Disc Golf Biomechanics

[polakos21]

Hi everyone,
I am a biomechanist by trade and work in an office with an 8 camera markerless motion capture system and tri-axial force plates.
I attached a video below which provides 2 camera views of a backhand standstill throw with time synchronized force plate measurements.
Figure on the left is the front foot force plate ( Green line = Vertical Force (Ground Reaction Force Z), Red Line = Horizontal Force/ Direction of throw (Ground Reaction Force Y).
Figure on the right is the back foot force plate ( Green line = Vertical Force (Ground Reaction Force Z), Red Line = Horizontal Force/ Direction of throw (Ground Reaction Force Y).
Note: The Vertical Figure axis starts at the top (0N)

I will be collecting trials over the next few months and start collecting 3D kinematic (joint angles, joint angular velocity, joint angular acceleration) and 3D kinetic (Forces in all three planes of motion). Please let me know if you have any ideas, theories, or cues you would like tested. Once I develop the variables to assess and accompanying pipeline, I can begin sharing data from different capture sessions.

I am aware of my rounding problem and will be working on fixing it. Since I am currently rounding, I will be interested to see how I can leverage the data to make technical advancements (e.g., try to produce more lateral force vs vertical force, increase or reduce elbow flexion angle at certain points of the throw).

I will also be working on defining the backhand throw based on phases:
e.g.., back leg touchdown, peak reach back, front foot touchdown, Power pocket). If anyone has suggestions as to when to biomechanically define "power pocket" I'm all ears.

Looking forward to fruitful conversations

 

[Brychanus]

Oh. Yes.

If you weren't aware, Dr. Kwon has data sponsored by Chris Taylor from multiple top PDGA pros, but there are reasons that has been slow to move forward. Just mentioning in case there is an opportunity to team up. Not sure what your specific area is if you'd like to say more about that.

Do you have interests in developing holistic motion/predictive models, or just testing property->outcome relationships?

I'm going to nudge you toward one of the main topics people debate about, and then try to reduce it down to a single mathematical hypothesis abstracted from the details of the motion itself.

One model has "power pocket" at the theoretical center of mass and/or gravity of the body.

Another model (which is not necessarily mutually exclusive) emphasizes the "deep pocket."

Power pocket at "CoM"



But I have a bigger fish for you to fry.

One related deep idea lurking underneath these endless debates is whether or not the idealized swing path approximates Fibonacci's sequence in motion (the motion is 3D, but this would be the theoretical motion along the swing plane. The research hypothesis would be that building form that increasingly approximates this pattern would be (1) more efficient and (2) more powerful once momentum/athleticism are added.

I have no investment in one view being right or wrong. I am a scientist, and I just want to know the truth, or some approximation of it.

If someone tested these things appropriately, I would be grateful to modify Fundamentals in my link below and to stop writing so much. Happy to support you in any way I feasibly can or clarify what I am asking you to consider.



Plenty more you could look at specifically. Would be good to not just experiment on yourself but get some high PDGA rating person(s) in there in this case. Does a "Fibonacci quotient" describe the best forms?

Sidewinder first pointed this concept out to me/the community. Was he right or wrong, or somewhere in between? Someone like you with some "peer" input might be able to figure it out.

 

[BenjiHeywood]

Really interesting. I did some similar stuff in Ultimate a few years ago, but the data never got processed (and I'm not able to access any in raw form). The guy in charge of the lab lost interest.

So I'm super keen to see what you find.

Re: pocket - I'd look for the time that the forearm starts to rotate outwards (ie the point at which the wrist starts traveling faster than the elbow, obviously ignoring any times that happens during the wind up etc). That might happen with the disc behind the body if someone is rounding badly, or it could be from left pec or centre or right pec. That doesn't define what a pocket is, but it does (IMO) define the point at which your pocket has done most of what it's going to do. It's certainly an interesting way to compare different throwers.

And i think comparison is by far the most exciting thing - get a bunch of throwers and see what is similar and what is different and how it all correlates to results.

Is Simon moving faster than me throughout the throw, or is he kind of similar through most of it but getting more disc speed out of the back end of the throw? What does the acceleration curve look like for wrist, elbow, shoulder, hip for predominantly vertical or horizontal bracers (v + h separated by force plate ratios i guess). That sort of thing.

And of course comparison of the same thrower in different reps - can we see differences that might explain why one throw is just a tiny bit faster than another? But i think a lot of that would need to be sense-checked by also looking at how different throwers do it. The combination of the variation between people and the variation within one person's reps might help us nail down some of the correlation versus causation stuff.

