Overview of a simple pulley system

Let's start simple.

I don't know about you, but when I start looking at diagrams of complicated pulley systems and 5:1 rescue setups, my eyes get crossed and my brain starts to fog. Good news is, we don’t need to analyze a 5:1. At least not right now. Let’s break this down by starting at the beginning and then working up to a simple MA system, so you can really see how this works.

---

The basic 1:1 pull

Sticky the climber needs to haul a 100 pound load up to the ledge. Sticky ties a rope onto the load, and starts pulling. To even budge it off the ground, Sticky needs to pull up with 100 pounds of effort. Sticky pulls 1 foot of rope, and the load rises 1 foot. Sticky has no mechanical advantage or progress capturing, so Sticky’s arms get tired pretty fast!

In this case, it's probably not a good system. But in other situations, it might work just fine. Got a crevasse rescue with five people on top ready to pull? Great! Put prusiks on the rope for everybody, have them clip in, and start pulling, ideally with their legs and bodyweight. Probably no need for anything fancier than this.

---

Bluehat thinks, “Hmm, how about clip a carabiner to that bolt, and clip in the rope? That way, I can pull DOWN with my bodyweight instead of lifting UP with my arms, and I won't get so tired. That should be easier, right?”

The 1:1 pull with a redirect carabiner

Does Bluehat gain any MA with this setup? No. He changed the direction of pull, but because the direction change is on the fixed anchor, he did not gain any MA. He’s still pulling a 1:1, just like before, just with the rope now moving down instead of up. He pulls 1 foot of rope, and the load rises 1 foot. This carabiner on the anchor is called a “redirect”, because it, umm, redirects your direction of pull.

Does he get an easier pull? Maybe. He can now use gravity and pull down using bodyweight rather than lifting up with his muscles. Pulling down is usually easier than pulling up! But the redirect adds a lot of friction. By running the rope through a carabiner, which is only about 50% efficient, he'll have to pull down with 150 pounds of force to move the 100 pound load.

Which is better, 100 pounds lifting straight up, using your arm muscles, or 150 pounds, pulling down, using your bodyweight?  There's really no right answer. It depends on how far you need to move the load, your weight, and your strength. (Personally, I'll take 150 pounds with the redirect pulling down, thank you very much.)

Does he need a bomber anchor for the redirect? Yes! When Bluehat is pulling, the force on the anchor is approximately twice the force they’re applying to the rope, or about 300 lbs. Which introduces a good general rule of thumb: a redirect on the anchor increase the load on the anchor.

---

The 1:1 pull with a redirect carabiner and progress capture prusik

Bluehat thinks, “Well, it is easier pulling down with my bodyweight, but if I ever let go, this load is going to zing all the way to the ground again. How about I put a prusik on the load strand so I can take a rest?”

Excellent idea! This is known as a progress capture (aka “ratchet”). It allows the load to move up, but whenever Bluehat wants to let go and rest, the prusik keeps the load from sliding back down. (If you want to get fancy, you could use a progress capture pulley here, such as Petzl Traxion.)

---

The 2:1 pull

Bluehat thinks, “OK, time to start working smart instead of working hard!” He clips one end of the rope to the anchor, puts a pulley on the 100 lb. load, runs the rope through the pulley, and starts hauling. Now he’s getting somewhere!

Does Bluehat gain any MA with this setup? YES! He’s now pulling with a 2:1 mechanical advantage. Look how the load is distributed on the rope. 50 pounds goes to the anchor, and 50 pounds goes to him. So, if he pulls with 50 pounds of force, the load will rise! He has to pull 2 feet of rope to move the load 1 foot.

Does he get an easier pull? Well, it depends how you look at it. In theory, he only has to pull with 50 pounds of force to move the load, which is good. But, he needs to pull twice as much rope, which is not so good. In the end, he's doing the same amount of “work”. Would you rather lift 100 pounds 10 times, or 50 pounds 20 times? In the end you’ve still moved 1,000 total pounds, it's all the same.

Here's another way to think about it: work equals force times distance. You're doing the same amount of work in the end, lifting a given load the required distance. But with a mechanical advantage system, you use a lower force to move the load over a longer distance. 

How’s he doing for efficiency? Great! By using a quality pulley on the load, he can lift the load with much less effort. Way better than the carabiner with a 1:1 redirect.

---

The 2:1 pull with redirect and progress capture prusik

Bluehat thinks, “Well, this is definitely easier to pull, but my arms are still getting tired. Let's put a redirect on the anchor, and a prusik on the load strand.”

Now we're getting somewhere. He’s lifting with 2:1 MA, , and added a ratchet prusik so he can take a break whenever he needs to without dropping the load. Nice!

---

This, right here, is the foundation of mechanical advantage systems. All the fancy stuff in the rock or crevasse rescue books that makes you go cross-eyed? It's all just adding and stacking additional redirects and pulleys in different variations on top of pretty much what we just saw.

---

Now, the above diagrams might appear to be overly simplistic. But if we break them down, we can learn some important principles that apply to any flavor of simple or compound pulley systems.

• Changing the direction of pull at the anchor does NOT add mechanical advantage.

• Changing the direction of pull at the load (or the load strand) DOES add mechanical advantage.

• Even if a change of direction at the anchor does add friction, it might make your pull easier, depending on your own personal strength, body weight, and the weight of the load you need to move.

• A redirect on the anchor increase forces on the anchor. Be sure your anchor can handle this.

• Try to minimize friction at every change of direction by using a pulley rather than a carabiner whenever possible.

• Adding a progress capture / ratchet means your load will not slide back down if you stop pulling.

• In the end, the “work” you do is the same with an MA system. You move the same amount of weight over the same distance. So, in a sense it's not necessarily “easier” to move the load all the way up, you just get to pull less weight on each stroke.

A 3:1 “Z” drag, step by step

The “Z” drag (so named because the rope looks like the letter “Z” if you turn your head sideways) which gives you a theoretical 3:1 mechanical advantage, is one of the fundamental setups of crevasse and rock rescue. After you've done it a few times, most people get the hang of it. But if you haven't rigged it in a while, or if you're doing it under the stress of a real rescue situation, setting it up efficiently and correctly can be a challenge. (I’ve seen some quite experienced climbers have a complete brain fade trying to do this if they’re out of practice . . .)

---

Here's a step-by-step walk-through. Hopefully this will help if you're new to rope rescue, or to dust off this skill if it’s been awhile. So get a rope, 2 prusiks, a few carabiners and a pulley if you have it, and follow along.

Note 1: This shows the basic mechanics of how a Z drag is set up, not all the possible nuances of gear and technique. Prusik minding and progress capturing pulleys, rope grabs, backup knots and releasable hitches, and other fancy rope tricks can be added after you know this foundation inside and out.

Note 2: Don't pull furniture around inside your house as it's tough on the floor and carpet, ask me how I know this . . .

---

Step 1 - Construct a bomber anchor. Add a locking carabiner. Clip this carabiner to the rope with the load. You now have a 1:1 system (zero mechanical advantage) with the rope redirected.

Step 2 - Add a “capture” prusik on the load strand of the rope, and clip this prusik to the anchor. (This prusik loop “captures” your pulling progress, holding the load if you let go of the rope.)

Step 2A - The way it's set up now, when you haul on the rope, the prusik will pull through your carabiner. Not good. There are a few ways to prevent this. One is to add a quick link, as shown below, which should block the prusik from sliding through. The effectiveness of this depends on the few variables such as size of your quick link, and diameter and grip-tion of rope and prusik cord. Give it a try and see how it works. (You can get quick links that are actually CE rated for climbing from CAMP, discussed here.)

Another way is to have a second person “mind the prusik”, keeping it loose when you're pulling, but letting it go tight on the rope to hold the load when you stop pulling.

If you have a fancy and somewhat expensive “prusik minding pulley”, this is where you’d put it.

(And yes, clever reader, I know the trick of adding a tube belay device here, we're not covering that today.)

Step 3 - Add a second prusik, called a “travelling” prusik, onto the load strand of the rope. It's called the “travelling” prusik, because it moves when you pull. If your prusik cord is a little long, like the one I have here, tie an overhand knot to shorten it up. Shorter is better.

Step 4 - Put the free end of the rope through a pulley, clip a carabiner to the pulley, and clip that carabiner onto the travelling prusik.

If you don't have a pulley, use a carabiner here. A pulley is better. If you have only one pulley, put it on the travelling prusik to increase your hauling efficiency.

Sweet, you now have a 3:1 and you’re ready to pull! Pull on the rope until your load is where you need it, or until the travelling pulley touches the anchor. If this happens and you need to pull some more, set the capture prusik to be sure it can hold the load, and then reset the travelling prusik by sliding it as far as you can down the rope toward the load. Continue pulling.

Alpine vs. big wall - different needs in MA systems

In general, there are two basic user groups for MA in climbing. They each have distinct needs and equipment.

The first one, for rock/ice and glacier climbing, involves a rescue situation. Here, you typically need to haul a big load a single time for a short distance with improvised materials, the size, bulk and weight of gear are important factors, and you may likely be using a more complicated mechanical advantage system, like a 3:1 or 5:1, to raise your patient. It’s also unexpected and stressful, meaning you may have to rig a more complex haul with improvised materials, quickly, without much practice.

Big wall climbing can also involve hauling big loads. But instead of 30 feet, you’re doing it for maybe 3,000 feet! Bulk, weight, and cost of equipment are factors, but not nearly as important as in alpine climbing. You are unlikely to be doing a 3:1 or greater haul, 2:1 is typically all that’s needed even for honker haul bags. And, because you are repeating the same movement so many times, you’re willing to take some time and invest in a perfectly dialed system that uses more expensive and slightly heavier gear, if it increases your pulling efficiency even a tiny bit. To summarize:

In alpine climbing . . .

