How the comparable market analysis works

THE METHOD

How the comparable sales are chosen

One rule proposes the sales for every home. A visitor may choose among them, and a request then shows Julian both lists: the rule's and the visitor's.

The same kind of property

A single-family home is compared with single-family homes, a condominium with condominiums, a townhouse with townhouses.

The same water

A home on the water is compared with homes on the water, and a home off it with homes off it.

A similar size

A living area within 10% of your home's, larger or smaller. Only where fewer than 3 sales are found is the band widened, to 15%, then 20%, then 25%, and an analysis that states a range states the step it used.

A similar age

Built within 10 years of your home's year, earlier or later. The year leads: it is held at 10 years while the size widens, and only once the size stands at 25% are the years widened, to 15, then 20, then 30, and never past 30. A home built in 1950 is not read against new construction. Where a home was rebuilt or renovated whole, its owner can state that year: the sales are then compared by it, and it is marked as the owner's word. A sale's year is the one the MLS records, and the page does not know which sales were renovated.

Recent

Closed in the last twelve months.

Near

No farther than 1 mile, the nearest first.

Up to 12 shown, 6 read, never fewer than 3

Up to 12 sales are shown. The 6 nearest are checked: they are the rule's. You can take a check off or check another, and the range is read from the checked ones alone, 3 at least. With fewer than 3 sales found the page states no range. After the step the rule stopped at, the sales that stand one step further are shown after its own, unchecked, so a home read from few sales has others to add: none of them is read unless you check it, and a request says which ones you did.

How the range is computed

The range is arithmetic. A house in one of the site's markets is read from the 15 sales of its own market most like it, each brought to the home's size and lot by what those differences have been worth in that market's own sales. The page says the figure they point to and how sure that reading is, from how many sales there are and how close together they closed. Where the reading is high, it states the range that 7 homes in 10 like it sold within, as measured on 703 houses on October 29, 2026; otherwise it states the figure alone, and Julian refines it. A sale far from the others is shown unchecked. Every other home is read as follows. Each checked sale is compared by its size: what was paid for each square foot (Sq.Ft) of the home sold. The lowest and the highest of those figures are multiplied by your home's living area, and rounded to five, ten or fifty thousand dollars: the low end down, the high end up. Where the home or the sales state no living area, it is what the sales closed for. For those homes every checked sale counts: the page sets none aside for being high or low, and one that closed far from the rest is marked, not removed. No language model writes a figure on this site.

What the sales point to

Beside the range the page says one figure: what the sales most like your home point to. Each sale counts more the more it looks like your home: the same size, the same age, closer, more recent. The figure is the middle of those sales by what was paid for each Sq.Ft, the ones most like yours weighing most, applied to your home's size. It is a reading of the sales, not a price for your home. When 5 sales or more sit close together, the page says they agree; when they are few or far apart, it says so and asks you to take the figure as a first idea.

How well it reads, measured

Tested on 976 houses and 2,451 condos sold in 2026 in the site's markets, each read only from the sales closed before it: for houses, the figure stood within a tenth of the later sale price for 31% of them and within 17% for half; for condos, 40% and 13%. Read on October 13, 2026. That is why a person has to look.

What your answers about the home do

Two answers describe the home: how its kitchen and bathrooms are finished, and how it has been kept. They move no figure. The sales cannot show a home's finish or its upkeep, and the page has no measured rule to put a price on them, so the answers go to Julian with the analysis, and he weighs them when he reviews it. The answers are the visitor's own: nobody has seen the home. A last question asks when the owner plans to sell: it moves no figure either.

What the page says of each sale

Beside the range, the page sets each sale against your home: its living area, its lot, the year it was built, its bedrooms, its bathrooms and how long ago it closed, as the public records state them. It puts no dollar amount on a difference: what a difference is worth is the part of an analysis a person does. It marks a sale with Unusual price — sold for a price very unlike the rest for its size — where what was paid for each Sq.Ft is under half the median of the sales shown (the median is the middle price: half the sales paid more for each Sq.Ft, half less), or more than two standard deviations from that median, a measure of how far the sales spread, where leaving it out would move the lowest or the highest of the sales by more than a tenth. A marked sale is still read unless you take its check off or your answers leave the other half of the sales. The page does not compare a pool, a view, the condition or the finishes: the records it reads do not state them reliably.

Where the data comes from

The sales are closed sales of the MIAMI REALTORS® MLS, from the copy this site refreshes each night. A home's facts are Miami-Dade County's public property record; the site asks that record nothing about who owns a home.