Running Footwear and the 2 Hour Marathon | Dr Wouter Hoogkamer

Hello and welcome to Lecture 5 of the Sports Biomechanics Lecture Series, supported by the International Society of Biomechanics in Sports and sponsored by Vicon. I'm Stuart McCurlane-Naylor from the University of Suffolk and today I'm joined by a very special guest. I have with me Wouter Hoogkammer from the University of Massachusetts Amherst and Wouter is going to be delivering a talk that I know a lot of people have been looking forward to since this series was first announced. Wouter will be talking on the influence of running footwear on marathon performance with a specific hint towards the sub-two-hour marathon. Wouter has a really interesting research portfolio on this topic covering areas such as running shoes, drafting and even course design, so I think he's going to touch on a few of those topics today but like me I'm sure you're looking forward to that. I'll hand over to Wouter and just one last thing, sorry, if anybody has any questions, if you can either use the comment or chat section on YouTube or use the hashtag at the bottom of the screen now on Twitter, I'll keep an eye on both of those and then once the talk is finished I'll direct your questions about it. So over to you, thank you. All right, thanks Stuart for the kind introduction. So yeah, my, there we go, yep, there we go. Yeah, so I'm assistant professor at the University of Massachusetts, sorry about that. So I'm on Twitter and my lab has a website that you can find there and I will be talking about running footwear and the two-hour marathon. So this all kind of started for me about five years ago when I was a postdoctoral researcher at the locomotion lab run by Dr. Crum at the University of Colorado and Nike reached out to us with a very specific question about whether we can use lab measured improvements in running economy to predict improvements in distance running performance and that I was going to do that research and do similar research over the next five years. If you would have told me that six years earlier I would have laughed at you because six years before I was back in the Netherlands studying human movement sciences, spending a lot of my time on the track running, dreaming about studying the biomechanics of running footwear but not necessarily taking any steps that would get me there. I was happily in my comfort zone and I was lucky enough that at some point this opportunity arose where I could spend some time abroad for my master's thesis research and I had to be pushed a lot by people in my clothes around me that said like this is great for you and I was like no I just want to run on the track and stay where I am but people luckily kept pushing and I made the decision to go to the U.S. to do my master's thesis and that sort of started everything for me and it was probably the best decision I've ever made. So eventually that got me here and that also got me in Colorado in 2015 specifically trying to answer this question. So there we go we got our first take home question for today or the stay home message. Dream big but be ready to leave your comfort zone when an opportunity arises otherwise you don't get anywhere. So going back to that question I don't have the time to go into the details of the research we did. I can just tell you the conclusion that yeah at that time we learned that if we do a lab test where we see that people use less energy to run we can use that to predict how much faster they will run and at that time we even thought that was a one-to-one relationship. And sort of that finding allowed us then to go back to the literature and look into all the different studies done in the lab where people try to improve running economy by either different scenarios and then we could see if we could apply those to breaking the two-hour marathon all with our goal within the rules for record eligibility. So we looked at aerodynamics like drafting and tailwind. We looked at course design and also about shoe technology obviously. But like I said important for us for this study was that we kind of wanted to acknowledge the rules and try to do this within the existing rules and basically our conclusion was that it was possible. So one of the the big conclusions we had from this paper was that was possible and one of the scenarios you see on the screen right now is we have for example on the top corner we talk about forerunner drafting which is a cooperative drafting strategy. So different from what happened at Monza in Vienna where Kipchoge was running behind a fresh team of pacers every lap we sort of suggested that if we bring together the best runners in the world and they sort of work together as a team and all of them are going the full distance but they could still alternate who's leading and who's drafting they could still gain a lot of time. And after this review we actually followed up with more detailed simulations of these scenarios and two papers on that are actually below in the link but I won't have time to go into details about them right now. Another thing is was specifically to those rules so the IAAF rules at the time said to the maximum allowable downhill you can have during a marathon should be about 0.1 percent of the overall course distance. So in this case we're talking about a marathon of 42 kilometers so you can drop about 42 meters. So we then said okay let's try to find a course or design a course that exactly drops 42 meters and then we quantified how much time we could gain with that which was about 30 seconds. And then that later became relevant again when the INEOS team actually reached out to us to get some advice about the course design. So we then ended up writing up our analysis of the Vienna course including the uphills and the downhills and the turns and that's also a pre-print that is freely available in the link downstairs. And then finally we suggested that