Why we land in front of our center of gravity

Pete Larson, once again, had an excellent blog post on foot strike and center of gravity (COG) (click here to read it). The notion that all runners should land underneath their COG (or hips, they’re often used interchangeably) has been repeated so many times by so many experts that it’s essentially become dogma. However, as Pete so aptly demonstrated, it doesn’t actually occur. None of the elites Pete had video of landed underneath their COG, and when I went and looked at high speed video of myself and several other good runners, none of them did it either.

The question becomes, why do we have to land in front of the COG and not directly underneath it like Pose or Chi running advice?

Getting Quick

Many well intentioned ideas or concepts come about because we have no patience. In regards to running form, that means minimizing the braking forces on the ground. In theory this is a great idea, minimize braking so we can spent more time in the propulsion phase. However, in practice it doesn’t turn out so well.

The idea of the pawback developed based on this theory. Coaches thought that if we could start the backwards movement of the leg before we even hit the ground then braking forces would be minimized as the foot would already be moving backwards. Similarly, the land right underneath your COG idea probably developed because people noted that you couldn’t have propulsion until the foot was past the COG and on its way backwards. They reasoned that if we eliminated the time between when the foot touched the ground and when it was under the COG, we’d minimize braking and useless time spent on the ground waiting to be able to push off. In essence, both ideas came about because of wanting to get quick.

Theory and practice don’t always work out so nicely unfortunately.

Loading up-Why patience is a virtue
Many of the proponents of the landing directly underneath your COG point to the fact that propulsion can’t occur until the body is over the COG. Only then can the body begin to push off and propel it self forward. They claim that time spent from landing in front of the COG to when this propulsion can actually occur is just time spent waiting. Their principle idea is to eliminate this needless waiting and get on with the propulsion phase, or in the case of POSE, the pull/fall phase.

The problem with this line of thinking is that it ignores the impact forces and elastic energy usage. If you recall that the bodies tendons, ligaments, and muscles act in a similar way to a spring, it only makes sense that there has to be a compression in that spring system before it can rebound and release energy. Put another way, there has to be energy storage before there can be energy release. Along these same lines, the initial impact of the collision of the foot to the ground has to be absorbed.

When you land slightly in front of the COG it allows for these process to take place. The body has time to absorb the impacts and move into a propulsion phase. If somehow you were able to land directly under your COG, you’d waist part of the time in which you could be applying propulsive forces to the ground because part of that time would be spent absorbing the passive forces.

To help clear this up, let’s look at some data.


(source: http://www.latrobe.edu.au/podiatry/documents/podbiopdfs/BioofRunning.pdf)

The above chart shows a typical ground reaction force graph. The top graph is vertical GRF, while the bottom one is anterior-posterior (forward-backward) GRF. I like this illustration as it shows exactly where passive and active forces take place and is easy to understand.

The anterior-posterior (bottom) graph shows us the crossover point where we go from deceleration to propulsion. The first half of the graph there is a frictional force backwards against our foot, preventing it from continuing to slide forward as we land. On the latter half of the graph, the frictional force is forward to prevent us from slipping backwards and we push behind us. We can use this transition point as a guideline to show what’s going on in terms of propulsion.

If we combine this graph with the vertical GRF, things get very interesting. Note that the transition point corresponds with peak GRF, and as pointing out in the picture, this is due to active muscle forces and the rebound of that stored elastic energy.

Finally, take a look at this screenshot from the Lieberman study video on youtube.


While it’s not exact since it’s an incomplete picture, note that the peak vertical GRF corresponds nicely for when the center of pressure on the foot is underneath the approximate CoG. The red line in the picture marks the runner’s approximate CoG. By approximation the Center of pressure in the foot is on the ball of the foot at this point, so it falls pretty close to directly underneath the CoG.  While it’s hard to get an exact mark, it’s obvious that the foot is under the CoG sometime around or just before peak vertical GRF.

