Coordinates: What Is More Accurate Etrex Vs Google Earth

[+bobkeenan]

I set up my first cache yesterday. I took the advice on-line and walked back and forth 10 times and took 10 readings. Then I averaged them.

 

But I then input the coordinates into google earth and I was about 15-20 feet off. So changed the settings until google put the cursor right on my cache. It was pretty close to one of my 10 readings but 15-20 ft from my average.

 

So which is more like to be accurate. Google Earth or my averages?

[+humanloofa]

Google earth is usualy off, go with the average.

[+badlands]

Try setting the GPSr down at the cache site and let it settle in. It should do averaging on it's own. If you've got a lock on a lot of birds you should get a good read.

[+alexrudd]

15-20 feet is excellent accuracy for a handheld GPS. Your coordinates are fine; usually up to 35 feet is acceptable.

[+Prime Suspect]

I set up my first cache yesterday. I took the advice on-line and walked back and forth 10 times and took 10 readings. Then I averaged them.

 

But I then input the coordinates into google earth and I was about 15-20 feet off. So changed the settings until google put the cursor right on my cache. It was pretty close to one of my 10 readings but 15-20 ft from my average.

 

So which is more like to be accurate. Google Earth or my averages?

I wish they would remove that bit about averaging. Unless you're going to take hundreds of reading over many hours, you're just as likely to make your results worse, rather than better. Just let your GPS sit for a few minutes, then take a reading.

[+JohnnyVegas]

I set up my first cache yesterday. I took the advice on-line and walked back and forth 10 times and took 10 readings. Then I averaged them.

 

But I then input the coordinates into google earth and I was about 15-20 feet off. So changed the settings until google put the cursor right on my cache. It was pretty close to one of my 10 readings but 15-20 ft from my average.

 

So which is more like to be accurate. Google Earth or my averages?

I wish they would remove that bit about averaging. Unless you're going to take hundreds of reading over many hours, you're just as likely to make your results worse, rather than better. Just let your GPS sit for a few minutes, then take a reading.

I agree, averaging on your own is a huge waste of time. Most every GPS made does the averaging as a function of the GPS and is going to be more accurate.

[+hurley_108]

In general, it's a bad idea to place a lot of faith in google earth. Take a look at Calgary, Alberta. Right up the middle of the city, there's a split in the imagery. One side is hopelessly, hideously off. Turn on roads to see what I mean. The imagery is misplaced by hundreds of feet.

[+fizzymagic]

I wish they would remove that bit about averaging. Unless you're going to take hundreds of reading over many hours, you're just as likely to make your results worse, rather than better.

I wish people who don't understand statistics would quit making uninformed comments about averaging.

 

Averaging is not just as likely to make your results worse as better. And three or four measurements taken after walking around in different directions can significantly improve your coordinates.

 

Seriously: if you don't understand statistics, please don't make statements about it, OK?

Edited July 6, 2006 by fizzymagic

[CoyoteRed]

So which is more like to be accurate. Google Earth or my averages?

 

I wouldn't trust Google Earth. I've seen coords off by a good 100' or better.

[DLH Fenrick]

I wish they would remove that bit about averaging. Unless you're going to take hundreds of reading over many hours, you're just as likely to make your results worse, rather than better.

I wish people who don't understand statistics would quit making uninformed comments about averaging.

 

Averaging is not just as likely to make your results worse as better. And three or four measurements taken after walking around in different directions can significantly improve your coordinates.

 

Seriously: if you don't understand statistics, please don't make statements about it, OK?

 

Gee, I just don't get it.... could you be a little less vague. What is it that you're trying to say?

 

Perhaps it you told us three or four times, we could average your disdain and we could figure out what you really mean.

 

[+Tharagleb]

I wish they would remove that bit about averaging. Unless you're going to take hundreds of reading over many hours, you're just as likely to make your results worse, rather than better.

I wish people who don't understand statistics would quit making uninformed comments about averaging.

 

Averaging is not just as likely to make your results worse as better. And three or four measurements taken after walking around in different directions can significantly improve your coordinates.

 

Seriously: if you don't understand statistics, please don't make statements about it, OK?

 

What he said.

[+briansnat]

I wish people who don't understand statistics would quit making uninformed comments about averaging.

