A simple adjustable loop.

[xarax]

   With a tensioned standing end, if you pull the tail towards you, it becomes quite easy to adjust the loop. Once you adjust the loop in the desired position, if you leave the tail free to point to whatever direction, the working end of the line tends to make an L, or even a U turn inside the knot s nub, as designed, so the nipping structure of the knot can hold it without much difficulty.
   ( I have tried something like this elsewhere in the past (1), but now I use an even simpler structure, without any of the collars used there.)

1) http://igkt.net/sm/index.php?topic=2996.msg17831#msg17831
    http://igkt.net/sm/index.php?topic=2996.msg17841#msg17841

[xarax]

   A second variation on the same theme, is shown in the attached pictures. It is expected that those knots are sensible to the width of the bight, the strength of the pull, as well as to the used material. In comparison to the first, I have found this second variation to be more stable, but less effective in securing the tail . The interested reader is kindly requested to report his experiences with the loadings and the material he used.

[Dan_Lehman]

Interesting that you took the dbl.overhand knot in the form of
a *reversed* anchor bend ... !?  Heinz Prohaska some time ago
presented a similar knot, using the anchor bend right-side-around,
with the idea being that it should grip the tail adequately (for some
rockclimber incomplete tie-in) without further precaution --but
some further tie-off was expected.  (This was motivated by the
rumored uncompleted-fig.8 eyeknot-tie-off of Lynn Hill's,
leading to her frightening fall (though in fact, as per her book,
she simply tied a **nothing**, intending a bowline ).

--dl*
====

[xarax]

   Thank you Dan Lehman,

the idea being that it should grip the tail adequately (for somerockclimber incomplete tie-in) without further precaution --but some further tie-off was expected.

  I do not expect such a simple two-coils knot to grip the tall "adequately", just "satisfactory well", until it is secured even further by two half hitches, for example. The main idea is to device a mid-line knot that, when tightened, forces the main line that goes through it to bend, to make an L or U turn, so it becomes far easier to block its movement/slippage. I have done something like this for an adjustable binder, in (1),(2), using similar, but more complex, collared knots.( See attached picture). Now I have seen that the collar is perhaps redundant altogether... Of course, those simpler knots are not so stable as their collared versions, (shown in (1), (2) and next two posts), but I was surprised of how well they do almost the same job, with even simpler, almost elementary, means...
   As you have probably guessed by now, my main purpose was always a knot for a binder like the one shown in the attached picture....because there we have the mechanical advantage offered by the presence of the two connecting lines, so the grip of each knot on the line that passes through its nub becomes even more effective. This two-connecting-lines configuration can be replaced by a more naive two-interlinked-loops one, because there we have also two knots each griping a line that bears half the total tension.  
   I would nt characterize this knot, functioning like this, as a "reversed anchor bend", as it can not be a bend : it is the gripping structure of a "jam" noose. (in the sense of ABoK#1228)
    
1) http://igkt.net/sm/index.php?topic=2996.msg17831#msg17831
2) http://igkt.net/sm/index.php?topic=2996.msg17842#msg17842

[xarax]

   Instead of (adjustable) loops / "jam" nooses, we can use pairs of those knots as base for binders. They hold surprisingly well, given the simplicity of the nipping structures. The first binder is based on the double overhand knot, and the second on the "8" shaped crossed-coils double nipping loop.(See the tying diagram)
   The crossed-coils double nipping structure holds very well AND it is very stable (much more than the double overhand knot !) when loaded from all the three ends (the standing end and the two legs of the bight.)
   (See the attached pictures).

   Warning : The tails in those adjustable binders should be left much longer than those shown in the pictures- because some slippage is expected in the initial stages of the tightening process, before the knots "lock" around the tails that pass through them. In the pictures below, I show very short tails, for presentation purposes.

[xarax]

   The second adjustable loop (the "B") presented in this thread, is based on a crossed-coils, double nipping loop, aTIB knot. So, we can transform this loop further, add a collar, and get a crossed-coils double nipping loop bowline. In the attached pcitures, one can see another interesting and secure loop, where the bowline collar is replaced by a hitch, that goes through - and is interlocked with - the nipping structure.