Xarax, that first pic in Reply #88 is freakin' sweet! Why didn't I think of that before?
Because you don't think so well? -- or because
thinking isn't
the best way ... .
It kinda takes the fun, the serendipity out of finding ELFEKS
(empirically less-frequently-encountered knotted structures(s))
--a "new"-in-knotting denier's long-winded term-- ,
but one could develop a rigorous checklist to employ for each
new
tangle encountered, to be sure to touch all the bases.
E.g., with an end-2-end knot, one has 4 exit-parts ("ends"?),
in two connected-through-knot pieces, A & B, and assume that
the given knot has a canonically defined loading A1-v-B1,
ends A2 & B2 free; then one can go down the list and check:
end-2-end knots A2-v-B2 (unless highly symmetric like Reef knot),
A1-v-B2, and if asymmetric A2-v-B1 ;
eyeknots A1-v-A2+B1, ...+B2, A2/B1,B2, and if asymmetric B1/B1, etc.
eyeknots by my
"twinning" method (for each case above (!!)),
stopper knots A1->B1, A1->B2, A2->B1, A2->B2, and if ... ;
and that's w/o variances in dressing to consider.
(Such rigor should, e.g., discover
Ashley's Stopper from the
Sheet Bend (or from the
Bowline); or "should" => "might"/"could".)
It quickly gets overwhelming!
(Mind you, I'm not suggesting that one will have a high ratio
of "keepers" in these discovered ELFEKS, but they will have at
least been considered (though even here there is risk of missing
value).)
It's been said that the Honda Loop provides the most perfect circle. Correction, I think we have a knew winner.
Another "round" one is the
Anchor bend as the base, tail brought around
through it, and secured in various ways -- pretty secure w/o tying, which
was part of its creation for a rockclimbing tie-in eye (to guard against
situations in which one forgot to finish the tying). But the Honda's eye
is small and not really in need of roundness; the knot is compact (one
I found has I think just a melted-strands stopper of its end (nylon (VERY
hard-laid))).
And the
Eskimo Bowline. And similar treatment w/
SmitHunter's bend.