Girth hitch masterpoint

I'd love to see a test of the shelf on both the girth and clove MP. I suspect the clove shelf may perform better but that's not much more than a guess. 

drew Awrote:

I'd love to see a test of the shelf on both the girth and clove MP. I suspect the clove shelf may perform better but that's not much more than a guess. 

+1 for shelf test. Also wondering if you tested the Girth-X. According to Yann Camus it slips at a higher KN.

Start at 9:45 time stamp

drew Awrote:

I'd love to see a test of the shelf on both the girth and clove MP. I suspect the clove shelf may perform better but that's not much more than a guess. 

Just picturing it in my head, I can see a load on the shelf of a girth hitched masterpoint pulling the cord/webbing down around whatever is hitched, totally defeating the set-up.

Gunkiemikewrote:

Just picturing it in my head, I can see a load on the shelf of a girth hitched masterpoint pulling the cord/webbing down around whatever is hitched, totally defeating the set-up.

That's what I was thinking. Maybe that wouldn't happen with the clove. 

Maybe with a hanging or semi-hanging belay there might be enough force on the MP biner to avoid that. Not sure. 

Anyway the shelf isn't necessary at all. 

I think the wet results are pretty interesting seeing as most people seem to prefer this for ice/alpine. 15cm avg slip and such a high std dev feels pretty uncertain; plus whatever personal safety margin. Granted are you going to FF2 on your anchor in these situations, probably not? Cool data though, I've been wondering this for a while with so many guides and instapros posting about it.

Walker Wwrote:

I think the wet results are pretty interesting seeing as most people seem to prefer this for ice/alpine. 15cm avg slip and such a high std dev feels pretty uncertain; plus whatever personal safety margin. Granted are you going to FF2 on your anchor in these situations, probably not? Cool data though, I've been wondering this for a while with so many guides and instapros posting about it.

I was confused on these results on my first read through but *hopefully* have it figured out now. Testing with wet/frozen slings presented in the report was performed over a range of other conditions. Notably, those tests were performed with both tightened and untightened girth hitches, with the larger slippages presumably occurring with the untightened girth hitch. The discussion states that the authors noticed tightening may be a significant factor, so one (maybe more?) test was performed with a wet sling and tightened girth hitch, resulting in minimal slippage (~2cm if I recall correctly). Presumably the variance of slippage should be lower with the tightened girth hitch as well, but can't really tell anything about the distribution with results presented.

I'd be interested to see wet/frozen tests repeated with just the tightened girth hitch to get a better look at those distributions, but I also understand the cost and especially time that goes into this kind of testing.

Re: wet slings: I wish I had more data to offer. There was a pretty wide variance of slip among the wet slings. The dry slings had some variance, but stayed near enough to central tendency to be more encouraging. However, we were still left attempting to explain outlying large amounts of slip. The only thing we thought might explain it was how tight the hitch was. So we increased the tightness from "hand tight" to "body weight" and that seemed to do the job. Sorry I don't have more on that. It ultimately boiled down to material and time constraints.

Re: the shelf, I don't have or know of data there. But here's my analysis. Potential problems with the shelf might be:

1. Anchor component(s) failure or unclipping.

2. Anchor leg(s) cutting.

3. Unclipping masterpoint carabiner.

4. Rolling/inverting the girth hitch.

Taken in order, #1 shouldn't be a problem (short of total anchor failure) since the carabiner on the shelf should eventually catch on the end of the sling, pro, etc. even if it slips completely. #2 is addressed similarly by the remaining anchor strands and the fact that the shelf carabiner is blocked by the masterpoint carabiner, which shouldn't slip more than a few cm.

#3 is not directly catastrophic in the bit of playing around I did with the rigging. The shelf carabiner ends up clipped to at least one strand of the sling loop, though depending on how it's clipped as well as the number of anchor components, removing the masterpoint carabiner can result in significant (and possibly catastrophic extension).

#4 isn't much different than how a shelf might load the traditional overhand on a bight or the less common inline overhand. Regardless, the masterpoint carabiner effectively prevents this problem.

