Analyzing Climbing Technique with Force Plates!

[@mickey_rose]

Okay, technical question: Do these force plates measure in only one axis or in multiple axes? The point being that outward force could be one of the major differences in force disparity since this wasn't tested on a vertical wall. Billy could be pulling in the Z-axis more than Josh, and Josh could be producing more force in the X-axis during drop knees. I'd love to see more of this! Specifically, off balance throws and cut loose moves to highlight the difference between intermediate and advanced climbers.

[@digitalsmear]

+1 for this question. I had the same thought pattern.

[@jonnes__4657]

I think the most interesting on the hand holds is the z-axis. Because that takes your most energy. With the drop knee, the foot holds had no measurement. Of course Josh used better the x-axis on the foot holds to get tension. .

[@lukasvilim8499]

I would say that it is very unlikely that the devices used in this experiment are suited for multi-axial usage. Guessing from how these sensors look like I assume that at best their construction mitigates the outward forces all together or worse it skews the downward pull measurement by applying bending strain to the measured medium. But hands down this is a fun experiment and since they are shooting for very rough measurements and mostly entertainment this is completely fine. I'm pretty sure that they are aware of what limitations they are dealing with. 👍

[@latticetrainingleodefrette8601]

Can confirm these force plates measure in only 1 axis. Still so much to learn, I agree omnidirectional forceplates have huge potentail to help here!

[@Frog_Wizard801]

I also figured the devices were limited to a single axis, which is a shame because like everyone else is saying, I think the z-axis is the most interesting part, I’m not sure how omni-directional plates would work but would definitely be interested in seeing it

[@nickhenscheid369]

As a scientist/mathematician and climber this makes me all giddy. So many interesting studies one could do to quantify movement patterns, key in on specific muscle group weaknesses, analyze injury-prone movement, the list goes on. Wish I had studied kinesiology!

[@elliotrichards6948]

Now we need a full training board that has a pressure sensor for each hold 🤣

[@dark-o]

Hahaha the ultimate "sucking the fun out of climbing" 😂 as Josh would put it.

[@meh6513]

I would love this! Some people like the puzzle solving or the co-operation or collaboration process of the boulder problems but I just like finding the best technique such that if no one at my strength could achieve something but I can distribute my strengths well enough and achieve my best instead of just doing poorly while doing my "best" is very empowering and you get the best efficient stats then people stronger than you or even for women can see "oh there is a person at this height and this gender and if work on my techniques and distribute my energy properly I can achieve this goal once my efficiency is optimized enough... it could give hope to the weak minded people who need to see other people do it...but not want to wait for an hour for someone good to solve it...

[@th3l05t-f1z]

@@dark-o as an engineer I have to disagree it would make it even more fun fore me xD

[@jonnes__4657]

The best way to get frustration. 😜 .

[@groghnash]

also make the sensor sense force in all directions, not just downwards pressure! for example when you do the sideways deadpoint and you weight your foot more you actually need to pull on the handhold more, but sideways! This is not captured by the forceplates, because they only capture downward pressure!

[@TomOHalloranAus]

This is amazing! I made a video about this a couple months back. I think there’s so much amazing data to gather in movement efficiency and technique analysis. Can’t wait to see this expand bigger and better! Good one team!

[@LatticeTraining]

Thanks Tom. We had a bunch of comments say you'd spoken about this! I'll go find it. Let us know if you've got any specific ideas/requests we try do for you.

[@michaellam6612]

Now imagine sensors that can handle more than one direction. The angle you pull at does impact how well you seem to stick to a hold, so I’m sure that info could be useful as well

[@adrianvramulet]

Yes, finally! More of these videos, please! Loved the split-screen comparison on the first move. Would have liked the same for the other moves. I have been thinking of this digital board for 2 years now, and I can finally see how useful it could be to identify strengths and weaknesses and to coach accordingly.

[@areteclimbing]

I've discussed this exact setup before with people in a 'wouldn't it be interesting' way. Stoked that lattice have actually done it😊 One day there will be a Kilter board with a force sensor in each hold...