This fits in nicely with the data from the anterior/posterior transition point and shows that the stride can likely be seperated into an absorption and a propulsion phase.  Although hard to tell in the graphs, the absorption phase takes less time than the propulsion phase, as the foot extends backwards from the CoG more than it extends in front of the CoG upon landing.  This can be seen in the Lieberman videos.

For some rough info, I ran some data on myself.

As you can see the time and distance from initial contact to when the approximate center of pressure (approx average of where on the foot the load is) is under the CoG is less than the time and distance from when the CoG is crossed to when toe off occurs.  Initial foot strike occurs ~27cm in front of the CoG and it takes .086sec for the foot to get under the CoG.  On the other hand it takes .133sec to go from the CoG to toe off.  Toe off occurs ~50cm behind the CoG.

What about footstrike?

If you are a keen observer, you’d notice that the first graph is of a heel striker. Let’s take a look at a midfoot striker and see if the premise holds true.

The above data is from a separate study by Peter Cavanagh (1995) (link here). The left column is heel striking and the right column is midfoot striking. The top row is mediolateral GRF (side to side), the middle row is anterior-posterior GRF (front to back) and the bottom is vertical GRF.

As you should be able to see, even in the midfoot strike, the anterior/posterior transition point corresponds to the peak vertical GRF.

So what’s the takeaway message?

Have patience. We need to land just in front of the COG to allow for absorption of the impact. This doesn’t mean reach out with your leg and land way out in front of your COG. If that was done then the delay between when impact occurred and when propulsion occurred would be longer, thus more energy dissipation and less use of elastic energy.

In an earlier post I mentioned using cue’s to improve form. Landing under your CoG still may be a reliable cue at first as you try and overcorrect the problem of landing way out in front of you. The problem arises when we mix cues and what’s actually going on.

So be aware of what’s a cue and what actually happens.

Get My New Guide on: The Science of Creating Workouts


    1. Anders on August 23, 2010 at 4:14 pm

      About pawback, I still think it exists, but perhaps it should not be actively thought about or trained. The knee does drop before foot-strike, and when the knee drops the foot starts moving backwards.

      If you look at high speed video on a foot strike of a good midfoot/forefoot runner, you will see that the foot stands almost still relative to the ground at impact. If it would still be traveling forward you would slide forward in the shoe (which happen if you combine forefoot strike with overstriding).

      I have this video of a decent runner 😉

      Look at the forefoot touchdowns in vibram fivefingers around 50 seconds. You see a slight "pawback" and a nice "stand-still" footplant with no friction braking, right? It would be interesting to measure, but I think with such a foot strike that there is very little force forward in the foot, instead the force forward (coming from landing slightly in front of the CoG) is absorbed by bending at the knee (thus loading it with elastic energy). To make sure the force forward is used to bend the knee and not sliding the foot, one must land with the foot under the knee, roughly perpendicular lower leg.

      Landing too close the CoG (for example by actively bending at the knee to get the foot under the body, very hard to do but probably not impossible) would lead to poor use of the stretch-shortening cycle and thus poor pushoff.

    2. Anders on August 23, 2010 at 8:00 pm

      Additional comment about pawback. The first time I came in contact with the term was via Dr Yessis texts about sprinting (in the book Explosive Running), I thought he invented term? Interestingly enough, he talk a lot about pawback but nothing about landing under the CoG. For long distance running he says that foot contact is made "slightly in front of the body".

      While pawback phase looks a bit tame in long distance running (=relaxed as of passive mechanics), it indeed looks powerful in sprinting. I'm no sprinter so I don't know the relevance of thinking about it actively though.

      Could it be the case that all this pawback talk is the good old problem with sprinting cues being transferred to long distance running without taking into account that there actually are some differences?

    3. Anonymous on August 24, 2010 at 6:33 am

      Herb Elliott seems to paw back a little:


    4. Anonymous on August 24, 2010 at 8:30 am

      Why do you use treadmills to examine running technique?

