Lets make a list of our worst knows (from all of them - hitches, bends etc).
Choose minimum 1, according to your experience and if u like, write something about it.
My choice:
I never worked for me. I tried it with different kind of ropes, and it was a fail everytime.
I’ll contribute.
ABoK #1875 Blackwall hitch.
Not truly a “worst”, but not one I’d trust, even temporarily.
SS
I fully agree on the Tautline hitch, I can’t recall it ever not having slipped under tension.
The Farrimond friction hitch is far superior to it in my opinion.
!!? Well, I use it often, and it works often. (Absent
ABOK now, I don’t know if the cited knot is one that
essentially makes an adjustable eye knot, or is purely
hitch to rigid or flexible object.) I frequently tie it in
small cords, binding this or that to a backpack for
transport 'a velo. I even use it in lubricated polyester
webbing, and here it’s certainly a YMMV sort of thing,
where one considers needed security vis-a-vis forces;
it might need some help.
One can finish the knot with a slip-knot stopper (not
trusting that finishing half-hitch to hold); one can add
an additional turn, or even guard the knot with a
round-turn nipping loop taking the main load.
To you naysayers, what do you (try to) use it for that
it fails, and with what do you replace it?
As for my worst knots, there are probably many candy dates
among the innumerable new things I keep fiddling!
I used rolling hitch http://www.animatedknots.com/rollinghitch/index.php to hold my tarp once, near my house and it wasnt holding too long 
Everytime I tie tautline only for testing/learning stuff, it fails miserably..
The construction of this hitch is not closed/compact. Last turn, that is separated with first two, makes this hitch weird and not trusty imo. Even tying this one is complicated for me because of that floppy structure. You have to pack and correct it a few times. Maybe I am doing it wrong..
On such ocasions, I use adjustable grip hitch https://en.wikipedia.org/wiki/Adjustable_grip_hitch or for best security blakes hitch.
To my mind, knots are not good or bad in and of themselves - merely suitable or unsuitable for a given application.
I’ve used the tautline hitch (when appropriate) with success countless times over the past 40 years, and I’ve employed other knots when I felt it was not appropriate.
Regards,
Eric
Same here, many hundreds of times. Now, for slippery stuff one needs to add a turn (or two) to make it hold. That said, the Farrimond looks interesteing.
I’ve also use the Midshipmans hitch (ABOK 1027), sometimes doubled, with a hitch at the end, and that has worked fine.
You don’t say in what materials, but that matters.
(I do see slippage in that polyester cable-hauling tape
cited above. Oh, and for that --and more often than not–
I use the version of building a coil and not the one you
link to where the follow-on turns are cast up into
the knot, and which form is recommended for hitching
to cordage vs. rigid objects.)
Extra security can come from putting in a full (2 x 360degrees)
turn and leading the tail up to THEN tie off with some version
of rolling h..
Yes, that closing & spanning half-hitch doesn’t give good
security; to this, put in a stopper in the tail, possibly by
half-hitching with a bight and so making a slip-knot in
the bight tail (where pulling the slipped tail spills first
the stopper then the hitch!).
On such occasions, I use adjustable grip hitch https://en.wikipedia.org/wiki/Adjustable_grip_hitch or for best security blake's hitch.I've had [i]Blake's/Prohgrip hitch[/i] fail to hold, initially, and to remedy that I *loosened* it a little, so to enable the near/entry coil to open a bit (larger helix). The thing about such "coil-away" hitches is that they do depend on such *expansion* to build grip, whereas the coil-back ones depend on tightening the coil by tension through the line, and are better at holding --if they can-- right where tied.
I agree on tautline. Blake Hitch is far superior.
Bowline is not one of the worst, but it is overrated. There are a thousand different add-ons to prop up this weak knot that cannot stand strongly on its own. Bowline is useful to me as a quick and easy loop, but I do not trust it for anything important. Because it is so popular and easy to tie, the general public mistakes this knot for being great.
For more important tasks, Double Dragon is a better option. Not only is the base structure more secure. It can also be tied in the bight or as an end loop. It holds a small tight bundle and does not jam.
I like the tautline hitch. I use it in my initial pass to tie up a stack of newspapers. It does not slip when i bring the cord around perpendicular to that first pass to complete the bundle tying process. Tony
Taut Line Hitch is awful I agree
The Adjustable Grip Hitch is the best friction hitch though
I use it for everything
Clove hitch is terrible as a standalone hitch on a bar or other round anchor point. By terrible, I mean there has always been a better hitch for me to use in its place.
Clove is a fundamental knot that is taught to everybody learning about knots. However, Clove is only useful to me as a structure within another knot (e.g., working end of Buntline, working end of Roundturn and Two Half Hitches, etc.).
The rolling hitch (I take “tautline” to mean the eye
structure formed using it) has the advantage of gripping
immediately where it’s set, whereas for coil-away
hitches such as the Prohgrip (Blake’s) there often needs
to be some extension of the coil to build gripping force,
and this can make setting it snug to something difficult.
Except that it does stand strongly on its own over
a vast amount of human experience! You think you
can just ignore that, given its challenge in some
materials? (And as for “weak knot”, YMMV here, too.)
.:. One cannot attribute everything to the knot,
to the structure/shape, independent of tied materail!
I should contribute a candidate to this thread which
is something I’ve (& Roo) argued against (!) ::
the highwayman’s hitch.
This supposed used-for-a-quick-getaway tying
is but one of myriad ways of forming such a hitch,
and it suffers from a vulnerability to capsize,
as the knot brings its heaviest loading upon
the “slip-bight” toggling the structure,
which can be folded and drawn through
the “frame” it’s braced against.