• You start your climb not knowing if you will need to haul (and hoping you won’t!)

• Need understanding of various MA systems, 2:1, 3:1, 5:1, etc.

• Probably won’t have the exact right tools for the job, creative gear improvisation may be required

• Weight and cost of gear may dictate what you choose to carry

• Usually need to move a load only a short distance, so highly efficient systems are generally not so important

In big wall climbing . . .

• You start the climb knowing you will have to haul

• Typically only using 2:1 MA, more MA not required

• Carrying the exact right tools for the job is a priority, even if they are expensive, bulky, and weigh more

• You are moving a heavy load for a very long distance, so well practiced and efficient systems becomes very important

A few basic questions and answers about MA systems

When I start looking at some of the more crazy pulley diagrams of 5:1, 6:1, 9:1 . . . my eyes glaze over. As a climber, what's the most basic system(s) I need to know?

Learn how to do a 2:1 and a 3:1 pretty much with your eyes closed. These are the foundation systems that are used most in rescue and hauling scenarios. Any system that creates greater MA than a 3:1 is just a combination(s) of 2:1 and 3:1. Once you have these dialed, you can play around with combining them in various way to make a 5:1 or maybe a 6:1. That should be everything you should ever need for a rescue scenario. Learn these basics first and don't get confused by the fancier stuff.

Note that it’s not likely you’ll be able to lift someone just with a 3:1 unless they are actively assisting. To be really sufficient for various rescue scenarios, add a 5:1 or 6:1 to your engrained memory. Here’s a great way to make a 6:1, for example.

The good thing is, you can practice these at first on the floor and in the comfort of your warm, dry house. But then, please go try them in a more realistic setting.

---

What are some real world climbing situations where I need to know this stuff?

Alpine climbing - In a crevasse rescue, you may need a 2:1, 3:1 or even 6:1. Or, perhaps various rock rescue scenarios where you might need to haul your second past a difficult move or two with a 3:1, and less commonly, setting up an “alpine block and tackle”, which can be a 4:1 or a 6:1 theoretical MA, which we cover in this tip here.

Big wall climbing - You’ve got two big honker haul bags for you and your partner to spend a week on El Capitan. The total load is going to be well over 200 pounds, and is going to absolutely suck to haul 1:1. Time to rig a 2:1 haul system. The scope of setting this up properly is a complicated topic and is covered in a separate post here.

---

I don’t quite understand the math around pulley efficiency. Can you explain that?

Sure thing. One way to think about it is how different pulley efficiencies determine the effort needed to lift a load. Let’s use 100 pounds as a nice round number.

The table below pretty much spells it out. 

What do we notice? With high pulley efficiencies of 70 to 90%, you’re not going to notice much real world difference when lifting a load. When pulleys start to fall below about 60%, you’re definitely going to start noticing a difference. If any component in your hauling system is much below 50% efficient, you need to ask yourself why you’re using it at all.

---

I don’t quite get how a redirect increases force on the anchor in the real world. Can you explain that?

Sure thing. Say our friend Sticky decides to raise her 100 pound load through the carabiner redirected through an anchor point. The carabiner is roughly 50% efficient. From the table directly above, we can see that for her to lift 100 pounds, she needs to pull down with a force of 200 pounds. That force gets applied to the strand of rope she’s pulling. At the same time, the 100 pound load is weighting the other strand of rope. So together, her pulling force of 200 pounds plus the 100 pound load add up to be a 300 pound load on the anchor.

Now, let's say she has a pulley that's 80% efficient, and she runs her load through that instead. From the table above, we can see that she would need to pull with 125 pounds of force. We add this to the 100 on the other side, and get a total of 225 pounds on the anchor. Hopefully this convinces you to use a pulley whenever possible!

(I did this myself with a 10 pound barbell weight, a spring scale hanging from the ceiling, and a 9 mm climbing rope. Sure enough, set up exactly like below, it took 20 pounds of pull to lift the 10 pound barbell plate off the ground.)

---

If a 9:1 is easier to pull than a 3:1, why don’t we use a 9:1 for everything?

Well, first try pulling with the 3:1 that you already have set up. If that's getting the job done, don't make it more complicated. Remember, every additional redirect and pulley that you add increases friction and decreases the real world MA. It also increases the amount of rope you have to pull through the system to raise your load, requires more gear like carabiners, prusiks and pulleys, and it may increase how often you need to reset the haul pulley. It can even add some more unexpected weird variables, like ropes twisting, ropes rubbing on each other, and prusiks slipping. Remember the law of diminishing returns from that long ago economics class; adding more input does not always get you a good return on the output.

A good rule of thumb: use the minimal mechanical advantage system that you can to get the job done. The best system is not necessarily the one which creates the greatest MA.

---

How about something like a DMM Revolver carabiner, that has a little wheel in it, or the Petzl “ultra legere” orange plastic wheel thing. Can I use these instead of a pulley?

Probably not. The Revolver carabiner was really designed to minimize rope drag when lead climbing, not serve as a proper pulley in a block and tackle system. (My real world tests showed a Revolver carabiner was pretty much the same as a regular carabiner in hauling efficiency.) The legere I personally have found quite difficult to use in a crevasse rescue scenario, because the rope does not properly stay in place and wants to skip off of it at every opportunity. It also has it safe working load of only 1 kN, so that means it's appropriate for lifting your pack, but not a body. Get a real pulley (or two). See some real world test results here.

image: dmmclimbing.com

image pinterest.com/pin/369295238177239123/?lp=true

---

What kind of pulley should I get for crevasse / rock rescue?

There are two main types of pulleys, which you could refer to as fixed plate and swing plate. With the fixed plate pulley, the sides of the pulley look like a letter “U” and are made from a single piece of metal. Because of this wider shape, these can usually only be used with a oval or HMS belay carabiner. With the swing plate pulley, the two sides are separated, allowing one side to swing down to more easily insert the rope. The swing plate pulley is probably the most useful one to have for alpine rescue. The advantages are, it's easier to put the rope in, and it works with just about any kind of carabiner.

Having said that, it's usually better to use an oval shaped carabiner with a pulley if you can. If a D-shaped carabiner tilts the pulley off to one side, you're going to lose efficiency, because the pulley bearings will not be properly sharing the load.

image: https://smcgear.com/crx-pulley-orange.html

The American Alpine Institute recommends this model pulley for their glacier travel and crevasse rescue classes: the CRx, made especially for crevasse rescue. (“CR” Crevasse Rescue, get it?) It's a solid pulley, from the respected company SMC (Seattle Manufacturing Company), and best of all it's only about 15 bucks. I have one, it’s great.

Note that the wheel (aka sheave) in the CRx pulley is plastic, not metal. This makes it a hair lighter and less expensive, both of which are good for a lighter duty rescue pulley. But if you’re looking for a big wall hauling pulley, you want a larger diameter metal sheave with sealed bearings, both of which will increase your hauling efficiency, which is important when you're doing it 3000 times. More on big wall hauling systems and pulleys is in this post.

Petzl, SMC and CMC all make quality pulleys. Pretty much any small pulley by a name brand climbing company is going to work fine. You can find a lot of small inexpensive pulleys on Amazon; some of these are probably great, and some of them probably suck, so personally I'd stay away from those. There are ways to skimp on gear - rescue equipment is probably not a good one.

There is also a flavor of pulley called a “prusik minding” pulley, or as some manufacturers call it, a “PMP.” These pulleys have a wider faceplate on either side of the wheel, which is designed to keep the prusik from being sucked into the wheel during a crevasse rescue. If you're going to get one pulley for crevasse rescue, you very likely want to get a PMP. Like I said, don't skimp on rescue gear. PMPs vary a lot in price. That’s why this CRx is such a good deal.

Where to put the pulley?

I only have one pulley. Where should I put it to get the easiest pull?

Excellent question! We often have to improvise with limited equipment, and the location of the pulley can make a difference in the efficiency of your hauling system.

You should use your “good” pulley on the position that’s closest to your pulling force (aka, your hands).

A simple explanation, in the words of rigging expert Richard Delaney: "...the best place is closest to where the effort is applied, as this preserves maximum effort moving into the system rather than wasting it at the first bend."

Or, to say it another way, any inefficiency at the first pulley is compounded throughout the system, so you want your most efficient pulley closest to the pulling force (that’s you).

For a 3:1 (below), your pull is closest to the travelling pulley, so that’s where the good pulley should go.

For you engineers and physics folks out there, an Alpinesavvy fan on Instagram (@jared_vilhauer who's way smarter at this stuff than I am) calculated that:

• If you have a 50% efficient carabiner on the tractor, your real world mechanical advantage is 1.95.

• if you had a 90% efficient pulley on the tractor, your real world mechanical advantage is 2.35

---

If you want to take a deeper dive into this, here's a nice video from The Rope Access Channel that walks you through each step. His example shows a redirect off the anchor (good practice if you need to lift the load vertically instead of horizontally) but the principle is the same. If you put the pulley closest to the load from the pulling force (aka you) that’s optimal.

image credit : the rope access channel screen grab from https://www.youtube.com/watch?v=sIMFvf_E5Y8

---

A lot of folks think the pulley always should go on the moving part of the load to gain easiest pull, but this is not always true. Below, in the 2:1 with a redirect, the pulley should go on the anchor. Again, it’s because the anchor is closest to where you’re actually pulling on the rope.

