So in attached image I made a 2 inline fig8 bend. It would seem to me this would enhance security a great deal, what say you?
//
A better, more elegant knot is one Fig 8 Bend backed up with Strangle knots on each side
Firstly, please observe that there are two distinct "fig.8"
knot-dressings shown in the two images above --and the
URLink'd one is slightly different in its exact dressing from
the asymmetric one of the pair shown embedded above!
--this so-"easy-to-recognize-that-it's-tied-correctly" knot!!
(And, of the embedded pair of knots, we can't tell which
ends --upper & lower-- are (to be) loaded, which ... tails.)
It would be a worthwhile research project to study various
user who tie "fig.8"s both "re-threaded" & "in the bight"
to see if they end up with the same dressed knot --or not.
(As we see in the examples above, we got botched fig.8s,
by one tying method and even with one tyer!)
As far as
elegance is concerned, to my mind the
"twin
fig.8 knots" end-2-end knot has the edge --not much
to simply tying off tails, esp. at such remove from the
base knot (which might be an un-re-tucked
thief knot(pointing out that the
fig.8 end-2-end knot can be seen
as a further tucking ... !), and then some tension would be
brought to the
strangles).
Note that, in making a "
twin" <eyeknot> end-2-end joint,
the loading in the two eyeknot components (which might be
different --not sure how this will affect the
elegance quotient

)
will be that of, well,
eyeknots, not end-2-end knots.
Here, i.p., there will be the loading on both end-most collars
and so possibly easier untying; possibly greater strength, too.
(But, oddly, in one shade-tree tester's set of results, where
in each test specimen he tied competing end-2-end knots
and then the specimen's ends were tied in
fig.8 eyeknots,
(a) the eyeknots
never broke/failed; and the
(b)
twin fig.8 knots end-2-end knot wasn't as strong
as some others!? (I.e., it is essentially two eyeknots and yet
it
did break, but never did the equal eyeknots of the
specimens' ends break!? --might point to Knot4U's doubts
about setting with equal tension (or there might have been
some differences; w/o examination of the actuals I cannot
say (though I vaguely recall raising this point and there
didn't seem to be a good answer)).)
Further, as to
"this would enhance security a great deal",
in order for there to be a great deal of that, there must be
a great deal to do! Given that the
fig.8 end-2-end knotis not known to be a security risk, it's arguable that one
cannot make a great deal of improvement. (And a similar
point can be seen in some cases of strength : given some
85% strength efficiency, how much better can it be?!)
--dl*
====