other than Dennis Kim Meadows adidas shoes which weigh about 250 grams it should probably be possible to make a shoe that is as good but only weighs 150 grams and that would save him about 30 seconds during the first and 30 seconds during the second half of marathon. And even though in the lab we were already testing this new Nike prototype shoe that eventually became the 4% shoe we couldn't really talk about it in the review because we hadn't published this data yet. So it's sort of like two stories that we had to tell. The one that we could tell and the one that we already know was going to be probably way more important than all of the other ones combined. But that's just how the timeline was at that time. So I'm going to be talking about running footwear in a two-hour marathon. So there might be a lot of you that hope I will be talking a lot about the Nike AlphaFly next percent shoes. Well there's not a lot of scientific data out there that is freely available so I can't really talk about that. Then I can't also not talk about the specific vaporfly elites that were made for the Monza breaking to attempt again because there's no data. I also won't necessarily be talking about the next percent shoe that Kipchoge was wearing when he set the official world record in a marathon in Berlin. But we're going to go all the way back to 2016 to when he won the Rio Olympic marathon. And at that time people didn't really know but he was actually wearing one of the first iterations of what became the vaporfly 4% shoe. So that's actually also the shoe that we were testing in the lab. So they look very similar and this eventually became sort of the vaporfly line starting with the 4% elites the next to the AlphaFly. So we in Colorado studied these shoes like I said and we set out to compare the Nike prototype shoe against sort of like the state-of-the-art top-of-the-line running shoes at that time. So we went with the Nike Streak shoe and the Adidas Boost Adios Boost 2 shoe that basically if you looked at the world all-time list in marathon times at the time we started the study the top 10 of times set were all running either the Nike shoe or the Adidas shoe. So we were really starting at top of the line and use that as a baseline to compare the prototype shoe against. And obviously since we're in a sports biomechanics lecture series we want to talk about the biomechanics. Before I can start talking about the biomechanics of the Nike shoe I want to talk a little bit more just about the mechanics of the shoe. So we were able to determine the mechanical properties of the footwear. So the way we did that is in two ways. One of them was just looking into sort of like the foam cushioning and energy return properties. And we do that by vertically loading the shoe in compression the midsole using this instrum machine. And we in this case made sure that we not just loaded it we also made sure that we applied about a similar peak load as what a runner would experience during mid stance while running at a very fast marathon pace. We also made sure that the loading cycle was short kind of also similar to simulate the time that the shoe typically spends on the ground during a step. And when we loaded the shoe with that loading rate that you see in the top graph we can then measure the deformation. So we get this force deformation curve which shows how much a shoe deforms when loaded under specific form. And here you see what would happen if you would load the shoe in this case. And it's a foam so it's not necessarily a linear coil spring. So you can see it's not a perfectly straight line but for this foam it's pretty close. And then what we can do from that the slope of this line is actually what we call the stiffness which is force over the displacement. And we can rewrite that to show that the displacement is equal to force over the stiffness. Now the interesting part about this force deformation curve is that the area under this curve that is actually the mechanical energy that is stored in the shoe when you load it. And you can see that because we know that if this would be a perfectly straight line this would be a triangle and the area of this triangle would be equal to half times force times displacement. And you can rewrite that in different ways. One of the ways would be that the energy stored is equal to half times the spring stiffness times the displacement squared something that you all know of. Another way to rewrite it which I personally prefer is sort of substituting the displacement by the force over the stiffness because what we see there is actually that energy storage goes up with the inverse of stiffness. So if you want to store more energy in a midsole which you know the force of you want to take a less stiff midsole not a stiffer. So same for a spring if you want to store more energy you want to have a more compliant spring because stiffness is actually in the denominator of the equation. Now this is when we load the shoe. If you then like you saw in a diagram we first increase the force then we decrease the force then we can also measure how quickly the shoe sort of bounces back. So here you can see that when you unload it it's not following the same line it's actually coming down slightly lower which means that this is what we call hysteresis and it's the energy the area between those lines is the energy that is lost during the loading cycle. This also means that the area under the bottom line is actually the energy that is returned every loading cycle by the shoe. So we quantified all these properties so both the stiffness or the total amount of deformation and the relative amount of energy return. So looking at the three shoes we tested here we have this graph which I basically just showed you for the streak shoe which deforms about six millimeters and returns about 65 percent of the energy during the loading cycle and then we can quantify that in joules