Informal testing finds Apple Maps arrival times are 'intentionally conservative' to provid...
A months-long, quasi-scientific test of the three leading car navigation apps -- Apple Maps, Google Maps, and Waze -- found that Apple's software regularly overestimates trip times, which may help drivers arrive on time or even early.
Relative to Google Maps, Apple's estimated trip times were typically 8 percent longer, according to Artur Grabowski, a senior manager at Adobe who said he spent "the majority" of 2017 collecting data on the three apps. On the other end, however, real arrival times with Apple Maps were on average 1 percent faster than the software predicted.
"In other words, Apple sandbags its estimates so that users on average arrive at the predicted time or slightly sooner," Grabowski commented. He argued that the company is being "intentionally conservative" in order to provide a good user experience.
Apple's error-adjusted estimated trip times were on average 5 percent higher than Google Maps, though Google's own Waze had a 6 percent gap.
"If you want to get to your destination most quickly, use Google Maps," Grabowski concluded. "If you want an accurate prediction from your navigation app to help you arrive at your destination on time, use Apple Maps. If thinking you'll get to your destination quickly helps to ease your commuter anxiety, use Waze."
The executive admitted though that all of his testing was done in the San Francisco Bay Area, and that a third of the routes were for his city commute. He also acknolwedged that companies are constantly changing their routing algorithms, and that different driving behaviors could impact times.
Over five years since its introduction, Apple Maps largely remains in the shadows of Google Maps, which is the most-downloaded navigation app on the iOS App Store. Waze is number two.
Relative to Google Maps, Apple's estimated trip times were typically 8 percent longer, according to Artur Grabowski, a senior manager at Adobe who said he spent "the majority" of 2017 collecting data on the three apps. On the other end, however, real arrival times with Apple Maps were on average 1 percent faster than the software predicted.
"In other words, Apple sandbags its estimates so that users on average arrive at the predicted time or slightly sooner," Grabowski commented. He argued that the company is being "intentionally conservative" in order to provide a good user experience.
Apple's error-adjusted estimated trip times were on average 5 percent higher than Google Maps, though Google's own Waze had a 6 percent gap.
"If you want to get to your destination most quickly, use Google Maps," Grabowski concluded. "If you want an accurate prediction from your navigation app to help you arrive at your destination on time, use Apple Maps. If thinking you'll get to your destination quickly helps to ease your commuter anxiety, use Waze."
The executive admitted though that all of his testing was done in the San Francisco Bay Area, and that a third of the routes were for his city commute. He also acknolwedged that companies are constantly changing their routing algorithms, and that different driving behaviors could impact times.
Over five years since its introduction, Apple Maps largely remains in the shadows of Google Maps, which is the most-downloaded navigation app on the iOS App Store. Waze is number two.
Comments
Not clear tho if he's saying Google Maps typically offers a faster route than Apple maps, particularly since he doesn't say it's not offering an accurate travel time, nor that of Waze either. Only that he's not arriving later than predicted by following Apple Map routing.
EDIT: OK it's clearer now after looking at the source. Google Maps is creating more time-efficient routes for him...
in San Francisco.
"Adjusted for prediction errors, not only does Google Maps outperform its competitors, Waze is actually the worst performing of the three."
IN SAN FRANCISCO.
If you're not traveling there than all bets are off.
So that’s what Adobeman quasi-scientifically concluded after a year, and he wished to put his name to it? Completely anecdotally, I just used iPhoneX and AppleMap for driving directions in a strange city (Charleston, South Carolina). Quite pleased with both voice and quickly legible graphics detailing what lane you should be in well in advance. Made rush hour less of a stress. Easy to quit app too. Much improved from old version, when I envied my friend’s GoogleMaps.
If I want an accurate prediction, I consult Apple Maps. It’s consistently more conservative than either Google Maps, Waze or NAVIGON. And mostly more accurate as well.
The efficiency of the ride is about the same for either one, but I like the accurate rush hour detour of Apple Maps. It’s mostly even better than Google Maps.
However, Google Maps is still more aware of names of public places, although Apple Maps improved greatly after adding the Foursquare database.
It clearly estimates the times using the posted speed limits. So more often than not, I arrive ahead of the estimate.
I went through a phase where I was blindly following Waze as the "most efficient" way to get from point A to point B. At least it felt that way, but I noticed that for no apparent reason I would be sent on a "scenic" route, especially while the most direct route was available. I felt like a tool collecting data for Google. Then within past couple months Waze started acting laggy in announcing the turns at the time of the turn - so in frustration I checked out Apple Maps (again).
After a lengthy absence, I am glad I gave Apple Maps a another try. Announcements are with enough lead time to execute the maneuvers, voice is clear and pleasant and if you want to shut it up - there is big "end" button on the screen. No pesky ads.
Been with Apple Maps ever since, Waze and Google Maps have been retired.
So Apple might not have been the lead horse in the beginning, but now, in my opinion, it is a most pleasant choice for navigation in Los Angeles.
Apple will never even come close to the vast multi-dimensional business, geographical, tourist infobase Google has.
Because it's not their core business (and they'd better concentrate on that, first)
Even 6000 Apple maps "engineers" in an Indian Lab won't change that
Shows a total lack of understanding of basic statistics:
I got the impression that the initial estimate might be based on posted speed limit, because rush hour arrival times seemed a bit optimistic (not factoring in slow moving traffic). I've learned to allow for this before I leave, but once I am on my way, the actual speed seems to then be factored in, and the estimates are revised accordingly.
However, the traffic report ahead of me does not seem to be a major factor in revising the estimate even while en route. If I am stuck in slow moving traffic, the estimated arrival time goes later, but when traffic opens up, the arrival estimate returns to the earlier time.
Second and I disagree with the assessment that Waze arrival time is off, I have been using it for 5 yrs and it is dead on ever time. The only time I have seen it off is when traffic becomes unpredictable. Waze claims they do not do this, but I think Waze learns your driving habits or style and tend to route you base on how you drive. In my case, my driving style is path of least red lights, even if is means going a little farther than the direct path. I rather keep driving than sit at red lights. Waze tend to route me this way unless I am in the new place altogether. Also avoiding red light means you are saving on gas. Because of this I tend to drive back road more often so less traffic and able to maintain a good speed. If Waze is not learning my driving habits it is definitely figure out my paths are fast since I am not spending 1 to 3 minutes sitting at red light or stuck in more traffic.
I know google maps and Apple maps does not do this, I have used them and they always want to route you the most direct way even though backroads could be much faster or easier to drive.
I know Google has fixed this for the most part, but when I used in the early days it had serious issues finding locations. I figure out why, it was mostly due to the fact they associated a street address to a GPS positon of the house or build and if the build was set off from the road a good bit and some other road near by was actually closer to that GPS position it would route you to the road next to that location, not the actual entrance to the house or build. I use to report these to Google and Apple and they never fix most of these. Waze did not do this, they associate the address to the actual spot on the street where that address is suppose to be, kind like where the mail box is at. Believe it or not when I was in the Silicon valley on business Google maps has more issue funding business locations and hotel locations then my stand lone Tomtom. You would think Google would have gotten their own home town correct. I know they still have some of these issue the last time i was other there.
My experience with Apple Maps is similar - the estimates are on the conservative side, but I much prefer that. It’s far better to have a few extra minutes and arrive a bit early - less stress and also less likely to lead to aggressive and more risky driving to arrive in time.
Google on the other hand is 100x better on place and street names but will direct me to congested main roads as it doesn't recognise uncategorised roads (of which there are MANY around here). I haven't tried Apple maps for using b-roads as it would probably direct me to another country as usual.
Garmin was well-known to do that, assuming that traffic on a major highway will be traveling perhaps 10% faster than the posted limit, and using that in computing travel times. While Google doesn't automatically assume that 10% they too will allow travel times that might require traveling faster than the posted speed to actual arrive when estimated, using the speed that traffic is actually moving at that hour over what the speed limit sign on the roadside says they should be. Then of course there will be those times when traffic is at a crawl which again is included in routing algorithms.
Too, if your driving is nearly always in or around major metros your experience may be different from those who travel long distance and in more rural areas. I would not hang my hat on one service offering you better routes only basing it on what the estimated travel time is anyway. That's something you can only determine from using it. For me I've found I can generally assume I'll arrive sooner than my nav service says I will. I found TomTom's times to be less accurate than Garmin's for my driving style, and still find TomTom (in a Mazda) less accurate than Google, and depending on time of day not as route efficient as Google either.