This may seem a little counterintuitive (it was to me!), but it's easy to set up a test and prove it to yourself. Get a pulley, a carabiner, a rope, something heavy, and an anchor point. Set up each way and notice the pulling force needed in each set up. In this case, a pulley on the anchor is better.

Confession: this did not intuitively make sense to me, so I did a little observational study to prove it to myself.

I set up a 2 to 1 system, redirected through a top anchor point, as in the diagram above. I had a 10 pound barbell weight, and attached an inexpensive spring scale to the pulling strand. I pulled at a slow steady rate, and noted the most common whole number reading on the digital scale while I was pulling.

• 2:1 - pulley on ANCHOR, carabiner on load: 8.5 lbs. of force needed

• 2:1 - pulley on LOAD, carabiner on anchor: 11.3 lbs. of force needed

• 2:1 - pulley on BOTH anchor and load: 8.2 lbs. of force needed

Clearly, putting the pulley on the anchor is the best approach. I almost didn’t need a pulley on the load, as the force needed with a pulley or a carabiner on the load was almost the same.

Pulley vs. carabiner - What’s the difference?

You may be wondering, does it really make that much difference if I use a pulley or a carabiner?

Short answer, it can make a lot of difference. Use pulleys whenever possible. We had a look at this above with the Sticky diagrams, but it's important to get this, so let's have a quick review.

Say you need to lift 100 pound load, with a 1:1 system redirected through a high-quality pulley which is 90% efficient, which is pretty typical for a standard rescue pulley. Here, you need to apply 111 pounds of pulling force to move the load. (The math for this is 100 divided by 0.9).

However, let's redirect that same 100 pound load through a carabiner, which has an efficiency of roughly 50%. Here, you need 200 pounds of pulling force to move the load. (The math for this is 100 divided by 0.5).

So, use a pulley and pull with 111 pounds, or use a carabiner and pull with 200 pounds? Easy choice!

---

Here’s a table of pulley efficiencies. You may have seen this in another post, but it’s important, so I’ll include it here again. This is for a 1:1 haul through a redirect point. (If that last sentence made no sense to you, read this post first.)

---

OK, so that's for 1:1 redirected pull. How about for a 3:1 hauling system?

Great question. Check out the series of four photos below.

• Top left: the most efficient system, with 90% efficient pulleys. MA of about 2.7 to 1, about as good as it's ever going to get.

• Top right, carabiner for progress capture, pulley on tractor. MA of about 2.4 to 1. Still not bad!

• Bottom left, pulley for progress capture, carabiner on tractor. Disappointing MA of about 2 to 1.

• Finally, bottom right, carabiners in both places. Lousy MA, about 1.8 to 1.

A few thoughts on this . . .

• My calculations use a pulley efficiency of 90% and a carabiner deficiency of 50%. (Yes I rounded off in a couple of places, don't beat me up on the math.)

• It's clear that having a pulley closest to the hand that is applying the pull is the best way to rig.

• Many people think that you should always put the pulley closest to the load. That is obviously not true.

---

Here’s a chart takes a little deeper dive into this for different systems.

With an MA system of 3:1 and only use 50% efficient carabiners, your real world MA is going to be about 1.75:1, ouch! (Plus, you still have the dismal progress of a 3:1, with only 1 foot of lift for every 3 feet of pull, even though you're pulling harder than you should have to.)

In this case, you may be better off using a 2:1 with one good pulley than a 3:1 with carabiners! We can see from the chart that a 2:1 with 20% friction (i.e., a 80% efficient pulley) gives us an MA of 1.80:1. But, a 3:1 with carabiners gives us an MA of 1.75:1.

So, use real pulleys whenever possible.

image: https://roperescuetraining.com/physics_friction_raising.php

---

If you have to use a carabiner, which kind is best?

I’ve heard over the years that generally, a carabiner with round metal stock is is going to be more efficient than the new style “I-beam” construction with a narrower cross section. But is it really? if so, how much?

Here’s a Camp Nano carabiner on the left, and an old school Petzl Attache carabiner on the right.

---

What about the DMM Revolver carabiner?

The DMM revolver carabiner is a cleverly designed piece of gear. It’s a standard snapgate carabiner with a tiny roller wheel in the bottom. The Revolver carabiner was designed to minimize rope drag when lead climbing, not serve as a proper pulley in a block and tackle system.

Many people think (hope?) they can use a Revolver to lighten up their rescue kit, but unfortunately it doesn't work. I don't know precisely why, but I think the pulley wheel is so small that under any significant load, it's sort of compresses and you end up with an efficiency pretty much the same as a standard carabiner. I actually tested it and found about 50% efficiency.

If you want a proper combination carabiner and quality pulley, check the Petzl RollClip (or Edelrid Axiom). It has a more substantial pulley in the bottom and works as advertised under load. It’s not used much by recreational climbers, but it's common equipment for rigging and rescue professionals.

I couldn’t find any sort of formal testing online that showed this, so I decided to try a little observational study myself. 

Components: 

• 10 pound barbell weight

• Digital spring scale (about $11, I used this one)

• 9 mm dynamic rope

• Old style Petzl Attaché carabiner (rounded)

• New style Camp Nano carabiner (I-beam)

• brand new rescue pulley

I tied the barbell onto the end of the rope, ran the rope through the carabiner on a bolted anchor to get a 1:1 with a redirect, clove hitched another carabiner in the pull strand and clipped the spring pulley to the carabiner.

I tried to pull straight down in a slow steady haul, and noted the most common reading on the scale. Any force over 10 pounds shows the inefficiency of the system.

• Force needed with rescue pulley (baseline): 13 lbs - 77% efficient

• Force needed with rounded carabiner: 20 lbs - 50% efficient

• Force needed with a “I-beam” carabiner: 23 lbs - 43% efficient

(Math: 10 / 13 = .77; 10 / 20 = .50, 10 /23 = .43)

So, the rounded Petzl carabiner gives a slightly easier haul. (Note, this result was spot on with the often stated 50% efficiency rating of carabiners.)

Would you notice that extra bit of inefficiency in the real world? I’m not sure. But, if you have the option to use a round stock carabiner over an I-beam type carabiner use the round stock. Every little bit helps, right?

(Also, just for fun, I rigged two identical Petzl Attache carabiner side by side. The force needed to lift these was just 21 pounds, basically the same as the single Petzl Attache carabiner. In this case, adding one more carabiner really did not change the friction one way or the other.)

Now, something that definitely ventures into engineering-land that is beyond the scope of my expertise is something called “coefficient of friction”, which is a technical way of measuring how “slippery” something is. From my limited reading on this, the coefficient of friction for steel is different than that of aluminum, so apparently a steel carabiner will offer less friction than an aluminum carabiner. I don’t have a steel carabiner or else I would’ve tested this, it would have been interesting.

So, in summary, use a pulley if you have one!

Progress capture options

The progress capture (aka ratchet), is a critical part of a hauling system. It allows you to take your pulling tension off the rope to rest or reset, without the load sliding backward.

There are a few possibilities for the ratchet. Here are some common ones, listed in increasing order of cost and/or weight:

• Garda hitch

• prusik

• plaquette style belay device (like a Black Diamond ATC Guide or Petzl Reverso)

• Grigri

• Petzl Traxion

Let’s have a look at the pros and cons of these. 

Note: especially for crevasse rescue, it’s really important to practice these different ratchet systems in the real world. It’s one thing to have them work in your living room floor, it can be completely different to see how they work under a lot of tension with snow being jammed up inside them. 

I measured the efficiencies of these systems and wrote about them here.

---

Garda hitch

Pros: free and weightless. (So far, so good!) Cons: even when it’s set up correctly the carabiners can get a little wonky and fail to lock up, so it’s not the most reliable system, in my opinion. Plus, it adds a HUGE amount of friction, making your haul a lot harder! Personally, I would maybe use the Garda hitch for some non-critical tasks like hauling up a backpack, but not in a rescue scenario unless it was really the only option. Read more on the Garda hitch here. (And yes, I know it’s best not to use screwgate lockers for a garda hitch, sorry about that in the photo . . .)

Prusik loop

The classic method, and one still often used by guides, rescue teams, fire departments, etc. Pros: inexpensive, lightweight, can be improvised out of almost any kind of sling material. Cons: If it’s cinched down hard on the rope, it can add friction to your pull. You always want the prusik to be loose when you’re pulling, but in the confusion and stress of a rescue this can be an easy step to overlook. Unless you have a prusik minding pulley, (also known with the great acronym of “PMP”), or an extra person sitting next to it with the unenviable title as “prusik minder”, or someone who’s coordinated enough to do both the hauling AND the prusik minding at once, the prusik can get sucked into the pulley and cause all kinds of problems. Plus, every time you slack off from pulling, unless someone slides the prusik back toward the load, the load is going to slide backwards the length and stretch of the prusik loop, which can mean when you reset your pulley you’re losing a foot or so of hard-earned lift. (Your partner stuck in the crevasse will NOT appreciate being a lowered a foot or two when this happens.)

An old-school Crafty Rope Trick (CRT): if you don’t have a prusik minding pulley or just a carabiner at the anchor: run the rope through a tube style belay device like an ATC before you clip it through the carabiner. The belay device keeps the prusik loop from getting pulled through the carabiner. It actually works surprisingly well, give it a try. See photo below.