My initial solution to redress this knot’s problems
was to reverse the orientation of the 2nd-cast
bight (a common tying method works by bight
upon bight with a finishing tucked bight qua toggle) :
rather than put this 2nd-cast bight around
the first (which is formed in the SPart and will
deliver the hard force upon what it surrounds),
I put the 2nd bight into the first, and then the
toggling 3rd, “slip-bight” was tucked into this
2nd one.
.:. The SPart bears heavily upon the “frame” bight
and less heavily against the slip-bight.
(There is now before me a variation in which the
SPart’s bight is crossed, ergo a “turn” --indeed, like
the bowline’s nipping turn–, and then the
tucks are as above in my revision. For this hitch,
one needs essentially to pull out the slip bight
toggle AND the surrounded frame --the nipping
turn can grip it as with a sheepshank on its
own, so one will need to pull it out.
(I’ve not played with this enough to assess its
stability --something a quick check appears
to question!)
My candidate is tying two ropes together by making a single overhand stopper knot where those two ropes is parallelled. Just go to any store and see what knot is used - that one.
That is actually probably the knot one will use if not learned how to tie any other kind of knots.
Hmm, if I am not misinterpreting what you are saying, (overhand bend), that knot is what I and others have used for a climbing rope bend for many years. I don’t think it is so bad… 
SS
A pic or link is needed. I can think of 3 different bends that might fit your description.
Taut-line is the basic starter friction hitch for many; especially most olds-cool tree climbers.
.
But, we don’t use it piggybacked on to a host line; in a system only offering a mono-support leg.
.
We terminate the end of a line to carabiner to connect to saddle; and leave a long bitter end.
The Standing Part serves up and over a support branch, and then back to climber.
Then, with the long bitter end/tail on climber side, tie Taut Line to control side of line, and give stopper after Taut line.
Should place stopper at end of control line so Taut Line can’t come off line
.
The mechanic is totally changed when shared support; now when move Taut Line to allow more rope in loaded part of system
the load switches over to the ‘solid’ leg of support, unloading the Taut Line to slide easily-er.
Also, the line ‘below’ Taut Line is unloaded/fat; in relation to the loaded/skinny part of line, offering shelf too i think.
.
i don’t have my ABoK with me ; but believe the Taut line is pictured as a tree climber uses.
Materials and these mechanism make the difference i think; also i think best if support is frictional, not pulley.
.
Another mechanic hidden here is can grab control leg of line and pull self up and have 2/1 (+friction) over own self.
Watching descent, can clearly see 2 legs of support to climber, feeding from 1 leg of free line under climber.
100# climber places 50# on each leg to him/her; and then only has to pull another 50 to control side to lift 100#..
in descent once again favor frictional / not pulley support, so less friction/heat in Taut Line especially on long descents, spiriting down nicely!
.
edit: to #1856 style usage of Tautline concept as a Rolling Hitch, i can see some problems; especially in slicker lines compared to Natural fiber. 1 less turn and not the same mechanical force flow thru rope architecture.
ABoK lessons 480/1/2:
Shows Taut Line usage as i’ve known it and so many tree climbers start this way i have noted it as taught line
especially in slicker lines of today, would say to have stopper after Taut line so can’t pull out especially in the clove-ish continuous direction form 480 (as opposed to the cow-ish reverse turns of 481), and another on the end of line as a stop to the slide so you don’t run off the track/ nylon highway!
Ascending is a 2 handed operation holding Taut Line and pulling out more ‘bitter end’ ; unless place small pulley or carabiner under Taut Line ; so just pull ‘tail’ /bitter end and device acts as ‘knot tender’ to serve friction hitch higher 1 handed.
i always prefered a DBY with long bitter end that becomes friction hitch to control side of the folded S-Part.
Other call-outs:
479: Timber Hitch nip carries to HH best nip zone properly close to peak of the ‘radian’ of pull equal and opposite to the Standing Part pull.
478 the great friction power of RT in lowering (but don’t stand under the load!) i think in force line terms is a ‘triple radian’
2 radian make circle; and see 1 as a combined direction of pull to it’s equal/opposite opposing twin radian’s combined pull direction
in a straight line load concept single point is equal/opposite support of load pull
radian simply to me a ‘cap’ the centers of which form best seating/best nip zone as the circle’s curved answer to an equal/opposite support of loading pull
Also, friction reduction of pull can be seen as the amount of frictional path,but this is nominal compared to the more geometric friction reduction of 2+ radian ‘lines’ as each other’s equal/opposite etc. similar lowering thru Dbl.RT /5radians can sieze VERY easily unless HEAVY load.
All in all the greatest friction to control lowering the loads doesn’t so much come from the diameter of support branch as an increasable friction path distance,but rather the amount of turns/radians as opposing/competing force lines in this imagery.
http://mytreelessons.com/images/ABoK_tautline.png
straight from the book: “The Tree Surgeon
478. There is unsuspected virtue in a few turns of line.
A single ROUND TURN on a branch will allow a man to lower several times his own weight.
The device is much simpler to manipulate than a tackle but, of course, will not serve for hoisting.
479. The TIMBER HITCH unties readily and is one of the most practical of hitches for slinging cylindrical objects.
480, 481. These are tree surgeon’s variations of the MAGNUS HITCH (~1734)’ They work on the same principle as the CAMEL HITCH (~215) and the steeplejack’s SAFETy-BELT HITCH ('/1.452). All five knots may be slid up and down with the hand, but they remain firm under a pull on the standing part.”
.
i purposefully call these ABoK lesson # to myself ; named knot itself is just an example of the mechanics the lesson presents.
I was thinking about the first one mentioned on this page:
http://www.bigfootmountainguides.com/2013/06/saturday-night-live-clinic-recap-knots.html