Why Do Mar Vista Pages Show Such Different Medians? Reading Small-Lot Homes, the Hilltop and the Census Building Types

Why do the Mar Vista medians differ so much?

Start with the geography. The brokerage tracker says it cut its figures to the postal area, not the neighborhood. Redfin uses its own neighborhood boundary. A brokerage page that quotes a median near $1.84 million says it is for the postal area as well, and its figure, down about 9%, is almost the same as Redfin's $1.8 million, down 9.06%. They may share a source. The postal area in the Census data has 27,045 homes and 56,740 residents, which is far more than a single neighborhood, so none of these pages is a clean count of one place.

Then the windows. Redfin's $1.8 million covers the three months ending June, although its page title says July. The tracker's $2,265,000 covers the 90 days to early October, and its twelve-month median of $2,017,610 covers a year. The brokerage articles describe winter, spring and March. A 25.8% gap between $1.8 million and $2,265,000, computed here, is much larger than the gap any one-quarter shift in the market would produce. If the Redfin figure is right for the spring and the tracker is right for the summer, the median rose between them, which the tracker says it did, up 10.9% over the prior 90 days. The pages do not settle it.

The movements disagree in sign. Redfin has the median down 9.06% over a year. The tracker has it up 10.9% for 90 days and up 3.9% for twelve months, and the article's March median is up 8.2%. Zillow's value estimate is up 5.7% on its page and 6.1% in the article. A 9.06% drop to $1.8 million implies about $1.98 million a year earlier, and a 10.9% rise to $2,265,000 implies about $2.04 million, both computed here, so the two pages put the year-earlier median within about 3% of each other and disagree only about where the median is now.

The counts raise another question. Redfin shows 89 homes sold in June, up from 78. The tracker shows 89 closings in the last 90 days, which is about 30 a month, and 342 in twelve months, about 28 a month. Redfin's single-month figure is three times the tracker's monthly pace, computed here. The article's March count is 47, down from 78 a year earlier, so the figure 78 appears as the prior-year count for both March and June. One of them may be mislabeled, and this brief cannot tell which. Counts this different mean the pages count different sets of sales.

What are small-lot homes doing to the median?

One article builds its whole argument on them. It describes a repeating pattern: a single 1940s house or a 1950s fourplex is torn down and the lot is split into five, six or ten parcels, each with a new home of roughly 1,443 to 2,200 square feet. It says the pattern has run on the same few blocks for over a decade, and that recent state housing law makes this kind of subdivision easier on a standard lot. It lists several projects by street and developer. This brief does not repeat those names or addresses, and has not checked any of the filings.

The article's reasoning is that five new homes from one parcel are worth more together than one old house, and that each new home enters the sales record at new-construction prices. A median across a quarter's closings then rises or falls depending on how many of those new homes closed, with no individual house changing in value. The same article says price per square foot rose 11.5% in March against 8.2% for the median, and reads the gap as partly new construction. That is a reasonable inference, labeled as an inference here, and it cannot be tested without the sales.

A second article reads the opposite effect in spring. It says the median fell to about $1.9 million because smaller homes, fixers and entry-level product traded, while the average home that sold was bigger, about 2,250 square feet against 2,150 a year earlier, on a smaller median lot of about 5,750 square feet against 6,100. It also says closed dollar volume rose slightly, to about $226 million from $224 million, which is 0.9% more, computed here. Its two figures for the winter-to-spring drop are $260,000 in the text and $300,000 by subtracting its rounded medians, so the text figure probably uses unrounded values. Both articles are one-sided in the same way: each explains a number by a change in mix, and neither shows the mix.

The first article also names two separate stories inside the neighborhood. It says a turnkey three-bedroom postwar bungalow in Westdale trades between $1.9 million and $2.4 million, and that entry-level homes on the Hilltop start near $3.2 million, on lots that average about 7,500 square feet, a quarter larger than a typical Westside parcel. Those are one writer's ranges. A seller can ask for closed sales on the same kind of lot, built the same way, from the last six months.