Sounds like you'll have a ton of fun!

 

[Brychanus]

I am aware of my rounding problem and will be working on fixing it. Since I am currently rounding, I will be interested to see how I can leverage the data to make technical advancements (e.g., try to produce more lateral force vs vertical force, increase or reduce elbow flexion angle at certain points of the throw).

Will help if you loooooad the bow baby!

 

[polakos21]

Oh. Yes.

If you weren't aware, Dr. Kwon has data sponsored by Chris Taylor from multiple top PDGA pros, but there are reasons that has been slow to move forward. Just mentioning in case there is an opportunity to team up. Not sure what your specific area is if you'd like to say more about that.

Do you have interests in developing holistic motion/predictive models, or just testing property->outcome relationships?

I'm going to nudge you toward one of the main topics people debate about, and then try to reduce it down to a single mathematical hypothesis abstracted from the details of the motion itself.

One model has "power pocket" at the theoretical center of mass and/or gravity of the body.

Another model (which is not necessarily mutually exclusive) emphasizes the "deep pocket."

Power pocket at "CoM"
View attachment 335362


But I have a bigger fish for you to fry.

One related deep idea lurking underneath these endless debates is whether or not the idealized swing path approximates Fibonacci's sequence in motion (the motion is 3D, but this would be the theoretical motion along the swing plane. The research hypothesis would be that building form that increasingly approximates this pattern would be (1) more efficient and (2) more powerful once momentum/athleticism are added.

I have no investment in one view being right or wrong. I am a scientist, and I just want to know the truth, or some approximation of it.

If someone tested these things appropriately, I would be grateful to modify Fundamentals in my link below and to stop writing so much. Happy to support you in any way I feasibly can or clarify what I am asking you to consider.

View attachment 335363

Plenty more you could look at specifically. Would be good to not just experiment on yourself but get some high PDGA rating person(s) in there in this case. Does a "Fibonacci quotient" describe the best forms?

Sidewinder first pointed this concept out to me/the community. Was he right or wrong, or somewhere in between? Someone like you with some "peer" input might be able to figure it out.

I have read the forum post related to the data collection sessions conducted by Dr. Kwon and have been eagerly awaiting the results. From what I gathered, they utilized marker based motion capture, which is the gold standard for motion capture. I'm not sure what the hold up is, but I'd be happy to help if needed.

My background is in power-based sports. I was a competitive javelin thrower for 12 years and I have published a research paper on biomechanical predictors of the javelin throw and currently coach some javelin athletes. My technical expertise is in track and field throwing events (e.g., javelin, discus, and shot put. I have a solid understanding of the hammer throw).
My other research areas focused on predictors of explosive jumping performance, comparisons between olympic weightlifting variations and relation to jumping performance, and resistance training biomechanics.

I have the following research ideas with respect to the disc golf backhand and forehand throws:
1) Establish biomechanical definitions for key landmarks (phases) in the disc golf backhand and forehand throws.
- Once consistent phases have been established, we can compare similarities/ differences in mechanics between elite/ novice and within elite throwers.

2) Determine what are the biomechanical (kinetic and kinematic) predictors of disc release velocity
- in terms of angles and release parameters of the implement, I believe Tech Disc would be more adequate. Combining a tech disc with motion capture and force platforms would give a robust overview of input and output.

3) Test technical cues and provide clarity as to what is happening mechanically to provide sound definitions

The idealized swing path is an interesting rabbit hole. I would need to read up on the Fibonacci quotient.
Did Sidewinder make a post about the Fibonacci quotient/ idealized swing path that I can review?

 

[polakos21]

Really interesting. I did some similar stuff in Ultimate a few years ago, but the data never got processed (and I'm not able to access any in raw form). The guy in charge of the lab lost interest.

So I'm super keen to see what you find.

Re: pocket - I'd look for the time that the forearm starts to rotate outwards (ie the point at which the wrist starts traveling faster than the elbow, obviously ignoring any times that happens during the wind up etc). That might happen with the disc behind the body if someone is rounding badly, or it could be from left pec or centre or right pec. That doesn't define what a pocket is, but it does (IMO) define the point at which your pocket has done most of what it's going to do. It's certainly an interesting way to compare different throwers.

And i think comparison is by far the most exciting thing - get a bunch of throwers and see what is similar and what is different and how it all correlates to results.

Is Simon moving faster than me throughout the throw, or is he kind of similar through most of it but getting more disc speed out of the back end of the throw? What does the acceleration curve look like for wrist, elbow, shoulder, hip for predominantly vertical or horizontal bracers (v + h separated by force plate ratios i guess). That sort of thing.

