Google Maps oddities/bugs

[yorkie]

Google Maps suggests a 13 minute walk from Pontefract Baghill to Monkill.

Yet when you plan a longer journey, such as Sherburn in Elmet to Castleford, instead of simply suggesting a walk between the stations, Google suggests an 8 minute walk to a bus stop, a short bus ride, and another 8 minute walk. Overall that's more walking than going direct, not to mention the cost and reliability implications of adding a bus into a rail journey.

Does anyone know why Google does this? And have you seen any more oddities?

Another bug I discovered is that when you get many itineraries, it is often the case that when I select one, I get the wrong one.

 

[A Challenge]

It is to reduce walking when there is time on the connection, but it no slower on end-to-end tines, but it is on journeys just between the two.

 

[A Challenge]

Another one was last week when I went to get a bus and it said I needed to leave at 14:41 for a 4 minute walk to the bus stop. Nothing wrong with that, except the bus left at 14:41!

On another bus route it was saying leave 3 minutes after the bus left for the three minute walk!

 

[tony_mac]

It is to reduce walking when there is time on the connection, but it no slower on end-to-end tines, but it is on journeys just between the two.

But it doesn't reduce walking time - two 8 minute walks is more walking than one 13 minute walk.

In the case of a connection, the time difference wouldn't matter as long as the overall journey time is the same, so it could just be a case of how the ties between equally valid journeys are broken.
I wonder if it is preferring journeys with a shorter maximum walk, rather than the overall walking time? That makes some sense if there is a long gap in the middle, but not for a 2-minute bus ride.

 

[BestWestern]

What I find amazing is that Google somehow knows the times of all these buses, seemingly from every stop. A lot of the time you can't find that much detailed information on the operators' own websites! Where do they get it from?!

 

[telstarbox]

It comes from Naptan data.

 

[Bevan Price]

What I find amazing is that Google somehow knows the times of all these buses, seemingly from every stop. A lot of the time you can't find that much detailed information on the operators' own websites! Where do they get it from?!

Try this site:
https://bustimes.org/

 

[BestWestern]

It comes from Naptan data.

Say what!?

 

[NorthernSpirit]

Google has also marked a section of the Pennine Way wrongly.

If one looks at where Wessenden Lodge is near Marsden, Google marks it as running along the edge of Wessenden Reservoir wall, past an air shaft, round past some waterfalls, past a weir and up to some grouse butts which is near some transmitter thing.

The actual route, going by the Ordnance Survey, goes past Wessenden Lodge passing Hey Green and down to the very bottom of Blakeley Reservoir where there is a fingerpost stating Pennine Way before a sharp left hander down a hill and over a footbridge before climbing to the aformationed transmitter thing. I'm giving it a good 15 years before the Pennine Way magically gets rerouted many thanks to the balls up that Google is showing.

 

[robbeech]

I suppose there are so many possibilities for the algorithm and in the database that it’s not going to get them all right. From my house to “London” it will frequently suggest a couple of minute walk, a bus to Chesterfield (takes an hour) then a train rather than a 10 minute walk to the train station, a train to Nottingham then to London. And it rarely suggests a Worksop/Retford/King’s cross combination despite that being the quickest route at certain starting times per hour.
All in all I think it does really well with things, though our discrepancies on Bradford buses between the iOS app and the mobile website was interesting.