Anyone else notice that Bing and MapQ have more accurate locations than Google?

Yeah, my house is 3 blocks off.

Interesting topic. Maybe we should include our 5 digit zips as well?

My area is 20904 (older MD suburb of D.C.) and the streets that were planned and only partially completed have been corrected over the past 6 or 8 years anyway. It's also usually a hi-res photo area as well. (I long ago gave up worrying about 'them' watching me )

Comparing the 3 with my address-
Google (web)- proper side of the one lane street, exactly on my driveway apron.

Bing- Marker dead center on my chimney! Not sure if that's good or bad since my house is about 300 ft from my mailbox/driveway apron.

MapQuest- No closest level imagery, but looks to be correct side of the road, but ever so slightly southwest of my driveway apron & probably on my neighbor's.

I'll try Google Earth when I get a chance, and also the reverse (by coords rather than address) just for the sake of completeness.

N.B.- In all cases the map roads overlay the imagery very accurately.

I'm in Louisville, KY. Several times (in town or out of town) I've found Google and Garmin) to be off, but Bing or MapQ to be accurate--more Bing than MQ for some reason.

Likely it varies from region to region, but Bing seems to be the most accurate of the three mentioned above.

Google puts my address right at my drive about 10 or 15 feet from the street, but I think I may have corrected the original location a few years ago (a nice feature in Google maps). MapQuest puts it in the middle of the street in front of the house, so you wouldn't be able to tell which side of the street is my house unless you knew the even / odd scheme. Bing puts the address right at the side of my house on the opposite side from the drive. Garmin puts my address right at the corner of my property bordered by the streets forming the corner I live on. They are all within about 100 feet of each other. I think any one of them would bring you right to the correct house.

I do not have the answer, but for this discussion to make any sense in the context of GPS units, there ought to be some agreement as to what would constitute accuracy.

To set the stage, we really do not want a lot of the POI files to give us the coordinates of the front door of some establishments. Otherwise, if we had alerts turned on, the "door" might be more than 75 feet or so from the "road" on which we are traveling and the alert would not sound.

If someone knew how each of Bing, Google, etc set the coordinates which they use when identifying an address, we would be more able to compare the services.

In MY mind, the gps is to get you to a reasonable proximity of the address, not invite you in for a drink...! That said, it's way better than a paper map!

People expect way too much these days.

Holy Cow! The chimney top or the middle of the street somewhere in front of the house isn't close enough? What is wanted? Nobody probably could agree on what is close enough.

Juggernaut you hit the nail on the head on that reply.I just came off a 7000 mile cross country trip.I was more them amazed to find the locations to be pretty close for the areas I was looking for.Only had two problems on the complete trip that the locations were way off and it wasn't the built in pois.These units do a pretty darn good job.You may find a few problems sometimes but the overall results are more then satisfactory.

My OP was not about which side of the drive Google, Garmin, Bing... places a location. I too am amazed at the millions of locations that are correct. Simply mind boggling. I just don't get those times it's way off when other info on the same street is correct--not how close it is to the hot water in the NW of the basement.

I'm just surprised that, and wonder why the same info given to some mapping companies is correct but is missing from others or way off. Off the top of my head I know several incorrect addresses (mine, in-laws, two friends, my work, a rehab) and these are just the ones I've found since using a GPS in the last three months. All on streets and addresses have been there for years.

I just find it odd that some mapping gets it correct while others get it wrong and wondered if others noticed that Bing seems more accurate than say Google. Do some use old data or just have inferior or incomplete data, or don't buy new data as often as others maybe?

Different algorithms produce different results. It may be a 30/30/30 mean average spread across them all for accuracy.

Have federal surveyors agreed upon the doordinates of some particular locations? And, hopefully marked it with a concrete monument?

If so, try one of those and use street views to see how close the mapping service gets it.

