top of page

Monkey fists don't work, or get stuck if they do

In the Czech Republic, there is amazing climbing on sandstone that isn't really held together well. Climbing protection like nuts or cams can damage the rock and there are very few bolts or "rings" to clip too. It has been a tradition there and is the actual rule, that you can only place textile protection. Tie knots in ropes and webbing and wedge them in like nuts or hexes. But does it actually hold a fall???

Thank you Jenny Fischer for narrating this episode and explaining where and why the rules exist, how it isn't ideal to be whipping on them, and how they can get stuck.


We tested overhands, double overhands, figure 8s and monkey fists wedged in our adjustable crack jib with granite slabs inside. This was in the drop tower and slow pull slacksnap machine to see if we got wildly different results, and the slow pull machine is easier to get a camera to show what's happening inside the crack. DON'T take these numbers as gospel. this was backyard science and the jib was flexing and the load cell app may not have been fast enough since it isn't the full 1280hz like the LineScale3 device itself is. It also really depended how constricted we made the crack. HOW they broke, got stuck or slipped out is way more interesting than the number itself and I encourage you to watch the video to see that. Timestamps will help you skip around.

SO WHAT KNOT TO DO - Test Summary

1" webbing tend to not slip as easily and the loop didn't break until 7-10kN. The surface area is pretty good as friction plays a big role, but also probably because we could pinch the jib pretty tight below it. The 10mm knots almost all came out because the tightest the bottom of the crack can be is 10mm but one held 8kn which would hold a typical lead fall. Roll your dice. Paracord isn't strong enough at 3.5kN and only held when a rock was inside the monkey fists which you can't use. 6mm is barley good enough because when it did hold we would get around 6kN. Our slow pull 8mm monkey fist was impressive at 25kN but it slipped in the drop tests.

10% Supports us

Buy pro that works! Totems, Aliens, C4s and more are in stock.

How KNOT to Test

Decker lost an arm

Many people wonder why we do certain things in our test. You can see in the video we started with a 90lb dummy named Decker. We can move him around easy and it's more entertaining to see a dummy than steel weights but as you can see, that didn't work out. We aren't trying to see the force if A, B or C is done, we are trying to find out the breaking strength of each of these tests, so steel weights it was. We only do 2x 75lbs for a 150lb "dummy" because we just need to break the gear and really heavy weights free falling on only 9" of rubber gets complicated fast.

When we use a "static" rope on the samples, it is too stretchy, believe it or not, to break most of the stronger gear 8kn and up. We have to go with more static or more weight so we chose a more static item - an industrial sling. It holds up to abrasion well and is plenty strong enough. Most ropes break around 16-20kn in a figure 8 so even if we went with more weight, ropes are quite limited.

8mm Monkey Fist slipped

Load cells have to be FAST if using a STATIC attachment. The pro itself stretches enough that 1280hz, or 1280 readings per second, is probably super fast enough, though we will be testing this with a 10,000hz load cell to find out how short our peak force is soon. The app however is only able to feed at 40hz and the first few samples was only the app's number until we discovered that. Based on the numbers we got, it was super close enough because HOW this stuff broke or slipped was more interesting than the actual number itself. I think it's super helpful to know the flaws in any test because most testing has some flaws in it. If a polished graph is put on instagram or in an article, it is a mess with lipstick on it. We want so badly for a single number to explain everything and that is the problem with even MBS or minimum breaking strength, a number that has all variables removed so sterile that it doesn't even represent reality anymore. Reality is messy and that's ok. Understand your gear holistically and don't lean on 1 number to explain the entire story.

Another Test

The most referred to testing regarding textile passive pro is this ARTICLE by Jörg Brutscher from a 1997 test where they used a clamp and pulled various knots to see if they would slip or break. We found that our knots would only stay if the constriction was 1 diameter wide (pinching the rope) but they found with the figure 8, you could go up to twice the diameter of the rope. My rock jib is laughable because you can see it literally flexing on the drop tower, but this is also not very realistic in that friction plays a huge role in how these knots do. Their jib doesn't replicated a realistic scenario very well as it doesn't consider friction at all. Their machine only went up to 10kn. On their graph, you can see 8-10kN is "hard fall" which we could only achieve 8kn with TANNER who is 290lbs and his grigri was bolted to the wall. You would have to be doing something gnarly with someone heavy to get above 6kn. If trad gear is holding above 8kn, it's super good enough.

In 2010, they tested knots in real rock which is pretty cool. The PDF shows their testing set up. They tested a figure 8 knot with two strands in an 11-13mm constriction. They hung a 50kg "belayer" and I believe the wei