waypoint averaging?

[+6coondawgs]

One of my anxieties in hiding caches has been making sure that the coordinates are good. I have been taking 10 to 15 readings by coming into the location form different directions and taking a series of reading at the cache site. What's the best way for averaging a series of readings on a location? Can my Vista HCx average a series of points somehow? Do I just do the math? I have tried putting the points into EasyGPS and then somehow figuring and average point from looking at them on there. So how do professional Geocachers figure their cache waypoints?

[+briansnat]

I make sure I have a good sat lock, let the GPS settle for a minute or so and take one reading and go. Averaging is a waste of time. To do it right you'd have to do it over several days. And after all that what does it buy you? Maybe an additional 5-10 feet of accuracy if you're lucky, which is insignificant in the geocaching world.

 

230+ caches placed and I've rarely had complaints about coordinates and often get compliments.

[yogazoo]

Averaging is a waste of time.

 

It may be a waste of time in NJ, but in mountainous terrain and places that your signals may be weaker and deflected sporadically (canyons, deep valleys, rocks, cliffs) say in the Rocky Mountains, averaging is definately not a "waste of time". Averaging for "days" to do it "correctly" is also innacurate in mountainous terrain. The whole point of averaging for most of us who live in the west is to avoid taking a snapshot during an aberrant signal event. Averaging for a minute or less would be sufficient for this purpose. My two cents.

Edited May 28, 2008 by yogazoo

[bmirak]

Question:

What's the best way for averaging a series of readings on a location?

 

Answer: Don't use a Garmin Colorado. Sorry Colorado owners, just a playful jab. The Colorado actually is a great unit for geocaching - except for not being able to waypoint average when placing a cache, and sometimes not being to find caches because of accuracy bugs. It makes geocaching more fun by increasing the challenge!

 

Seriously, if you've got a Vista HCx, why not simply use the waypoint averaging feature?

[+Gipsie]

On the Vista Hcx, when you mark a waypoint, you can navigate to the "Average" box and hit enter. I (also from NJ, sorry, Brian) set my Vista down and allow it to get between 100-150 readings (it doesn't take a long time) and then use those coords. I rarely have trouble with my coords and check them out a few days later and they are usually spot-on.

[+Berta Nick Zoey]

I placed a cache the other day in a moderately forested area and I sat the GPS down and after letting it average for a minute or two it got down to about 17 ft accuracy. I then lifted it over my head and held it vertically and the average dropped down to 11 feet. I'd venture to guess that if an external antenna (which the 60Csx can use) were employed, and held higher the accuracy may even drop some more. But, 11 feet was more than I had anticipated.

[+briansnat]

It may be a waste of time in NJ, but in mountainous terrain and places that your signals may be weaker and deflected sporadically (canyons, deep valleys, rocks, cliffs) say in the Rocky Mountains, averaging is definately not a "waste of time". Averaging for "days" to do it "correctly" is also innacurate in mountainous terrain.

 

I've placed caches in rocky areas, in valleys and along cliffs and I still think averaging is a waste of time. If your unit has an averaging feature, then by all means use it for a minute or two. Can't hurt. But if you're going to do it manually then it's a waste of time.

 

If your unit experienced an "aberrant signal event" when you marked your waypoint you'll know it the instant you save it and hit Go To. If the area just has a poor signal, then you are simply averaging bad data.

Edited May 28, 2008 by briansnat

[yogazoo]

...and the waypoint averaging debate rages on.

 

One point I'd like to make though. Seems averaging is, or was, valuable enough to include in EVERY HANDHELD GPS UNIT line that has rolled onto the market in recent years by every major GPS maker minus the Colorado series.

 

"Rocky" areas in NJ are, to say the least, very different than the Rocky Mountains. I've seen signals jump wildly standing in valleys and canyons where satellite reception appeared to be at least OK. We get alot of bounce and sometimes positions jump wildly often by dozens of meters. The signal strength and array are all calculated into the equation Garmin uses to perform averaging calculations. I'd sooner allow the unit to take all that data into consideration before I'd just let myself arbitrarily choose when a good time to press the button is. It comes down to personal preference and what works best for each user. I just happen to come down on the side of averaging under my own personal conditions and for mostly why I take data (mostly GIS applications). If forgetting about averaging works for you, god bless, go forward.

Edited May 29, 2008 by yogazoo

[+BlueDeuce]

Take a couple-three marks, walk away and test to see which one brings you closest.

 

(I have never understood how taking 50 readings of the same location got the final figure to go closer towards my cache location. Sats detect my gps?)

[yogazoo]

-------"(I have never understood how taking 50 readings of the same location got the final figure to go closer towards my cache location. Sats detect my gps?)"-------

 

It's simple mathematical averaging weighted with the position error of each reading.

 

The same reason pollsters survey 1000 people and not 50. The more data points the higher the accuracy of the final reading. There are variables (sat. array, deflection etc.) that affect the precision of the final outcome, but the mathematical rule of averaging, the more data points the better, still apply's.

Edited May 29, 2008 by yogazoo

[Hertzog]

 

It's simple mathematical averaging weighted with the position error of each reading.

 

 

I don't think they weight them. I could be wrong, and am only speaking for the 60CS/CSx models. I took a close look at some data once, and the averaging looked to be a simple average of the points involved.

 

By the way, it may be useful to know that when you mark a waypoint the elevation recorded is the barometric elevation, but when you average it switches to the GPS elevation (again, I'm talking about the 60CS/CSx models in particular, but I think it is probably true for all Garmins with waypoint averaging -- except the non-barometer models would of necessity always use the GPS elevation). Probably not a crucial consideration in most geocaching locations, but good to know -- and something Garmin doesn't mention in their manual.

[+briansnat]

"Rocky" areas in NJ are, to say the least, very different than the Rocky Mountains. I've seen signals jump wildly standing in valleys and canyons where satellite reception appeared to be at least OK. We get alot of bounce and sometimes positions jump wildly often by dozens of meters

 

No different here. It doesn't take a 10,000 foot mountain to create multipathing errors. Just standing between a pair of 2 story buildings can do it. Everyone sees their GPS fluctuate wildly when near cliffs and in canyons whether in the Rockies or Maryland. I've experienced fluctuations of a few hundred feet at times when close to a cliff or in a ravine.

 

Of course when hiding a cache in difficult reception conditions one should take extra care with his coords , but we're really talking a small percentage of caches.

 

The same reason pollsters survey 1000 people and not 50. The more data points the higher the accuracy of the final reading. There are variables (sat. array, deflection etc.) that affect the precision of the final outcome, but the mathematical rule of averaging, the more data points the better, still apply's.

 

Obviously taking more data points will increase accuracy, but in reality how much accuracy is needed? We're not surveying, we're geocaching. Close counts in horseshoes, hand grenades and geocaching.

Edited May 30, 2008 by briansnat