    5. Anders on August 24, 2010 at 12:29 pm

      On an example of landing (almost) under the CoG, I found this video on the tube:

      If you ask me, the overall style is better than most recreational runners (who generally overstride badly), but I see signs of taking Pose cues too literally (landing under CoG, lifting foot actively) causing a shortened stride. For recreational running up to fairly high levels I don't think it matters too much though, you can trade some between stride length and stride rate without significant changes in efficiency. But if you really make the effort to study technique it is nice if it is sane stuff you learn…

    6. stevemagness on August 24, 2010 at 12:47 pm

      Anders- First off, great post on runblogger, very informative and thought provoking.

      On the pawback- I think your point on it not being active is the best idea. What I'm trying to get across is that it should be very minimal at best. As the lower leg unfolds, the antoginist muscle works to slow down the unfolding. Are we going to get it to stop perfectly on a dime right underneath the knee? No, they'll be a slight overshoot due to momentum. The key is that it's not something you should try and do and it shouldn't be a let your leg swing straight out then pull it back rapidly like Yessis proclaims.

      In the study above by Cavanagh look at the anterior=posterior graph (2nd down in the 2nd column) that shows the frictional forward backward movement of a midfoot strike that occurs on the outside edge of the foot.

      Anonymous- A lot of the pawback phenomenon is occuring because of a simple lowering of the foot and the fact that were traveling in the same direction. In the Elliot video, the foot drops down as the thigh begins to move backwards, not because of a rapid pulling back of the lower leg in my opinion.

      Anonymous-I don't like using treadmills, but in this particular case I wanted to see changes in the velocity of the limbs during the recovery phase. Because I'm not moving, you can do these calculations easier.

      Anders-Interesting video. I'd agree with your analysis. First thing I notices was the shortened stride. You make a good point as for a lot of people POSE might work fine because they aren't really trying to maximize performance or optimize stride length/frequency.

    7. Pete Larson on August 24, 2010 at 2:44 pm

      Great post Steve – amazing how well the kinetic changes match up with the position of the foot relative to COG. Would be interesting to see a similar analysis on someone running pose. It almost looks like in the video that Anders linked that the foot also leaves the ground earlier after passing by the COG. Hard to say for sure though.


    8. Anders on August 24, 2010 at 4:48 pm

      Oh I did not realize the importance of the frictional GRF graphs at first, sorry. Now when I look closer I am a bit surprised that they are so large. I am also surprised that it looks so messy in the midfoot strike compared to heelstrike.

      I started out as an overstriding heelstriker and I have run quite much on slippery ground (winter climate). When I changed to midfoot without overstride, I feel much safer, less risk to slip with the foot at touchdown. Looking at the graph, it seems like midfoot would be the more slippery style, but I guess one needs to combine all three dimensions to understand something about grip. I guess low vertical GRF combined with high horizontal GRF and fast rise times leads to large risk to slip.

    9. stevemagness on August 24, 2010 at 5:01 pm

      Anders- Check out the center of pressure graph from this study (page 5)

      It should make the frictional forces make more sense if you look at the way the center of pressure moves and the anterior-posterior forces.

      It might show why midfoot is less slippery. Probably because you have two changes in direction instead of a constant frictional force all going one way at the beginning.

    10. RICK'S RUNNING on August 24, 2010 at 7:38 pm

      Great articles by both you and Pete.
      I think good running form starts with good Posture.
      So many runners spend so much time sitting at a desk-working on the computer or driving or even watching TV that their bodies are so out of balance and suffer with Lordosis!
      The effects of this are, well arthur lydiard called this 'Sitting in a bucket' butt sticking out the back and leaning at the waist a position which not only makes fast running impossible but will cause all sorts of injury problems.
      Lung stretches and butt strenghtening exercises will hell correct the posture as well as Arthurs advice to 'run tall' think of yourself running two inch taller than normal!
      You will now be able to glide over the ground instead of collapsing into it and waisting energy, also you will be less likely to heel strike!
      Read article on Lordosis

    11. Mill Runner on August 25, 2010 at 4:04 am

      First off, great pose!

