Averaging

[Clan Riffster]

Hey Gang! Me again...

 

I've been experiencing some accuracy difficulties with some of my caches, and I often hear folks saying they "averaged" the location X amount of times. I'd like to describe how I create a cache waypoint, so y'all geoexperts can tell me;

1) What I'm doing wrong,

2) How to do it right.

 

First I scout out what I think is a likely location for a cache, then after placement, I set my GPSr, (Magellan SporTrak Color), over the spot upright. I verify a 3D reading, and confirm at least 4 satelites. Then I let it sit nice and still for 1 minute before hitting the "mark" button. After I've created my waypoint, I walk off about 50' and do a "Go To", back to my waypoint. At this point, my accuracy is typically within 10', which I find acceptable, concidering how often I've dropped, submerged and run over my Magellan.

 

So what's the problem you ask? Well, the next day, (week/month/etc), I'll return to check my numbers, and on many occasions I've found them to be over 50' off. Third trip will sometimes have me on top of the cache, or far away. Typically I'll stop about 50' from the cache and let my GPSr get better readings, but it doesn't seem to help.

 

I figure if'n I could learn how to "average" a set of coordinates, I would increase my accuracy.

 

On account of my not being the brightest crayon in the box, please try to keep your replies monosylabic.

 

Thanx!

 

Sean, of Clan Riffster

 

(BTW, why does monosylable have so many sylables?)

[+JohnnyVegas]

How long is your GPS runing before you take a reading, I like to keep mine turned on while I look for a place to hide a cache. Just turning a GPS on when you find a place to hide a cache and leting in run for a few minutes is not enough. My guess would be to lhave it running for at least 15 minutes, but the longer the better.

 

Also, heavy tree cover or being around lots of tall buildings can be a problem.

[+briansnat]

I'm not sure you're doing anything wrong. When they say they are having difficulties is it because of the coordinates? And if it is, how far are they off? 25-50 feet is very common.

 

Personally, I don't average my coordinates. If you have a good sat alignment and signal, all you need to do is take one reading and run with it. At one time I would mark numerous waypoints at the cache site and average the results. My coordinates were no better then than they are now that I don't average. I've placed over 125 caches and have had complaints about the coords of maybe 5, some of which were averaged and some of which weren't and in no instance did the "bad" coordinates keep anyone from finding the cache.

Edited September 19, 2005 by briansnat

[+Prime Suspect]

What briansnat said.

 

Few people have the time and patience to make the number of reading needed to have a statistically significant beneficial effect. If you want to camp out and take readings every 15 minutes over a 24 or 48 hour period, then yeah, you might get marginally better results.

 

One thing I've noticed is that once one person reports that the coordinates are off (even a perfectly normal 20 feet), then suddenly everyone else who logs that cache feels compelled to also report how much "off" the coordinates were. It makes it appear that there's something wrong, when usually there's nothing wrong at all.

[twjolson & Kay]

Yea, averaging isn't a big deal. Even if you do get the most accurate coordinates you possibly could, the very next person that will come will still have an error box of 25 feet at best.

 

However, if you want to do it, here is how I do it. When you have the spot for the cache, put the GPS over that spot. Wait five minutes then mark. Get up, walk around, or walk out to the trail, then come back and put the gps over the spot again, wait 5 minutes and mark. Get up and walk somewhere else, then come back and mark it again. I do this 3-4 times a day, for 3 days, depending on how accurate I'm trying to get it. Then when I get home, it's just simple math averaging to give me my final coordinates.

[+bilgeratt]

Well, as others have said, I'm not really sure you NEED to average. However, right or wrong, this is what I do.

 

Once I pick my spot, I place the GPS there while I ready the cache. Since I prep my caches on location, this gives the GPS at least 5 minutes of sitting perfectly still. When the cache is ready, I WRITE DOWN the coords showing on the GPS on a seperate sheet of paper. Then I pick up my GPS, take a walk away some 75 to 100 feet, and return and write them down again. (usually just the last 2 #'s, nothing else will change).

 

I repeat this excercise at least twice for all 4 points on the compass, if at all possible. So I wind up with 9 or more readings counting the original. I throw out the highest and lowest reading on both lat and long. Then I add up the #'s and divide by the amount of #'s I added to get the average.

 

The funny thing is, the average is usually within 1 or 2 digits of the original reading I took. That's why I say I'm not sure you need to average. I still do it to be safe, mainly because my GPS is a very low end one. But I really feel like I'm just overkilling it.

 

Anyways, good luck.

[+tands]

Also, some folks have newfangled GPS units that will average for them. But they don't realize that 5 minutes of averaging is still not enough time to really get things right. Yet they still say their average shows your coords are off. Not cool.

