> ... I was able to get full scale images,
> which you really need to have in order to do the test.
That's an understatement : in the grapevine images,
I'd say that many of them satisfy the question --it's
a dubious determinant to lean so hard on slight
differences in *bulges* and so on (they are not exploded).
Also, it's arguable about excluding the "wrong" orientation
of the opposed
overhands, though I concur in choosing
the common, "concordant" (C.Warner's term) one.
Btw, I was especially pleased to see the one trick
"fig.8"(rightmost, IIRC), in which one line makes not that but
an
overhand --and the structure is asymmetric,
so one should wonder about which loading gets what
behavior pluses/minuses! (I myself have long thought
that that knot makes a good legend-buster about how
the intended knot is so readily recognizable : I surmise
that one could present that bogus knot to many who
assert this characteristic and prove
them wrong!)
--dl*
====