Very nice link. That example is in perfect agreement with what I have witnessed in my neighborhood (see a few posts above). Precisely the same problem. Two houses that are identical, were valued nearly the same a year or two ago, and are apart by more than $200K now.

I offered an explanation for the discrepancy (maybe Zillow 2.0 gives too much weight to prices of nearby houses). That’s really the only way I can wrap my mind around this. Everything else (lot sizes, year built, etc) is the same and can’t account for the difference.

Yes, Zillow’s got a bit of a problem. To be fair to them, what they want to do (produce an accurate estimate on each and every house) is very hard work. Close to impossible. In contrast, the Case-Shiller algorithm is a piece of cake. They average over an entire metropolitan area, kinda hard to get that wrong.