 

- T of TandS

Edited September 19, 2005 by tands

[+briansnat]

However, if you want to do it, here is how I do it. When you have the spot for the cache, put the GPS over that spot. Wait five minutes then mark. Get up, walk around, or walk out to the trail, then come back and put the gps over the spot again, wait 5 minutes and mark. Get up and walk somewhere else, then come back and mark it again. I do this 3-4 times a day, for 3 days, depending on how accurate I'm trying to get it. Then when I get home, it's just simple math averaging to give me my final coordinates.

 

You need a hobby!

[Clan Riffster]

Thanx guys & gals! I'll try it & see what happens. If all else fails, I'll just blame my Magellan.

[CoyoteRed]

...I set my GPSr, (Magellan SporTrak Color), over the spot upright.

A lot of people don't understand the problems associated with the SporTrak. You have the color version which I'm not familiar with, but if it's like the Map or the Pro the unit actually acts very sluggish. In a way, this is a good thing.

 

These units auto-average a couple of levels. It is always doing a bit of averaging which I guess gives it its sling-shot effect. That's the effect where if you simply walk up to the waypoint according to the unit you will find you've overshot the mark.

 

The other averaging is what most modern units can do, but the SporTraks have theirs turned on by default. All of this "take a waypoint, move away, come back, take a waypoint" mess is actually not a good way to get a mark with the Pro or Map--maybe any Magellan that uses the same or similar algorythm.

 

Long story short: make sure you have good signal as you are doing. Sit the unit down--and still have good signal (I hang it from a high thin branch)--and let it sit for 5 minutes. Move the unit just enough to break the autoaveraging and then put it right back. Let it average for 10-15 minutes if you have a decent number of satellites, longer if the signals are a little sketchy. Mark the waypoint. You're done. That whole time the unit is gathering waypoints, one per second I think, and averaging them for you.

 

You can do the same thing finding the exact--as close as the unit can tell you--the coordinates the other person got. Use this on a cache where someone else did the same thing and it's almost like cheating--if you consider having a very high confidence that when you zero out the cache should be within an arm's span distance.

 

I encourage anyone who is concerned with their marking skills to do whatever exercise they choose in an area close by, like their own backyard, so you can easy do this after dinner or whatever. Do both taking a waypoint and finding one. This will boost your confidence you are not sending someone on hunts with 50' errors.

 

I've personally done this experiment where I took waypoints of a couple of fence posts in my backyard and was getting 0.001 repeatability (which is different from, and more important than, accuracy). I did this several times over a period of a few weeks to simulate marking a cache and others coming to find it.

 

Hope this helps.

[+reveritt]

I have a Sporttrak Map, and has been mentioned, WAAS averaging is the normal mode. The unit goes into averaging mode automatically whenever I am standing still. I find that 1-2 minutes is plenty of time to let the unit average when taking a waypoint.

 

An exception would be when you have poor reception, which often seems to happen under a heavy canopy of pine. In that case, I would average for much longer (maybe 10 minutes).

 

See your manual for a discussion of averaging, and how to tell when the Sporttrak is doing it.

[+Yamahammer]

One thing I've noticed is that once one person reports that the coordinates are off (even a perfectly normal 20 feet), then suddenly everyone else who logs that cache feels compelled to also report how much "off" the coordinates were. It makes it appear that there's something wrong, when usually there's nothing wrong at all.

Pretty much is the case.

 

[+JohnnyVegas]

Yea, averaging isn't a big deal. Even if you do get the most accurate coordinates you possibly could, the very next person that will come will still have an error box of 25 feet at best.

 

However, if you want to do it, here is how I do it. When you have the spot for the cache, put the GPS over that spot. Wait five minutes then mark. Get up, walk around, or walk out to the trail, then come back and put the gps over the spot again, wait 5 minutes and mark. Get up and walk somewhere else, then come back and mark it again. I do this 3-4 times a day, for 3 days, depending on how accurate I'm trying to get it. Then when I get home, it's just simple math averaging to give me my final coordinates.

Every time I had gone looking for a cache when someone has post that they took several readings and then averaged them out to get the BEST reading I have found that those caches have always had the worst coordinates . With all the caches to be found, Now when I see some one post that they took several readings and they selected the best one, I just do not look for that cache. Because I know the coordinates are going to stink

 

Just make sure that if your GPS has WAAS that you let your GPS run for at least 15 minutes to allow it to build an almanac, then you will have very good coordinates.

Do not turn off WAAS in your GPS unless you want to by into the junk science that WAAS creates errors .