I personally feel the shelf is okay (based on the above). However, I rarely use the shelf in general unless I really need to create distance between something in the masterpoint and whatever I'm using the shelf for (which for me only tends to happen in some cases when redirecting the rope for lowering, or in certain specific rescue manipulations).

Thanks for posting.  Gee whiz!  I've still been using the figure-8 on a bight or overhand on a bight. I'm way behind.

Love seeing this kind of examination. Very interesting the difference in results between drop tests performed here and static pulls that others have done previously. Do we have any information on how cordalette performs in a girth hitch or girth-x compared to dyneema?

Spider Savagewrote:

Thanks for posting.  Gee whiz!  I've still been using the figure-8 on a bight or overhand on a bight. I'm way behind.

Maybe your way ahead.

The real need for the girth hitch masterpoint is incredibly limited. I'm kinda dumbfounded at people using it at bolt or otherwise solid anchors.

Randywrote:

Maybe your way ahead.

The real need for the girth hitch masterpoint is incredibly limited. I'm kinda dumbfounded at people using it at bolt or otherwise solid anchors.

GHMP requires less material and can be built with shorter slings / be higher and more ergonomic.

Of course, with tricks and longs cordalettes you can achieve similar results but I digress

Noel Z wrote:

So, to combine what Yann and Derek have kindly shared:

Don't tie a girth hitch masterpoint, but instead a Girth-X Hitch Masterpoint. For the girth hitch part itself, cinch it down tightly and weight it prior to use. Yann's method allows for slippage all the way to the sling's end should the carabiner fail.

Derek's data suggests that the sling isn't going to slip all the way to the end, even if the carabiner does fail.  Why add complication to the knot, when the basic knot already handles that situation fine?

Spider Savagewrote:

Thanks for posting.  Gee whiz!  I've still been using the figure-8 on a bight or overhand on a bight. I'm way behind.

Wait, isn't that the right takeaway for wet/icy? There was high variability in the amount of slippage when wet. There was only one measurement made of wet and tight (that's what she said). So we don't know if the amount of slippage is highly variable or not in wet-and-tight girth hitches. And it sounds like no one else has a set of data to see if that's the case.

Until the variability has been measured, it is unknown wether the girth hitch is redundant while wet. But, if you're stuck tying girth/clove for other reasons (sling too short?), at least tighten it. No?

Kyle Tarrywrote:

Derek's data suggests that the sling isn't going to slip all the way to the end, even if the carabiner does fail.  

I’m not sure you can draw that “suggestion” from the data that was presented, especially for wet slings where the amount of slippage was erratic.

If additional wet sling tests are performed to a point where the population became large enough to become statistically significant AND the stdev remained close to what Derek found, I would have a very hard time calling this a viable anchor setup in wet conditions. 

I think it’s sometimes a good idea to put away the kilonewtons and use some common sense.  Is it a good idea leave oneself open to the possibility of a sling pulling through a girth hitch at an anchor?  I think not.

Plus who knows what kind of weird forces are being applied when one leg of an anchor is getting blown out, cut, crushed by rockfall, etc.  I’m sure it’s different than those pull tests!

In a lot of ways I find it a little comical how a few instagram posts from some celebrity guides can create a lot of hubbub and notoriety to techniques that have been around for a while. David Coley published about using clove master points back in 2016 at least but mostly wrote them off because of the dangerous shelf issue.https://people.bath.ac.uk/dac33/high/6TheBelay.htm

I've used the clove version for a few years here and there, but most of the time would rather NOT sacrifice a locker just for the master point. IMO the best anchor system out there is a mini-quad made with a 120cm sling. Perfect length for a bolted anchor, and for 3 pieces just add a trad draw or clip it in series... super versatile, lightweight and doesn't waste a locker on the master point! 

Personally, I like the clove/girth anchor in some applications but don't see the benefit for the amount of attention it's getting. That said, Derek thank you so much for your invaluable research on the topic! I really appreciate your approach and level of critical thinking and detail. 

Tim Dolanwrote:

I think it’s sometimes a good idea to put away the kilonewtons and use some common sense.  Is it a good idea leave oneself open to the possibility of a sling pulling through a girth hitch at an anchor?  I think not.