And of course comparison of the same thrower in different reps - can we see differences that might explain why one throw is just a tiny bit faster than another? But i think a lot of that would need to be sense-checked by also looking at how different throwers do it. The combination of the variation between people and the variation within one person's reps might help us nail down some of the correlation versus causation stuff.

Sounds like you'll have a ton of fun!

That's unfortunate. What was the hypothesis of that project?
In terms of defining the "pocket", I would speculate it would be the point of peak elbow flexion. I'd have to review some of the data to confirm.

I think if I first come up with phases for the Disc golf throw, mechanics will be much easier to breakdown and compare within and between individuals.

 

[Brychanus]

I have read the forum post related to the data collection sessions conducted by Dr. Kwon and have been eagerly awaiting the results. From what I gathered, they utilized marker based motion capture, which is the gold standard for motion capture. I'm not sure what the hold up is, but I'd be happy to help if needed.

My background is in power-based sports. I was a competitive javelin thrower for 12 years and I have published a research paper on biomechanical predictors of the javelin throw and currently coach some javelin athletes. My technical expertise is in track and field throwing events (e.g., javelin, discus, and shot put. I have a solid understanding of the hammer throw).
My other research areas focused on predictors of explosive jumping performance, comparisons between olympic weightlifting variations and relation to jumping performance, and resistance training biomechanics.

I have the following research ideas with respect to the disc golf backhand and forehand throws:
1) Establish biomechanical definitions for key landmarks (phases) in the disc golf backhand and forehand throws.
- Once consistent phases have been established, we can compare similarities/ differences in mechanics between elite/ novice and within elite throwers.

2) Determine what are the biomechanical (kinetic and kinematic) predictors of disc release velocity
- in terms of angles and release parameters of the implement, I believe Tech Disc would be more adequate. Combining a tech disc with motion capture and force platforms would give a robust overview of input and output.

3) Test technical cues and provide clarity as to what is happening mechanically to provide sound definitions

The idealized swing path is an interesting rabbit hole. I would need to read up on the Fibonacci quotient.
Did Sidewinder make a post about the Fibonacci quotient/ idealized swing path that I can review?

He's shared it publicly before so I don't feel bad relaying it here - you could reach out to Chris Taylor directly and see if there is a way to access what you would need. Worth a shot. I would just like to see this work make progress myself.
[email protected]

Well, sounds like your background is going to mean disc golf would love your input! I am a Jan Zelezny fan.

Your expertise about lifts and jumping will be relevant to debates people have about "one-legged" vs "two-legged" throwing mechanics.

From a novice biomechanics enthusiast perspective like mine, your goals seem interesting and relevant.

Keep in mind that tech disc has a different center of mass than "real" discs, but otherwise makes sense and there is evidence that people are benefitting from working with them at this point.

If you didn't know, Sidewinder contributed the lion's share of work over the years and I served as editor and coauthor of the "Fundamentals" guide in my link in my signature below. Even if you don't feel like reading the rest, here are a couple of the core ideas that can possibly be computed from motion capture,
Fundamentally:

The Fibonacci quotient/path is mentioned just briefly there as an "idealized" hypothesis about the motion along the swing plane for a given throw. If the swing plane can be rendered on a single 2D plane form the mocap it seems possible to try and do a function fit or AUC or maximize some objective function. I don't know that Sidewinder intended it literally, but it's an interesting simplifying idea about the overall path a swing should take relative to a player's center of mass. That concept is either close to or exactly right, or wrong in some fundamental way (or somewhere in between). One other related idea is whether there is efficiency and peak power potential in the motion of the center of mass follows something like a brachistochrone curve in advanced form.

You'll notice that in a lot of Sidewinder's conceptualization there are analogies to all kinds of throws and swings, and one-armed hammer throws are exactly one model for part of the backhand. So if you prefer to focus on the "parts based" elements of the motion capture and model among them, there are plenty of things to ask and clarify to the benefit of DG. There are also various questions about individual differences and adapting different bodies for optimal form. I am happy to help clarify anything since I know his content well, as well as where some of the controverisal areas emerge between coaches. Chris Taylor himself will likely be willing to point out his takes if you talk to him.

 

[BenjiHeywood]

That's unfortunate. What was the hypothesis of that project?
In terms of defining the "pocket", I would speculate it would be the point of peak elbow flexion. I'd have to review some of the data to confirm.

I think if I first come up with phases for the Disc golf throw, mechanics will be much easier to breakdown and compare within and between individuals.