I understood that dagrev was just curious as to which mapping service would be more consistently accurate enough to get you to where you wanted to go. Years ago (pre-GPSr) I got directions from MapQuest which took me to who-knows-where, and I never did find the place I was looking for and missed a meeting. I've never used it sense. I hope it's better now, but I just have a hard time trusting it. I'm sure the same could have happened with Google or any other service. I've even had it happen with directions from locals. I drove up and down an interstate one time looking for an exit that at least 4 locals said was the one I wanted but never found because the name all the locals knew the exit by wasn't the way it was labeled on the exit sign. The main question to answer is "Did it get you to where you wanted to go?".

I guess you could check them all every time you needed directions, but who is going to do that? And they could all be wrong.

For me, the question was interesting. Not that I care which one pinpoints my chimney.. but mainly because I've seen complaints of inaccurate files and it often gets blamed on web maps or geocoding. Also because I built geocoding & reverse geo-coding using Google into my Excel spreadsheet, as does EPE and GeePeeEx.

I did some further checks ('cause I'm retired & look for diversions) and I've come to the conclusion (as many of you have) that none of the services are totally accurate for all addresses, but all of them will get your GPS and you to where you want to go. Avandyke mentioned the ability to submit corrections to Google. That along with the tools and image resolution in Google Earth, plus it's programmable geo/reverse geocode ability puts it over the top in my book. However.. Google was off by the largest amount when pinpointing by street address (other than mine, which I also corrected earlier in Google).

For sheer accuracy using coordinates all of them were dead-on in mapping a USGS National Control point 1/2 mile from me at a high school. All meaning Google web, Google Earth, Bing, and Garmin's CNNT 2009.1 map in both MapSource and BaseCamp. In MapQuest I couldn't find a way to search it using coordinates. Google & G.E were also the only ones where I could actually make out the exact spot of the control point on the school grounds. If I hadn't actually seen it in person, it wouldn't have been so recognizable, though.

I also found out that my two nuvi's (a 200W and a 765T) report the same accuracy levels, but do NOT report the same positions, all other things being equal. They are off from each other by 35ft Lon, and 70ft Lat. Each reported 18 ft accuracy. In my case that's the difference between my driveway and my neighbor's on the opposite side and up the street.

The answer to the OP's question seems to be (imo) in the databases used and also how well they align to the 'base maps'. Google was odd, and seems to use a different data set for street addresses than it does for named locations. When I searched the address of the school in both the web version and G.E. they placed it rather far off (but still locatable). When I searched using the school name and zip code only both put the marker dead center of the building, but the coordinates listed in the "link" on the web gave the street address coords (wrong). G.E. gave the correct coords. None of them including Garmin put the position at either of the school's entrances. Close, but no cigar. It's hard to miss a High School, though.

All of the road lines in the area of the school were located properly, and registered with the imagery very well (where they had imagery.. Garmin, you listening?).

Google HAS changed within the last year or so. It's not as accurate as it once was for street address queries. But, I think the better imagery and base map accuracy make up for it, as long as you actually look at the images and adjust as necessary.

All in all, this has been interesting & enlightening.
..and a nice Friday diversion.

For finding certain ships in my POI file, I found it handy to utilize Bing's bird's eye view, when it was available. For the most part, I used both Google and Bing in building it, and EPE to fine tune things.

Now that said, for my parent's house Google puts a pushpin on the end of the driveway attached to the street, which really isn't accurate since the house is set some 100 feet back from the road. I didn't try Bing but I don't think it will fare much better.

For my apartment building, both put the pushpin in the parking lot, which is pretty good.

But when I look up an address on google, I would like to know exactly where it is and not have to guess. Right now it's just a guess since it seems to be a house or two away.

Using the 3 services mentioned here, I looked for my house using my address.

Google Maps had my house correctly identified and uses a later image than either Bing or MapQ.

Bing Maps identified my next door neighbor's house incorrectly as my address.

MapQ put a mark near the edge of my driveway between the two houses... call it close enough.

I tried Google Earth too and it lasered in right on the center of my roof!

The most interesting thing to me about my experimenting was that my two nuvi's were so far off from each other. I never expected to be able to use them to mark my property lines or anything, but I was really hoping that they would be at least close to the reported position by the indicated accuracy amount. At least one of them was off by more than double it though. I know that's why surveying GPSr's are so costly & even then they use differential GPS for real accuracy, but it's sort of a moot point about accurate waypoint addresses if your unit's reported position can vary like that.

