3.6 Crunch Time

[TerryRohrer]

That's because the AVM doesn't give a fixed point of value just a range with a probable point include the deviation range.

On Large cookie cutter tract developments like we have out here it's often a very tight 2%-5% range which is less than appraisers net and gross adjustments and there is no need for an appraisal.

That is baloney. You can't develop a "median error rate" without establishing a "median" which is, by definition, a point. The "range" is that half the results are above that median and half are below that median.

 

[TerryRohrer]

How accurate are automated valuation models in real estate?​

AVM accuracy varies by property, market, data freshness, and use case. Current vendor figures show that listed homes can receive median errors below 2 percent, while off-market homes may be closer to 7 percent. The difference exists because an active listing provides newer and richer property information.

Redfin currently reports a median error rate of 1.85 percent for on-market homes and 7.24 percent for off-market homes. It says the model uses more than 500 data points across about 92 million US homes. These vendor figures are product claims, not universal AVM benchmarks.

What does a 7.24 percent median error mean? It means half of the relevant estimates were closer than 7.24 percent to the final sale price, while half were farther away. It does not mean every property estimate falls inside that range.

For a $400,000 property, 7 percent equals $28,000. For an $800,000 property, it equals $56,000. An automated valuation model real estate user should examine both the percentage and monetary consequence.

Automated Valuation Models (AVMs) in Real Estate: How They Work and How Accurate They Really Are

Learn how an automated valuation model in real estate works, what accuracy rates mean, where AVMs fail, and how to use them confidently. Read the guide.

noseberrydigitals.com

 

[TerryRohrer]

Traditional AVM testing compares model estimates to sale prices after those prices—and the listing data that precedes them—are already available to the models. As we’ve documented extensively, many AVMs incorporate MLS listing prices into their estimates, a phenomenon known as listing price anchoring, which inflates apparent accuracy in ways that don’t reflect real-world performance. When a model knows the listing price before producing its estimate, the resulting test measures something closer to consistency with known information than true predictive accuracy.

The anchoring effect isn’t theoretical—it’s quantified. Zillow reports a median error of approximately 1.7% for on-market homes but 7.2% for off-market homes. Redfin reports nearly identical figures: roughly 2.0% on-market and 7.6% off-market. That gap—a three- to four-fold difference in accuracy—represents the influence of listing price anchoring on AVM performance. Independent analysis by the AEI Housing Center corroborates this: their 2024 evaluation of five AVM providers found that “springiness”—how much a model’s estimate jumps when listing or sale prices become available—showed the widest variance in scores of any criterion tested, with some providers receiving failing grades. Since the vast majority of AVM use cases involve unlisted properties—refinances, home equity lending, portfolio assessments, loss mitigation—testing that includes listing data paints a misleading picture of how models will actually perform when it counts.

Blog - AVMetrics

Use this static Page to test the Theme’s handling of the Blog Posts Index page. If the site is set to display a static Page on the Front Page, and this Page is set to display the Blog Posts Index, then this text should not appear.

www.avmetrics.net

 

[TerryRohrer]

Another viewpoint:

"Translation The term “median” splits the results in half so half the properties analyzed could be off by 500% and the other half could be off by 3.2% making the median accuracy rate 3.2%. The consumer reads this as accurate within 3.2% since they don’t understand the application of the word “median.” If I was appraising a home and told the client, half the time I am usually within 3.2% but the other half of the time I am off an average of 500%, I would never be taken seriously. The Zestimate is literally so bad that when the AVM punches out what ends up being accurate results, it comes down to luck."

Bank AVMs Are As Wildly Inaccurate As A Zestimate – But Will Be Regulated As Legitimate Values

AVMs are incredibly inaccurate and are being misused in property valuation

housingnotes.com

 

[OSU Beavers]

AVMs from public and Realtor data ARE a disaster. Many Oregon tax assessment offices have stripped out the interior information from all home records because it was causing calculation errors with their new software. They no longer factor HVAC, bathroom count, materials or extras like fireplaces or spa tubs in their market value. Just quality class times sqft minus depreciation.

Then you have Realtors that will list a split level home as 1000sqft upper (second) level and 1000sqft basement, leaving the main level entry blank.

My favorite was a house with 1000sqft in the main level with a 1000sqft basement that was half finished. Reported in the MLS seveal times over the years as 1000sqft main & 1000sqft finished basement, 750sqft main with 750sqft finished basement, and 1500sqft main & 1500sqft finished basement!

 

[J Grant]

The accuracy of AVM's is measured by how close or far the estimate was from sale prices or a sale price.

The accuracy of an appraisal is measured by how credibly it developed and supported a market value opinion.

 

[Airstream]

If AVMs use public data, good luck. One of the reasons I refuse to do drive by appraisals. Data stinks.

 

[Steelers74]

Realtor data is going to really screw up AI and the AVMs around here. Our agents routinely include enclosed porches in the GLA as well as finished basements, not the bi level, tri level, type, the typical finished basement, that's if it really is finished, to some a rug on the floor and painted joists makes it a rec room.

 

[Sandra Koutsopoulos]

Realtor data is going to really screw up AI and the AVMs around here. Our agents routinely include enclosed porches in the GLA as well as finished basements, not the bi level, tri level, type, the typical finished basement, that's if it really is finished, to some a rug on the floor and painted joists makes it a rec room.

Same here... probably same everywhere. They fudge the numbers to make the property look better than in reality. I've even had a Realtor who listed a house, when I measured the modest house short by 400sf, then had the SELLER measure it with me on the second trip out there, deny reality and insist to the Buyer that there was an extra 400 sf in that house. Musta been invisible, or existed in a parallel universe.

A few months later I did an appraisal on a property listed by someone else, and that goofy gal had the buyer. Listing agent mentioned her uneasy feeling about what the agent stated as fact regarding the Buyer's situation. I told her of my experience with that other agent, and Lo and Behold! the "facts" were fiction and the deal fell apart. So much for Realtors being honest and believable! I always felt that if you had to create a Code of Ethics, it was because too many times ppl got ripped off and burned, and the the organization or government or whatever decided to make their members swear to do the right thing to allay public concerns. ...but still, not everybody is honest and transparent.

 

[Airstream]

Same here... probably same everywhere. They fudge the numbers to make the property look better than in reality. I've even had a Realtor who listed a house, when I measured the modest house short by 400sf, then had the SELLER measure it with me on the second trip out there, deny reality and insist to the Buyer that there was an extra 400 sf in that house. Musta been invisible, or existed in a parallel universe.

A few months later I did an appraisal on a property listed by someone else, and that goofy gal had the buyer. Listing agent mentioned her uneasy feeling about what the agent stated as fact regarding the Buyer's situation. I told her of my experience with that other agent, and Lo and Behold! the "facts" were fiction and the deal fell apart. So much for Realtors being honest and believable! I always felt that if you had to create a Code of Ethics, it was because too many times ppl got ripped off and burned, and the the organization or government or whatever decided to make their members swear to do the right thing to allay public concerns. ...but still, not everybody is honest and transparent.

MLS data: 3600sf. My measurement 2800sf.
MLS data: 10,000sf My measurement 5200sf.
MLS data: 4 car garage. My data: 2 car garage.
MLS data: built in 1970. My data: built in 1920.

I came in 30,000 short of the sale on one occasion. Agent called me and asked, and I said the sf was off by 440 sf. They even had the plans. But they included the garage in the GLA.. I’ve had this happen numerous times.

We were going to buy a house when we moved. Agent had 1700sf. I didn’t think it was right, and I checked, and sure enough off by 200sf. I withdrew the offer. And then she tried to tell me some nonsense excuse.