The following is taken from my book Down, as well as the stripped-down Down-Core version. The aim of the Core version was simply to get the information out there on the basics of safe rappelling (the Kindle version was as cheap as I could set it, which is around $2), but I thought I would also begin sharing sections from the book here on substack. It would be great if people could share this as widely as possible.
This may be the most contentious part of this book, which is funny, as really it should be one of the easiest sections to write, as any climber really only needs to know one or two knots.
Instead, due to the clash of old and new ideas, and a host of confusing and conflicting information and data, I’m going to dig deep into this subject in the hope we can come to some simple conclusions.
I’ve chosen to ignore a few knots that work OK, but are pretty esoteric (for a good reason), as the aim of any technique is to find the simplest yet effective way to solve a problem, not some show-off trick technique. Before we start, let’s just set some ground rules:
#1 Well Dressed
Carefully tie and dress all knots. Take your time when you do this, and take pride in creating perfect knots.
The knot should be tightened down, with each strand pulled tight to lock the heart of the knot in place.
#3 Forearm Tails
The knot should always have a tail at least a forearm in length (30 cm) after tightening, but no longer.
Offset Knots and Inline knots (Fig 150)
An offset knot (sometimes named a ‘flat’ knot), has the abseil knot offset from the loaded strands, which translates as a knot that will self orientate on meeting an obstacle: the knot rotating due to the loaded strands taking the easiest path, rotating the knot out of the way of obstructions as it goes.
With an inline knot, the knot is held within the loaded strands as the force is passing directly through it. This creates a stronger knot, important for high rescue loads, for example, but is also a knot the is harder to untie.
The inline knot, on meeting an obstruction, is often forced to overcome it, and so are more likely to get stuck or hung-up (Fig 151).
I should point out here that it’s wrong to assume every rap is hazardous in regards to hang-up points and that every rappel should require offset knots. In reality, most rappels have very little chance of a hang-up, especially with older established rappels, where the climbers used what we’d class as unpredictable knots (like a reef knot backed up with a fisherman’s knot) for decades without much trouble.
Most of the time an inline knot would just be as effective, while generally also being stronger, but most climbers will simply adopt a knot and just use that for all descents. What knot you choose depends on several factors.
The Ideal Abseil Knot
To sort through all the knots that are on offer, it’s worth considering the following points:
The knot must be child’s play to tie, so easy that it cannot be tied incorrectly, both in ideal situations, but also in the pitch dark, with thick mitts on, when injured or disoriented, or hypothermic (your fingers don’t work when you’re that cold).
An abseil knot must be practical and applicable to its intended use, i.e. able to hold the highest expected load with a good safety margin.
The knot should be one the user has 100% confidence in, has been used over a long period (has an established pedigree) so that any mistakes will not occur, and both climbers can identify errors if they are made, which is something you can’t do if you’re pulling some rabbit out of a hat knot you saw on the internet. If you’ve dreamed up some new knot, or seen one on a forum, trying it out in a situation where a mistake will kill you is stupid.
The knot must not only be easy to tie but also easy to untie, either when loaded by one, two or twenty climbers.
The knot should remain stable over multiple rappels without needing to be monitored or re-tied (the fisherman’s/reef knot comes undone far too easily).
The knot is as small as possible to reduce the risk of hang-ups.
The knot is offset to reduce the likelihood of problems.
Be applicable to any rope combination you might use or be forced to use.
When you trawl around the internet, you will find the big focus on knots is always on testing strength; people spending hours and hundreds of dollars pulling and breaking knots in every way they can be tied, and with every diameter of rope, all in search of the Holy Grail of knots.
Yes, such tests appeal to the male obsession with quantifying the world (and filming it and putting it on Youtube), but personally, I’d like to see a climber dumped in a bath full of ice for ten minutes then asked to tie a rappel knot when they can’t feel their hands (I can guarantee it would be an Offset Overhand Bend, not a Double Fisherman’s or some German guide’s uber knot).
So how strong does a knot have to be? Well, the first thing to remember is that the rope dictates the strength of the knot, as in most cases it’s the rope that breaks, not the knot.
Secondly, most often, the knot is only taking half of the load, and that load is static; so if a climber weighs 100 kg, then that’s just 50 kg, something a shoelace knot or piece of 2 mm cord could hold.