I didn't get out to the school control point with them yesterday, but when I get my 295W in a few days I'm going to take all 3 as well as my Explorist 210 and see how far off of all 4 of them are from the known position.

Both Google and Bing images seem spot-on with the control point so I'd say they give very accurate coords for any given visual point (exception was the Google web link, but turning on the 'drop marker' feature was accurate). This probably varies from region to region?

I took away several key points-
-How much is your unit-reported position in sync with any map?
-How much is your unit map in sync with with the lookup map?
-How much are the lookup and unit maps in sync with the same (accurate?) control points?

All three (Google, Bing & Mapquest) put the pushpin for my house in the middle of our 10 acre field.

Using the Satellite/Bird's Eye view:
All three had very old images.
Very close to my house there is a development of homes that are at least 6 years old.
Google had the area as trees/woods.
MapQuest had it as cleared
Bing had it cleared with equipment.

The cars in my driveway have been gone for over 3 years. The trampoline in my backyard has been gone that long as well.

That being said, yesterday I looked up my daughter's address in a new part of Sahuarita, AZ and did a street view with their little yellow man. It brought me right in front of her house. Based on the trees in her front yard, the image was about 3 years old.

While my Nuvi gets me to my driveway, my house on

Google Maps is my next door neighbors'

Bing maps is also my next door neighbors'

MapQuest puts my house 1 1/2 houses away

I really appreciate JD4x4's taking the time to do this and report.

My feeling on this subject is that we would debate "accuracy" without some "Point of Reference"that was Known.

We can all look at our home chimney or mailbox but we really do not know the true coordinates - so we cannot say whether a GPS unit or a mapping service is "accurate".

Does anyone else know of some location in their town that has been accurately surveyed to specific coordinates and would you be willing - as JD4x4 was - to go check it out?

Can be found by State/County here - http://www.ngs.noaa.gov/cgi-bin/ds_county.prl

and by distance from your home coords here- http://www.ngs.noaa.gov/cgi-bin/ds_radius.prl

You want to filter the list for "GPS Sites Only" to get places you can walk to (some may be on private property though) & use with your GPS. Also be aware they use dd mm ss.ssss coord format.

A good overview is here- http://www.geocaching.com/mark/default.aspx

The Geocaching site also has a search, but includes a lot of lesser accuracy markers.

Thanks for thos links,

When it stops raining - and I hope it does not for some time because we are over 10 inches short this year - I will go see if I can find some markers on the ground nearby.

I think this is a great topic, the more I'm learning and the more I think about expectations of accuracy.

Not so much accuracy for driving-related POIs, but for some of the walking City & other 'attraction' files here, like Williamsburg, VA.

As long as it gets me close enough to use my eyes to see the sign/business, I'm happy with the accuracy.

Me too, as long as I remember that on a good reception day and an 'accurate' poi the thing I'm looking for could be on the back side of a building in the middle of a block.

Agree with that.The only time a big problems comes in is when you are on a busy highway with 4 lanes on one side and cars everywhere.The gps says on right and the wife hollers there it is on the left.

I have noticed the same thing. As a truck driver I try to lookup where some of the warehouses are at and google might show a corn field and mapquest will show the whse...

Fine, we can find our target anyway.

Bing puts my house 4 lots away from its correct position, Google gets it spot on.
M.S. Streets and Trips puts it at the complete opposite end of my street. Both my Nuvi's get me to my front door.

MapQ,google,MapSource & Bing all mark my driveway 60 yards further north, and Bing shows the house back from the street about 30 feet. Searching for address on my Nuvi, it comes up perfect, and about 30 feet back from street.

Google might not be perfect but once you send them a map update you get a response a lot quicker than Navteq

In case you were wondering how Google is improving it's location accuracy ...

http://www.fiercebroadbandwireless.com/story/googles-wifi-co...

Probably one of the telling comments is: "Google could be secretly tracking its Android users to map these WiFi locations without users' knowledge."

