How I Learned To Stop Trying And Finally Fixed Knee Collapse

Rather than starting with a thesis statement, I’m here to tell a story. And that story is about knee collapse and all my struggles to fix it. This story plays out gradually in the background for almost a decade, as I was progressing along my HEMA journey. Hopefully you can learn from my experience and have your students “skip to the last page”, so to speak.

To begin, let’s look at what I mean when I say ‘knee collapse’. The image on the left shows what I consider to be a collapsed knee. It also has the whole body going over along with the knee, which is something that may or may not happen when the knee collapses.

Right is an example of core rotation without knee collapse.

Act 1: Engaging The Core

I, like most others doing any sort of sport or martial art, started out doing things using traditional coaching approaches. That is to say, you work on the skill in a more controlled and isolated setting and then start adding variability/unpredictability after competency in the skill is learned.

I, unlike most others in HEMA, actually had a coaching background going into HEMA. So while I was new to HEMA, I wasn’t new to coaching, and I very quickly had a role instructing other students and trying to bring the benefit of my knowledge of competitive canoe/kayak. One of the really important things in paddling is engaging your legs/core to rotate and deliver power to the tool (aka paddle) in your hands. So I’ve always been interested in how to transmit power through the body and into the sword.

What sprint kayaking looks like, FYI

To get the most into our swings we need to have the body engaged, which will typically mean rotation of some fashion. This was something that I was comfortable teaching. But what was outside of my experience was the footwork – something noticeably absent from kayaking. What I quickly found was that as people learned to engage their body in the cut, they also tended to start collapsing their knees… a lot.

Act 2: Correcting The Form

To get people to stop collapsing their knees, I went to traditional coaching 101: you give people a form correction. This on the whole proved… basically useless. When given a correction people tended to be able to keep the knee in place when thinking about it, but went right back to collapsing as soon as they focused on something else. (Like the sword swinging at them.) 

Of course adaptations take time, and the conventional belief would be that through prolonged effort and correction there would be eventual improvement. And I wouldn’t say that there was never any gradual improvement, but it was very slow, and very marginal for the effort put into it.

To make matters worse, I was also constantly getting better at teaching people to engage their body with the cut. Which meant a beginner’s ability to learn the thing that aggravates the knee collapse was now vastly outstripping their ability to not collapse. And I was sure I could do better.

Act 3: Action Capacity

My next thought was that perhaps it was an issue with muscle conditioning to resist the collapse: by strengthening the muscles and coordination paths, the process of learning-to-not-collapse could be accelerated. (This is something I would now call an Action Capacity.)

I started by using bands to strengthen the musculature required to keep the knee from collapsing in. By lashing the knees together, the fencer (or me) would have to fight to keep the knees out, and thus build up muscle strength, endurance, and “memory”. 

I can tell you from first hand experience and second hand whining that this was definitely a workout for those lesser used stabilizers. And if you’re making a muscle sore through training, it’s a sure shot you’re making it stronger. Which correlated to… no real improvement in the knee collapse issue. Go me!

Act 4: Movement Pattern

Ok, so if conditioning wasn’t the answer then maybe it’s coordination? Next I pursued the idea that people just needed to learn how to coordinate their body better, and establish/engrain the proper movement pattern so that it can come out properly when actually fencing.

I devised a series of exercises to help build this movement pattern, and started rolling them into the regular warmup exercises. In the one illustrated below, the body is lined up just slightly off the wall and something, like a book, is pressed against the wall with the knee. The upper body is rotated back and forth, while keeping pressure out against the wall to hold the book up. If you let the knee follow the rotation, like it wants to do, the book will fall.

It was a real pain in the ass to try to do this with stick figures from multiple angles, because it’s already pretty ambiguous what direction they are facing.

This seemed to have some improvements, but one of the exercises in particular seemed to work much better than the others, which brings us to…

Act 5: External Cues

As I was developing and implementing these movement pattern exercises I was also trying to re-align a lot of my teaching to be based on external rather than internal cues. (Internal vs External Cues)

The exercise above is technically external, as you’re focused on pushing the object into the wall rather than focusing on the position of your knee. But that’s somewhat of a cop out because the external focus is not on achieving a goal (keeping the knee inline where it belongs), but on pushing laterally as hard as you can while you do rotations.

Instead of focusing on fighting the knee from collapsing (aka, not going where I don’t want it to go) laterally, my next thought was to focus on having it move forward (aka, moving where I do want it to go). So the exercise

  1. Take a dagger and hold it on your hip, pointing forward.
  2. Take a passing step, while trying to stab the dagger forward as aggressively as possible.

Using this there is a clear goal: to have the dagger stab forward, and not sideways. Which is a much better focus for the body, and having people focus on the motion of the dagger did a much, much better job of keeping the knee under control. And it even seemed to have some good carry over with people’s normal footwork, leading to a big decrease in the amount of knee collapse observed in students.

Normally I would call this a big success, but it was actually nothing compared to what was on the horizon.

Act 6: Ecological Approach

If you’ve been paying attention to my recent writings, you already have a pretty good idea where this story ends, which is me moving into using Ecological Approaches to coaching. Which means many things, but it also means that rather than teaching people to step or cut directly we instead jump to activities like Direct Attack or Soviet Foil immediately for all new students.

This was done for many reasons, but I noticed something interesting after a while. Since I was no longer teaching people footwork, I was no longer seeing the knee collapse issue. Which is about as anathema to a traditional coaching perspective as possible. How is this possible?

The reason is that all skills and movements are context dependent. When you have people doing footwork for the sake of footwork, they are going to adopt movement solutions that reflect this lack of purpose. If you are stepping to reach something you know you don’t have any difficulty reaching, you don’t need to step well. So if you are doing a partner drill where your partner lets you get to the distance you need to do the technique, you have no reason to extend the step to your potential. 

Footwork exists to facilitate a goal, not as an end in and of itself. When the goal is doing some technique prescribed by the training activity, the knee collapses to allow greater body rotation and make the technique easier. If the goal is actually getting to the opponent, the knee stays forward because that’s how you translate your body forward to achieve this goal.

Moral Of The Story

One of the most common worries about switching to a more ecological approach is that people will learn bad habits, or some other version of “learn it wrong”. Which I don’t want to invalidate at all, because based on the coaching approaches that we’ve all grown up with it is a very real and apparent problem with the approach.

But this is not the case, and if the training activity has proper goals and context you can actually learn better form in less time. And there are a whole generation of new students who never had to go through this learning struggle to prove it.

This article was just a case study to show how you can learn good form without focusing explicitly on form correction, and that many of the “bad habits” we struggle to fix are actually the result of training divorced from fencing context. And more contextless training is not the answer! If you’d like to learn more you can follow up with: