EDK, again. I know, I know.

[HPTW]

I know that the European Death Knot debate is always rehashing the same old stuff, but...
I was with a friend about a month ago who said he refused to use a EDK (or offset ring bend) for rappelling, because he had known somebody who had known somebody etc... who had it fail.  While I think the EDK is fine if tied well, especially with long tails, I was messing around with ways to make other simple, small, offset bends.  I capsized a loose butterfly bend away from the ends, essentially reversing the knot, and then tightened it very well (this is essentially, I think, like using the "tails" of the butterfly bend as the tensioned portion).  If well tightened this creates an offset bend.  However, if it does capsize, it capsizes into the butterfly bend's normal confirmation.  Has anyone ever messed around with this?  Just wondering.  Thanks

[Dan_Lehman]

I know that the European Death Knot debate is always rehashing the same old stuff, but...

Sometimes as though people are talking to walls, it seems.
And things are rehashed partly for want of the Search function.
But there are some URLinks to point to, at least.

cf http://www.postimage.org/image.php?v=PqAodEJ

There is a good presentation in two photos of two orientations of one
particular Offset Water Knot w/Overhand tie-off (for want of brevity).
I.e., the very knot shown at left/right could be just rotated into the
other's orientation, even after some loading.  (Now consider whether
any of the knot testers realized this (I strongly doubt it --some are slow
to realize even if it is THE topic of conversation!).)

The thinner rope is positioned so as to be hard to get pried out
around the thicker; and the thinner rope's tail (given this position)
is what is tied off around the other's --this rope's part of the
knot is what must be opened and pried around the thicker
rope's for the knot to "roll", and both the thin/thick difference
AND the particular tail tie-off impede if not preclude such rolling.
(And given the asymmetry of the knot, there is no need for both
tails to be tied of, as in the promoted back-up EDK; so, spare that bulk.)

... because he had known somebody who had known somebody etc... who had it fail.

Myths grow from this sort of hearsay, devoid of hard evidence & analysis.
But one can, as shown above, secure the basic knot, just as one can do
with the venerable Bowline --and keep the benefits of both knots.

While I think the EDK is fine if tied well, especially with long tails, ...

I object to this qualification : it is fine, or it is in need ... --take those tails
and DO something with them to rid the recommendation of this caveat!
(One can read of errors made with the long tails, mistaken as main lines.)
(("This road is safe, just carry a machine gun & bullet-proof vest ... ."))

I was messing around with ways to make other simple, small, offset bends.
I capsized a loose butterfly bend away from the ends, essentially reversing the knot,
and then tightened it very well (this is essentially, I think, like using the "tails" of
the butterfly bend as the tensioned portion).  If well tightened this creates
an offset bend.  However, if it does capsize, it capsizes into the butterfly bend's
normal confirmation.  Has anyone ever messed around with this?  Just wondering.
Thanks

Fortunately, yes, I have --with Ashley's Bend #1452 and its lanyard form
qua offset bend.  And I learned in a safe though accidental way an
important lesson : the knot can partially capsize --i.e., one side
goes while (or whilst) the other doesn't, and <slippppwhishshshs!> the
ropes are separate !!!  THAT was a surprise, with a bang up into a table
beneath which I'd hand-vs-foot tensioned this knot.

So, stick to the simple OWK with a back-up (if desired), which can be
well tied in the common materials at issue (more concern arises for stiff
rope which will leave too much air in the knot and enable too much
collapsing & shifting & treachery.

For other like offset knots that are secure, consider these:

http://www.postimage.org/image.php?v=PqAoxBS

The upper one is the infamous "EDK" but where the choking line (thinner,
orange) makes a full turn before being tucked out --it is thus
making a "Figure 9".  (And whereas an Offset Fig.9 has been tested
ages ago by then Black Diamond's Chris Harmstron and did well,
here too one can pare bulk by recognizing that it is only the one
rope that needs the extra turn, as it guards the *inner* other knot.)

The lower one is the EDK with just a half-turn extra, making a Fig.8
rather than Fig.9 --and hence the tails exit in opposite directions.
Yes, the Offset Fig.8 has a bad reputation, as having been the actual
"EDK" that failed and killed one climber (British man in Zion Nat. Park, USA),
though that is likely overstated.  With the ends exiting as they do in
this hybrid, you'll find the prying-out-&-around capsizing action
greatly impeded.

With your own ropes, you can give whatever knot you are considering
a pretty good test using something to give a 2:1 (crudely) mechanical
advantage.  Consider that abseil use should split your weight and some
added force from movement across the two lines, and the knot being
in one side of the double-line abseil will see half the force.  Using a
crude 2:1 (actual MA about 1.5:1 say) pulley will approach triple
the expected load on the knot (and you can bounce, deliberately
spiking the load).  (Break tests are NOT much to the point, here.)

--dl*
====