Black Diamond ATC Guide

(or similar plaquette-style device). If you set these up in autolocking belay mode, the rope will slide through as you pull, but when you let go, will lock down immediately. Pros: no loss of progress when you stop pulling. If you’re already belaying your second from the ATC, it's very simple to set up a 3:1 left. Cons: An extra piece of gear you may not have with you, especially on a glacier climb.

I did some informal testing on this, and believe it or not, found it does not add any significant friction. This surprised me a lot, because when I tested the ATC simply as a redirect for 1:1 progress capture, it had a terrible efficiency of only about 15%, and I assumed this would also transfer over to doing at 3:1. Happily, it does not!

I tested this with both a 10 pound load and a 100 pound load. Real world mechanical advantage was the same in both cases: about 1.5 to 1. This is pretty much the same as using only carabiners and no progress capture at all.

I'm not entirely sure, but I think I know why this is happening. When you start to pull, you generate a little bit of slack at the ATC, and this slack means the rope passes through the ATC with minimal tension, so no extra friction. I am no engineer, and this is just one observational study, but it seems to me that’s why it works reasonably well.

Grigri

The Grigri functions much the same as the plaquette style belay device, with pretty much the same pros and cons. You may well have one with you when rock climbing, but not for a crevasse rescue situation. A Grigri also has the benefit of being able to release under load, which can be a great help if you need to release tension for some reason.

As mentioned just above with the ATC Guide, I did some testing on the Grigri to check the real world efficiency. It turns out to be the same as the ATC guide, which is pretty much the same as just running the rope to the carabiner. In other words, it does not introduce any significant extra friction. I tested agree with a 10 pound weight and a 100 pound weight; the results were the same in both cases.

The real world mechanical advantage with the Grigri as a progress capture is about 1.5 to 1. This is a good thing! It means you can use commonly carried devices as a progress capture with no friction penalty of decreased efficiency.

Petzl Traxion

These little suckers have a high efficiency pulley combined with an ascender type spring loaded rope grab. They give you a great easy pull combine with zero loss of progress . Pros: work perfect.  Cons: cost about $100, ouch! (Kind of a lot for seldom-used rescue gear, IMHO . . .) Traxions come in various flavors: the Micro, the Mini (the one I have, in the photo below, now discontinued) and the Pro. For alpine climbing, you want the Micro.

If you’re doing a 1:1 haul of fairly heavy bags on a big wall climb, you probably want a slightly larger diameter pulley wheel to get a small increase in efficiency. One popular ratchet pulley for big wall climbers is called the Kong Block Roll.  (I don’t have one, but word is they work great.)

MA in the real world

In a previous post, I did a little testing of 1:1 pulls redirected through various devices. Now, I wanted to see how things worked with some actual mechanical advantage.

I tested various rigs of 3:1, 2:1 and 2:1 “far end haul” systems.

Here’s what I used, and here’s what I found.

Components:

• Anchor about head height

• 10 pound barbell weight

• Two reasonable quality rescue pulleys 

• 9 mm dynamic climbing rope

• Sterling hollow block or WIld Country Ropeman as a prusik

• Digital spring scale

• Various carabiners (Petzl Attache rounded stock, Black Diamond Neutrino and DMM Revolver)

I tied the end of the rope directly through the barbell weight for the 3:1, ran it up through the anchor, and set up pulls using different gear. For the 2:1 tests, I ran a short runner through the barbell plate and clipped the rope to that.

I did several slow, steady, pulls on the rope and recorded the most common number from the scale. 

This is hardly a scientific test by any means, but I think I can give a pretty good idea of relative efficiencies. Try to replicate this yourself. All the extra gear you need is a 10 lb weight and a spring scale, about $11 on Amazon.

My testing set up looked like this:

What do we see here?

• 2 pulleys in a 3:1 gives the most efficiency, but it’s still basically a 2:1 in the real world.

• If you have just 1 pulley, put it on the load strand of a 3:1 to get the most efficiency.

• Not much difference between round stock and non-round stock carabiners, use what you have.

• A 2:1 with a pulley required the same pulling force as a 3:1 with carabiners. You may want to use the 2:1, because you get more lift with your pulling strokes.

• Use a pulley with a far end haul - using carabiners gives a less than 1:1 mechanical advantage.

Progress capture - efficiencies of various devices

So, we have more than few options for Progress capture / ratchet devices. But which one’s best, in terms of minimizing evil friction in our MA systems?

I did a few studies on the efficiency of different pulley, carabiner and ratchet systems, and found some to be dramatically better than others.

It's important to note that the efficiency of the ratchet varies a lot on your rigging. If it's a progress capture for a 1:1, the efficiency might be terrible. But, if it's a progress capture for a 3:1, the efficiency can be much better.

Here’s how I set up my 1:1 study.

• All 1:1 pull with a redirect through the anchor, no mechanical advantage

• Fixed anchor point around head level, could be pretty much anything

• About a 9 mm dynamic rope

• 10 pound barbell plate

• Various types of pulleys / ratchets clipped to anchor point

• Inexpensive spring scale from Amazon, about $11, cloved to the pull rope, LInk: https://www.amazon.com/gp/product/B00ZWNGZFO/

The set up looks something like this. The scale is cloved hitched to the left side “pull rope”.

I set up the pulley or ratchet, then slowly pulled the spring scale to raise the weight and noted the scale reading during the steadiest pull I could manage. The measured force is approximate as I used a cheap spring scale, but I think it’s accurate enough to give a rough idea of efficiencies.

I tested pretty much every flavor of pulley or ratchet mechanism that I owned.

All of the pulling force listed below is for a 1:1 redirected pull of a 10 pound weight.

---

Here’s a summary of the raw data.

and here’s a bar chart:

Here are some takeaways.

• Never use a garda hitch or ATC in guide mode as the ratchet in a 1:1. Friction is HUGE, it took about 60 lbs of pull to move a 10 lb weight! (If this is your only option as a progress capture, you’re probably better off setting up a separate 2:1 on the load to lift it, and then when the rope has some slack, use the hitch or ATC to capture the progress of the loose rope.)

• Pulling force of round stock vs “I-beam” carabiners is pretty similar, not really noticeable in the real world.

• DMM Revolver carabiners did not seem to reduce friction very much, comparable to a plain round stock carabiner in this study. (They were actually worse.)

• 2 identical carabiners side by side did not change the friction much compared to a single carabiner.

• If a prusik if jamming in your pulley even a little (as I had), it adds noticeable friction.

• Using a 5.5 mm Spectra cordelette gave better efficiency than a 9 mm diameter climbing rope. This gave the best efficiency of 87%.

• Pulley ratings from manufacturers are probably calculated under ideal lab conditions, and not under real world testing conditions like I tried to model.

---

As I covered in this article, “Progress capture options”, things are quite a bit different with a 3:1. Surprisingly, using an ATC guide or similar device, or a Grigri as a progress capture introduces no significant amount of friction into the system, at least according to my limited testing.

I tested this with both a 10 pound load and a 100 pound load. The efficiencies were just about the same with each one. A 3:1 Z drag set up with just carabiners at the change of direction gets a real world mechanical vantage of about 1.5 to 1.

Check out the photo below. You would think that ATC Guide would had a ridiculous amount of friction, right? In fact, it doesn't really add any at all.

---

Finally, we have this interesting chart created by Yann Camus of BlissClimbing. (Shared here with permission from Yann. He’s an expert in rope soloing, and if you want to learn this from someone who's been there done that, Yann would be an excellent choice.)

It shows a few interesting general concepts: The smaller diameter cord, the greater the efficiency. The largest diameter pulley, 3 inches, gave the highest efficiency. The DMM revolver carabiner was slightly better than a regular carabiner, but not nearly as good as a proper pulley. (If you're squinting at this graph on a phone, it's easier to see on a desktop / larger screen.)

image: https://blissclimbing.com/en/

Does an MA system make hauling easier?

This might sound like a trick question. The first response might be, “Duh, of course it makes it easier, you can pull more load with less effort!”

Well, that's true, MA does multiply your pulling effort. But this increased lifting force comes at a cost of increased lifting distance. As the economists say, “There’s no such thing as a free lunch!”

Example: Imagine you’re on a big wall climb with a 200 pound haul bag, and your climbing partner is a big burly rugby player who weighs 250 pounds. RubgyDude leads the first pitch, which is 100 feet and happens to be a little overhanging. (This is great for hauling, because it means no friction between the haulbag and the rock.) RugbyDude decides to rig a 1:1, because he knows he outweighs the haul bags and just wants to get the pain over with. Besides, it's only the first pitch and he’s still feeling pretty fresh.

The next pitch is yours. It also is 100 feet long and overhanging. You finish your lead and start thinking about your hauling set up. At this point, you have to ask yourself two questions: 1) Do I weigh less than the haul bags? and 2) How much pain do I want to suffer? If you weigh under 200 pounds, then obviously a 1:1 with just your bodyweight is not going to move the haul bags, and you’re going to have to use some kind of MA to get that bag up the cliff. If you happen to weigh a bit over 200 pounds, it might technically be possible to do a 1:1 haul, but you know you're going to be a wreck when it's over. You decide to rig a 2:1 haul.

When the bag reaches the anchor, you have moved the same amount of weight over the same distance as RugbyDude did on his pitch, you just pulled in 200 feet of rope to his 100 feet. In the end, the same amount of “work” was done, even though the 1:1 would've hammered you, and the 2:1 allows you to still feel pretty good when you have finished. HINT - This is why the 2:1 hauling system is popular on big walls when you have a serious load.

There’s not a correct answer here; it has to do with your own strength, bodyweight, and willingness to suffer.

Think of it this way. Would you rather lift 200 pounds once, or 100 pounds twice?