 

Averaging is not just as likely to make your results worse as better. And three or four measurements taken after walking around in different directions can significantly improve your coordinates.

 

Significantly? What is significant? If I can get from 11 ft accuracy to 7 foot accuracy by averaging then I guess that's statistically significant, but is it really significant to a geocacher?

[+Team Cotati]

Google Earth will normally be much more accurate.

[+briansnat]

Google Earth will normally be much more accurate.

 

Please don't give out erroneous information. The OP asked an honest question and there is no need to confuse him.

[+StarBrand]

Many local roads map 75 - 100 feet off on the sat image vs map and reality.

[+budd-rdc]

I would trust the eTrex more than Google Earth in most cases.

 

Most cachers will find errors of 15-20 feet acceptable if the container is large enough, or if the hide location is easy to figure out. What you want to do is stand still as close to the cache as possible and wait for EPE to drop reasonably low for the environment. I usually watch the coordinates as I stand to wait for it to stop drifting.

 

Here's the exception: I've noticed from personal experience that eTrex's don't like your standing right next to obstruction (trees, walls, cyclone fences) when you take coordinates. You can triangulate/project by taking a reading some distance away, or walk to the hide from several directions to see if you get consistent readings at the cache.

 

Keep in mind that older eTrex's don't support the averaging function, which is the recommended way to take your reading.

[+GPSlug]

I think an eTrex is closer than this (1/4 mile)

[explorerboy]

Google Earth is always off by alot or more. Sometimes it gets close but not often. I only use the kml for when I'm trying to see the geocaches in a current area in general.

[Keystone]

Google Earth is always off by alot or more. Sometimes it gets close but not often. I only use the kml for when I'm trying to see the geocaches in a current area in general.

Using the Geocaching kml is an entirely different subject. The cache locations there have the coordinates "fuzzed" by design. In contrast, if you enter a set of coordinates in Google Earth, or if you open a GPX cache location file in Google Earth, you will see Google Earth's best guess at where those coordinates take you. Then, you are back at the subject already being discussed.

[explorerboy]

I will continue to say then that if it's off by that much with the kml then I will not trust it with marking waypoints. Happy now?

[Keystone]

I will continue to say then that if it's off by that much with the kml then I will not trust it with marking waypoints. Happy now?

No, I'm not happy, because you are misunderstanding how the Geocaching kml functions. I'd be happier if you acknowledged the difference between that kml and "regular" Google Earth. The Geocaching kml is *designed* to be inaccurate, for good reasons. Google Earth is capable of providing better accuracy without the Geocaching kml. And, when the issue is checking the coordinates for a new cache, the Geocaching kml is rather irrelevant, is it not? You would just input your coordinates and see where Google Earth takes you.

[+Team Cotati]

Google Earth will normally be much more accurate.

 

Please don't give out erroneous information. The OP asked an honest question and there is no need to confuse him.

 

Was that significantly erroneous information or normaly configured erroneous information? Nevertheless, thank you very much for your kind and thoughtful feedback. I'll be sure and keep this in mind going forward as I tip toe through the forums assessing the accuracy of all postings. Thank you.

[+fizzymagic]

Google Earth will normally be much more accurate.

Please don't give out erroneous information. The OP asked an honest question and there is no need to confuse him.

Nevertheless, thank you very much for your kind and thoughtful feedback. I'll be sure and keep this in mind going forward as I tip toe through the forums assessing the accuracy of all postings.

No need to tiptoe. Common sense works just fine. A couple of very simple rules to live by:

• Don't post "factual" information on subjects you don't understand.
  
• Don't post incorrect information in "getting started" forums, no matter how funny and clever you think it makes you look.
  

That isn't so hard, is it?

[+Team Cotati]

Google Earth will normally be much more accurate.

Please don't give out erroneous information. The OP asked an honest question and there is no need to confuse him.

Nevertheless, thank you very much for your kind and thoughtful feedback. I'll be sure and keep this in mind going forward as I tip toe through the forums assessing the accuracy of all postings.

No need to tiptoe. Common sense works just fine. A couple of very simple rules to live by:

• Don't post "factual" information on subjects you don't understand.
  
