Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
So finally this week Zillow admits that Zestimates are inaccurate as they launch a $1.2 million competition to see who will come up with a better AVM algorithm in the next six months. But as is usual with Zillow its all about self promotion and yet another gimmick to attract users to its website as they discover that constant TV advertising to attract audience is expensive and soon as the TV Ads end the audience plummets. When they paid Stan Humphries,the Zestimate algorithm architect $7.7m in 2015 its loose change for Zillow to try and find someone who can do the job better.
Zillow announced Wednesday that it was launching a competition aimed at awarding $1 million to the first person or team who could most improve the Zestimate algorithm. “Zillow Prize” hopes to attract data scientists everywhere to try their hand at tinkering with “one of the highest-profile, most accurate and sophisticated examples of machine learning.”
OK just fine tuning a bit... divide by the angle of the front most roofline; add .08 x height of all doors (minus .0036 for each brass knob); subtract the speed limit of nearest roads surrounding the property; minus .467 per insect outside the structure, minus 1.23 per insect inside; add/subtract randomly generated small factor; add 1 if the pictures are pretty; plus 400 if the homeowner is smoking hot; [call an appraiser]
So finally this week Zillow admits that Zestimates are inaccurate as they launch a $1.2 million competition to see who will come up with a better AVM algorithm in the next six months. But as is usual with Zillow its all about self promotion and yet another gimmick to attract users to its website as they discover that constant TV advertising to attract audience is expensive and soon as the TV Ads end the audience plummets. When they paid Stan Humphries,the Zestimate algorithm architect $7.7m in 2015 its loose change for Zillow to try and find someone who can do the job better.
Zillow announced Wednesday that it was launching a competition aimed at awarding $1 million to the first person or team who could most improve the Zestimate algorithm. “Zillow Prize” hopes to attract data scientists everywhere to try their hand at tinkering with “one of the highest-profile, most accurate and sophisticated examples of machine learning.”
What a strange tone by the OP. Are you hinting that anytime a person or a company attempts to improve something, that it is an admission that what they already had was "crap"?
Imagine where we would be as a society if everyone in history took that approach to improving on things. It wouldn't be pretty- very little innovation and very little change.
What a strange tone by the OP. Are you hinting that anytime a person or a company attempts to improve something, that it is an admission that what they already had was "crap"?
Imagine where we would be as a society if everyone in history took that approach to improving on things. It wouldn't be pretty- very little innovation and very little change.
Oh, I think Surbiton is most definitely not "hinting." I do think he is "grinning," though.
He has well-documented many issues with Z.
I would assume that he would actually applaud improving on a great many things, and there is common and routine improvement on a great many things that don't suck like Zestimates, "Estimates of market value," do.
Value is between the seller and buyer and nobody can do better.... Zestimate is the second best thing because nobody has anything better over a wide swath of the country... if you don't like something then make something better otherwise Zestimate is the best thing available...
No algorithm will tell you whether there's a gas station next door or a killer ocean view, both of which will significantly affect value. Lenders have been trying to eliminate us appraisers with AVMs for years, but it won't happen. An AVM can give you a ballpark figure, but that only works for a low LTV loan. For an 80/20 loan you'd better use an appraiser. As for Zillow, it's just a parlor game that people like to play.
Value is between the seller and buyer and nobody can do better.... Zestimate is the second best thing because nobody has anything better over a wide swath of the country... if you don't like something then make something better otherwise Zestimate is the best thing available...
Well... For $1 million, I offer instant improvement over current AVMs, all of 'em.
Make all commercial ad sites and real estate firm site AVMs be required to allow property owners to OPT IN.
I can't see there ever being an algorithm that will be perfect. Every market is different, right down to the neighborhoods.. then you have house quality issues, what needs fixing, etc. Best to just leave the field blank and let homeowners or local people fill in the numbers based on local market data. I pretty much know what my house would sell for after 3 years searching the local real estate market on an off. The realtor was nice enough to send me a letter telling me the house across the street sold for $160K after being listed only 2 days.. based on that and the condition of that house I can further narrow down what mine would sell for.. I bought it at auction for $120K in October and it would sell for $175K right now. I lucked out and got 2010 prices if not better for it.
My zestimate says $156,016. Houses have sold above city appraisal price for years around here because of tight supply.
Please register to post and access all features of our very popular forum. It is free and quick. Over $68,000 in prizes has already been given out to active posters on our forum. Additional giveaways are planned.
Detailed information about all U.S. cities, counties, and zip codes on our site: City-data.com.
Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