Or, as Sticky discovered above when she set up the 2:1, “Do you want to work hard or do you want to work smart?” Talk to RugbyDude in a couple of days and see how chipper he's feeling at the top of pitch 15; he might think that 2:1 haul is sounding pretty good!

Does an MA system put more load on the anchor?

Warning to non-enginerds: this question gets a little technical, feel free to skip it if you want. The takeaway is YES, it usually does, so make your anchor extra stout. If you want to know why, read on.

Ahh yes, this is a very interesting question, and one subject to much debate on the inter-webs.

Lots of folks think that an MA system always magnifies the load on the anchor. For example, if you have a 2:1 system lifting a 100 kg load, then the anchor is holding 200 kg. A 3:1 system, the anchor is holding 300 kg, etc.

This is NOT correct!

• Theoretically, an MA system does not put more force on the anchor.

• In the real world, with friction, you can magnify forces on the anchor.

Let's first have a look some ground rules, and then we’ll get into the friction part.

---

The higher the MA of the system, the more force goes to the anchor.

Let's have a look at the diagram below, made with the clever software vRigger. Assuming no friction in the system, and a load of 100 kg, there's quite a large difference in the load that gets transmitted to the anchor with the 2 to 1 compared to the 3 to 1. This is a pretty straightforward rule that applies to all MA systems.

Why is this? The lower the MA of the system, the more of the load you’re supporting with your hand, and the less goes to the anchor. With a 2:1 system, about half the load is on your hand and half is on the anchor. With a three to one system, roughly 1/3 of the load is on your hand and the other 2/3 of the load is on the anchor.

made with vrigger software

---

A redirect on the anchor for pulling increases the force on the anchor.

This is a great rule to keep in mind when you want to reduce forces on your anchor, such as crevasse rescue with one buried dead man.

Check out the example below with a 2:1. With the standard set up on the left, about 0.5 times the load goes to the anchor. But when you redirect as shown on the right, about 1.5 times the force of the load goes onto the anchor.

made with vrigger software

---

In the real world, friction can magnifies force on the anchor.

Let's revisit Sticky, who is pulling a simple 1:1 pull, redirected through the pulley high up in the tree. Only this time, instead of the tree branch conveniently hanging out over the edge of the cliff, the tree is set back, so the rope is running over a rock ledge. For this discussion, let's say that Sticky needs to pull with an extra 50 pounds of effort to overcome the friction of the rope running over the ledge. For her to lift the 100 pound load, she needs to generate 150 pounds of effort.

If she pulls with 150 pounds of effort to raise the load, that means there is now 150 pounds on the strand coming out the other side of the pulley. Which means the anchor is holding 300 pounds rather than 200 pounds. Remember, when you redirect your pull, that redirect point will receive twice the force that you apply.

Note that this does not have anything to do with the actual mechanical advantage of the system. Instead, it's an example of how friction in your hauling system can result in increased forces on your anchor, regardless of the mechanical advantage you’re using.

Is this a problem? Maybe, maybe not. Do you have an anchor on a stout tree limb or three well equalized points of rock protection? Probably not.

Or, is your anchor a a vertical snow picket, and you're about to set up a 3 to 1 haul with two strong people trying to pull somebody out of a crevasse? Then, that lone anchor might be a significant problem. Take a few more minutes and make a deadman anchor from that picket, or better yet, two that share the load.

---

With this in mind, let's consider a real world crevasse rescue scenario.

• One person from your 3 person rope team fell into a crevasse. The rope going to them is cut deeply into the crevasse lip, adding a lot of friction.

• You only have 50% efficient carabiners instead of 90% efficient pulleys.

• On your alpine climb, you’re using small diameter, stretchy dynamic ropes.

You set up at 3:1 Z drag, and you and your partner both start pulling with your entire body weight. To move the load, your pulling force has to overcome all the extra friction from the carabiners, the rope against the snow, (and the inefficiency of the stretchy rope) and that pulling force has to be transmitted to the anchor. How much force? Hard to exactly say, but two people pulling as hard as they can on a Z drag with a lot of real world friction can generate a BUNCH! That extra force has to be absorbed by something - most of it’s going onto the anchor. (Better bury another picket as a deadman!)

You may be thinking: “Here I am with my buddy, and we’re both pulling as hard as we can on this 3:1 Z drag, trying to lift our 150 pound partner out of the crevasse. Me and my partner weigh a combined 300 pounds, so in theory, if we’re pulling with a 3 to 1 we should be creating 900 pounds of pulling force, but we’re still only barely lifting my 150 pound friend. This is way harder than it should be . . .”

Here's another way to handle that crevasse rescue scenario.

• Your three person team is carrying high-efficiency pulleys, a Micro Traxion or two, and enough extra rope with each and person to be able to drop a loop down to the person in the hole.

• After your buddy falls in, can you drop a loop of rope to them which they clipped to the harness with a Traxion. You also prepare the lip on top, knocking down some loose snow and putting an ice ax underneath the dropped loop to help reduce friction.

• The victim, who is functional, can greatly help in this process by pulling down on one of the draft strands. This effectively reduces his weight and friction on the other strand that’s being pulled.

• Now, the two partners on top can easily pull up the victim, and have a minimum load on the anchor.

To use a more extreme example, let’s say you get your car stuck in a ditch, and you rig a 9:1 with a big tree as anchor to try to pull it out.  Now you have basically a tug of war between the tree and your car. When you pull onto 9:1, the anchor (in theory) gets your pulling force multiplied by 8.

Now, say someone else steps up to help you pull. How much force on the anchor components can two strong people apply? That’s now the force of two people pulling multiplied by eight. Now you’re probably getting pretty close to the safe working limits of some of your equipment. There’s a chance the weakest links on the system could start to fail, like prusiks sliding or breaking or even hardware failing. Hopefully the car moves before anything breaks or slips, but the point is, your anchor and the components of your pulley system need to be stout enough to handle these magnified forces.

So, here is the final answer - In the real world, mechanical advantage systems often result in extra force on the anchor, because of the extra effort needed to overcome friction. The greater the MA of your system, and the heavier the load you’re trying to lift, and the more friction is involved, the stronger your anchor needs to be.

This is discussed in the excellent book “The Mountain Guide Manual”, by Marc Chauvin and Rob Coppolillo, pg 276.

The 3 kinds of pulley systems

I'm reading up on pulley systems, and I’m hearing about “simple”, “compound”, and “complex”. What do these terms mean, and which one should I use?

Now, this is heading into slightly more advanced territory, but if you’ve read this far you hopefully still have an interest in slightly esoteric things like this. :-) As to which one to use, there’s not a quick and easy answer.

The one potential issue with the compound and complex systems is that you usually have to reset the pulleys more often as they “collapse” (or, are pulled into each other) when you pull. If you have a large working area, like the top of a crevasse, this is probably not a big deal. If you have a tiny working area, such as a hanging rock belay, then it might be more of a problem.

If you want to geek out on this, look at the YouTube video links at the bottom of the page on compound pulleys and start playing around on your living room floor. That's really the single best way to learn this. You can do it with some parachute cord and a few carabiners, you don't need pulleys or a even a real climbing rope.

1 - Simple system

When you pull the rope, the pulley(s) move in the same direction and the same speed toward the anchor.

As the rope is pulled, the pulley moves toward the anchor at a constant speed. There are three strands of rope going to and from the load and load strand, so this means it's a 3:1 MA. This is also known as a “Z drag”, because the shape of the rope is a “Z”. (If you tilt your head to the left . . .)

• In a simple pulley system, when the rope end terminates and is attached at the anchor, then the MA will result in an even number (e.g. 2:1, 4:1, 6:1, etc.).

• When the rope end terminates and is attached at the load, then the resulting MA will be an odd number (e.g. 3:1, 5:1, etc.).

In the photo below, the rope end attaches to the load, so we have an odd MA number, 3:1.

A 3:1 simple system. The pulley moves at a constant speed toward the anchor.

2 - Compound system

When you pull the rope, the pulleys move in the same direction, but at different speeds toward the anchor.

This can be created by building a 3:1 Z drag, and then adding a 2:1 onto the strand you’re pulling. With a compound system, the mechanical advantage of each separate pulling system is multiplied.

Below, we see a 3:1 on the white rope, and a 2:1 on the black rope. Together, the two systems are multiplied to get a 6:1. Note that the white rope will move the load 1 foot for every 3 feet of rope you pull, while the black rope moves upwards 1 foot for every 2 feet of rope you pull. Therefore, the black rope will reach the anchor point before the white rope, meaning you need to reset the system more often.

Note - If you have 3:1 set up and and need more pull, making a compound 6:1, as we see below, is often a great idea. An example would be crevasse rescue on a two person team, when one person on top may have to do all the pulling. If you have a lot of friction from the rope running through the snow, and/or your partner in the crevasse is not able to assist you, the 3:1 is probably not going to work. Then, the 6:1 is going to be your best friend. Adding the 2:1 only requires one additional pulley and carabiner. Sweet!

Note: For a compound pulley system, you can add the very Crafty Rope Trick (CRT) of building a second anchor that’s farther away. This can allow you to completely collapse the 3 to 1 system before the 2 to 1 system collapses, which means you need to reset the system less often. Granted, this trick is probably more appropriate for professional riggers or maybe search and rescue teams, and not so much for climbers, but it’s still a pretty cool trick.

6:1 compound system, 2:1 on a 3:1.

Two different pulleys move at two different speeds in the same direction.

