Many websites and emails offer to give you the DISTANCE between your home and whatever they’re selling. These are always VERY low, like 30-40%, including sites that should know better, like CarTalk and AutoTrader. What’s the bit? Why the deliberate and consistent underreporting of these distances?
Are these distances specified as the crow flies, or as driving distances?
I would guess that the websites are using IP addresses to determine location and the best they can do is general area. So the distances given are from a generic point, the centre of the city say, to a known point.
Concur.
Or possibly from the closest point in a city to the business. If the nearest Acme location is a half-mile from the Urbansburgh corporation limit, and the IP detector says that I’m in Urbansburgh, then it’ll tell me that there’s an Acme a half-mile away from me (never mind that I’m clear across on the other side of Urbansburgh, but the website doesn’t know that).
Mine always tell me where things are in Manhattan, though I’m deep in Jersey. This is a consequence of using my employer’s VPN and appearing to be in the city.
AutoTrader requires you to input a ZIP-code, and since distances change radically if you change Zip-code I expect that’s how they guesstimate a home address.
It still seems like they miss by a lot though. I checked three listings and distances in Google maps:
1: 46 miles away - shortest suggested driving route 55 miles, not possible to cut it down much
2: 38.4 miles away - shortest suggested route 55 miles, down to 46 miles if not caring about time (Increases from 1h04 to 1h20)
3: 13.1 miles away - shortest suggested route 17 miles
My zip code is only 4-5 miles along the longest axis, and I live close to the center, so that doesn’t seem like the source of most of this mismatch.
@naita, are the distances given by AutoTrader close to the actual straight line distances (that is, ignoring roads)?
I find that distances are often given from the location of the Post Office when they use Zip Codes. For me, the location of the Post Office that serves my Zip Code is no longer within the boundaries of the Zip Code, but a lot of sites still use the old location of the Post Office that used to be within the boundaries decades ago.
But there is a lot of inconsistency.
In my experience, it’s sometimes crow-flying distance, sometimes driving. If you are working with a road map, it may be driving, as they count up the legs from point to point. However, if you are looking up the nearest [whatever] store, they use the crow.
How do I know this? Because if I search for, say, a nearby UPS dropoff point, the one showing nearest at ~20 miles is actually more like ~130 miles. Why? Because the mapping software ignores large water obstacles like lakes or bays and computes point-to-point. If I followed the 20 mile route, I’d either have to launch my boat first, wait for the surface to freeze, or fly.
Ok, I just tried a few examples on AutoTrader
- Reported distance 11.1 miles. Shortest road distance by Google 16.6 miles. Straight line distance 10.9 miles.
- Reported distance 27 miles. Shortest road 41.5 miles. Straight line distance 27 miles.
- Reported distance 12.1 miles. Shortest road 17 miles. Straight line distance 11.8 miles.
My straight line measurements were using the Google Maps tool, which may be slightly off depending on how accurately I clicked the map. But it seems clear that the web site is using a simple latitude/longitude distance calculation, and ignoring actual roads.
Wait, are you saying those lonely single women waiting for me don’t live just ten miles away? Because I’m not driving five extra miles.
I actually wrote one of these calculators for my old employer to identify the closest in-network facility to a given workplace.
Since it was an offline kind of thing, we ended up finding a table of the lat/long coordinates of the geographic center of each zip code and using that to calculate the distances between zip codes. Since it wasn’t customer-facing, it didn’t really matter.
But what made it interesting is that there were small, but significant distance variations between calculating the distance as a flat plane vs. spherical. For example, the distance between my home and work is 9.62 miles on a sphere, and 9.88 miles on a flat plane.
I’ll bet you wouldn’t walk a mile for a camel, either.
If you do a Google search, your assumed location is listed at the bottom of the page.
In may case, it’s just over three miles from where I really live. Cox must have a distribution center there.
I have no desire to correct their mistake.
That’ll depend on how you’re projecting your map onto a flat plane. There are projections which will make it longer, and also ones that will make it shorter.
I’m happy with Skynet having the wrong coordinates when it sends its hunter-killers after me.
I don’t know if the mapping issue with Google is the same as the one I experience with Waze, as they’re part of the same company, but I have experienced some underreported distances. Last night we were in an unfamiliar part of town and craving In n’ Out. I checked Waze and the nearest one is supposed to be 3.8 miles away. First direction, go two miles and make a left on Whatever Road. Stopped at the light at Whatever Road, I check Waze and it’s now…3.6 miles to the In n’ Out.
I thought I’d compared to fairly straight routes, but I wouldn’t be surprised if there was enough zigging and zagging for my results to equal yours if I tried it out.
Yep, it will. What I was getting at is that it’s entirely dependent on whatever API or toolkit their developers are using to calculate the distance, as well as what exactly the starting and ending points are.
I’d think that would be the case, even if it’s calculating road distance, with some additional variance thrown in for differences in the route choosing algorithm.