Accuracy is another relative term that depends on someone's personal definition (and hardware/methods)!

In my area, Google's point coordinates are accurate to within inches of the USGS coordinates.

Addresses are another animal though, and not entirely Google's (or others) fault I don't think. 'Address' is another definition that varies depending on who you ask. Post Office? Tiger data? State/County land records? Location owner? Phone book/web site? ..That's the rub, and causes the most 'inaccuracies' I've found.

Google seems to be very good in providing address info, but you need local knowledge sometimes to know 'how to ask' for it, and it's always being updated (by user input).

Take this example of my closest USGS survey control point- go to this url that I searched using the USGS coords:
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=...
The green arrow is within 6 inches lat/lon of the actual marker. The map should also show Point A, which is the closest address in their database to the arrow.

Now, click on "Search Nearby" and type "High School". Point A is now the correct mailing address of the school that the marker is on the grounds of, and the purple marker is the original 'geocoded' address Google returned. If you have a Google sign-on, drop a lat/lon marker on the arrow to also see the six digits that Google rounds the 9 digit USGS points to (reasonable, but it all contributes to 'accuracy').

Finally, go to your nuvi and search for either address and see what IT comes up with. On my 2011.30 map set it's very close to Google, but earlier maps were not.

Lastly, neither of my 3 units (200W, 295W, 765T) will give me the same coordinates at the same time & day when standing on the control point.

Moral of the story? In coords we trust (give or take!), anything else will get you close but won't be as universally 'accurate' as coords, and NOTHING will replace your brain and some luck.

Personally, I'm thankful if a poi file tells me there might be/used to be a poi in the general vicinity. It beats the heck out of not having a gps and all of the collective info handy. Doorstep directions are a nice plus if it turns out that way. Which is a nice goal, but it ain't going to happen universally anytime soon, imo.

as long as it gets me close and recognizes the poi or my address>

The point being made in the news article is that Google may be using your device's connection to a wifi node that it has an address for, but the coordinates may be off. By comparing the wifi node data against its database, Google will be able to approximate an address much closer than assuming there are 100 addresses in a block and this one is X% of the total possible.

In related news, Google has agreed to turn over its method of collecting wifi data along with emails, passwords and other information collected in its drive-by street-level activities to the FCC for investigation of Google's possible wiretapping of the signals. It gets more interesting as most of the people advising the administration on wireless and Internet are executives on loan from Google.

I get it. I guess I was just thinking in terms of the thread topic, since none of the Google/Fed connection surprises me.

Especially since my son in law told me about the military version of Google Earth, and I'm old enough to know how AT&T got it's start as well as seen some of it's previously clandestine facilities from the 50's onward.

..Now ponder Google's Picasa facial recognition stuff, and how they're going to recover those development costs.

Gotta keep us safe, eh?

But it will still take quite a bit of time before anyone has a very accurate universal address database, imo.

They aren't always very accurate. Heck Google recently messed up the border between Nicaragua and Costa Rica bringing the two countries close to war!
http://blogs.abcnews.com/thenote/2010/11/google-nearly-start...

You were right! I tried Bing maps and it did put the marker in the middle of my roof. Now I have to go up and get the arrow out,LOL

I just got back from a Doctors appointment,the trip took 30 minutes for a 10.62 mile trip with my Garmin Nuvi 1450 it had me making strange twists and turns.Mapquest shows a more direct route at 18 minutes for the same trip.

I think none of the 3 are close to perfection. Depending on the address you input, one or the other two may be right on (or close enough). Other times all three are lousy. I guess I just have to use all of them and see which one is the most accurate based on my input.

I prefer Google because it is the only one at this time that I can see my house with aerial view (100 ft), bing & MapQ can't get any closer than 2500 ft and the aerial view is poor, Google I can still make out were my house is at 2000 ft. But I must admit that MapQ is closest to my driveway when I put all 3 in map view.

I haven't noticed much of a difference in my area here in Northern California. Most places are just a few feet difference - top of the house vs into the driveway, etc. - in the places I'm trying to find.

Guess I'll start looking a little closer in the future. Thanks for the heads up.