xpyttl wrote:

Random length, of course, is random. you COULD get lucky. Or not ....

I guess I'm lucky. I've never had a problem matching a random-length
dipole on any frequency I've tried. That does not imply, of course,
that any random length can be successfully used on any arbitrary frequency.

And of course, "random length" should have read "as long as possible,
given the placement of the trees."

Thanks to everyone who chimed in on using an external balum and an
automatic tuner. My gut told me that the consensus ("it won't work")
was correct, but it was nice to get some more factual backup for that.