      I must say that there is a vast discrepancy between how one perceives and how thing really is. As much as I would like to believe that I am a mid-foot lander( clearly, I feel the mid-foot touches the ground first instead of the heel, I wonder the build up of the shoe in the heel has anything to do with it), high speed video clearly shows that I touch down with the heel first, also with shin fully swinging forward, then bringing it down and back(pawback) before the touch down. I have tried some cues that you suggested, like bring down the foot right behind the knee, or not swinging the shin all the way out. Problem is, I do "feel" that the stride is shorter (if I consciously limit the shin forward swing, not sure whether the stride really is shorter as I didn't measure it) and I am really afraid of stumbling especially running on the treadmill at high speed( trying to bring the foot underneath the knee).

      One thing that I observe in the video you posted previously is that the runner (Cody) does not unfold the shin totally before the touch down. Think a little about this, I wonder whether how much one can unfold the shin and still landing "decently" just a little in front of the CoG depends on how much propulsion force we can apply and projected our body to flight. I mean if the body can "catch up" with the shin during the flight time, we would still landing close to the CoG, and this may suggest that the degree of the shin unfolding is a function of the speed of running. Could you comment a little on this?

      About the stumbling feeling in touching down too close to CoG, is there a guideline on how close it should be. I remembered you (and Tom Sellez in his video ) have mentioned that one should land with a shin roughly perpendicular to the ground as a guideline, but I wonder whether you can elaborate a bit further on how much the knee flexion or hip flexion should be at touch down. Clearly, there should be difference between a touch down of a knee flexion of say 150° and 100°, both with the shin perpendicular to the ground. (I know, I am a bit nerd about this, but as the feeling – the cue- doesn't really tell a precise story, I have to rely on something more measurable).

      Thank you in advance.

    12. Anders on August 25, 2010 at 1:27 pm

      Mill Runner, I have the same question as you have.

      The foot is swinging forward and then stops and swings backward a little before touchdown (this is synchronized with the knee swinging forward stopping and changing direction), this happens for all runners. This backswing is called pawback by those that want it to be active.

      Anyway, when the knee/foot is about to change direction, the lower leg is as unfolded it gets, the knee flexion will be nearly the same in this position as it is at touchdown. I have observed that different runners (all landing with perpendicular lower leg) have different amount of knee flexion. Sprinters always have quite little, but among distance runners there is a much larger variance. My guess is that if the distance runners with the more bent legs would increase cadence in a sprint, their knee flexion would reduce.

      I truly don't know what to say about that, if it is better with a straighter leg or a more bent, or if it does not matter.

      With a more bent leg you land further ahead of the CoG, you get greater breaking force but also longer time to passively load the leg, which perhaps is better for some individuals. You will probably also "ride lower", that is have more knee flexion in the whole support phase, and keep ground contact longer before pushoff as well. Probably, short ground contact and fast force production is best, but if you don't have that ability?

      Perhaps there is a strong individual component to this, depending on your muscular capability.

      (The purpose of having a perpendicular lower leg instead of a forward-leaning angle is of course to avoid breaking forces travel up the leg and smash the joints. With a perpendicular lower leg the breaking force will just increase flexion in the joints and load the leg with elastic energy.)

    13. RICK'S RUNNING on August 25, 2010 at 3:12 pm

      Six Harvard 'Grouchos' between the ages of 25 and 43, all of whom were well-trained runners, were filmed by McMahon while running on treadmills with both their normal and Groucho running styles – and also ran over force plates using both techniques. McMahon and his research crew noted that Groucho running decreases the stiffness of the legs by about 20 per cent during running and also – as one would predict – expands foot-contact time ('Groucho Running', Journal of Applied Physiology, vol. 62(6), pp. 2326-2337, 1987). From an injury-prevention standpoint, Groucho running reduces the amount of shock which is transmitted up through the body during footstrike by as much as 80 per cent, but there is a price for this trauma minimization: the oxygen cost (economy) of running at a particular speed goes up by about 50 per cent during Groucho running, compared to regular ambling! While compliant Groucho legs might keep you away from injury, they will also slow you down!