I can't actually remember what the project was officially looking at, though i do remember having to do an awful lot of paperwork (the facility was in a hospital, so it must have had clearly defined parameters, participant waivers etc). I described it informally here though - Top throwers wanted!

 

[yossarian]

I've tried to extract timing landmarks and swing phases from pro form videos and my own form reviews. So far my only conclusion about "good timing" is that max reach back (coil) should coincide with the moment the plant foot becomes braced.

The Fibonacci analogy feels like a bit of a red herring to me, like a geometric description of a problem that is so much more than geometry. The brachistocrone curve idea seems more promising - and following it to Wikipedia points on to optimal control theory, which might be a good mathematical framework for the analysis. Or just throw a pile of data at a machine learning system and see what comes out.

Very interesting research, I look forward to hearing more!

 

[Brychanus]

My guess is that the actual optimal curvature is slightly different from the Fibonacci analogy, would just be curious if something like that pattern emerges along the swing plane. Closer to how a mathematician or physicist would summarize aspects of objects in motion with a description independent from the inputs-outputs and try to "reduce" the problem into simple representations.

I would fully agree that the problem is so much more than geometry, just want to point out how academics in different contexts might attack aspects of problem. I work with some folks in optimal control theory too and they would certainly have something to say.

 

[sidewinder22]

The idealized swing path is an interesting rabbit hole. I would need to read up on the Fibonacci quotient.
Did Sidewinder make a post about the Fibonacci quotient/ idealized swing path that I can review?

Blake_t on DGR brought up the golden spiral as a way to throw with a 2 finger grip so that you don't lose leverage or slip early.

 

[Mocheez]

Blake_t on DGR brought up the golden spiral as a way to throw with a 2 finger grip so that you don't lose leverage or slip early.

JR,

i'm not really going to respond to all of this as I'm tired and I'd rather just start writing up the finished product.

this is a different technique that is about maximizing angles.

all power throws have one thing in common: they bring the outer edge of the disc around.

there are different techniques that do this in differing ways. e.g. swedish technique can do this with hardly any elbow or wrist bend with an emphasis on driving the shoulders hard. the technique i am working on maximizes the impact of the elbow and wrist and minimizes the impact of shoulder rotation. there doesn't need to be any backwards pull since it can make the wrist unload with only forward motions.

the trick to this is to accelerate the outer edge of the disc on an INCREASING radius. e.g. a golden spiral. the behavior of the arm abruptly lengthens the radius of the arc at the last second and causes a tremendous amount of angular velocity on the edge coming around.

on these throws my wrist is actually completing a 120-140 degree extension (2" of motion is about 45 degrees), which leads to an exceptionally hard and heavy disc pivot.

 

[Brychanus]

Hi everyone,
I am a biomechanist by trade and work in an office with an 8 camera markerless motion capture system and tri-axial force plates.
I attached a video below which provides 2 camera views of a backhand standstill throw with time synchronized force plate measurements.
Figure on the left is the front foot force plate ( Green line = Vertical Force (Ground Reaction Force Z), Red Line = Horizontal Force/ Direction of throw (Ground Reaction Force Y).
Figure on the right is the back foot force plate ( Green line = Vertical Force (Ground Reaction Force Z), Red Line = Horizontal Force/ Direction of throw (Ground Reaction Force Y).
Note: The Vertical Figure axis starts at the top (0N)

I will be collecting trials over the next few months and start collecting 3D kinematic (joint angles, joint angular velocity, joint angular acceleration) and 3D kinetic (Forces in all three planes of motion). Please let me know if you have any ideas, theories, or cues you would like tested. Once I develop the variables to assess and accompanying pipeline, I can begin sharing data from different capture sessions.

I am aware of my rounding problem and will be working on fixing it. Since I am currently rounding, I will be interested to see how I can leverage the data to make technical advancements (e.g., try to produce more lateral force vs vertical force, increase or reduce elbow flexion angle at certain points of the throw).

I will also be working on defining the backhand throw based on phases:
e.g.., back leg touchdown, peak reach back, front foot touchdown, Power pocket). If anyone has suggestions as to when to biomechanically define "power pocket" I'm all ears.

Looking forward to fruitful conversations

These are slightly more concrete and can perhaps come more directly from the motion capture, potentially:

1. Modeling the "top line" dynamics of the arms and shoulders: There are plenty of theories but few models I'm aware of talking about the lever length, sequence, and angulation when maximizing force/exit velocity. In other sports and abstract compound pendulum modeling minor differences in the system can cause large differences at the end of the action. I'd be very interested in how such a model could be used to (a) explain individual differences in high-level form and (b) used to "retrofit" or adapt the model to a given player to inform their development.

