I think you should always avoid using a single property when trying to go through these things and that was my point in the first place. I’m sure both the specific 2000 property and the 2001 have their considerations and I wouldn’t latch onto either one. That’s why I offered averages for 2000 (again, what’s supposed to be the working year for the exercise). I ran it for 2001 and here’s what I’ve got:

Year Sales Median Average
2000 30 $267,250 $285,800
2001 18 $322,750 $331,361

My data source is Redfin and as I mentioned before, at least one of the shortcomings is that it only provides information about the last sale of any given property. This will exclude properties that have sold since. One thing that’s really nice is that Redfin let’s you export the data, so it becomes easy to slice and dice it. That makes it easy to come up with metrics (like the ones above) quickly, but it’s important to realize that it’s also not the same thing as detailed property analysis (which is kind of in between the picking one or two houses, i.e. model matches and specific properties that are still pretty close). The latter could, of course, be used to achieve more tightly tuned (i.e. accurate) numbers.

You should check it out, it’s kind of fun. What I do is pick a zip, set basic criteria for maybe square footage and if I’m trying to weed out newer stuff, maybe a lower maximum price. Then I see what I get, zoom in and start exporting data. For the numbers above I got back 492 results and from there I could drill into years. There are flaws with the method to be sure, but seems like a reasonable proxy for tract homes.