      As you can see from the Groucho investigation, the stiffness of your leg springs can play an important role not just in terms of speed but also in determining your running economy – the oxygen 'cost' of moving along at a particular pace. Although serious runners and coaches often extol the merits of heightened flexibility, if your legs are too compliant your foot-strike times will be sub-optimal because you'll spend too much time uncoiling your permissive springs just before take-off. That's why research exploring the link between flexibility and economy has actually shown that higher musculoskeletal tightness is often related to lower aerobic demands during running, i.e., somewhat stiff individuals tend to have better economy than very flexible runners ('The Influence of Flexibility on the Economy of Walking and Jogging,' Journal of Orthopaedic Research, vol. 8, pp. 814-823, 1990). The investigators in this particular study indicated – in line with the arguments presented in this article – that elastic energy contributions may be enhanced in less-flexible runners and that the need for neutralization of unproductive leg movements may be reduced when one's 'springs' are somewhat tight.

    14. Anders on August 25, 2010 at 3:37 pm

      Rick, thanks for that information!

      However, when I see the definition of Groucho running, it seems to be a bit more extreme than I was thinking.

      The differences I've been looking at is in a smaller range, I would call Ryan Hall one with more bent leg (closer to Groucho): http://www

      and Haile Gebrselassie with a more straight leg:

      (Unfortunately no slomo on Haile, but you can see the leg becoming almost straigt in the front position, and it does not seem to bend much on its way down to touchdown). I've seen Haile in slower running too, and he seems to keep the leg as straight there.

    15. RICK'S RUNNING on August 25, 2010 at 4:28 pm

      Anders,I find I run faster and smoother with an almost straight leg, the study says the trade off for faster running is increase in shock from the ground, I guess as I've run this way for many years [ due to damaged knee ligaments from cycle racing 20+ years ago i find landing with a bent knee hurts my knees!]my body can handle the increased shock, infact bones ligaments and tendons adapt and get stronger with increased weight baring!
      The opposite effect to this would be living in space and losing bone density and muscle.
      End of the day it is what feels right for you.

    16. Anders on August 25, 2010 at 4:58 pm

      (Okay, I've been talking a little bit of rubbish above saying that knee changing direction and maximum unfolding happens at the same time, I don't know what got into me :-). Of course knee changes direction before maximum unfolding… duh.)

      Anyway, for some runners the lower leg unfolds more than for others.

      From the above videos I put together this image:


      Haile's leg unfolds more, and does not bend as much during support. Obviously Haile is the faster runner, but would Ryan's technique be improved with a straigther leg?

    17. RICK'S RUNNING on August 26, 2010 at 4:44 pm

      Well I do remember Ryan Hall joking that after a visit to the Asics R@D department in Japan last year they told him that "he runs like a girl"!
      See video here

    18. Post Paint Boy on August 26, 2010 at 8:53 pm

      Nice graphic, but I'm not sure that it's a fair comparison. Geb is running on an extremely soft treadmill, while Hall is running _slightly_ downhill on the road, 17 miles into a race.

    19. Anders on August 27, 2010 at 7:46 am

      I should say that the example Ryan/Haile was not as different as I thought it would be. The knee bend at touchdown is about the same, and thus the distance to CoG. I thought it would be even more straight for Haile, and thus he would land much closer to the CoG, but that does not seem to be the case (a little closer though).

      The difference seems to be that if you get a near-straight leg in the front, you need to bend it some before touchdown, while the other alternative is to have as bent in the front position as it is at touchdown. I think that this difference is due mostly due to passive mechanics though, that is if you swing forward your knee faster and change direction faster it is more likely that the leg will be near-straight. So it perhaps does not make so much sense to analyze it, the important part is that there's no active reaching out with the leg.