[@TheMymast]

I would love to see a force plate that not only shows total force, but also the vector, in other words, the direction of force applied over time. I think that would give us so much more insight into "efficient" climbing, sadly I have no idea how that could be achieved, besides maybe a halo of strain gages that all connect to a "floating" gold (held in place only by the gauges). Also, the force plates on the feet would have been a great insight in the twisting move by the end, seeing how the force per feet changes through the twist would be so interesting!

[@helpjack8843]

Could have a force gauge for both the x and y axes, and use some trig on the backend to calculate a full vector :)

[@jonnes__4657]

Should be no problem, but the sensors are more expensive. .

[@danschmidt5189]

​@@helpjack8843 You also need the Z axis - pulling into the wall is a critical technique. I think it's under-appreciated and, generally speaking, more important than actively "pulling" up. It gets the COG into the wall where the legs can actually drive it upwards.

[@helpjack8843]

@@danschmidt5189 Great point, It'd be interesting to see the data on that! Personally I quite like this data driven approach to climbing

[@jasonhuang1447]

Imagine a complete training board where every hold is force-plated, with multi-axial sampling, and able to record every attempt a climber smashes out. Wouldn’t be surprised if we see something like that in the near (or maybe not so near) future! Exciting stuff in sports science :)

[@LatticeTraining]

Yeah this would be so cool. Though I think we'll need a big commercial gym chain to fund this! Sounds super expensive.

[@CamiFL593]

Really great video! Here are some studies I would love to see: 1) Compare force at start vs end of a big session. How much does tiredness change the force you can generate? How much does tiredness increase the force required due to loss of technique? 2) Compare hold types and angles. How much more force does it require to hold a crimp at 40° vs vertical etc? How does this compare for the angle of pull (Directly beneath the hold vs pulling out)? 3) Add force sensors into big moves on a board. Could you use data to predict the max hang performance needed off the wall for specific contact strength required for moves on the wall? If these studies gave usable results then you could get Will on replicas like burden with the force setup and have a measurable target for finger strength required for the moves (Obvs not all you need but a good baseline for athletes before a trip).

[@LatticeTraining]

Great ideas!

[@lars4058]

Oh my god I wanted this for so long! Awesome you are doing it. 3-dimensional obviously is the next step that has to come!

[@mikerockell8590]

SAME !

[@Terminx666]

From a viewer perspective it would be cool to see a setup with LEDs rigged to the holds (like the kilter board holds) that light up in a color gradient which corresponds to the force produced on that hold.

[@Rmikeyhow]

@KilterClimbingGrips where you at on this?!

[@Tferdz]

The area of the curve is the impulse. You can integrate it to directly compare efficiency (normalized by body weight). Or maybe compare the Jerk? Or plot jug vs foot, instead of always by time.

[@Nefif]

Exactly! Looking not only at the peak force might lead to better conclusions!

[@zsoltpszota]

There is something wrong about the 1st and 2nd measurement. Starting with the 2nd one: If we look at not only the weight measured by the sensors, but also the timing, you can see that red line goes to zero right after the green jumps up, while there's still much weight on yellow. The red is indicated as "Foothold", yellow as "Middle jug". It is not reasonable that the weight would disappear from the foot (red) as soon as the climber catches the top hold. Similarly as soon as the climber catches the top hold, the "middle jug" is no longer used so the yellow should have gone down. I think that both in the chart and in the video explanation the yellow/red were mixed up. Also I think there's a similar mixup in the first comparison as well. It is most easily noticable if we look at Josh's chart. He is either using the yellow or the blue, never use both at the same time. Which means that those 2 must be the handholds and red is the foothold. It is also mixed up both in the chart and in the explanation. (In the video yellow is indicated as leg, but that can not be the case for this move.) An other interesting thing that's worth mentioning is the 1st comparison. In this Josh has a huge "Overall" drop in the middle. Basically he stops using the bottom hold before starts using the top hold, which shows that he does the move more dynamically. I really love this experiment, but the explanations and the charts need some adjusting.

[@LatticeTraining]

Might have made some mistakes on this. We are generally pushed for time on these videos and hopefully this is just the "pilot study". We'll take more time on the next one and double check the graphs.

[@carolinedegroot5604]

Thanks for the video & analysis! Really interesting. A couple of ideas for experiments (hopefully not duplicated in the comments): - Test the force output in four kinds of flags in an advanced vs. intermediate climber: forward vs. backward flag, inside vs. outside flag. - It seems like what you quantitatively mean by "efficient" is "work done". Is that right? It would be really interesting to actually compute the work done between different athletes on the same problem, see how low it can go while still completing the move, and investigate the factors which go into that.

