Thank you DDK,
The WEs must be configured as in the Whatknot/Grief Knot ABOK 1407 as mentioned previously. Otherwise, this bend becomes a very simple and insecure interlocking of overhand knots.
After tying and trying almost*
all the possible bends based upon interlocking overhand knots -, I have only this to say : Even in the case of those "elementary" bends, we can not predict, in advance, if a "simpler" bend will be more secure than a more complex one - and we can not even define what "simplicity" means in this case... The apparent. pictorial simplicity is, in fact, misleading : the mechanism is more complex than it looks ( probably because we can not "see" and evaluate the contribution of
torsion ). We need
tests, and we need
people able to perform those tests...otherwise all predictions are condemned to remain highly hypothetical and uncertain.
Anyway, in the above mentioned bend, my aim was to retuck the working ends in a way that would leave the collars as large as possible . So, the original purpose was to tie a Hunter s-like bend that will be difficult to jam and easy to untie - not a more secure interlocked-overhand-knots bend.
* I say
"almost", because we do not have a systematic method of enumerating all those bends. Even the "mathematical" approach of Roger E. Miles, based upon the devise of symmetric planar diagrams and the
"invention of suitable lattice walks"(p. 143), is in fact heuristic, not exhaustive.