Thank you, SS369,
The picture shows a tight knot where the tail has settled in a position "over" the returning eye leg, not "under" it, as it should have had. That may explain the dragging of the tail under high load you report. This 8 Loop is a very simple knot, and each and every element of it should be utilized, in order to work properly.
The tail should be squeezed in between the standing part s first curve and the returning eye leg s first curve - otherwise it will not be locked properly. That was the initial idea : to have a horizontal, "Eskimo"- like tail, something like a "handle" or a "step", that would be "lifted" by the standing part s first curve, from below, and, at the same time, would be "trampled" by the returning eye leg s first curve, from above. Missing this favourable position, the tail at the knot you had tied was left without adequate support.
... as I stress it, the tail moves to a sort of parallel position
Why is this happening ? Why the tail is twisted like this inside the knot s nub, and settles at this position parallel, or even "over" the returning eye leg, on top of it, instead of remaining "under" it, below its belly ? My first attempt to an explanation is that this twisting is due to the constant contact of the tail with the standing part, that spins / revolves the tail during the tightening phase, and inserts some considerable amount of torsion in it. So, as the knot is still shrinking but has not become 100% compact yet ( there is still some room in it, so the various segments can "move" relatively to each other ) this torsion manages to force the tail be displaced inside the nub, away from its initial "correct" position, to this final "wrong" position. I guess that one should pull the tail downwards, once or even twice during the lightening phase, to counter-balance this tendency.
With the rather stiff, in-compressible, round ropes I use, I have not encountered this problem. ( See the B&W pictures of a loaded knot, tied on a 8mm cord, loaded with mine s and my son s weights
- the tail remains well underneath the returning eye leg ). However, we wish knots that can be tied in most materials, even on this pink slick compressible rope you show - so this displacement of the tail during the tightening of the knot should concern us, indeed. I can only say that the full, "finished" 8-8 knot ( where the returning eye leg and the tail, moving in this wave-like motion, penetrate both openings of the shape "8" nipping structure ) would not behave like this.
Now, this 8-8 loop may be more robust, and very beautiful, but deviates substantially from the knot I had in mind when I tried to stabilize the "double, crossed coils nipping loop". The returning eye leg is tucked once more, and so it pulls the whole nub downwards - so the favourable horizontal position I was seeking for the L-shaped segment of the returning eye leg and the tail, the "handle" or "step", now runs the danger to disappear altogether ! The knot resembles more a "double Myrtle" loop, that a fixed "double crossed coils" loop. What can I say ? You win some, you lose some !
In such simple knots, even the slightest unbalance of forces can force the nub to rotate towards the one or the other direction / side. The mere presence of the bulk/volume of a rope segment in one place, that forces another segment going around it to be displaced a little bid, can cause a rotation, which might twist one segment, or even the whole nub ! Moreover,
when we do not have a genuine collar ( a "proper" bowline collar or even a "Myrtle" collar), but a so-so "collar", as in the present knot, that turns around the standing part within the knot s nub ( and not "before" / over it, as it happens in the common (+)bowline, or "after" / under it, as it happens at the "Eskimo" (-)bowline ),
we also do not have the required leverage that would stabilize an elongated nub in the required horizontal position.
Returning to the original 8 loop, I think we should test it on a variety of materials, before we kiss it good bye !
I would nt like a knot that would require a constant pull of the tail, or a careful dressing
during the tightening phase, in order to lock correctly. Myself I have tied it on about two dozens of different ropes, without noticing such a severe twist of the tail inside the knot s nub - but I use only rather stiff, torsion-resistant, round, kermantle climbing ropes, so my tests are not very inclusive / conclusive.
When somebody presents a "new" knot in the Forum, what does he wish for it ? I can not tell what other knot tyers are thinking, but every time I happen to do it, I do it with the same hope ; to persuade somebody to test it in detail, or find a reasonable argument, that would suffice to make me dismiss and forget it !
We already have too many knots, and I always hope we have already reached the end of the line . It would be nice to know that there is not any other good practical knot looming under our nose - so we would be forced to concentrate on the most important thing we still miss / avoid :
KNOT TESTING !