it comes out to about 3.3 joules. Then we did the same thing for the adidas shoe it's about the same compliance or stiffness it deforms about six millimeters when we load it with these two kilonewtons and but what you can see in this graph clearly is that the area between the lines is smaller which means that there is less energy dissipated in the adidas shoe than in the streak shoe. If we quantify that less energy dissipated more returned in this case it's almost 76 percent that is returned. Again we can calculate that in joules and we get about 3.6 joule energy return. Now what about the prototype? The prototype sort of used that concept that I just showed you if you want to store more energy under the same force you want to make less stiff or more compliance midsole and when loaded with two kilonewtons this shoe deforms about 12 millimeters so it's less stiff. It means that the area under the top curve is a lot bigger it means a lot more energy gets stored. At the same time this new midsole material is also more resilient so it's it's wasting less energy or losing dissipating less energy and there's less hysteresis so the area between the lines is also relatively smaller. In this case this more resilient bouncier midsole foam returns about 87 percent of the energy that is stored. So combining those numbers we'll see that the total energy return every step is about twice as high in the prototype shoe than in the other two shoes. Obviously that's not the only feature about the shoe. We also looked into sort of the bending stiffness of the shoe because obviously we all know that it's enforced by a carbon fiber plate which has a purpose to provide stiffness in bending. So to determine that you can generally use a similar setup where rather than compressing the full shoe now you put the shoe on two supports and then compress it right in between those supports and see how much it bends. From that test we can calculate how much torque we apply and how much it's resisting flexion and that gives us the bending stiffness typically expressed in either newton mediums per radian or sometimes based on this three-point bending stiffness test in the amount of displacement in millimeters divided by the force applied. So we applied a similar test to these shoes and found that adding the carbon fiber plate to a shoe makes it about twice, based on the properties of this specific plate, makes it about twice as stiff as the other two shoes. So we got a shoe that when loaded vertically returns about twice as much energy and when flexed is about twice as stiff. So how does that affect running performance or running economy? So before we go into detail about the biomechanics based on the mechanics I just explained to you, we're first going to look into what is the overall response of the human runner with the shoe. So we measure that by looking at their metabolic rate or we call it running economy. So running economy is defined as a specific metabolic rate that we measure at a specified sub-maximal running velocity. We do that in the lab, we measure oxygen uptake and we make sure that the runners are sub-maximally because we want to, we're only measuring aerobic contribution to their metabolism, so we want to make sure that when they're sub-maximal they're not relying on any anaerobic sources. And that also helps us to test multiple shoes in a series without having the runners building up fatigue from the start to the end of the protocol. And we measure that for the three different shoes and this would be the place where I would go into a lot of details about how much care we took to all of the details of this study, but I don't really have a lot of time to do that here today. I just want to give you that take-home that in sports science you need to be paying attention to the details of everything you measure because that's how you get the results and that's how you get reliable results. So another take-home, if you want to build a career in this field, pay attention to the details and be careful with all your measurements and be careful with all your analyses. So what did we find after what I think was a very careful study? We saw that when we looked at energy cost or metabolic rate or running economy that between the Adidas shoe and the Nike streak shoe, the energy consumption was basically the same. Metabolic rate was on average the same. You can look at the gray lines. Each gray line represents one individual. Some people end up, some people went down, but then when we started adding in the third number for the prototype shoe, you can see that all of the gray lines are going down, which means that each of our 18 individuals in the study is using less metabolic energy while running in a prototype shoe. On average, the group average was four percent less and then we didn't just test it at this velocity of 14 kilometers per hour, we also tested it at 16 and 18 kilometers per hour and we saw that those savings were consistent. At all speeds, all runners use less energy in the prototype shoe and at all the speeds, the difference for the group was about four percent. That is then what eventually inspired Nike to name the shoe that they, the next version of the shoe, the Nike Vaporfly, four percent. Well, that's all great, but we are biomechanists, so we are mainly interested in how does that work. We know the mechanical properties, we know that it works, people use four percent less energy, but how does that come about? So that's why we need biomechanics. Here you see a picture of the lab setup when we did the second follow-up study where we specifically focused on the running biomechanics. So our subject is running on a force measuring treadmill which has allowed us to measure the forces, ground reaction forces, the forces exerted by the runner on the ground or the treadmill. And also in the corner of the pictures, hopefully not hidden behind my face, is a FICON camera. So