3 - Complex system

A complex pulley system is one that doesn't quite meet the definition of a simple or compound. A complex system has a pulley(s) that moves in the opposite direction of the load. Complex MA systems are okay, but a simple or compound system is usually a better choice, because they are generally easier to rig and require fewer resets.

Below we have a 3:1 simple system. With the addition of a friction knot (red) and carabiner, we now have a 2:1 pulling on the 3:1. Because this is a complex system, the two components are added together, giving a 5:1.

This is now a complex 5:1 system. When the rope is pulled, both pulleys move toward one another. When the pulleys touch (aka “collapse”), you need to reset the system. Probably not a problem if you have a large area to work in. But if you’re on a tiny rock ledge, you’ll only get a foot or so of lifting until the pulleys collapse, which is going to be a hassle.

Compare this with the compound 6:1 diagram just above. With the 6:1, you get a little more MA, plus avoiding the collapsing pulley problem, so that's why the complex system is usually not the top choice.

5:1 complex system.

A basic 3:1 with a red friction knot added, and the pull strand redirected through it.

Two different pulleys move toward each other at different speeds.

---

What's the MA of my system?

I get confused when I look at these fancy diagrams with ropes running in all different directions. How can I figure out what the real mechanical advantage is of a given system?

You're right, things can be difficult to figure out! At one level, you could say it doesn't really matter. If the 3:1 isn't working, you can add or multiply a 2:1 on top of this, and hopefully the resulting 5:1 or 6:1 gets the job done. The number doesn't really matter in the end.

But, since you asked the question . . . Remember our discussion simple, compound, and complex systems from this post? (Go read it now if you have not seen it.)

The answer of “what’s my MA” varies depending on which one of these systems you’re using.

For a simple system, we calculate the MA by counting how many strands of rope are going to and from the movable pulley(s) on the load or load strands. And, always remember, any pulley or carabiner that’s fixed on the anchor only changes the direction of pull, and does not create mechanical advantage. Let's look at a few examples.

Also, in a simple pulley system:

• When the rope end terminates and is attached at the anchor, then the MA will result in an even number (e.g. 2:1, 4:1, 6:1, etc.).

• When the rope end terminates and is attached at the load, then the resulting MA will be an odd number (e.g. 3:1, 5:1, etc.).

There’s 1 strand of rope coming from the load. So, 1:1 simple system (no mechanical advantage gained).

There’s 2 strands of rope going to and from the load. So, 2:1 simple system. (Rope end attached to anchor, even number MA of 2.)

There’s 3 strands of rope going to and from the loaded strand. So, 3:1 simple system. (Rope end attached to load, odd number MA of 3.)

---

For a COMPOUND pulley system, the hauling systems are MULTIPLIED together.

For example, this is a 2:1 on top of a 3:1, so multiplied we get a 6:1.

image: https://roperescuetraining.com/raising_6-to-1.php

For a COMPLEX pulley system, the hauling systems are ADDED together.

For example, the image on the right shows complex system. The 2:1 on top of a 3:1 added together gives a 5:1.

image: https://roperescuetraining.com/raising_5-to-1.php

For more on simple, compound and complex pulley systems, see post “The 3 kinds of pulley systems”.

---

There‘s also a more math oriented way to calculate your MA. It’s called the “T-method, aka “counting tensions”.

Here are two excellent videos on using the T method to not only calculate the theoretical mechanical advantage but to also account for friction in your system. If you want a deeper understanding of this, this is a great place to start.

---

Pulley size and rope stretch

What effect does pulley diameter have on efficiency?

A larger diameter pulley wheel (aka sheave) is technically more efficient than a smaller diameter pulley. But it’s a trade off: a larger pulley has increased bulk, weight and cost. For example, in a 1:1 haul, you gain about 7% efficiency going from a 1.5” pulley to a 3.75” pulley. This efficiency increases with really big loads (600 lbs+) and larger mechanical advantage, such as 6:1 and 9:1. So, it's probably of interest to mountain rescue teams, of moderate interest to big wall climbers, and not very relevant to alpine climbers. As long as you use a good quality pulley, the diameter doesn't matter much in climbing applications.

In the real world, on, say, a crevasse rescue, you're probably not going to notice the difference between a pulley that's 80% efficient versus a pulley that's 90% efficient. Get a small-ish rescue pulley from a name brand company and don’t stress about the actual efficiency rating.\

A trusted “workhorse” pulley is the Petzl Rescue - rated to 95% efficient with a 1.5 inch / 38mm metal sheave.

---

In the chart below, note the low carabiner efficiency - about 53%, ouch!

---

Is it better to use static rope or dynamic rope in a hauling system?

Steel cable has essentially zero stretch, and offers the highest efficiency. Next best is static rope. Third-best is dynamic rope. Alpine climbers may only have a dynamic rope available, so they may not have a choice. (However, this is one more argument in favor of using a static rope for glacier travel, see this tip for more on that.) Big wall climbers, however have a choice between a static or dynamic haul rope. Static haul ropes are more popular on big walls, and this is one of the reasons, greater hauling efficiency.

When you haul a big load with a dynamic rope, you have to pull all the “stretch” out of the rope before the load even starts to budge. As you might imagine, this is not much fun. Once all the “stretch” has been removed from the rope, if you then pull in a steady constant speed, all ropes are going to behave pretty much the same way. However, in the real world, you're going to have a pull that isn’t so smooth; you're going to accelerate and decelerate. When you do this, the dynamic rope is stretching and relaxing, back-and-forth, and this absorbs energy and lowers your efficiency.

See the graph below. Steel cable, with a large diameter pulley, is about as good as it gets, 98% efficiency. The other four flavors of static rope are all pretty darn close. Note again, the efficiency increases slightly when you go to a larger diameter pulley wheel. Too bad they didn’t test dynamic rope in this experiment, but it was done by a mountain rescue team, and they almost always use static ropes for hauling systems.

image: http://itrsonline.org/wordpress/wp-content/uploads/2014/09/McKently2011_ITRSPresentation.pdf

The MOFT - a very Crafty Rope Trick

Note - This post discusses techniques and methods used in vertical rope work. If you do them wrong, you could die. Always practice vertical rope techniques under the supervision of a qualified instructor, and ideally in a progression: from flat ground, to staircase, to vertical close to the ground before you ever try them in a real climbing situation.

---

I first saw this tip from IFMGA Rock Guide Karsten Delap. Connect with Karsten at karstendelap.com

---

“MOFT” stands for Munter Overhand Feed Through”. It’s a very #CraftyRopeTrick that allows you to pass an overhand knot that’s connecting two climbing ropes directly through a munter hitch (or a Super Munter hitch, which adds more friction) either when lowering or (less common) when rappelling.

There may be a few (probably rare) times when you may need to do this, such as:

• You partner is injured 100+ meters up. They are not able to rappel on their own. You tie two ropes together, and lower the injured person 120 meters at a time using an MOFT. You then pull up the ropes and rappel normally. (Or, if it's a serious injury, fix the rope, rappel the single line and come back later and get your rope.)

• You, experienced climber, are on an alpine route with two relative beginners. On the descent, there’s a long section of fourth class rock and 50 degree snow that the beginners cannot easily manage unroped, but you can safely down climb.  You tie both the beginners in close to the end of the rope and lower them both down two rope lengths / 120 m off of a bomber anchor, with a MOFT. When they are on safe ground, you then break down your anchor, untie the rope, toss it, and down climb. Instead of six rappels, you’ve done one lower and a down climb. Everybody down fast, no rappelling, no beginners having to pass the knot. 

• You and your partner are descending a big wall route with your haul bag. You’re within two pitches of the ground. You tie two ropes together and lower the pig to the ground, using an MOFT to pass the knot.  You secure one end of the ropes to the anchor. Partner 1 raps to the ground on a single strand, passing the knot on the way.  Partner 1 unties the rope from the pig. Partner 2 pulls up the rope and does two raps to get to the ground. Because, rapping with the pig is generally a hassle and best avoided when possible.

---

Notes . . .

• This is a fairly advanced maneuver that you absolutely need to practice before you try in the real world. It’s definitely strange, at least it was for me the first few times I tried it, and I did not find it very intuitive. In fact I had to do it a few times in slow motion to really understand what was going on!

• Practice both the lower and the rappel. The concept is pretty much the same, the execution is slightly different.

• First off, for lowering practice, you need some tension on the rope to really do this right. The easiest way is to have a friend just lean back with body weight on the rope a few feet away (on a flat floor please, not on a cliff the first time you try this!)

• Second, you’re going to be lowering and rappelling on a munter hitch. Be sure you know how to tie it, and how to lower on it. Use a friction hitch backup attached to the brake strand of the rope, and clip the hitch to your belay loop with a locking carabiner.

• Third, this requires a large diameter HMS, pear-shaped locking belay carabiner. Sidenote, if you want to learn what “HMS” means, click here.

• Good choices for a carabiner would be the DMM Boa, Black Diamond Rocklock (photo below) or similar extra-wide carabiner. Do not try this unless you have a wide HMS carabiner, or else the knot could get stuck.

• This works best on skinnier ropes. Any rope under about 9.5mm should be fine.

• Photo: CAMP Core Lock on the left, Black Diamond Rocklock on the right. You need a large HMS carabiner like this for the MOFT.

---

Here’s how it works for lowering. (It’s pretty much the same for rappelling.)  

• Tie two ropes together with a flat overhand bend.

• Tie your partner in one end of the rope.

• Tie a munter hitch onto the large diameter carabiner on the anchor master point, and start lowering your partner. Back up your lower: put a friction hitch on the brake strand and clip it to your belay loop. (For clarity, not shown in the photos below.)

