How accurate is the Walkscore app with respect to vacancies and rent costs?

In My Humble Opinion
1

For those of you who don’t know, Walkscore is an online service that lets you enter any address, and then calculates a “walkscore”…basically the area’s walkability and availability of good transit options. They have also released Android and iPhone apps that do the same thing. Areas with low scores are labeled car-dependent, while the areas with the highest possible walkscores–mostly some areas of Manhattan, are considered pedestrian/transit rider “paradises”.

More recently, Walkscore has added an apartment finding feature. Look up any spot in any reasonably urbanized area and you’ll find a score or more apartment vacancies in the space of a few square blocks. It certainly seems to be the case that you see a lot more vacancies in the app or on the website, than you do actual For Rent signs in front of the properties. This leads me to wonder how they get their listings and how current they are. For example, only yesterday a vacancy came up where I’m living now, and Walkscore already shows it in their listing. As far as I can tell, especially what with living in the same building, this listing is entirely accurate. In other cases, though, not so much. Looking at the overall picture for a given locale, it often seems that if I use Apartments.com or Rent.com to search, the rents are higher and the vacancies fewer. A cursory perusal of the WS website doesn’t tell me who they partner with, if anyone, for their vacancy listings.

I know there can’t be a definite answer here, which is why I’m posting this in IMHO, in addition to the fact that IMHO is the place for the personal experiences with Walkscore which I hope to read about here.

Bonus question: Given that so many people now search for apartments online, are physical For Rent signs beginning to fall out of favor?

2

It’s skewed in that instead of using my address, it uses the center of my zip code. I’m less than a ten minute walk from a grocery store (a Wal-Mart, but still), and six or seven restaurants (all but one chains, but not all fast-food), and–astonishingly, given the horribleness of Dayton’s bus system–30 seconds from two bus lines, and five minutes from a third. It’s certainly not a great walkable area, but it’s better than Walkscore makes it seem.

3

Walkscore’s transit algorithm is heavily skewed toward trains. Unless the location is close to some kind of rapid transit station, it won’t get any higher transit score than perhaps 25 out of 100.

It makes sense on one level, because a metro does provide much easier access to distant neighborhoods than buses do. OTOH, buses can work very well in some situations, and WS should recognize that.

4

While that’s true, it only provides access to places the train also goes. While in places like Toronto and Denver, I’d frequently take a bus to a train/subway, then another bus from the train/subway to my destination. I consider access to a reliable, wide-spread bus service to be just as good as having a train station nearby.

I just ran it again against the address I just left in Reading, Massachusetts and got an even worse score than here. (17 there, and 32 here.) It did the center of the zip code thing again (about three or four miles from where I actually was), and that was further from the train station that goes to Boston, and the variety of stores and restaurants on Main Street.

Walkscore seems to be a good idea that still needs some fine-tuning. I will say, I really do like the idea. It’s a good and useful concept.

Oh, and as far as apartments, I don’t recognize any of the ones listed for either location. Certainly the one I live in is not. I assume they’re also based on the center of the zip, and not the address.

5

I relied heavily on Walkscore when I was finding my apartment in Portland. That said, the actual listings provided by the app/website were basically useless. Not only were they not always very up to date, there just weren’t that many —far fewer than the neighborhoods I was looking in actually had to offer. My method: find the places on Craigslist, and plug the addresses into Walkscore.

6

I think it’s more accurate to say that the availability rapid transit adds an order of magnitude to the overall transit picture. Perhaps inevitably, WS is skewed towards urban environments where road traffic tends to be the worst, which are also exactly the same places where nobody will ride the bus more than a mile or two unless they have no choice. In a big city, the bus works best as an extension of your geographical range as a pedestrian, while beyond that the time spent stuck in traffic and waiting for new passengers to board and pay their fare usually makes it time-prohibitive. Granted, if your city has a metro it usually won’t go everywhere, but it can get you to dozens more places all over town, much faster than the bus system can.

