Clipping the Shelf of an Anchor with a Clove Hitch at the Master Point

Since a previous thread (see here) on this topic is old and has since been derailed, I am starting a new thread to throw out my observations and findings regarding the pros & cons of​ using a clove hitch at the master point of an anchor​​​.  I am sure there will be no shortage of passionately expressed opinions here on MP.  If this doesn't work for you, feel free to just disregard.  I am throwing this out there in case others find it useful.  I am always open to learning from constructive comments, but not too interested participating in a debate about the merits of one type of anchor vs. another.  To me, this is just one tool I keep in the bag.  With that said, I have found using a clove hitch at the master point to be a nice tool to have in my quiver for the following reasons:

• The ability of the clove hitch to reduce peak loads on the system as the knot cinches up under load is significant.  This is especially true for dyneema slings, where the more “slippery” nature of the material allows for more movement as the knot tightens.   See test results conduced by DMM here.  Of course building the anchor using the dynamic climbing rope is probably even better, but that isn’t always a good option, like when you don’t have enough rope to build an anchor after a rope-stretching pitch or when you won’t be swapping leads . . .
• The clove hitch takes up much less of the sling/cordelette.  This can be key when your anchor pro placements have to be spread out and there is either not enough cordelette to tie a figure-8 (or overhand) knot or doing so results in an unacceptably wide angle between adjacent anchor placements and the master point.
• The knot is very fast and efficient to produce.
• The knot is very easy to undo, even when fully weighted – you simply remove the master point carabiner and the knot is gone.

There are also some disadvantages with this configuration, including:

• The anchor has to be pre-equalized in the direction of anticipated load.  This is of course no different than using a figure-8 or overhand knot at the master point, but is a disadvantage to a “self-equalizing” Sliding X or quadralette type anchor configuration - but then they have their own drawbacks, see DMM test results for low load failure with an overhand knot to limit extension, or very high shock loads without the overhand knot (rows 1 & 3 in the table).  Using a clove hitch at each of the attachment points of the pro placements as well as at the master point can help add a small degree of load sharing due to the movement in the respective knots as they cinch up.
• Since using a clove hitch at the master point uses up less of the sling or cordelette, it may not be the best choice if you have a low anchor and are trying to increase the height of the anchor for convenience in belaying a follower directly off of the anchor.
• As alluded to by others in the referenced thread (see David Coley comment on page 1), there are some peculiarities in how you choose a shelf for tie in points.  If the “shelf” is setup in the typical fashion of clipping across one strand from each pro placement (which I will argue is not the best way to clip into this type of anchor), and you are using a two-piece anchor, then there is about a 50/50 chance that IF you or your partner further made the big mistake of pulling the master point carabiner (thereby undoing the clove hitch at the master point) while the anchor is still being relied upon, then one or both of the climber’s attachment points (via a tether) could become detached from the anchor with potentially fatal consequences.

I have spent a lot of time studying this shelf clipping failure mode and have figured out how to reliably reproduce a detachment of the carabiners tethered into the "shelf" for a two point anchor when the master point carabiner is removed prematurely.  For a 3 or more point anchor, a complete detachment is not possible if the shelf is clipped in the typical fashion (across either the upper or lower strands from each pro placement), but a sudden extension and subsequent shock load to the anchor is possible if the master carabiner is pulled prematurely.

If you change what you consider to be the "shelf" for this type of anchor, then this failure mode can be eliminated, even if one of the climbers is careless enough to pull the master biner while still relying on the anchor (like for example trying to grab an extra locker for building the anchor on the next pitch).  If you consider the shelf for this type of setup to be all the strands on either side of the clove hitch knot like shown in the photo below, then even if the big mistake is made of prematurely pulling the master carabiner, then both "shelf" connection points will remain fully attached to the anchor as shown in the second photo below.


If the "shelf" is clipped as in the picture above, then this is what happens when the master point locking carabiner is pulled (which you should never do unless you are not relying on the anchor anymore)


The shelf carabiners are clipped into all the same strands as the master carabiner was, just minus the clove hitch knot.  The load from the tether points is still distributed to all the pieces of the anchor.  A typical argument against clipping the shelf this way would be that if one of the anchor pieces blows, then it is possible that the blown piece could pass through the tether carabiner and the tether become detached.  However, this failure mode is only possible for a 2-point anchor, which one would argue must have fail-proof anchor points anyway.  For a 3 or more point anchor, if one piece in the anchor blows, the carabiner attached to the shelf will still remain attached to the anchor by at least one strand of the sling or cordelette as shown below:


Of course none of this matters if you never pull the carabiner out of the master point until you don't need the anchor anymore, but mistakes sometimes happen . . .

I almost always use a master biner with a clove. I just never clip the shelf and don't see what it gives you. Since the biner doesn't colapse like a knot, it is easy to belay a second and for them to clip right into the biner. I won't start using the shelf as you described because in a 2 piece anchor you have a single point of failure. In the sling + in 3 piece anchors it introduces a lot more complexity than just using the master biner. Why add the complexity if there isn't a problem to solve in the first place?