[@Ekid33]

One very interesting and useful application of these force-plates would be to set a limit move with them and track the force data as a climber projects it, until they send the move, say, 5 times. You can then compare the average force generated during send goes with average force generated during failed attempts. Until you get omnidirectional force plates, you could set a purely vertical campus/single foot move using this regimen, and might get some interesting results.

[@itzjoek]

would be interested to see area under the graph for total force produced and not just peak force

[@maurodibenedetto2706]

Definitively interesting and probably the tip of the iceberg. In particular for lead climbing, I think the actual efficiency of a move is related to the 'next move' and how long it takes to readjust the body to continue climbing. In other words looking at the integrals of those curves and not only at peak loads. Clearly this is very complex due to variety in climbing

[@EvanNathan-e2d]

Great video! How about the optimal way to rest on a hold, i.e. the body position that exerts the least force on one hand while you shake out the other (at various wall angles)?

[@robinoleary1592]

Love this! Been wanting to do this for movement analysis for ages. Filmed something using just a tension block and Tindeq recently, but this is excellent and really slick

[@pierrez8570]

It's really cool to nerd out on the data for sure. Love this! I do feel like for the twisting, Billy's position looks more unnatural than Josh's with the hold - the drop knee for him feels more awkward. Does he have longer legs than Josh? Despite being the same height this could also make this movement less efficient for him.

[@eurekaflows]

IIRC, Josh is Lattice's flexibility coach. It's not a surprise he's significantly better with drop knees.

[@LatticeTraining]

We didn't measure leg length. I do know that Billy has a larger Ape Index than me, so longer sideways reach. I suspect Billy's legs are stronger than mine so potentially better at rocking his weight over the foot in an open position. The twisting relies less on rock-over (hamstring) strength and more on hip position and pushing outwards on the feet.

[@AllegraClimbingPsychologist]

This could actually be a pretty useful tool to train different focus patterns and divided attention 👀

[@hemasgeir470]

Great video! One thing that would be super interesting for the biggest number i guess would be to increase the gap in climbing level between the climbers. I would be curious to see the datas compared between a 5'ish climber and a 8'ish one. Maybe on some 6th grade moves. I'm pretty sure we would see that the beginner is all about the upper body compared to the skilled climber. Anyway, it's always fun watching you guys, you're killing nothing at all!

[@amirzarie3631]

This is really cool! I wonder what's the difference in the overall force applied to the holds by Billy and Josh in the same move over time cause it seemed like this is comparing max force generated at a certain time point (if I understood it correctly).

[@akademicfedorov]

I love this video! Many times I've envisioned how interesting this might be to do the tests you did! So many nuances you could examine! Ok, so one question you openly avoided is what style is better (drop knee vs open)? As you mentioned, coaching does focus on working with particular persons style. However, I would add that there are real differences between how effective each style is in different situations. And from a coaching perspective you can influence both decision making and the degree to which someone practices their weaker style. I've always made a broad generalization that skill being equal, drop knees tend to be more efficient, while open climbing tend get you further in terms of distance/reach. Obviously this statement is very much situation dependent and also isnt a question isn't one thats best answered by force plates. I am interested, however, in a different comparison between dropknees and open climbing that could be informative in decision making and could be tested with the plates. After doing a drop knee or open move, what types of situations favor each style in post-move or flow to the next move? How much force x time is exerted on the target hold from the time the climber latches it until the time the hand(s) leave that hold? this would include the force x time used to adjust feet and hips, center of gravity etc.. I suspect in some cases we would see the dropknee making the move much easier, but the post-moves harder.

[@deminmax]

Awesome stuff! Do you have a video on how did you do the sensors?

[@kavali6320]

Oh no, I'm becoming a nerd 🙈. Seeing the data compared to the actual movements of both climbers was really interesting. Thanks to the whole team ;)

[@denislejeune9218]

Really grateful for this video as it's really driven the point home that my use of legs while climbing is pitiful. Now if you want to make a video on how to improve that... I'm now trying to be more conscious of the issue while climbing, but the problem as often is when the going gets tough, the brain reverts to type: pulling like a donkey that tires quickly.

