I fear I've opened a can of worms with my question. Very interesting read on all of your comments. If and when I get this together it will not be as hard as several have suggested. No math, just a suggestion as to where you might want to park and then go to approximate center. Thanks again, Bob Smith, 1/2 of Team Petite Elite David Thompsen wrote: > I've done a lot of puzzle caches, and one thing being left unsaid, > maybe it's obvious, it that if you want to work the math... convert > all waypoints to UTM. This gives you nice (x,y) coordinates to work > with, which respond nicely to geometry and trigonometry. > > As for the triangle, yeah, there's a lot of ways to define the > "center". Intersection of medians, orthocenter, and circle touching > all three points of the circle. I've seen puzzle caches using all of > the above. > > --Dave > Team Cowspots > > Steven Stringham wrote: > > When I did orbis - in Tucson, I used Autocad to calculate the > center of > that triangle. The radius of the circle is only about 2 miles. > And, that > had me off (because I was not taking the curvature of the earth into > account) about 900'. I then recalculated, and was still off, but > came up > with another way to figure it out (because of the way the points > are layed > out, it allowed me to do some triangulation on it). > > Good luck. > > http://www.geocaching.com/seek/log.aspx?LUID=e0d67a76-96ce-4797-a186-0de15bfff37a > > > Robert & Linda Smith wrote: > >> Thanks Brian, > >> The page you listed is a very interesting read. I really enjoyed it > >> (didn't understand all of it) but enjoyed the read. I don't > need to be > >> this close and accurate. A few feet this way or that is fine. I > just > >> wanted to be close. > > > > Hopefully you are talking about 3 points that are fairly close > together. > > The error actually gets large rather quickly. > > a couple points that are half the height or width of arizona > apart can > > have an error of 1/4 mile if you don't calculate it correctly. > > > > Brian Cluff > > Team Snaptek >