Ah…I think I’d worry if the apartment I lived in showed up as a vacancy. I think the map locations are accurate but I’m more skeptical about the asked rent, and availability. For that matter, when somebody rents an apartment, how quickly does the vacancy disappear from the apartment hunting websites? (On that score, I believe WS gets its data from the other apartment sites, but I’m not sure which ones, or what kind of delay might be involved.)

7

The bus in such situations would not be any slower than a car, and much cheaper. If I lived in a location such as Manhattan, central Boston, or central Chicago, I wouldn’t even see a point in owning a car; I’d just rent one for travel to the Catskills, or New Hampshire, or Wisconsin.

I’m really not understanding your defense of Walkscore on this point. There’s a trade-off between time and cost. In Reading, I could have bought a commuter pass including the train, bus, and subway for just under $200 a month, far less than a car payment, insurance, gas, and maintenance. If Walkscore is truly meant to show the use of a geographical area without a car, then it needs to include all methods of transportation.

It does not inevitably skew toward rail mass transit, or else Reading would be much higher than 17. It seems to inevitably skew to the center of a major city with rail transport, and that, in my opinion, is a failing.

8

P.S. The building I live in is not listed, and there are vacancies. No buildings near me that I know the names and addresses of are listed. Same for what I am familiar with in Reading.

The map locations are accurate, for a certain value of accurate. They nailed the center of my zip code in both Kettering and Reading, which is not, in either case, even close to my actual entered street address.

9

Walkscore is weird. I got 4 different results.

At the exact address I used to live at in zip code 6065-78 score
At the center of the zip code 60657, 98 score
At my new exact address in zip code 60613, 94 score
At the center of zip code 60613, 88 score

Now I can walk between my old address and the new one in about 20 minutes or so, it’s about a mile apart. I can probably easily walk between the centers of zip codes 60613 and 60657 in about the same time. Both addresses have access to several CTA bus routes, and a CTA train station.

10

I would do the same thing in that situation. However, I disagree that the bus would be just as fast as your car at times of high traffic; remember you have more people getting on or off at each stop; at less busy times the driver can skip some stops entirely. More people boarding means more time spent collecting fares or verifying passes and occasionally answering questions, since for most people the bus operator is their transit company representative.

Now the bus finally gets to pull away from the stop, but it takes a lot more time to accelerate to speed than your car…etc., etc.,

Nobody’s saying Walkscore is perfect, least of all I; another deficiency is that it doesn’t factor crime statistics. It is a good start, though, and I happen to agree with their transit bias towards rail, because generally a good metro means better transit generally. That’s just my opinion for what it’s worth; I’m not a professional in the field. It’s my further opinion that in a typical large city you do need both trains and buses to have effective transit. Having just buses in a city like San Francisco or even L.A. is about as futile as having only a rail system with stations miles apart from each other.

It seems to skew against suburban rail in comparison to urban metros, and I agree that can be a failing, but I think a lot of that comes down to personal preferences and it probably isn’t possible for the app or website to please everyone in this regard. I do know if you are only along one rail line or near the end of it, it doesn’t give a very high transit score. It seems to me too that Reading should get a higher score. To improve on this point, perhaps the service should allow users to weight their preferences in this area. For some people, “good transit” might just mean they can commute by train without regard to how they get around in their leisure time; for others it might be just the opposite. For them driving to work is fine, but they’d rather not drive to the museum or ball game and get gouged for parking.

Another difficulty is that the website and app, from what I’ve seen, aren’t always intuitive. When you look at one of the apartment listings, you may see a transit score or you may not, depending on what, I don’t know. I’ve seen listings near the Glendale (CA) station that don’t seem to have any transit score, even though that station must be served by at least a couple of dozen Amtrak and Metrolink trains each day. There are settings in the app but, again, they’re not entirely straightforward.

So yeah, it’s definitely not perfect.