3 ms·
Pretty cool visualizations, but I would take the accuracy of the data with a grain of salt until I saw the source data. There's no universal map of neighborhood
by rmxt 12y ago
Pretty cool visualizations, but I would take the accuracy of the data with a grain of salt until I saw the source data. There's no universal map of neighborhoods in NYC, and I think that every New Yorker has a different sense of where certain neighborhoods start and begin. I'd also question incentive alignment regarding the real estate sites they used to cull the data. If they are using the real estate site's self-reported neighborhoods for these listings, I think that everyone can relate to the situation whereby a real estate agent tells you that a certain place is a "steal for the Upper East Side/(generic, more upscale neighborhood)," but when you go out to the apartment it's actually a few blocks into Harlem/(not so upscale area). Such vagaries in the listings would distort the prices per square foot for these boundary areas, but leave the solidly expensive ones alone, which is perhaps why areas like Central Park South (well defined geographically, and buffered by upscale areas) stand out.
- untog 12y agoI would imagine that the listings have lat/lngs (or at least geocodable street addresses), and the author has assigned them a neighbourhood by seeing which polygon the point intersects with. So real estate agents lying about what neighbourhood a property is in wouldn't matter.
- asciimo 12y agoTrulia uses lat/lon to determine a property's neighborhood. While neighborhood borders are often subjective, they do the best they can to meet the expectations of public consensus.