• Don't post incorrect information in "getting started" forums, no matter how funny and clever you think it makes you look.
  

That isn't so hard, is it?

 

OK, will do. Thank you.

[+Mosaic55]

Around here, south Chicago suburbs, I'm finding Google earth to be very accurate. Streets drawn on the map match up perfectly with the street images. If there's something visible from the sky to use as a landmark, like a sidewalk or bridge, I can just print the map and use it to walk right to within 10 to 20 feet of the cache.

 

Of course, this tactic doesn't work well in the middle of the woods, all the trees look alike from above.

[+alexrudd]

Why not test the accuracy of Google Earth in your area b comparing it to a benchmark of known coordinates or something? I've gotten pretty good results with programs such as USAPhotoMaps and GE in the Chicagoland area.

[+sbell111]

The best advice was given by badlands way up near the top. Rather than worrying about google earth or manually averaging the coordinates, just set the GPSr down for a little bit and let it crunch the numbers. I've found that this gets you the best coordinates. Other cachers and mathmeticians may disagree, but I don't really care.

[+Jhwk]

The best advice was given by badlands way up near the top. Rather than worrying about google earth or manually averaging the coordinates, just set the GPSr down for a little bit and let it crunch the numbers. I've found that this gets you the best coordinates. Other cachers and mathmeticians may disagree, but I don't really care.

ditto - but sometimes I do care, a little

 

Just use your average. Don't trust GE, if you zoom in and out with the geocaches loaded, you can watch them jump around...

[+Mosaic55]

 

Just use your average. Don't trust GE, if you zoom in and out with the geocaches loaded, you can watch them jump around...

 

They don't jump around if you enter the actual coordinates. They only jump around if you rely on the Geocaching.com .kml interface. (as Keystone already mentioned upthread)

 

[ I use GPS Visualizer to make a reliable .kml file from the .loc waypoint files you can download (I'm not a premium member, so I don't know if premimum members need to go through the same routine.) ]

Edited July 27, 2006 by Mosaic55

[+TheBeast]

ITake a look at Calgary, Alberta. Right up the middle of the city, there's a split in the imagery. One side is hopelessly, hideously off. Turn on roads to see what I mean. The imagery is misplaced by hundreds of feet.

WOW! Just had a look and that is REALLY bad!

[+MtnGoat50]

So which is more like to be accurate. Google Earth or my averages?

 

IMHO, you should always use the coordinates from the GPS. Virtually everyone hunting the cache will be using a GPS (I suspect a majority will use etrexs) not Google Earth, so the coords from a GPS will always work better. The Calgary example shows that Google Earth or any map program may be very accurate in one location and way off in another.

 

When I started, I was using an Etrex Legend and I would manually average multiple waypoints. Now I have a LegendC that will automatically average when I take a waypoint. In either case I will try to return a day or two later and check my waypoint against the cache location and if necessary take another average or adjust the waypoint slightly.

 

I like to have the very best possible waypoints I can get for my caches and I've always received good comments on this from the finders.

[+rambeaux]

I 'third' the comments above regarding jumping kml vs. entering actual coordinates.

 

I've done a ton of caches on GE (before I had a GPS unit) and have found it to be, as they say, "dead-on balls accurate" (apologies to Marissa Tomei).

 

The satellite imagery can be a little off, as we've seen, but most of the stuff around here has good data, so it helps. I'm betting the roads are more accurate than the pictures, too...

[gpsvisualizer]

[ I use GPS Visualizer to make a reliable .kml file from the .loc waypoint files you can download

Another useful feature of GPSVisualizer.com: the ability to get alternate map/image overlays in Google Earth. See this page for details: http://www.gpsvisualizer.com/misc/kml_network_link.html

 

For example, up on Mount Hood, I find that GE's fancy color satellite imagery is off by tens of meters; you can easily see that the shadows of ridges don't line up with the ridges themselves! But when I use the GPS Visualizer network link to bring in a USGS 1m aerial photo overlay, it's spot-on.

 

Not as high-resolution as the more recent GE imagery, but far more accurate -- and it's been my experience that the USGS aerials are accurate almost everywhere.

 

(For those of you in Canada, there are a few NRCan overlays available too.)

Edited July 31, 2006 by adamschneider