[@JosephLaycock]

Assuming the force plate is a single axis, you could make it a "slab" to stand on and find/demonstrate the best techniques to maximise grip by showing how much force is applied to it.

[@bearelykoalified]

This would be a very very tall ask but I'd love to see the future of this sort of break-down where each hold in a climb is mapped to forces being applied throughout a climb and then mapped to a simple rendering of the climber with force diagrams throughout the climb.

[@LatticeTraining]

Would be very cool. Beyond our scope right now 😅

[@zacharylaschober]

while we are discussing the sensor detecting downward force, I think outward force may be a more interesting read on steep terrain and give a solid indication of exactly how strong a climber is as well as simply figuring out efficiencies for successful movements by reverse engineering the weight through the feet (outward force left hand+right hand higher indicates less efficiency).

[@LatticeTraining]

Yeah measuring outward force is a limitation at the moment. But as you say, an important factor.

[@TheAbd1233]

Maybe you could use this device setup to look at the coordination of different limbs. As you can see which body part is applying force if they are on different holds. From the data at 8:22 it seems josh starts off using mostly legs then quickly moves to using a lot of arms. Whereas Billy uses more arms at the start and his transition to the different holds seems to include force from all areas. I hope I am making sense. But I think if you have the time resolution on the device you can clearly see when people start pulling with a body part compared to others.

[@LatticeTraining]

Yeah we briefly discussed how the "shape" of the curve might be as, or more, important than peak force. And how these curves line up with each other. Though we'll need to develop a better system as currently they are out of sync. We literally had to press "go" on each of the 3 data channels in quick succession. Though I think we could line up the data retrospectively. It's just time consuming work.

[@TheAbd1233]

@@LatticeTrainingfair I guess you need to do a lot of software work to get everything cleaned before making very detailed analysis. Anyway cool video

[@Fred-oz3tw]

we need more of this. this is the future. but the very very far away future xD

[@DanielGarcia-rh9oj]

Would be interesting to see how ,much visualization/confidence/consciousness throughout/during the movement or even just thinking about a certain muscle does. Would be crazy to see the numbers on what just "feeling like a good climber" does!

[@LatticeTraining]

Yeah maybe we can take two groups, have one spend time between attempts visualising or being coached, while the other group have to complete a different task which distracts them from visitation/refection (so they get the same rest time). Then see how it effects the learning of a movement pattern.

[@lorenz07lt]

I think the numbers showed that also for josh doing the static move openly was more efficient

[@moonlight00001]

Yes that's how I understood the numbers as well: 45kg openly and 46kg while twisting, but they didn't address it so I don't know

[@crimped]

@@moonlight00001I was hoping they would address this. Although Josh's force on middle jug was 1kg greater on the middle jug with twisted position, it seems his overall force was quite a bit lower. I also wonder how much the muscle groups used for each method could affect the perceived difficulty/overall exertion to execute the move versus the actual forces applied.

[@colinweir2475]

Please do different dynos and pogos from different foot positions and look at the timing between force generated on the hands and feet (and how softly you catch the end).

[@LukeRockCimber]

Analysis 1: When you say the hold are you talking about the weight put into the handhold your catching? That seems like the most important piece

[@Bobonautti]

Wow! That´s so cool!

[@jacobgaylord9277]

I think it would also be interesting to see this but with a monitor on the foot hold(s) as well, to see how much emphasis they put on pressing through their feet and for how long

[@YSlimbo]

Maybe billy is a better climber but your the best coach of lattice training ❤

[@LatticeTraining]

You're making me blush 😍

[@IAmMarlonBrendo]

Might be a challenge, but I'd really like to see some counter pressure, particularly something where you're squeezing into a toe hook to make a hold engage.

[@Arthur-ru5mm]

Could you analyse the difference between doing a scorpion kick and trying to keep your foot close to the wall while dynoing?

[@DeathlyDiJ]

i am intersted to see someone repeating the move 100 times or so, and see the difference from when he gets the move 'dialed'. (with enough rest to not become fatigued in between the first and last attempt)

[@LePetitBat]

It would be really really cool to set an interesting boulder with these holds, and then challenge people to put as little energy as possible overall to climb it, proportionally to their body weight.

