I found the thing below on Jim's Sinclair's site at http://www.jsmineset.com and wonder if he's serious about this. It's posted on Dec, 1st - not April, 1st !
The belonging article is named "Rangold Analysis Using French Curve" and is in the archive; the URL is too long to be posted here.
Now, is this serious?
I usually just use the top of JJ Walker's hat as an "overbought" line.