      (In recreational runners I often see the leg becoming even more straight than Haile's, but then combined with overstriding and heavy heel-striking).

      Things are connected – more bent knee at touchdown – landing further ahead of the CoG – more bent knee during support – more degrees to straighten out at pushoff. A slower force production cycle with a longer range than the opposite. Good or bad? I guess most would consider it "bad", but perhaps for some it is the best way to make ideal use of their available physical capacity. Note that I compare only styles with perpendicular leg at touchdown, I consider positive angle with heel-striking a technique problem.

    20. stevemagness on August 27, 2010 at 4:31 pm

      Some great comments by all. Very interesting discussions.

      I'll try and comment more this weekend when I have more time.

      But to briefly touch on some things.

      -Remember that the stride is all connected and an integrated whole.
      -Remember passive/simple mechanics.

      How high the lower leg comes up to the butt and the speed at which the leg is coming through effect how much a person's lower leg is going to straighten. If there's more momentum with the lower leg unfolding, it's going to take either more muscle activity from the antagonist group to slow it down, or its going to go out further in front and be pulled back.

      As anders and rick have pointed out the degree of knee bend upon landing is going to impact stiffness, loading, and compliance. How bent the knee is also affects what takes the brunt of the forces.

      Also as post paint boy pointed out, the running surface plays a role. How much cushioning needs to be provided by the legs versus the actual surface might influence things. If the legs need to provide more cushioning, you're likely to land in a position that minimizes the impact forces. If the ground is softer, you don't have to worry about that.

      It can get really complex if we think about it.

    21. Petri on August 28, 2010 at 10:31 am

      Forward running consists of keeping the velocity AND in each step also countering gravity. The foot strike should ideally be exactly where the directional vector of your COG is heading to avoid braking, not directly below unless you are standing still.

      My 10 cents…

    22. RICK'S RUNNING on August 28, 2010 at 4:11 pm

      All you guys in the Uk should check out jon Goodwin's video's part one and two on the UKA website http://coaching.uka.org.uk/video/jon-goodwin-maximal-velocity-running-mechanics-part-2/
      He blows away a few myths on how to run fast.
      Gravity is the Number one thing a runner has to overcome it acts directly down on the athlete.
      Producing the most force downwards into the ground in the shortest possible contact time produces the highest speed.
      Overstriding either infront or behind has a negative impact on speed!
      Leg turnover and stride length are not something to be trained for, but are just a consequence of running speed brought about by contact force and contact time.

    23. Anonymous on August 30, 2010 at 4:52 am

      Rick, Goodwin's video / analyses support Weyand & Bundle's study of "Faster Running Speeds…" (JAP 2000). Mass-specific force is the number 1 contributor to faster running speeds. Every runner has to overcome the bounds of gravity…a simple matter of physics. Increasing muscular force, decreasing ground contact time, and increasing rate of force delivery will result in faster running speeds. Mechanics play a much lesser role in the process. As Dr. Weyand once pointed out, "Messing with mechanics would be akin to telling a piano player hitting many hundreds of notes a minute ot alter pinky flexion by six degrees to improve sound…the tuning is already so finely coordinated for balance and bouncing that the notion of intervening to improve and increase speed is quite naive."

    24. Anonymous on August 30, 2010 at 5:02 am

      Sorry, Steve, add Coach Robinson to the previous "Anonymous" post (but you probably already knew that).

    25. Anders on August 30, 2010 at 7:15 am

      Max speed running = sprinting. But can you really "overstride behind" in distance running? That is lift the foot too late from the ground. You can overstretch the leg (continue straightening the leg after the foot has left the ground) though.

    26. RICK'S RUNNING on August 31, 2010 at 1:25 am

      Video How to run really F>A>S>T

    27. RICK'S RUNNING on August 31, 2010 at 4:27 pm

      P.S. check out Rudisha's arm action [see video link above] he seems to be using a more downwards and up action than the normal forward-back arm action.
      Could he be generating more downward force into the ground?