When it comes to my anchors I prefer objectivity over subjectivity 

curt86irocwrote:

I’m not sure you can draw that “suggestion” from the data that was presented, especially for wet slings where the amount of slippage was erratic.

If additional wet sling tests are performed to a point where the population became large enough to become statistically significant AND the stdev remained close to what Derek found, I would have a very hard time calling this a viable anchor setup in wet conditions. 

Just don't use this anchor in the situations where this might be a risk, like a situation where the sling is wet AND a FF2 is possible AND the anchor is exposed to being cut or unclipping.  How often are all 3 of these things in play simultaneously anyway?

The specific scenario I was responding to was a carabiner "breaking" which is a very rare special case.  If your gear pulls and stays attached to the sling, it's not going to pull your cam or ice screw through the knot; a cut sling due to rockfall might pull through, and the suggested girth+x does nothing to help with that.

It seems like there's a desire to have this work in every possible worst-case scenario; it's just another tool for your toolbox to be deployed when useful and appropriate.

Pretty stoked to see a robust and reasonable discussion happening.

Re: cord, we did one drop with cord that had low amounts of slip. As noted upthread, the DAV and CAI also recommend the use of cord for the girth hitch masterpoint. I think cord is an appropriate choice for this tool.

Re: wet slings, there's definitely a fair amount of variance in those results, and more than for dry slings. Given the effectiveness of tightening under body weight in reducing slip, plus the slow pull result with slip occurring at loads above those seen in the drop tests, we felt comfortable saying that even with a wet sling the girth hitch is a reasonable technique. I feel particularly compelled by the fact that both a high factor fall and sling cutting must occur simultaneously for slip at the masterpoint to cause catastrophic failure.

(Note that practically speaking this also mostly only apply to 2 point anchors. With 3 or more anchor legs, it becomes nearly impossible to slip off the anchor unless the cut leg is very short. But in that case, it would be a miracle if the 4cm anchor leg got cut but the rest of the legs didn't, since the offending rockfall clearly would've impacted quite close to the masterpoint.)

Ice climbing certainly seems the likeliest place for this to occur, in my opinion (wet sling, icefall/rockfall hazards, less reliable climbing medium, many folks choose to run it out, falling leader's tools/crampons could cut the sling, etc.) Despite that, there's at least one anchor failure accident in the last 3 or 4 years that occurred while ice climbing and also featured a girth hitch masterpoint. The girth hitch was not the failure point. (It's cited in the paper.) Anecdotal evidence to be sure, but given the low probability of the combination of events, I'm okay with that risk and subsequent recommendation, despite my conservative risk management paradigm when it comes to recommendations for others.

HOWEVER, I'm not explicitly advocating one way or another. I think reasonable individuals can interpret these results differently, and I'm totally fine with that. It's not as though there's a paucity of anchor rigging options. Ideal risk is personal; I'm just adding a bit more data to the pile to help folks make their own decisions.

Full disclosure: I personally don't actually use the GH masterpoint that much. My main uses cases are:

1. Rappel anchor at 2 bolts cause I forgot to grab the other anchor rig before I started down, but I almost always have an alpine draw on hand. And I'm wearing gloves because I'm rappelling, so I find the GH less futzy than tying the inline overhand to create a wishbone rig.

2. Occasionally mixed climbing where I've got many components and they're not necessarily close together. Ex. A manky pin, the good but far away cam, the other cam in the snowy crack, the stopper, etc. that all need to come together. This is why the sudtyrol climbers invented the technique in the first place (analogous reasons for rock climbing).

I don't use it regularly on ice as I typically rig a banshee belay/fixed point for multipitch these days.

Thanks for studying (and publishing) this !

Regarding the result that body-weight tightening limits the the slip-length variance - I wonder if that result applies to the somewhat popular Mammut 8mm slings (as far as I can tell from the PDF, you mainly tested 10mm slings).

Another potentially interesting type of sling in this context is the Metolius 13mm Open sling, which has rounded edges and in my experience behaves differently from the more popular dyneema slings. Knots tied in it are actually easy to untie, which makes me think it might not cinch down as effectively under load.