How fast do Mar Vista homes sell, and what do listings show?

The clocks run from 9 to 42 days. The June article gives a median of 9 days on market in winter and 15 in spring, with 40-plus active listings, about 10 homes under contract at a median near $4.1 million and 6 pending near $1.8 million. A Zillow page, updated August 31, says homes go pending in about 31 days (Zillow, Mar Vista home values). The tracker gives 17 days to go under contract over twelve months. Redfin gives 42 days, up from 35, and a brokerage page repeats 42. The spread of 9 to 42 is a factor of 4.7, computed here, and it is the widest spread in this brief.

The supply counts differ too. A brokerage page of September 18 says there are about 88 active listings, roughly three months of supply, balanced and tilting slightly toward sellers (Local brokerage page, Mar Vista housing market). The June article counted 40-plus. The September page says 36% of homes sold above asking and the average sale-to-list ratio is about 99.5%. The June article says 105% in spring, and the tracker says 100.0% over twelve months. Three pages, three ratios, and none of them says whether it counts houses only or condos and townhomes too, which the September page says are all in its 88.

The September page also says the median briefly peaked above $2.13 million in late 2025 and gives about $1,140 per square foot. The tracker gives $1,091 for twelve months, and Redfin gives $1,160. The June article says $1,102 and $1,101 for two six-month windows. The four figures sit between $1,091 and $1,160, a spread of 6.3%, computed here. That is far narrower than the price spread, which supports the mix-of-homes reading, but it does not prove it.

The tracker adds a note that matters to sellers. Its lowest sale in twelve months closed at $395,000 and its highest at $10,000,000. A range of 25 times between the lowest and highest sale means no single median can stand in for a specific house. The tracker's twelve-month median is up 3.9% on a count of 342 sales, 9.0% fewer than the 376 sales of the year before, computed from its figures.

Taken together, the clock figures say less about demand than about what each page counts. A median of 9 days in winter is the pace of the few homes that sold quickly, and a median of 42 days is the pace of a longer list that includes homes that sat. Redfin's own page shows the median on market rising from 35 to 42 days while its count of sales rose from 78 to 89, so more homes closed more slowly. A seller who sees only the 9-day figure may expect an offer in a week, and a seller who sees only the 42-day figure may expect a long wait, and neither page describes a specific house on a specific street. The seller's own home, priced against its closest recent sales, is the only clock that matters.

Sources as cited. The windows, areas and measures differ and the figures are not directly comparable. Commercial content; not independently verified.

What does the Census say about Mar Vista homes, renters and vacancies?

The Census postal area that includes Mar Vista has 27,045 housing units. Detached houses number 10,651, 39.4%, and attached houses 743, 2.7%. Buildings of five or more units hold 13,940 homes, 51.5%, split between 4,139 in buildings of five to nine units, 4,104 in 10 to 19, 2,437 in 20 to 49 and 3,260 in 50 or more, 12.1%. Buildings of two to four units hold 1,687, 6.2%. Small lots with new homes are one part of this stock, and the apartment buildings are a larger part of it, which is a reason the postal area's numbers do not describe the single-family streets alone. The median year built is 1967, and 71.9% of homes were built before 1980.

Tenure runs the other way from most suburbs. Of 25,279 occupied homes, 9,680 are owned, 38.3%, and 15,599 are rented, 61.7%. Among owners, 4,002 moved in since 2010, 41.3%, 2,027 moved in during the 2000s, 20.9%, and 3,651 moved in before 2000, 37.7%. Another 2,349 owners, 24.3%, have been in the same home since 1989 or earlier. Among renters, 6,467 moved in since 2020, 41.5%, and 12,592 since 2010, 80.7%. Owners who moved in decades ago hold a large share of the single-family stock, and renters turn over quickly.