Bonus: two the concepts that I think people elsewhere also don't talk about quite as much as around here is (1) the effect of the mass of the levers, and (2) the contribution of the throwing plane and gravity.

Here's what I'm talking about in the context of a simulated pitching case:

Some of Paul Nyman's ideas are controversial and I'd be curious how biomechanics has advanced since this time and confirmed or debunked some of this, but there are a lot of cool related ideas in here:

Exactly what are effective throwing mechanics?

What the Lord giveth, the MLB hitter taketh away…

tht.fangraphs.com

Thanks again to SocraDeez for pointing me to some of this at one point.

2. Second and I'm aware more controversially is the role of the posture, spine, and pelvis. Is it more "golf like" as described here? If it is, how is it motified in high level throws thru the X-step? Is there a fundamental difference between on one extreme "blocking" forward momentum and causing rotation via force transfer at the hip socket vs. using the body in a natural locomotion pattern that involves "redirecting" the overall momentum of the move? Does the body's center of mass slightly "lead" the strides like walking, or hit the brakes completely like a truck hitting a concrete wall? How much player variation is there? Which one is "better," and for whom? I'm being slightly hyperbolic, just emphasizing a couple different concepts people throw around out there.

3. What's the relationship between things 1 and 2, if any?

 

[Mocheez]

I've tried to extract timing landmarks and swing phases from pro form videos and my own form reviews. So far my only conclusion about "good timing" is that max reach back (coil) should coincide with the moment the plant foot becomes braced.

I actually disagree with this. Max reach back (coil) coincides with front toes touching the ground. By the time the front heal hits the ground, they have begun to uncoil with some elbow bend.

 

[Brychanus]

I actually disagree with this. Max reach back (coil) coincides with front toes touching the ground. By the time the front heal hits the ground, they have begun to uncoil with some elbow bend.

One way people talk about that is as the toe contacts the ground it is already starting to resist/plantar flex and transmit force up the chain.

 

[yossarian]

I actually disagree with this. Max reach back (coil) coincides with front toes touching the ground. By the time the front heal hits the ground, they have begun to uncoil with some elbow bend.

For my own form review I was using Simon Lizotte's slow motion video footage, which to my eye shows full heel plant before pull. It may vary between individuals, I haven't reviewed a lot of examples.

 

[Brychanus]

Also watch the shoulder line relative to the foot pressure & think about forces transferring before they show up on camera.



Though I would agree you see some variability pro to pro, shot to shot, and some change within a given pro over time. A little wiggle room in the sequences/style. Simon is a good example of someone who has gone through a lot of changes in his transition move, compare Wiggins vs. Mcbeth, etc.

 

[419HondaGuy]

I recently discovered great sports-biomechanics podcast called Just Fly Performance.
There are many episodes that correlate to disc golf biomechanics, I will link to the current one I listened to during my workout last evening.
It's with David Weck and Chris Chamberlin on Rotation, side-bending and tensional balance in high performance training program.

David Weck and Chris Chamberlin on Rotation, Side-Bending and Tensional Balance in a High-Performance Training Program

David Weck and Chamberlin on bending and twisting integration into more traditional strength methods, concepts on the foot, and much more.

www.just-fly-sports.com

Adriann Barr episodes are great as well, and his work in rotation sports will apply to our world as well. There are about 4-5 episodes with him.
I have head the term " new form of athleticism" several times from a few guest now when talking about the fascia muscular system.

A New Paradigm in Biomechanics: Fascia, Rotation, and Waves - A New Paradigm in Biomechanics: Fascia, Rotation, and Waves
Also, worth taking a look at, you will notice the key similarities that we regularly explore and discuss here.

[h1][/h1]

 

[SocraDeez]

Excellent read. Thanks for linking that here. Sidenote: article author Kevin Foster is the reason the Jan Zelezny javelin throw gif gets posted here so often.

The notion that there is something fundamentally different going on in a standstill vs. regular throw because of "force couples" and "two feet on the ground" is so silly and ignores the meat for the bones. It even came up in the Overthrow AMA: "I'd love to talk about forced coupling in standstills..."

 

[Brychanus]

Excellent read. Thanks for linking that here. Sidenote: article author Kevin Foster is the reason the Jan Zelezny javelin throw gif gets posted here so often.

The notion that there is something fundamentally different going on in a standstill vs. regular throw because of "force couples" and "two feet on the ground" is so silly and ignores the meat for the bones. It even came up in the Overthrow AMA: "I'd love to talk about forced coupling in standstills..."