    28. Anonymous on September 1, 2010 at 2:22 am

      The arms provide balance in running and little else. Arm-assisted force has applications in jumping events. Consider athletes with arm deformities or even athletes with amputated arms. Do they apply less ground force than able-bodied runners? In American Football, I've repeatedly witnessed running backs and receivers carrying a football who outrun defenders that have the full mechanical swing of both arms. If the arms are important to ground force production, how is this possible? Answer: they aren't.

      Coach Doug Robinson

    29. Anders on September 1, 2010 at 10:26 am

      That's how armswing is discussed in sprinting, used to increase the vertical impulse so you get more force into the ground to get a stronger loading of the leg spring. While ground reaction force in distance running is about 2,7 times body weight, it can be as much as 4,6 times body weight in sprinting. I guess Rudisha's 800 meters is somewhere in-between.

      Apart from armswing, there's also the concept of "active foot-strike" (directly translated from Swedish, not sure what term is used in UK/US) in sprinting and to some extent in middle-distance, to increase vertical forces.

      I think it is important to note the differences between really high speed running and relaxed long distance running, since some things that may be good to focus on in 800 meters may not be suitable for marathon.

    30. stevemagness on September 1, 2010 at 11:40 am

      Rick- Anders did a good job summarizing the connection of the arms and legs.

      Additionally, at faster speeds the arms have to open up and have an increased ROM in part because the legs open up and have an increased ROM. In distance races, the stride length is still relatively short, so no need to open up the arms.

      As far as Coach Robinson's beliefs, I have to strongly disagree. Just to briefly address some claims. The anectdotal evidence is pretty weak. For example, the football example of carrying a ball ignores the fact that there are big speed differences between players. A fast wide reciever is still going to outrun a slower defensive back, even if he's carrying a ball. However, if the runners are relatively close in speed, the guy holding the ball will be slowed. For example, as a Houston Texans fan I remember the Texans return man a couple years ago Jerome Mathis got run down a few times. For those who don't know, Mathis was a 20.2 200m sprinter in college. There are very very few athletes who have that kind of top end speed in football. Yet, he got ran down by slower people because he was carrying a ball.

      Anyways, it's a simple test. Go to the track and run holding your arms by your side without moving. You'll run slower, no doubt.

      As for the research, studies have shown there connection. A study by Huang and Ferris (2004) showed that when using a recumbant stepping machine, arm swing increased the muscle activation (and thus presumably force) of the lower legs. If this occurs during recumbant stepping, is it a stretch to think it happens during running? I don't think so.

    31. RICK'S RUNNING on September 1, 2010 at 4:20 pm

      Thanks Anders and Steve for answering my question.

    32. Anonymous on September 2, 2010 at 3:50 am

      Hi Steve,
      Please don't confuse my analysis of arm function's benefit to running to that of running without arm movement ("Go to the track and run holding your arms by your side without moving.You'll run slower, no doubt.") I would never offer that as a comparative test of faster running speeds. Arms do provide a necessary function to running, although according to most locomotion experts, it is simply the role of balance. I know from experience that the position you take on the importance of arm mechanics to running performance is the same as most coaches accept as preferred form. I, too, held that same opinion for 35 years as a coach. I attended clinic after clinic, read copious coaching articles on mechanics, poured over the latest running research, repeated the mantras of coaching cues to athletes ("knee up, toe up, heel up, step over the opposite knee, dorsiflex, pawback, lock the elbows at 90%, hands at the crest of the hip on the downswing, hammer back"…ad nauseum). Dr. Peter Weyand initated a change in perception; but an observation later confirmed all that. I witnessed a HS sprinter with deformed arms outrun able-bodied sprinters at a commendable time for HS sprinters. If mechanics were so important, then how was it possible that this athlete accelerate from a standing start and win a sprint race? Could it be that perhaps arm mechanics weren't that important to force production? Consider Usain Bolt's 100m world record in the 08 Olympics. At approximately 80m he rotated his torso sideways and extended his arms, yet he did not decelerate. It was only after 90m when his foot extended beyond his knee in a braking fashion that deceleration occurred.

