parcel rates are not saved or published.
How long a property has been held
Years since the last priced sale is how long this property has been with its current recorded owner, as far as PROS knows. It says nothing about the age of anyone living there. On the Scarsdale file, among parcels with a last-sale date, a quarter have been held 4.63 years or less, the median is 9.18 years, and a quarter have been held 17.25 years or more. Long holds are common. That is why an 18-year hold is not scored above a 10-year hold.
The only holding period that points clearly forward is a recent one. A parcel that sold in the last two years is unlikely to sell again soon, because of transaction costs and the recent closing. After about eight years, parcels sell at roughly the town rate. Scoring a 20-year hold as "overdue" would mean making up a pattern the data does not show. The multiplier for 15 years and up is 1.0, the same as a mid-length hold.
What raises a street's rate
When a street has already had several recent sales, the remaining houses on it sell faster than the town rate. This is the biggest factor in the score.
A small street's raw rate is unreliable. Two sales on six houses works out to 33 percent, which is mostly chance. So each street's 24-month count is turned into a yearly rate and pulled toward the town rate, using a method called empirical Bayes. It treats the town rate as if ten more parcels on the street had sold at that rate. A street of six houses with two sales moves only a little. A street of twenty houses with six sales moves further. On long streets the model also looks at the eight parcels closest in house number, so a busy end of a long road is not averaged out.
Other factors
Rates by property class and by value band are worked out the same way, with a stronger pull toward the town rate (a larger prior). Condos and vacant lots do not sell as often as one-family houses, and the data is allowed to show that. A parcel whose assessment went up far more than the rest of its street gets a small increase in its rate. If it went up so much that the house was rebuilt, it gets a small decrease. If the owner name on the roll changed but no priced sale matches, that is treated as a change of title, and it moves the rate by about eight percent. For Scarsdale only, a kitchen or bath permit 6 to 18 months ago, after a hold of six years or more, raises the rate slightly. A major addition in the last three years lowers it slightly. None of these smaller factors can outweigh the street rate.
The factors are multiplied together and scaled so the average parcel rate equals the town rate. Each street is then put in a band by comparing its adjusted rate with the town rate. Quiet is below 0.75 times the town rate, typical is up to 1.25 times, elevated is up to 2 times, and concentrated is above that.
Rye's last 12 months show 3 priced sales against 177 in 24 months. A complete year should be closer to half of the two-year count, so the last 12 months on the file look incomplete. For Rye, the rate used is half of the 24-month rate (2.17%) instead of the 12-month rate (0.07%).
What a page can say
A ranking page can say that a street recorded four sales in 24 months where the town rate would predict 1.3, and that if last year's rate holds, about N of the M houses on it would sell in the next year. A property page can repeat that street sentence. No page says that a named owner is about to list.
Ardsley, Dobbs Ferry, Hastings-on-Hudson and Irvington have no PROS last-sale file, so they are listed as unavailable. We do not fill the gap with owner names or tax bills.
County listing climate · Scarsdale streets · How Dale Ledger is built
Unofficial. Confirm the deed on PROS or with the County Clerk. This is not a listing service, and it is not real-estate advice.
FAQ
Is this a probability that a house will list?
No. It is a street-level rate of recorded priced sales, adjusted toward the town rate. A sale is not a listing.
Why not score long-held houses as more likely to sell?
Many houses have gone 15 or 20 years without a sale, which says they do not tend to sell. Treating them as due is the same mistake as treating an older owner as about to list.
Why use empirical Bayes?
A raw street rate on a handful of parcels swings on one sale. Pulling it toward the town rate keeps a short street from looking busier than it is.
Unofficial. Confirm figures on the town’s official roll or in PROS, the county’s property records system, before you act on them.