If we were to have a safety factor of five, then a well dressed Offset Overhand Bend knot, which has a minimum strength of 5 kN, is still more than strong enough. Yes, you could go for a knot that is so strong that the rope will break before the knot, something you can demonstrate with graphs and screen grabs on a forum, but can you tie it with a broken arm?
Why Learn More Than One Knot?
A skilled climber needs to know a range of abseil knots, because although one will do, when it comes to fixing ropes, lowering ropes, abseil ropes of differing diameters, passing a knot through a monster munter, the bigger your knot tool kit, the better.
As covered above, the default tail on any rappel knot should be 30 cm. The reason for this is that it gives a good margin of error for badly tied or loose knots to tighten down without sucking in the rope end, which would lead to failure.
It is also long enough to allow an offset knot to roll several times in the unlikely scenario of high loading, and gives you just enough rope to tie a Clove-hitch, useful if you have to secure the ropes for your partner to return and have no hitch-cord.
Lastly, a 30 cm tail is short enough that it cannot be misidentified as the rap rope itself, which would be possible with a 1-metre tail (which I have seen happen), and so removes the possibility of a tired climber, or a climber working by touch in the dark, clipping their device into one or both tails rather than the rappel ropes itself.
So let’s kick off with the primary knots you should use.
Offset Overhand bend (OOB) (Fig 152)
Also Known as the EDK ( European Death Knot), flat Overhand, or just Overhand knot, this is the king of rappel knots, and when tied to iron law standards, ticks every box.
Being an offset knot is one advantage, but really, what makes this knot so good is it’s small size, making it a hard target for those rope gobbling cracks and fissures.
To tie, simply take both strands and tie them with an Overhand knot 30 cm from the end. How the knot fails is that under high loads (about 600 kg, depending on the rope) the knot begins to invert and roll, moving along the tail. Sometimes it will continue to roll as the load increases, other times it will stop, and the rope will break at the knot.
The kind of force that could achieve this during a rappel is hard to imagine, perhaps six climbers all going down at the same time (but even then the load on the knot would be below its actual threshold).
If used with ropes of different diameters, make sure the thinner rope is placed at the bottom of the stack, beside the loaded strands, not at the top, as this will increase the knots roll resistance.
Learn this knot and learn to trust it.
Backed-up Offset Overhand Bend (Fig 153)
This is the easiest way to increase the strength of the OOB knot, done by simply tying a second knot up against the first.
In reality, this is not necessary for any standard rappels, as the OOB is more than strong enough, but is useful in calming the nerves of those who don’t like the look of it. What you will find when using this knot is that it invariably becomes slack with each rap, and so needs to be re-tied, and so is actually not needed at all.
Gibbs Offset Bend (Fig 154)
This is a simple variation of the OOB that stops the knot rolling, tied by just taking the tails through the loop twice, creating a knot that is 50% bigger than a straight OOB.
This knot is ideal when you’re unsure about using an OOB, such as when your ropes are wet, icy or muddy, or when they are radically mismatched in diameters, such as a 10 mm rope and a 6 mm rope.
This is also an ideal variation to know if you’re looking at higher loads, but should only be used if you’re 100% sure how to tie it, as a mistake could lead to a lowering or even zeroing of strength.
Half Gibbs Offset Bend (Fig 156)
This is the same as the Gibbs knot, but you only pass one strand back around and through the knot, achieving the same strength effect, but without creating quite as large a knot.
Although less well known, this is perhaps the knot we should all be using, as it works better when dealing with miss-matched ropes and is also more forgiving, all for one extra loop around.
Double Fisherman’s Offset Bend (Fig 158)
This is a stronger knot than the OOB and might be worth considering by those who just don’t like any variation of it, but still want an offset knot.
To tie this knot, first tie a Double Fisherman’s around strand A with strand B, leaving a 30 cm tail A and 40 cm in B. Now tie a Fisherman’s knot around B with strand A. Tighten the knot.
I’ve also seen this tied with two single Fisherman’s knots, or with the rope passing through a single Fisherman’s knot, then finished with a Double Fisherman, but both knots are dangerous, or hard to untie, so stick to the standard double method.
If you found this informative it’s worth noting that it did not come free of charge, but was subsidised by paid subscribers of my Substack. If you’d like to see more content like this then please become a paid subscriber for only $1 a week, or just share this content so that others might.