When the knot connecting the ropes arrives at the carabiner, continue lowering (or rappelling) slowly. Try to assist the tails of the rope through the carabiner, but do it carefully so your fingers don’t get caught. Yes, the overhand bend will pass THROUGH the Munter hitch and carabiner! (Like I said, Crafty Rope Trick for sure!)

If you're lowering someone, if they can stand up for a moment and take the weight off the rope when you do this, it’s a lot easier.

If you did it right, once the knot passes through, you’re going to have what looks like a strange looking mess of three strands of the lowering rope coming down from the carabiner, looking something like this. Don't worry, that's what it’s supposed to look like.

The “U” shaped loop of the munter is caught on the overhand bend. Yes, this looks like a mess, but there’s an easy fix. When you’re practicing, this is the part you may want to do slowly to see what’s going on.

Try to keep this loop small. The larger it is, the more you're gonna drop your partner which they probably won't like.

Important safety note! Do NOT reach through the loop of the rope to grab the tails, you could lose a finger! Instead, reach below the loop, take the two tails, and carefully push them through the loop.

Once you have the two tails passed through the loop, pull on them. Warn your partner before you do this, they’re going to drop a little! At this point, if the person being lowered can lean into the rock or slightly unweight the rope for a moment, that makes is a bit easier.

The more the knot has passed through the carabiner, the more they will drop, so as seen below, ideally don’t let that overhand knot go more than about 6 inches / 15 cm below the carabiner.

Now, the munter hitch will magically POP back onto the carabiner, and you can continue lowering. Yes, it looks like an optical illusion, as in “what the hell did I just see!?” Try it a few more times in slow motion to see what’s really happening. It's quite amazing!

It might be a little exciting for your partner if they hear a popping noise and the rope suddenly drops a foot. This scenario will be a much less dramatic if you can have your second lean in on some kind of a stance to momentarily take their weight even partially off the rope. Actually, with dynamic ropes and your partner being 60 meters below, they may not even feel it.

Also, it's probably best to do this when using a standard dynamic rope. If you do this with a semi-static rope, the extra little drop can put additional force on the anchor, which is generally not a good thing.

Some people seem concerned that this will shock load the anchor. That's not gonna be a problem, because you've got 60 meters of dynamic rope below you. Yes some of the stretch is taken out, but it's still going to be very gentle on your anchor. If your anchor is more than halfway decent, this is not an issue.

---

Finally, here's a nice Instagram reel that shows how it's done. They’re using a double fisherman's to connect the ropes, which is even more bulky than an overhand, and it still works fine.

View this post on Instagram

A post shared by vladislav (@babikov_vlad)

Rappel backup: avoid a prusik above your device

Adding a friction hitch (such as a prusik or autoblock) as a rappel backup becoming more widely accepted. More conservative climbers might use one pretty much always. Other people prefer a back up when:

• beginners are rapping

• your hands are cold

• the rope is wet

• rapping on a single strand or a skinny rope

• if you need to swing or pendulum to reach the next rap station

• rapping with a heavy pack

• when you can see the rope is clustered / hung up and you need to free it

• if you’re not sure where the next rap station is

Or any combination of these factors. In other words, pretty much any situation other than a rap  in perfect conditions.

So, what’s the best way to rig a rappel backup? With an autoblock tied below your device.

The older school method was to add a prusik to the rope ABOVE their rappel device. So the theory goes, if they lose control of the brake hand on rappel, the prusik will catch them. This sounds reasonable, but this method has a few problems.

1 - If the prusik knot is above your rappel device, for it to lock up, it needs to hold all of your weight. With the knot below your device, it only needs to hold the same amount as your brake hand, which is minimal.

2 - Once it’s weighted, the rappeler must remove their entire body weight from the knot in order for it to be released, which if you don't know a few Crafty Rope Tricks, is actually kind of hard to do. (Bonus tip - One fast and easy way to remove your weight from loaded prusik is to pull one foot up underneath your butt, wrap the rope a few times around your foot, and stand up.) With the prusik below the device, you can very easily weight and unweight the prusik as needed. 

3 - For the prusik to slide freely, the non-brake hand must be on it or perhaps above it during the rappel to slide it along. To catch on the rope and stop the climber, the non-brake hand needs to be off the prusik. Problem: In the event of a loss of control, our instinct is to grab tighter on the prusik or the rope above it, not let go of it. This grabbing keeps the prusik loose, prevents it from cinching it down on the rope, and may cause the climber to accelerate down the rope and . . . SMACK!

4 - Another old school method was to attach the back up hitch to your leg loop. This is definitely not recommended, because your weight will end up hanging from the leg loop, which could flip you upside down, yikes!

(The one time when it might be a good idea to attach a prusik hitch ABOVE your rappel device before you start your rappel is if you know you’re going to be passing a knot. However, this is an very rare situation for most climbers, and usually can be avoided entirely if you know a crafty rope trick like this one.)

The image below is how you probably do NOT want to rig your rappel.

A better rappel backup method is to use an autoblock knot with an extended rappel. Here, the backup knot is tied below the brake hand rather than above it. If the brake hand comes off, the autoblock immediately grabs the rope and stops the climber. The auto block and extended rappel are covered in depth at this tip.

Here’s a photo from that post to show you the difference. Note the rappel device is extended away from the harness with a locking quickdraw (one of various ways to do this), and the autoblock knot is below the rappel device.

What’s a “closed” rope system?

---

You may have heard the term “closed rope system” in various books and websites. However, I’ve rarely seen it clearly defined, so let's talk about it.

A “closed rope system” means that both ends of the rope have a knot of some kind in them. It’s best practice to make this the default system for pretty much every climbing situation. 

A closed rope system can take many forms. Here are a few:

• A stopper knot in the free end(s) on a rappel or when top rope belaying,

• The end of the rope clipped or tied to a pack, rope bag or similar. This is helpful for single pitch top roping. If you start pulling the other end of the rope, you’ll immediately notice the other end is clipped to a pack. This avoids the extremely common problem of pulling the stopper knot up the route out of reach, whoops! Consider using a clove hitch instead of a bight knot. Once you unclip the hitch, the rope is free to pull. See above photo.

• Your retraced figure 8 tied directly to your harness,

• Both rope ends clipped to the anchor, as you might do with the “J loop” technique on a multi pitch rappel,

This is a simple habit that can prevent the end of the rope from ever going through a belay or rappel device, two common causes of climbing accidents.

Think of closing your rope system with knots the same as wearing a seatbelt when driving. 99.99% of the time you're never going to need it, but that one time you do, you're going to be damn glad you had it.

And, it's worth mentioning again, pretty much always close the system when rappelling. (The exception to this might be a one pitch rappel and you can clearly see the ends of the rope or on the ground.)

---

Comments . . .

Some people object to tying knots in the rappel strands, saying they “don't want the knot to get stuck”. I’ve never understood this. Unless you have unusually deep rope-eating cracks right below you, the first person down should easily be able to stop, pull the knot out of whatever crack it might be in, and toss it on down the cliff. A key rule of rappelling is to never go below any rope that’s stuck. The first person down should fix any issues and lower the rope properly.

If it's a crazy windy day and you're worried about getting your knotted rope stuck in some far-off rock crevice, you have some options. You can lower your partner with both ends of the rope, or the first person can saddlebag the rope.

Think of it this way: how many people have died from a knotted rope end getting stuck? Compare that to, how many people have died from rappelling off the end of the ropes?

Do you have a rope that's “long enough”? Don't be complacent. Say you’re climbing a 25 meter route and you have a 60 meter rope. No problem, you think, I have an extra 10 meters of rope. But, if your belayer backs up from the wall or walks downhill, or maybe the climber decides to pendulum off to one side and clip a redirect piece of gear, or something strange like that, you might suddenly end up short when lowering off. Having a knot in the end of the rope, or having the belayer tied to it, eliminates this potential accident. Always close the system, even if you're doing single pitch sport climbing and initially appears you have plenty of rope to lower off your partner.

When your rappel ends on the ground, have the first person down untie the stopper knots. You’re on the ground, so obviously they’re no longer needed, and they need to be untied to pull the rope. So, get into the habit of having the first person do this. This helps prevent that all-too-common mistake of starting to pull the rappel rope with a knot still in one end, yikes!

---

For a rappel, consider tying your stopper knot at least 2 feet from the end of the rope. By doing this, you give yourself enough extra rope to at least tie an overhand. This gives you something to clip your tether to, in case you screw up and reach the knot. If you don't do this, yes, you did prevent the catastrophe of rapping off the end of your rope, but now you might be kind of screwed, because you maybe can't do much else.

---

Think you’d never make a mistake like this? If it can happen to Alex Honnold, it can darn sure happen to you.

In 2016, Alex was dropped by his belayer because they were using a 60 meter rope on a 70 meter route, there was no knot in the end of the rope, and his belayer was not tied to the end of the rope. Whoops, open rope system!

While the belayer was lowering Alex, the end of the rope zinged through their Grigri and Alex fell onto some “gnarly rocks”. Luckily he only suffered mild injuries. Other climbers in the same situation have died.

Read the accident report here.

CalTopo pro tip - stacking map layers

(If you want to learn the basics of using CalTopo, start with this tutorial video.)

---

If you’ve spent any time around the navigation department of Alpinesavvy, you know I’m a huge fan of CalTopo.

Here’s a lesser-known CalTopo ninja trick you may find helpful - stacking map layers. 