[@BigTimeCharley1956]

This is cool, and is definitely not "sucking the fun" out of climbing. Define "efficient;" what exactly is the metric used in "quantifying movement efficiency?" Can Cam (sp?) publish a BOM for creating these types of holds?

[@sneakysnake123]

I think you should do a campus move as well, once dynamic and once slow locking off, just to see what influence dynamic movement has to the graph. I can't really tell, but I guess the peak force on the hold is higher during a very dynamic movement. I have a feeling, that interpreting the graph is not that easy, and the more I think about it, the more I get confused about the phyisics behind it

[@heididixon165]

I've always wanted to see a force plate analysis of how toeing in and body tension reduce the load on fingers on an overhang. I think for beginner and intermediate climbers, the whole body tension thing is really not intuitive. One of the instructors at our gym once told me that you can always toe in harder than you currently are. What does the force plate for your hands look like through that change. Or how about the difference between dropping your butt and hanging down versus arching your back and pulling your hips in toward the wall on a 45 degree.

[@RemusKnowles]

Very cool!

[@karikokkonen5564]

So nice! Try moves with small openhand edges (20mm)? I think it would require more nuanced movement and skill.

[@andrewhunter6536]

Which axis is the force measured in? All?

[@hj40]

this is amazing!

[@Michael-zz8yx]

Finally, this is what tom o'halloran was talking about.

[@JorgePacker]

Do another similar video, but where the climbers don't know about the measurements. Just to see if there is a placebo effect.

[@IdDuncan]

now it is time to go full nerd with direction of force and motion capture! pleeeeaaaase !

[@tanguylehoux4270]

we need that board

[@jameskase5639]

If you integrate the area under the curves then you'll know how much energy or work was done

[@codyratterman6553]

How do the force rates and ratios change before and after getting pumped? When you're fresh vs pumped you often use different beta and some people hold it together, while others fall apart.

[@LatticeTraining]

I really want to look at this. Another conversation I had was "how does fatigue effect movement and can this increase injury risk or skill development?".

[@zachrossi3590]

X, Y and Z direction force gauges for hands and feet. Repeat the test! Then model the data mapped across a sphere like a heat map that changes its signature over time for each foot/hand!

[@dillonliew5370]

So so so cool video.

[@constantinvoicescu1510]

Another factor that would be interesting to see is the anatomical differences between you two. Josh might have longer feet than Billy, while Billy could have longer arms than Josh and so on. P.S. Thanks for all the research that you do there!

[@LatticeTraining]

Billy has a larger Ape Index (longer arms) so I imagine that makes a big difference for some moves.

[@SyRyanYang]

The biggest problem with this is I don't think the force plate measure outward pulling force. If anyone watched one of the previous videos with Aidan Roberts about vacuum style climbing, outward pulling is as important as downward pulling when it comes to climbing at a steep angle. The other thing that didn't get mentioned is that a good, efficient climber knows more precisely the minimum amount of force required to make a move. It would be interesting to see if they repeat the same move over and over again, does the peak force decrease with practice.

[@LatticeTraining]

Yeah we are a bit stuck in measuring force in a single axis. But it would be really cool. I also like the idea of repeated practice to see how this effects efficiency. I think we can do this under different conditions too.

[@hugop7485]

Please,Try to put the feet in tow positions, normal and higher to see if is it harder for a taller climber.... sorry for my English

[@muumarlin1731]

Genius!

[@jamiepiper1995]

Cam cracking a joke about the amount of hair product the others have in, whilst clearly having the most hair product in

[@jonnes__4657]

🗽 To have a tool like this in a gym would be really nice to optimize your efficiency. On a monitor you can show in percent of total weight the force of the starting hand hold and the finishing hand hold. Mostly the force on the foot hold is too low, because of not capable to hold the core tension or the coordination is bad, or both. .

[@denislejeune9218]

Get Ondra, Megos, Ghisolfi and Bouin over and compare their results for various moves. So we have an idea of who is better at what, which may or may not explain why they prefer certains styles of climbing routes.

[@deltaflux2381]

Wouldn't it be more useful to express the force in a percentage of the bodyweight? Since bodyweight influences the absolute force exerted?