we had markers on the different joints of the legs and of the segments of the leg and we were able to measure kinematics. So we have ground reaction forces from the treadmill and kinematics from the cameras and that allows us to do a lot of joint mechanics. And one of the joints that was of a lot of interest to us was the metatarsophalangeal joint. So this is where I'm actually quickly gonna try to step out of my shared screen and go live where I'm gonna show you the metatarsophalangeal joint. So this is a foot model and these bones here are the metatarsal. So this is the midfoot. And then here there's a series of joints with the toes. The toes are the phalangeal bones. This is the metatarsophalangeal joint or in short the MTP joint. And it's important in walking and running because it allows us to do a nice roll-off with the foot. And since this model is actually pretty stiff I'm going to just show it with a shoe to make my point. So when we have a runner coming in landing on the ground during the roll-off there's a specific point in time where the midfoot is moving against the toe segment and it's applying a force in the other direction. And this part is actually where at this joint negative work is performed or energy is dissipated. Dissipated or stored? We actually don't know that. It could be dissipated, it could be stored in the elastic structures of the foot like the plantar fascia, things like that. Either way you can imagine that that is negative work and it might be dissipated and there might be active muscle contractions going on to restrict that movement. So what if we add a carbon fiber plate to a shoe to prevent that? So this is my analog for a carbon fiber plate and we put it under the foot. You can see that in this case there's no way that the toes aren't going to flex because they're supported by this very stiff plate. And so this is kind of what happens when you have a shoe like this. Like the normal shoe you have the flex, the stiff shoe you don't have the flex or you have a lot of less flex. So that is the theory for the specific MTP joint. So let's see if that actually holds up in the data. All right and back on the data screen. So what did we see when we measured at the MTP joint? Well first we measured the joint angle and you can see here in green we got in this case the Nike Streak which flexes a lot up to 30 degrees near the end face of the stance face just before takeoff. And then you see that the prototype shoe in orange is flexing almost half as much so a lot less. So yes the theory works if you apply a stiff structure to the foot you get a less dorsiflexion at the MTP joint. Now we can also, because we measure the ground reaction forces and know where the MTP joint is relative to the ground reaction force factor, calculate the moment. And we see here that the difference in moments is small. Well the next step what we can do is then use the joint angular velocity, the derivative of joint angle, multiply that by the during every instant during the stance phase and we can get a measure of joint power. And what is interesting about joint power is that the area under the joint power curve is actually energy that is either dissipated or stored. And when it's negative or when it's the area is positive that is energy that is either returned or generated. So in this case you can see that there's less energy dissipation at the MTP joint because the area the negative area under the orange curve is a lot smaller than the negative area under the green curve. So yes it works. But since we knew the actual mechanical properties of the shoes we can also calculate how much the shoe by itself is contributing to this. So this is measured at the MTP joint from a runner with his foot in a shoe. This is the overall combination of the foot and the shoe but we know how much the shoe is resisting flexion and we know how much it is flexion so we can measure that. So using the angular changes at the MTP joint combining that with our stiffness measures we can calculate how much the shoe is resisting flex and generating a moment. When we do that you get these curves where you can see that it's fairly similar which is interesting. It's sort of the result that we know that the orange prototype shoe is twice as stiff but is only flexing half as much. So it's about the same when you talk about the resistance as the other shoe. Then obviously the next step is we can then again combine the angular velocity with the moment to get the power that is specifically happening at the midsole of the shoe. So we can quantify the energy that is stored and returned. So again the area between zero and the line the negative area is the energy that is stored in bending and then the positive area is the energy that's returned from bending and in this case there's no muscles in the shoe so this is all passive elastic storage and return. So looking at this diagram we can confirm the carbon fiber plate does function as a spring. It stores energy and it returns energy. We also know that the other shoes seem to do a lot of it a little bit of it as well and you can see that too when I flex it. So the question then is, is this substantial? Is this whole four percent of metabolic energy savings that we see from running these shoes because there is some energy stored in flexion and returned by the carbon fiber plate? Well we can quantify it and we quantified it and it comes down to about 0.16 joules per step and then we still don't really know is that a lot but let me go back to what I showed you earlier when we had the numbers when we vertically loaded the shoes. From this test we saw that every loading cycle the shoe returns about 7.5 joules just from vertical compression basically from the midsole foam. Interesting was because we loaded the shoe in a way that represents the forces that it actually experiences when somebody is running every step we know that this 7.5 joules that