Looks like a great idea to me. I usually just tie a big fat figure 8 as a master point but this seems cleaner and and easier to untie.Thanks for the tip! I'm gonna practice setting this up while sitting at home.

Meh...! I just use the rope. No need to carry all that extra shit. 

I'm with Luke.  Master carabiner is the way to go when using a clove hitch or girth hitch masterpoint.  When everything is clipped to one carabiner, it's pretty clear that you can't remove it.

Thanks for doing the leg-work on this one, Jeff!  I don't know if I'll either use the "shelf" on the three piece anchor, but it's interesting nonetheless.  

Everyone else, let's for once have a thread that doesn't drift by the third comment.  Jeff said he knows there are other anchors out there.  He wanted to discuss clove hitch anchors, so let's talk about those and not tangent topics.  

Adam Fleming wrote: I
Everyone else, let's for once have a thread that doesn't drift by the third comment.  Jeff said he knows there are other anchors out there.  He wanted to discuss clove hitch anchors, so let's talk about those and not tangent topics.  

I was unclear what the rules would be for this thread. I am relieved that you covered it. Phew!

Adam Fleming wrote:I'm with Luke.  Master carabiner is the way to go when using a clove hitch or girth hitch masterpoint.  When everything is clipped to one carabiner, it's pretty clear that you can't remove it.

I like the simplicity of what you and Luke propose.  I have avoided clipping tethers directly into the master carabiner out of concerns with cross loading the carabiner.  As I think about it more, it does seem hard to think of a scenario where significant cross loads could be applied to the master point this way, but out of principle, I have always avoided it.  The same argument could be made for an anchor using a figure-8 master point - why not just have everything go into the master carabiner and forget using the shelf - unless you want to belay directly off the shelf to get the belay point a bit higher?  

Salamanizer suchoski wrote: Meh...! I just use the rope. No need to carry all that extra shit. 

I'm curious what you do when you are leading every pitch.

Jeff N wrote:
I like the simplicity of what you and Luke propose.  I have avoided clipping tethers directly into the master carabiner out of concerns with cross loading the carabiner.  As I think about it more, it does seem hard to think of a scenario where significant cross loads could be applied to the master point this way, but out of principle, I have always avoided it.  The same argument could be made for an anchor using a figure-8 master point - why not just have everything go into the master carabiner and forget using the shelf - unless you want to belay directly off the shelf to get the belay point a bit higher?  

Some people do use master carabiners for knotted anchors. Unlike a sling, a master carabiner doesn't collapse when load is applied to it, so you can fit a lot of stuff in it. 

I think your concern is actually tri-axial loading (multiple pulls from three directions), not cross loading (loading across the minor axis, the weaker orientation of the carabiner). Large HMS carabiners take loads from multiple directions better than small asymmetrical carabiners. Plus, unless you're in a hanging belay, only the follower is potentially loading the anchor so you're not actually tri-loading.

Jeff N wrote:
I like the simplicity of what you and Luke propose.  I have avoided clipping tethers directly into the master carabiner out of concerns with cross loading the carabiner.  As I think about it more, it does seem hard to think of a scenario where significant cross loads could be applied to the master point this way, but out of principle, I have always avoided it.  The same argument could be made for an anchor using a figure-8 master point - why not just have everything go into the master carabiner and forget using the shelf - unless you want to belay directly off the shelf to get the belay point a bit higher?  

I've never liked clipping biner to biner either but Europeans don't seem to mind.

Jeff N wrote: 
There are also some disadvantages with this configuration, including:

• The anchor has to be pre-equalized in the direction of anticipated load.  This is of course no different than using a figure-8 or overhand knot at the master point, but is a disadvantage to a “self-equalizing” Sliding X or quadralette type anchor configuration - but then they have their own drawbacks, see DMM test results for low load failure with an overhand knot to limit extension, or very high shock loads without the overhand knot (rows 1 & 3 in the table).  Using a clove hitch at each of the attachment points of the pro placements as well as at the master point can help add a small degree of load sharing due to the movement in the respective knots as they cinch up.

It is much easier to equalize and adjust a clove hitch than a figure 8 on a bight. I see this as a great advantage.

Do you have a link to the DMM test results you are referring to? 

I agree that "slippage as it tightens" can be a plus i.e. your first listed advantage. However, you're going to get much greater "advantage" from this with a masterpoint knot. Carry a extra long loop of hi-tensile 5mm cord, rig it as a classic cordelette, and tie a Figure 9 (or larger) knot to raise the masterpoint to a useful height. You'll get LOTS of slippage.  So I'd put the CH here in the DISadvantage category, as it offers less slippage than any alternative, knotted configuration.

Pete G. wrote:
Do you have a link to the DMM test results you are referring to? 