[@LatticeTraining]

That's why we weighed ourselves and we were only 0.5kg different. So our absolute scores should very closely approximate our relative scores. We didn't feel to need to work out the % in this case.

[@SonnyKnutson]

@Lattice Training Make a chaoswall with pressure plates on all holds and small displays by each hold showing the % of total weight through it. Then you can record a video and in slowmotion see how the numbers change vs another climber doing moves and full climbs :P

[@LatticeTraining]

That's going to be one expensive wall!

[@digitalsmear]

@LatticeTraining - In response to your question at the end of the video about how this could be used; I think the main functional benefit you're going to get out of this is helping people with less kinesthetic awareness and more "thinking" type learning styles to become aware of what their feelings and sensations in movement are telling them. This is in contrast with people who are predominantly "doing" or "feeling" type learners who will find this information obvious. I suspect that the human body is too complex for the data to be entirely conclusive of anything, in and of itself. For example, Billy's hip flexibility probably has more to do with his readings in the last test than anything else. And the fact that Josh and Billy are same weight, similar height (and hopefully similar proportions?) but Billy is a stronger climber, means that there are important differences in the physiological makeup of your two bodies - and it could be nothing more complicated than bone density. So how do you interpret the raw data when recognizing that? But as a teaching tool, it could be mega.

[@LatticeTraining]

I like the idea of analysing learning. We could have a group use reflection and visualisation and another group distracted from this practice and see how efficiency develops over repeated attempts. Or we could look at different coaching practices such as instructional and indirect coaching.

[@mrwhale7875]

@TomOHalloranAus

[@sightreader2507]

Exactly! Remember him talking about this idea and how it could change climbing Still a really simplistic setup and analysis, but it's fun that someone's trying to tackle this challenge

[@thenayancat8802]

Man should hit them with a patent infringement :D

[@dietmarpuchberger9612]

Question about the first test at kzbin.info/www/bejne/i6evlpynd8-Mg5o Why is the foothold force down to zero before the top jug measures any? That doesn't seem to line up.

[@Absolute1342]

Tom O'Halloran called it

[@dorveille1]

To my eye, the most obvious difference between Josh and Billy, revealed both in watching them and in the force data, is that (ignoring the drop knee), Josh's movement is jerkier. So I think you'd get much more revealing data if you looked at the jerk, i.e. the rate of change of force.

[@4D_SpaceTime]

im disapointed the thumbnail shot wasnt used as a thumbnail.

[@jcdyer3]

Did y'all mean Bad Technique Revealed?

[@LatticeTraining]

Yes we did 🙈

[@TestUserPlsIgnore]

This whole video basically yells "perfect your foot work!"

[@ericlin158]

Did you say Magnus Mitface?

[@sambeard4428]

Please be consistent in the choice of colour representations in your graphs

[@lukasschallibaum5046]

Now train an a.i. to simulate millions of different ways and find at the most efficient way ;)

[@mycose.toujours]

This just shows that focusing on metrics and data isn't the solution to determinate wether you should climb or not a said grade

[@LatticeTraining]

Yep, data is just a tool. Climbing performance is very complex.

[@cecilia_mackie]

Unbelievable! Let’s connect?

[@nurhasanah77]

Gimana cara mendapatkan alat itu

[@itsorbs2137]

dyno v static climbing!

[@LatticeTraining]

Yeah I think an all-points off dyno would be cool!

[@robertcreer8826]

For the algorithm

[@tristanmayfield4851]

Combine this with replica training for the ultimate hack!

[@KittenIgnition]

Was this Ollie's idea? I messaged him about this a long time ago! He stole my idea! 😠😠

[@LatticeTraining]

Not Ollie's idea, sorry. I think we've all had this idea at some point. It was actually in conversation under a Kilterboard in Sheffield, with a climber I'd never met before, that made me follow through on this idea.

[@muscularibuprofen69]

Seems like an extremely niche tool for very limited types of moves. Novel idea, but seems way to convoluted for any real world use for 99.99% of athletes you train.

[@sarahpemberton4089]

Super interesting! I've love to see how gender influences this e.g. due to women's lower centre of gravity.

[@narcsurvivors2445]

Rise above your peers and don't pronounce 'th' as 'f' , like saying strenf instead of strength , you will undoubtedly be a better every'th'ing.