is returned is returned every loading cycle or if you want every step. So here you can see that the carbon fiber plate does return storing return energy but it's almost about 50 times smaller than the total amount of energy that is returned in the shoe in vertical compression. So we know carbon fiber plate sort of is a good thing around the MTP joint but it's not a big contribution as compared to the vertical compression. But what about the ankle joint? So let me step out this again briefly to show you another concept. So I just showed you that we can stiffen up the MTP joint and so that there's less flexion there. I also showed you like we have an MTP joint and we can flex there and that's usually a good thing and during normal running that helps us a lot. So when we're running and we're flexing what happens we keep the moment arm between the ground reaction force and the Achilles tendon short. So let me show you that again. So in this case when we have our foot and we're flexing which means that we can apply the ground reaction force under the full toe segment. So the pivot point right now is kind of where the MTP joint is. You can imagine that Achilles tendons that have to lift the whole body over this pivot point have a lot easier job doing this than when they have to lift it all the way over this longer moment arm, right? And if you don't believe that what you can do at home is start stepping on a step, standing on a step, and when you're standing like this at the edge of the step it's probably really easy on your calves. But then you just go backwards and when you're standing like this you start feeling something in your calf muscles. You can imagine that if you're all the way forward your calf muscles are going to work a lot harder and even at the point of your toes that moment is even bigger and your calf muscles need to work even harder. So that is kind of what happens with the plate and the moment arms around the ankle. So let's see if that is something that held up during our test as well. There we go. So what did we see? Well, we saw small changes. We saw that the joint angle was slightly smaller, was a little bit less dorsiflexion at the ankle in the prototype shoes, but we saw that even though I just explained to you when this is stiffer we would expect a larger ankle moment, it wasn't really there. And that might be because what I just showed you was everything with a very straight plate and we had this pivot point under the ball of the foot. The way Nike went around that is by giving the shoe a quite substantial toe spring which is basically how much it comes up to the front. So you can imagine if this was a plate that was fully flat and we have to go around this pivot point that's a lot harder, which I just showed you, but because of this toe spring the roll-off doesn't get as much boost forward. Even though it's stiffer around the MTP joint it doesn't necessarily make it harder for the ankle. So that was a sort of a surprising finding, but we can explain it by the curvature of the shoe. We also sort of quantified how much negative work happens at the ankle joint. So this is energy that is either dissipated or stored and then the positive side is the energy that is returned or generated. And we can quantify that again. Remember for the plate we had about 0.2 joules. For the full shoe in compression we have seven and a half. If we just compare that to the full ankle push-up that's 55 joules. So that's a lot. So this spring function of the plate is so much smaller than what the ankle is doing. Well the thing is, let me go back to that one, we have a bigger push-up but is that necessarily or we have a smaller push-up actually in the MTP joint in the prototype shoe than in the other shoe. And is that necessarily a good thing or a bad thing? We don't necessarily know that because we do know at the ankle a lot of the energy fluctuations are not necessarily energy dissipations and generations that are actively generated by muscle forces, but a lot of that is energy that is stored in Achilles tendon and then later on returns. So there is this stretch and release of a tendon. So even though we can quantify these joint mechanics, it doesn't necessarily tell us everything about how much muscle fibers are active to sort of generate these forces because we don't quantify or able to quantify the storage and return of the elastic tissues. And we also are not capable to quantify or when you look at individual joints you don't take into account any energy that is negative that might be then transferred negative from the one joint to be positive at the other joint. So that's another thing to keep in mind when looking at joint mechanics. And there's a third phenomenon that I want to explain to you too about the plate which we didn't quantify but it's interesting because it might be an important contributor to the function of the shoe and that's related to the muscle shortening velocity. So I just said when we're looking at joint mechanics we don't know how much is happening at the level of the tendon and how much is happening at the level of the muscle fibers. So there's different approaches to that and first I want to show you the concept how we can look at shortening velocity. So here we have a foot and it's traveling at two hour marathon pace or 21 kilometers per hour and we can transpose the treadmill velocity into the ankle angular velocity by using the external moment arm. In this case between the ankle joint and perpendicular to the velocity it's moving. Then when we know the angular velocity at the ankle joint we know the internal moment arm between the ankle joint and Achilles tendon and we can calculate how much the total muscle tendon unit is shortening. And this was the scenario without a plate. If we add in a carbon fiber plate the concept still holds but things are slightly different. So if we can start comparing the two sides the treadmill