Yes, the word "here" in my post is a live link to the test results.  The results can be found here:

​https://dmmclimbing.com/Knowledge/September-2013/Slings-at-Anchors

Gunkiemike wrote: Carry a extra long loop of hi-tensile 5mm cord, rig it as a classic cordelette, and tie a Figure 9 (or larger) knot to raise the masterpoint to a useful height. You'll get LOTS of slippage.  So I'd put the CH here in the DISadvantage category, as it offers less slippage than any alternative, knotted configuration.

I don't know why DMM did not include a figure-8 master point knot in their testing (see link above) - I wish they had.  But they did test using overhand knots, and the clove hitch was shown to be far more effective at reducing peak loads (and more importantly not breaking the sling) than an overhand knot. It seems that the clove hitch knot has a unique ability to provide friction absorbing slippage while also maintaining structural integrity of the sling.  I think that this is in part due to the fact that a clove hitch is wrapped around a shaft (the carabiner), which both helps promote movement while cinching up (instead of binding up), as well as limits the tightness of the bend in the sling material, which is one of the key factors of knot strength.  

Lots of slippage alone won't absorb energy, you need slippage to occur together with lots of friction.  I expect that a Figure 8 (or 9+) would give improved results over an overhand knot in that it would not fail the sling as prematurely, but I doubt it would generate much more slippage under friction.  But without test data, it is just a guess on my part.  

Jeff N wrote: Yes, the word "here" in my post is a live link to the test results.  The results can be found here:

https://dmmclimbing.com/Knowledge/September-2013/Slings-at-Anchors

I thought that was the link for a previous thread. 

One thing I didn't hear mentioned in the video is how many tests they ran for each set up. I feel that's a pretty important piece of information to aid in interpreting the results. It seems like he only did one test for each set up.

I find the "bit of melting" result for clove hitch/dyneema for FF1 to be pretty concerning for multi-pitch climbing. After having a near miss, I could see someone missing the melt mark, mixing this sling with a bunch of others, and reusing the damaged sling on the next pitch. So I see slippage with dyneema as a bad thing.

I have always used the rope to tie into the master point with a clove hitch and this reinforces that decision.

Pete G. wrote: I find the "bit of melting" result for clove hitch/dyneema for FF1 to be pretty concerning for multi-pitch climbing. 

Keep in mind that this was a very severe test - 85 kg mass with FF1 & FF2 falls directly onto the anchor with no dynamic rope in the system.  The bit of melting proves that the hitch is effective at dissipating some energy.

Jeff N wrote: Keep in mind that this was a very severe test - 85 kg mass with FF1 & FF2 falls directly onto the anchor with no dynamic rope in the system.

I assume DMM used new slings for the test. I seem to remember there were some tests on one or two year old dyneema slings with some bad results. With the thought being that exposure to heat (like leaving the slings in your car on on hot day with the windows up) permanently degrades dyneema. I will see if I can find a website.

Unfortunately, 85 kg is not an excessive weight for a old trad climber like me. I agree though, FF2 is an unrealistically high test. 

Garry R wrote:

I'm curious what you do when you are leading every pitch.

Simple. If it’s bolts, they just slip in under my rope with theirs. If it’s gear, then they use runners and their rope to equalize while I’m racking and stacking. I have a sub 5 hour ascent on the RNWFHD using this method. 

Salamanizer suchoski wrote:

Simple. If it’s bolts, they just slip in under my rope with theirs. If it’s gear, then they use runners and their rope to equalize while I’m racking and stacking. I have a sub 5 hour ascent on the RNWFHD using this method. 

I've climbed that (once) and sub 5 hours seems unreal!

Anyway, you said you don't "need to carry all that extra shit" but I always carry a long sling so if I'm leading the next pitch I guess I'll just keep building the anchor with that instead of the rope.

I just watched the vid and would never have believed that the Cloves and loose overhand knot would have absorbed that much energy. 

Mark Hudon wrote: I just watched the vid and would never have believed that the Cloves and loose overhand knot would have absorbed that much energy. 

Actually, by just watching the video, the results were not nearly as clear to me as are the tabulated test results shown below the video in the webpage.  I am not sure what you are referencing with the energy absorbance of a loose overhand knot.  Below is a portion of the tabulated test results.  Focusing on the results for the dyneema slings (most commonly used?), note that the sling broke in several of the tests with the overhand knot, so the lower load value for those tests is just the point at which the sling failed.  A comparison of the FF2 load with a sliding-x setup (27 kN!) to the same test with clove hitches (10.8 kN)  indicates the level of shock absorbance that the clove hitches provided in this case.  All FF2 tests with an overhand knot resulted in sling failure.

The other big takeaway for me is this statement from the DMM analysis:

The results show how very high forces can easily be generated using slings to attach to anchor points if there is slack in the system. Clearly, it's important to be aware of this if for example, you are moving around at a stance while clipped into an anchor using a sling or rigging a multi-pitch abseil.

What I take from this is (1) be very careful to avoid having any slack anywhere between your harness and the bolts or pro placements of the anchor, and (2) always use the dynamic rope to tie your harness into the anchor whenever possible.