In CalTopo, you can stack two or more map layers together on the screen, and then with a slider bar or typed in percentage, change the opacity of each layer. (If you've ever used photo editing software like Photoshop, you're probably familiar with this idea.)

Stacking allows you to see features of interest in two different map layers at the same time. I rarely view more than two map layers at the same time, and definitely not more than three, otherwise you’re going to go cross-eyed trying to find the view you want.

Procedure:

1. From the main screen at caltopo.com, zoom to your area of interest.

2. Mouse over the map layer in the top right corner, and choose “Stack New Layer”

3. Choose a second map layer from the drop down to lay over the first one.

Let’s start with a satellite image base layer, for land cover, and overlay it with Open Cycle, which shows trails. Let's set the opacity of the Open Cycle later to 60%. You can do that with the slider bar, or by typing the percentage directly in the box as seen below.

In the screen grab below, we see the south side of Mount Saint Helens. You can clearly see which trails are in forest cover and which are not.

---

Let’s try a combination of satellite view, showing crevasses, stacked with slope angle shading, showing steep areas and possible avalanche terrain. 

The settings for that from the top right corner would look like this:

Here’s the SE side of Mt. Hood. You can see crevasses and areas of steep, moderate and low slope angle. This could help you create a route that avoids crevasses and stays off of steep terrain.

With both of these on the screen, you can draw a line that hopefully avoids most of both. In CalTopo, right click, choose “New > Line”, then hold shift, and click and drag draw a freeform line. 

You can then export this line as a GPX file for use in actual navigation in the field.

I think you get the idea. Dive in and play around with this.

And, if you find it worthwhile, I strongly encourage you to support CalTopo with an annual subscription of just $20 a year. Doing this helps ensure this great software continue to be available for us all to enjoy.

Avoiding snags when pulling a rappel rope

When rappelling, you always have a choice as to which side of the rope to pull. While on many routes it doesn’t matter, on some raps it may be the difference between having a stuck rope and/or a very difficult pull. Look carefully at every rappel you do, and try and identify potential problem points before pull the rope.

• Generally, try to pull the rope AWAY from any obstacles it may hang up on. For example, say you’re about to pull a single rope rappel, and you look up and notice a potential rope-eating shrub, block or crack below and slightly to the right of the anchor above you. Pulling on the left strand of the rope will cause the rope to fall to the left a bit, hopefully away from the potential hang-up point.

• If you’re doing a double rope rappel, you need to anticipate problems BEFORE you rig your rope. Get into the habit of looking down at the cliff below you before you thread the rope. Do you see a shrub or rope eating crack to the left? To keep your rope away from that, you probably want the knot on the side on the anchor away from that obstacle (or, to the right as you’re looking down.)

• Look up at the rope as you're pulling it. If the rope gets hung up on something, you might be able to see how it got hung up, which could help you solve the problem.

• If you’re on the ground when you pull the rope, you’ll almost always get a better angle of pull if you walk as far away from the wall as possible. If there's a snag point, say on the left as you're looking up, step away from the wall and to the right as far as you can. This can decrease risk from falling rock, and minimize friction.

image: From the book “down”, by Andy Kirkpatrick. Used with permission. andy-kirkpatrick.com

• On steep terrain, give the rope an extra flip or sharp tug just AFTER (not before) it slides through the anchor. This can cause it to fall farther away from the rock, further avoiding snags. Avoid the common mistake of giving a tug on the rope BEFORE it goes through the anchors. Doing this can cause the free end to whip around the other strand and make some random evil friction hitch that can block your rope. (I know this from painful experience . . .)

• But on lower angle blocky terrain, (or in high winds), it might be better not to do this and just let the rope slither down the rock under its own weight.

• If the anchor below you is off to one side, say the right, you should set your rope so the pull is in that direction. If you don't, the ropes can cross at the anchor and create a lot of extra friction.

image: From the book “down”, by Andy Kirkpatrick. Used with permission. andy-kirkpatrick.com

• If the wind is coming strongly from a certain direction, rig your rope to pull it into the wind. Or, in other words, if the wind is coming from the left, pull the left strand. Doing this makes the tail blow away from the pull strand, which reduces friction and hopefully avoids wind-induced rope mayhem.

• If you’re rapping in terrain where rope snag potential seem to be everywhere, you're probably better off doing a series of shorter raps using a partial length of the rope. It can be a better choice to have more rappels, but less chance of getting your rope stuck.

• Maintain control of the “up” strand until the last possible moment. This can be especially important on overhanging, traversing, or windy situations. If you get lazy with this, it's all too easy to have a stopper knot still in the end of the rope, you losing control of it, and having it swing out of reach. Big problem.

• In terrain with high snag potential, another option is lowering the first person and having the other end of the rope tied to them. This of course does not help with pulling the rope, but it can get it down more efficiently.

• If you’re doing short rappels in terrain with a lot of rope snag potential, try this: If you have plenty of rope at the lower anchor, pull down as much rope as you can before the final person goes on rappel. This reduces the chances for a hangup or rockfall when the rope is pulled and before it goes through the anchor, because less rope overall is moving through cracks, over blocks, around shrubs, etc.

Team of 3 in moderate terrain? Try end roping

Note - This post discusses techniques and methods used in vertical rope work. If you do them wrong, you could die. Always practice vertical rope techniques under the supervision of a qualified instructor, and ideally in a progression: from flat ground, to staircase, to vertical close to the ground before you ever try them in a real climbing situation.

---

Scenario: You have a team of three climbers and one rope. You’re in moderate technical terrain - low fifth class rock with good climbers, 4th class with occasional “steps” of harder moves, a open big slab, or fairly steep snow. The terrain falls lies that in-between spot, ranging from “No prob, I can solo this!”, to, “Hmm, I think I want a rope!”

This might be a good time for end roping.  

Look! It’s our friend Sticky! She’s back, and climbing with her 2 pals. Hi Sticky!

This is a technique often suited to alpine climbing, where you may have many pitches of easy to moderate climbing, and speed can become more important. It’s best not to use it on much anything higher than mid 5th class.

The leader heads out as usual, but the two followers tie in close to each other near the end of the rope. (End roping, get it?) The leader then belays up both followers at once.

The “end” follower ties in to the rope end as usual with a rewoven figure 8 knot. The “middle” follower, about 4-5 meters away from the end person, ties in with one of several ways, more below.

With the middle person on a large bight of rope, it gives them some freedom of movement. If the end person falls, the middle person may not be pulled off because of this loop. And, it means the followers do not have to be moving at the exact same rate.

---

A few things to note:

• You need to use a single rated rope, not a skinny twin or double.

• This is ideally for rock and maybe steeper snow. For ice climbing, you always want your seconds on separate ropes and end roping is not a good choice.

• The anchor needs to be especially stout, because it could potentially be loaded with the weight of two people with a little slack in the rope.

• If the last person falls, it’s possible they will pull off the middle person. Because of this, it’s best if the strongest climber of the two followers is on the end of the rope.

• This is really only a safe system up to low fifth class for most people. For more difficult terrain, you probably need a different rope system.

• If you're in an area where the rope is going to be potentially loaded over sharp rock edges, keep in mind that you could be putting a double load on the rope if the seconds fall together.

• This works best if the route is pretty much straight up. If you start any kind of a traverse, a slip by either of the two followers will probably result in a fall for both of them.

• Communication between the belayer and the followers needs to be clear and obvious, in case something goes wrong. If the route is long, it goes out of sight around the corner, it's a windy day and it's hard to hear each other, or all of the above, this may not be the best technique.

• The large loop connecting the middle person to the rope is a bit of a trade off. The loop can give you more freedom of movement, but it also can also be a tripping hazard. Use your judgment here. A smaller loop (1-2 feet) might be a better choice depending on terrain, skill of the climbers, if you have crampons or not, etc.

Use some common sense. If the consequences of a fall are low, meaning the terrain is a slab or snow or not very blocky, and it goes pretty much straight up, and your climbing team is skilled enough to mean a very low chance of a fall, this technique will probably work fine.

Anything more advanced or difficult than this, you're probably better off using a two rope system and bringing up each climber on a separate rope.

---

There are a few ways for the middle person to connect to the rope. Here’s one that's pretty straightforward - tying in directly to your harness with the rope. There are a few ways to do it. This example uses a bowline knot.

1 - Tie an overhand knot, leaving a large bight of rope about 5 feet long.

2 - Pass this bight through the two tie in points on your harness, and tie a bowline knot. (Yes, you use both strands of rope, and this might look a little funky if you've never done it before. It's the exact motions as tying a bowline with one strand of rope.)

3 - Finally, secure the tail of the bowline with a locking carabiner to the loop through your harness.

---

Another option is to clip this bight of rope to the middle’s belay loop with two carabiners, opposite and opposed, with at least one being a locker. (If you are feeling a bit bold, one good locker with the carabiner properly aligned with no chance of cross loading is probably okay too. It's really up to your personal level of acceptable risk. Me, I like two carabiners.)

Cross loading the carabiners is the main thing we're trying to avoid here. There are several ways to avoid this:

1. Use the belay loop rather than the harness tie in points, as this minimizes tri-axial carabiner loading.

2. You can tie a clove hitch in the bight loop and crank that down on your carabiners.

3. You could use a nifty new-school locking belay carabiner, like the Black Diamond Gridlock, that has a clip that captures the carabiner so it can’t be cross loaded.

Basic set up: overhand on a bight with 2 opposite and opposed carabiners.

Add a clove hitch and cinch it down to hold the carabiners in place.

Or, go with the sweet Black Diamond gridlock or similar carabiner designed to resist cross loading.

image: blackdiamondequipment.com