velocities are the same so we're still running a two hour marathon pace but in this case with this plate people get higher up on their feet and the external moment arm is slightly higher which means that the ankle angular velocity is actually slightly slower. So for the same running velocity the plate sort of allows the ankle to move at a slower velocity which then because the internal moment arm is similar between the two sides results in a shorter and slower shortening velocity of the muscle tendon complex. So we talk about the calf muscles which involves Achilles tendon. Then that is at the level of the muscle tendon the next thing we can then look at is how much the actual muscle vesicles in the calf muscles are shortening. And people have done that not necessarily for the Nike shoe and not necessarily for running. So Kota Takahashi did a nice study in 2016 where he looked at this for walking and what we see on the horizontal axis is increases in bending stiffness of shoe conditions and you can see that at peak force the muscle vesicle shortening velocity actually decreases with increasing shoe bending stiffness and that obviously can be expected to reduce the energy that is associated with generating those forces. Then recently and currently not published but it's out there in a pre-print Owen Beck did a study using similar footwear where he actually measured shortening velocity of the muscle vesicles and by the way you do that by applying ultrasound probes on top of the muscles and looking at the actual vesicle shortening length from the ultrasound imaging. And he saw that overall maybe you could see that for the stiffest condition the shortening velocity was slightly slower but this wasn't significant and it's also important to note that the Vaporfly shoes are expected to sit somewhere around that 45 kilonewton per meter bending stiffness that they assess. So there it might even be higher all even though none of these things were actually significant. So what's going on there we really don't know yet and there's a lot of things happening and this is a study done for a study with shoes and flat carbon fiber insoles. We know that in this shoe there's a different geometry there's a curved plate all those things might have different effects on all of these things. So before I wrap up quickly I want to touch on what's next. So we just saw that things are complex and we can measure things but they don't necessarily explain us how they directly affect energy cost and now the next challenge is how to look into findings from the one study that used the one shoe versus the other study that used the other shoe with an insole or another way of stiffening it versus other companies putting out different shoes. In this case you see five shoes that are either just came out or are about to come out where different companies like New Balance and Saucony, Hoka and Asics, Brooks all have their own sort of next version of the vapor fly where they have a carbon fiber plate with some sort of geometry or curvature or not and specific location in the midsole combined with a high resilient foam but all of these have slightly different properties. So what we're going to see in the next few years is maybe more studies into this but it's always important to keep in mind that the slight differences in geometry might have different effects on the ankle and the NGP joint and it's hard to predict the overall outcomes. Also what we're going to see is a lot more marketing on the one side and a lot more fake news on the other side. So this is an example where right after Kipchoge ran his sub two hour marathon people were speculating about the shoes he was wearing. Somebody dug up a patent from Nike and pulled up one figure and then started claiming that Kipchoge was running in a shoe that had three carbon fiber plates and two layers of air pots. Then somebody went ahead and used photoshop to actually add those second layer of air pots which were not even in the picture of the shoes. And guess what? Eventually came out that there's actually about 60 different figures in the patent application and some of them have multiple layers and multiple plates. Some of them haven't got any plates and then when it was released we know that the AlphaFly right now you know it has only one plate and only one layer of air port. So again fake news, utter fake news, people speculating about the cost. They were going to be $400 when the AlphaFly shoes came out and then somebody saw that video or image and just made their own version and made them $450. So there's a lot of things going out there that you have to be critical and careful about. Other things is like this is going to be next. We already have a two sub two hour marathon now but if we allow carbon fiber springs next thing we'll have it is people go going around and running a one hour marathon. Well if you think about it this guy is not running a one hour marathon and if he could he probably would be doing it. These things are really heavy so they're not going to save him any energy but make it even harder. So just as another advice and a take-home message to think critically about what you see out there on the web and think if it makes sense and see if you can find the original sources. And while I'm at it I also give you another advice that think critically about your own opinions. It's always an opportunity when there's new data or new insights to update your own opinions about things. All right I think I'm going to wrap it up. I quickly want to acknowledge the people in the lab in Colorado we did the study with and the people at Nike you see here. I also want to acknowledge that being a white male gives me privileges which gave me a step up to get where I am now. And finally want to acknowledge my current lab at UMass people in the lab we have our website up there and our funding sources Puma, BF and Nike. Thank you. Okay brilliant thanks