Vacancy is 6.5%. The Census counts 1,766 vacant homes: for rent, 847, 48.0%, rented but not yet occupied, 126, 7.1%, for sale, 98, 5.5%, seasonal or occasional use, 241, 13.6%, and other vacant, 454, 25.7%. Only 98 homes were vacant and for sale, in a postal area of 27,045 homes, 0.4%, computed here. The figures are five-year averages and do not measure how many homes sit empty at any moment, but they show that the homes visible on listing sites are a small slice of the stock.

The Census median owner value is $1,695,400, with a margin of error of $56,395, 3.3% of the value. It is owners' own estimate over five years, not a sale price. It is 92.2% of the Zillow value of $1,838,873, 84.0% of the tracker's twelve-month median and 74.9% of its 90-day median, all computed here. The median household income is $114,141, so the Zillow value is 16.1 times the median income, and the median rent of $2,279 a month is $27,348 a year, 24.0% of that income, a rough ratio of two medians. Readers comparing markets can also read the Menlo Park brief and the Burlingame brief.

Source: Census Reporter, American Community Survey 2020-2024 five-year estimates, tables B25024, B25003, B25038, B25004, B25034, B25077, B25064 and B19013. Percentages and ratios are this brief's arithmetic.

What should an owner ask, and what does a private sale change?

The first question is which Mar Vista the home is in. A postwar bungalow in the flats, a house on the Hilltop and a new small-lot home do not share a median, a buyer or a clock, according to the articles. A seller should ask for closed sales on the same kind of lot, from the last six months, with the days to contract, the gap between the first ask and the final price, and whether the comparable sales include new construction.

The second question is what the number measures. Ask whether a figure is a postal-area figure or a neighborhood figure, a median or an average, houses only or houses and condos, and what window it covers. The pages above differ on every one of those points, and a seller who anchors to the wrong page can price a home 25% off the market in either direction.

The third question is the commission, and the arithmetic is simple. Each 1% of $1,838,873 is $18,389. At 3% the figure is $55,166 and at 5% it is $91,944. At $2,017,610, each 1% is $20,176, 3% is $60,528 and 5% is $100,881. At $2,265,000, each 1% is $22,650, 3% is $67,950 and 5% is $113,250. These illustrate the size of each percentage point and are not quotes of any rate a seller would be charged. No page in this brief names a Mar Vista commission rate, and the medians are not the price of any one home.

The fourth question is exposure. In a neighborhood where one article says a new set of homes can come to market a block away at a higher price per foot, a public listing puts an older home beside new construction in every buyer's search. A private sale has no showings and no neighbors talking about a sale.

The benefits are plain. Maison Off-Market buys luxury homes and estates directly from their owners and speaks only to sellers. A private sale has no showings. It allows closing dates that give a seller time to find a new home. It has no commission costs and no closing costs, and it avoids inspection repairs, which matters for a postal area where 71.9% of homes predate 1980, though the Census says nothing about the condition of any one house. The company does not say that a private sale always beats a public listing, and in a neighborhood where one page shows $1.8 million and another $2.3 million, an owner is right to ask what each route yields. If you would like a private, no-obligation offer for a Mar Vista home, call 401-219-4207 or use the contact form on this site.

Methodology and limitations

The Redfin page is a commercial page whose title says July and whose data window ends in June. The brokerage tracker says its figures come from the regional listing service and are cut to the postal area. The brokerage articles are marketing content that quotes trackers and local reporting without links to the underlying sales, and neither the small-lot filings nor the Hilltop and Westdale price ranges were checked. A brokerage page repeats a figure close to Redfin's, so the two may not be independent. Counts of sales do not agree across pages and one prior-year count appears twice with different labels. The Census figures are five-year estimates for a postal area, not the neighborhood. Percent changes were taken as printed, and implied prior-year values are this brief's arithmetic.

Conclusion

The Mar Vista record shows medians from $1.8 million to $2.3 million, a clock from 9 to 42 days, new small-lot homes mixed into the sales and a postal area where 51.5% of homes are in larger buildings. The pages differ because they cover different geography, windows and mixes of homes, and the useful figure for an owner is the one for a comparable home on a comparable lot.