      As far as anecdotal evidence of football players carrying a football, your counter argument of Jerome Mathis deserves the same scrutiny. Without having seen your evidence (straight lines of acceleration vs. change of direction and deceleration to avoid contact, nor knowing the "slower" athletes who ran him down, nor knowing the angles of pursuit by the slower athletes), your counter claims are equally weak. Intuition tells me that in none of your observations of Jerome's returns did he run a straight line unimpeded to the goal line, nor did the defenders who ran him down run as far as he did before contact was made. It is highly doubtful that carrying the football was the significant difference of his being tackled by slower defenders. It would be interesting to see Jerome (carrying a football) line up in a race from goal line to goal line against those slower defenders. It is unlikely that the football would affect the eventual winner of that foot race.

      I think in many ways we are alike in our pursuit of the science behind faster running. In fact, much of my training philosophy mirrors what you are doing. However, based upon the evidence I have gathered, the importance of running mechanics effecting faster speeds is somewhat exaggerated.

      Coach Doug Robinson

    33. stevemagness on September 2, 2010 at 1:32 pm

      Coach Robinson,
      Again thanks for the comments. It’s always good to have discussion of different ideas.

      My point in the Jerome Mathis example was to show how fallible using example like people with arm disabilities or those who are carrying a ball are. There are way too many factors to consider. Yet you hear these examples from people like Barry Ross all the time as evidence. You can’t have it both ways, if people carrying a football run away from those without, then you have to account for fast runners who carry a football and get run down when they probably wouldn’t have without a football. The logic in the Barry Ross type examples is that if people can run fast like that, then the arms must not be aiding speed. The arms example at the track was another extreme example to demonstrate this. So the examples are useless. The real question is do they improve just balance or force production.

      I gave a study which showed that arm swing while stepping increases muscle activation in the lower body. So, the arms must be doing something beyond just providing balance.

      If you want another study, check this one out:
      When arms were suppressed, ground reaction forces DECREASED by 10-13%. So, that shows right there, that the premise of the football player/arm deformity examples are wrong. Arm swing does impact force production.

      It’s kind of amusing that almost all the coaching cues you named, which are prevalent, I disagree with.

      Finally, on your example. The important thing to remember, and the one that most researchers don’t grasp, is that performance is complex. It doesn’t come down to 1 factor like GRF. It never does. We’ve made this mistake with VO2max, lactate, etc. Yet no one learns.

      Arms play a role. Are they a major role? No, they are complementary. But every percent is important in high level running. As for your example of the sprinter, it simply demonstrates that arms aren’t the major role. You can overcome it, but could he ever run as fast? Who knows but he wouldn’t have everything going for him.

      So that’s my take. Be careful with falling in love with a single parameter that describes performance. It doesn’t exist. It’s too complex. It’s akin to distance coaches who believe oxygen utilization is ALL that matters. Well, have fun with that one…

      So, I’m not saying that force production isn’t important. It is, very important. What I’m saying is that it’s not the ONLY factor.

      One more thing. Although I don’t agree with much of the study, I thought you might find it interesting as they critique Weyand’s work.

    34. Anonymous on September 6, 2010 at 3:31 am

      Hi Steve,
      Thanks for the links to the studies. Interesting stuff.

      Coach Doug Robinson

    35. Anonymous on September 13, 2010 at 1:48 pm

      Hi Steve,

      Interesting post. In your post about running shoes you stated:

      "Looking at elite athletes, when racing and training, they generally have higher turnover, minimal ground contact time, and a foot strike that is under their center of gravity."

      Seems you have shifted your opinion that elites have a footstrike under their center of gravity. Care to elaborate on what brought about the shift in thinking?

    Leave a Reply