Wouter that was genuinely really really interesting. Even from before we went on air just chatting to Wouter I could talk for hours about there are so many kind of interrelated components and it's really difficult trying to tease them all apart. So just as kind of one maybe quick question I just wanted to throw in just one more component. How does foot strike pattern kind of interrelate with all of this? Yeah that's a great question and that might be something coming back to my last point too about to think critically and the fake news. So for some reason there's a lot of journalists that reach out to me talk about the shoe and they always bring up that the shoe is made for forefoot strikers and that they only work when you're a forefoot striker. And to be honest we don't really know that. We in our first study we looked at this we had these 18 runners and about half of them by random chance were heel strikers and the other half were mid-foot strikers. And when we looked at the data there was a small indication that it was actually the heel strikers that saved about four and a half percent and a forefoot striker mid-foot striker only three and a half percent. But that was just in our sample then other people have studied the same shoes and did similar analysis and they didn't really see that to be holding. So it seems that across the board independent of foot strike they seem to work. I mean I must acknowledge we tested very fit runners at very fast paces. We don't necessarily know that what happens when we're talking about a five-hour marathoner or four-hour marathoner. And then like I said in a biomechanical so all the results I showed you were from the biomechanical study where we tested only heel strikers because those are the most common type of runners. But like I said throughout all the different studies performed with these shoes there is not a clear trend that it works better for either forefoot strikers or rear foot strikers. And I think it definitely there's this feel that when you put the shoes on you're sort of pushed more forward and things like that. Again something to think about things critically take into account physics when people start making claims about shoes that are breaking less and propelling more because overall when you're running at a consistent velocity and there's hardly any air resistance you're going to be breaking as much as you're going to be pushing up because otherwise you just keep accelerating and that's not what you do at constant speed. So things like that be critical when people start making claims and think about it and sometimes you just need to acknowledge that we don't have the data to answer the question. Okay yeah brilliant thank you and yeah just one hopefully last question I think. If it holds across kind of different running speeds different foot strike patterns everyone's around about four percent saving in running economy why are we not seeing kind of a four percent improvement in elite marathon times? Yeah that's a great question and that is some of the articles that are also in the links below that we discussed that I didn't have really time to go into detail there today but to be true what they claim to be holding if we want to have a four percent saving translating to a four percent faster time that needs that can only happen when the relationship between energy cost and speed is perfectly linear and goes through the origin. It's a directly proportional relationship because we're talking about relative changes on the one side and they're only translating to relative changes of similar magnitude on the other side if we have a directly proportional relationship which we don't. For a long time we always thought that we can estimate the relationship between running speed and running economy to be linear but now we got more data and we look at this carefully we see that even on the treadmill this relationship is not linear we see that at higher speeds to go a slightly bit faster you need a lot more energy than you need at lower speeds and then that is on the treadmill when you then translate it to an overground racing situation we also have air resistance which generally is not as important in running but if we talk about Elliott Kipchoge to our marathon pace it becomes a thing. So the combined effect of those two mechanisms the one increasing air resistance which becomes more and more important at higher speeds and the fact that at higher speeds you need more energy to make a small improvement in speed results in the fact that improvement that people can get in times with a four percent savings if we would assume that it was there for everybody for every runner with any foot strike then still the people that run slower would see a bigger improvement in time than the people that run really fast. I think we we did the map in the paper that is first author is Shalaya Kip and there we we sort of show that for a two-hour marathon a four percent savings would only make you about two and a half percent faster if you're slower than a four-hour marathon that same four percent metabolic savings could even allow you to run more than four percent faster just because of that relationship being not directly proportional. Okay so while you're answering that we now have a couple more questions on YouTube so do you have time for a couple more? I have time for sure yeah. Okay brilliant so Mason Coffey has asked says you touched briefly on weight of the shoes in the first Nike prototype study did you normalize for weight? Yeah that's a great question and and and that is something that we did so we know that shoe mass is really important for metabolic rate so we know that every 100 grams saves you that you add to a shoe makes you about one percent slower or increases your metabolic rate by about one percent so making shoe 100 gram less heavy would improve running economy about one percent. Just when we started this