All of those knots backwards are the same as those knots forward.
Whoa, bowlines are in that list, and they aren't ... .
[re bowline] Tail outside is weaker. So what. If it matters, your rope is too weak to begin with. You would back the tail up anyhow - assuming you even use bowlines any more - and that also changes the story <grin>.
As Bob noted, those doing tests usually report that there's no difference.
I.p., the once-hosted-@NSS-now-missing Dave Richards report made
the note that the so-called "Dutch"/"cowboy" bowline was tested and
found to be the same strength (but he didn't include this in the
data tables --though DID include "re-threaded" & "on a bight" Fig.8s,
as does CMC's
Rope Rescue Manual!?).
The tail outside is more likely to get snagged and loosen the knot.
Frankly, one can dress the bowline to put the tail in various places
vis-a-vis the plane containing the eyelegs; the draw of the mainline
will move the tail, typically, like one crossing one's legs after sitting
--one can anticipate this in dressing ... . The tail-on-outside knot
resists ring-loading; the common bowline does not. But as Gary
notes, there should be further securing to either of these knots
in kernmantle-rope applications.
Apparently for the knot purest, there is a "right" and "wrong" way to tie the figure 8 bight depending on which side the standing part of the rope exits the knot. Exhibit A - a 6 year old post from our old friend rescueman.
And to different "knot purests" (knot knowledge garners no praise
),
different rights/wrongs : noted was OnRope1.com's Bruce Smith's apparent
denial of Dave Merchant's assertion of how to tie the Fig.8 (a rumored 10%pt.s
strength difference and easier untying), which differs from what Rescueman
Riversong was trying to point out --which might be more a matter of loading,
though Fig.8 dressing is much varied. Now, were Richards & CMC testing two
different Fig.8 versions vs. different ways of tying, we might have some data
to fit into the assertions. (Bruce needs to look more carefully.)
But to knots recognition, I'll borrow my old lamet:
I'm rather dismayed that the clear image of the red Fig.8 loopknot has taken now
several responses to figure out that it is "right", in THIS forum--this knot that
is soooo easy to check!
There is a discussion of the "Sideways Bowline" on this forum that involves the wrong way to tie what looks like a bowline,
... or, rather, the right way to tie
that but the wrong way to tie the bowline!
There are many variations, and I'm having some experimentation
now with TheraBand PT tubing, which is a novel material for me
(doing some PT).
There is a right way and a wrong way to tie the knot called the retraced overhand bend, ring bend, or water knot in rope. I am posting pictures of the right way, the wrong way, and the top and botom views of how the wrong way can distort.
I'm reminded of Heinz Prohaska's article about the supposed vulnerability
of the water-knot-in-tape being snagged & untied (supposedly verified
by testing done by Pitt Schumann?) and (y)our reaction that such knots
once set/weighted sometimes required tools for loosening; I have
similar thoughts about this knot in rope --at least that this supposed
vulnerability to deformation is more theoretical than experienced!?
This knot is commonly used for webbing, for which there is only one way of tying.
Actually, no, there is a way to make it symmetric in tape (too),
which I think you know of. After Tom Moyer reported on the
cyclical slippage seen in
some tape (not in
new 1" tubular nylon!),
where the slippage came only in the tail lying
exterior, I found a way
to have both tails lie
interior (w/o Prohaska's awkward twist);
this makes for a slightly more elongated, less chubby knot.
And so on & on & on ... !
*kN*