study we weren't sure about the differences that we were going to see between the shoes we didn't want to find a one and a half percent and then have a mass difference between the shoes that would have explained half of that so to be sure we made sure that all the shoes weighed about 250 grams that these shoes are so we added about 50 grams to or 60 grams to the street shoes and about 50 grams I think to the Nike prototypes so in that case the comparisons the four percent was purely based on the geometry of the midsole the foam and the plate and the mass was out of the equation in hindsight we could have also not normalized the mass because the difference was so big that people wouldn't claim that was just because of mass and that would be a more fair comparison or ecological where you can actually compare the two shoes as they are out there in the field so that's sort of these decisions that you can sort of weigh on which one do you take do you really have an applied question do you want to go with the most ecological validity and take shoes that are out there as is or do you also have an interest in sort of like the fundamental underlying principles and then do you need to start controlling some of these parameters as much as you can and obviously everything I showed you about the carbon fiber plate and comparing it to those shoes ideally what we would have done is have this exact same shoe with this exact same geometry and have a shoe that is the same foam but doesn't have the plate and have another version of the shoe that is exactly the same and does have the plate but doesn't have the fancy foam and that would give us more detailed fundamental insights but it also is less of an applied because in reality we don't choose don't exist so yeah it's always the trade-off between those kind of decisions do we go really applied or do we go fundamental. Brilliant and I think that ties in with your second stay-home message as well of meticulousness and tick all the boxes there. Last question at the moment at least is from Fabio Lanfardini who says thanks Wouter can you comment a little about the effects of drafting for the two-hour break in America? Yes so like I said in this talk we explored this concept where we would have a team of runners drafting together so what happened during Monza and Vienna was Nike took a different approach which obviously didn't meet the regulations for record eligibility but they made it a separate event and then both events actually had a different approach where we saw in Monza this sort of aero formation where there was the paces were in aero formation and Elliot Kipchoge was behind them and then in Monza it was sort of this inverse aero and all of those sort of scenarios were done by guys who know a lot more about computational fluid dynamics so in our simulations or in calculations we typically need to go back to some data by Griffith Pugh that was he collected in the late 60s published in the 70s which is an exercise physiologist from the UK and he had a wind tunnel and he had a runner in a wind tunnel and actually measured running economy of the runner running at a set speed in the wind tunnel on the treadmill and then had the same runner run in a wind tunnel behind another runner and in that study they saw that running economy was about six percent better when drafting behind somebody else so that is kind of the number that we generally use to simulate the effect because right now we have the technology to use a lot of computational fluid dynamic simulations and we can theoretically simulate any scenario of air resistance but the outcome of those simulations is always a drag force so if we have a specific configuration of runners and drafters and pacers if we simulate that the outcome is the drag force experienced by the runner then we can see changes in drag force but how does that translate to differences in running economy and how does that translate to difference in running performance those are two additional steps that we always need to make so we typically shortcut that and just go with drag force towards metabolic rate but actually in Colorado we have been doing a study that we're currently writing up where we did measure the relationship between drag force and running economy we saw that it varies a lot interestingly enough the one guy that was tested back in the 70s that was a study on one runner back then and that guy is right in the middle of the 12 people that we tested in Colorado so that sort of works out so this five to six percent savings is still what we can expect in Colorado we didn't have a wind tunnel we were using elastic bands and hanging weights over pulleys to pull people backwards with very tiny forces great study to do but we hope to follow up on that and try to recreate that scenario in a wind tunnel and then validate our findings and validate our simulations okay brilliant i think well i could keep talking for hours but i think we'll leave it there for now and then if anyone does think of any other questions we'll kind of they can take to twitter or kind of ask and we'll see if we can try and get a response sounds good okay thank you and yes just hopefully everyone enjoyed that as much as i did and got as much out of that and kind of if you did then please share with friends students colleagues anyone that you think will find that talk useful and don't forget that's now five so we're halfway through the 10 talks that we initially announced we've since got more lined up that we'll hopefully be announcing over the coming weeks but if you haven't already go back and look at those first four and keep an eye out for the five coming up which are all scheduled on youtube as well i think last thing for me is just a huge thank you to voucher which i know everyone else on youtube and twitter is agreeing with me that was excellent and yeah hopefully i'll see you all again soon thanks right thank you