Weyand v. Pose: it’s the push that counts

A few weeks ago, I presented an overview of my plan to get fit enough to run a ‘good’ marathon again in 3 years time. Each 3-4 month period will have a specific goal in addition to the continuing basic objective of improving my aerobic capacity. The specific goal for the first 4 month period is recovering some of my lost muscle strength, largely by resistance exercises and drills, together with a small amount of sprinting and uphill running.

Although I have now completed the first three weeks of the program, I have not fully settled the question of what leg strength exercises and drills are likely to be most beneficial. At present the exercises that I am doing are mainly body-weight exercises performed standing on one leg (such as one leg squats , hip swings, calf raises etc) since the major muscle actions of running are executed while standing on one leg. I am also doing a very small amount of plyometric exercise to increase my capacity to deal with eccentric contractions but in view of the risk of long term damage, I am being quite sparing with these exercises. I plan to test for gains in leg muscle strength (assessed by a hopping test, and by measuring my sprinting speed) after 6 weeks and if there has not evidence of substantial gain, I will modify the types of exercises and drills in my program.

In preparation for this, I have been once again reviewing the question of running technique in order to identify what exercises and drills are likely to be most beneficial. Two apparently coincidental happenings in the past two weeks have also served to re-focus my attention on running style.

 Gravitational torque

One has been a series of comments by Simbil on a post entitled ‘if gravitational torque is a red herring, how to do we run fast?, which I had posted in February. Simbil is a runner and thinker based in Sheffield, who contributes thought-provoking comments to the Efficient Running thread on the Fetch website. He is an advocate for Pose technique, though he is prepared to re-examine the tenets of Pose in a thoughtful manner. Gravitational torque plays a crucial role in the theory of Pose. Simbil has been challenging my assertion that the gravitational torque that generates head-forwards and downwards angular momentum in the second half of the stance phase is balanced by an oppositely directed torque acting in the first half of stance. We have been lobbing arguments back and forth for some time. We have not yet reached any definitive conclusion, though we are in agreement that the gravitational torque must be counteracted by an oppositely directed torque acting a some point in the gait cycle, and furthermore, that whatever the mechanism, the forces involved in generating gravitational torque and compensating for it are far less important that the forces required to get airborne. I will return to this issue a little later.

Pose Tech

The other coincidental happening came to my attention when I noted a little upwards blip in the number of hits on this site a little over a week ago. At present this site gets about 2000 hits per month (far below the league of the Huffington Post or Belle de Jour, or, I suspect, Younger Legs for Older Runners, but nonetheless, satisfying). Most of the hits come from Google searches seeking information about various aspects of running technique and training, though the flurry of increased activity in the few days after each posting indicates that there are also some faithful followers. It always gives me a little thrill to see these flurries of activity and I hope that those of you who read the posts find them interesting. In addition there are occasional flurries of increased activity at other times indicating that some other website has drawn attention to something on my site. One such flurry occurred last week, and inspection of the stats provided by WordPress revealed that the hits were referred from one of the forums on the Pose Tech website. Those forums are protected by password, but as runner interested in Pose, I have been registered with the site for several years. So I was tempted to have a look.

When I logged on I was intrigued to find that there was a quite active thread discussing my blog. Those of you familiar with the Pose Tech site will probably be aware that many of the people who post on the forums are strong believers in Pose. Therefore, I was more amused than surprised to see comments such as ‘burn the heretic’. However a gratifyingly number of the postings were in agreement with my views, especially my view that landing on the ball of the foot, in combination with a short time on stance places large forces on the foot, and is associated with risk of injury to metatarsals and connective tissues of the foot and ankle.

I was a little saddened by a passionate posting from Lana, the wife of Nicholas Romanov [correction: Lana is his daughter, see comments below], lamenting the fact that I and others did not give due credit to the amount of work and expertise that her husband has devoted to the Pose technique. However, while I am happy to acknowledge the work and expertise of Dr Romanov, I think there is a serious issue arising from the fact that the 2002 edition of Pose Method of Running reproduces the diagrams and photos from the first edition showing the heel held high off the ground at mid stance. In the early years, it was quite common for Pose novices to report Achilles or calf muscle problems, which I think are largely due to holding the heel off the ground. The importance of these problems was reinforced by the findings of the Capetown study (Arendse et al, Medicine & Science in Sports & Exercise: Vol 36 pp 272-277, 2004) demonstrating that Pose is associated with increased stress around the ankle, together with the reports by Ross Tucker (who assisted Dr Romanov in coaching the Pose technique in that study) that the runners in the Pose group suffered a high rate of injuries to connective tissue around the ankle (http://www.sportsscientists.com/2007/09/running-technique-part-ii-scientific.html).

When I expressed disappointment that these figures had been reproduced without anything in either text or illustrations (as far as I have been able to ascertain) describing how the risk of these injuries might be minimized by relaxing the ankle to allow the heel to touch the ground, Lana suggested that ‘you need to “zoom out” a bit – because you are not doing yourself any favours by using a magnifying glass all the time’. In an email, a good friend has suggested that it is unlikely that I will be on Lana’s Christmas card list this year.

Despite this little storm in a tea-cup on Pose Tech, I have a fairly positive view of how Pose works in practice, based not only on looking at the evidence through a magnifying glass; but also by talking to runners who have applied it; and by experimenting with it myself. I am less impressed with the theory underlying Pose, but the evidence that the technique can be beneficial makes me want to understand more clearly why it works

A subjective overview of Pose

Many of the principles of Pose such as high cadence, short time of stance; avoiding reaching out with the leading foot, and keeping the pelvis forward, make mechanical sense. I discuss these features in my own account of technique (‘Running: a dance with the devil’ – listed on the side panel). Many elite athletes exhibit these features despite never having been trained in Pose, though I think Nicholas Romanov deserves credit for publicizing these principles among amateur runners.

Other features such as landing on the ball of the foot (BOF) also make sense, but are attended by risk. As I had mentioned in my comments on Pose Tech that provoked the anguished response from Lana, BOF landing places substantial stress on the bones and connective tissue of the foot and lower leg, and creates a risk of repetitive strain injury to metatarsal, planate fascia, Achilles tendon and peroneal tendons. At least when running long distances, I consider it is best to relax the ankle so that the heel brushes the ground in mid-stance

I also consider that Dr Romanov’s claim that gravitational torque provides the motive force for running is misleading. According to the law of conservation of energy, it is impossible for gravity to provide net energy when running on a level surface. I accept that the unbalancing due to gravitational torque during the second half of stance does help promote the muscular actions necessary to swing the leg forwards in time to support the body at subsequent footfall. However I am not convinced that this is plays a major role, and I therefore consider that gravitational torque is probably a red herring. Nonetheless, I am quite happy to continue debating this issue with Simbil.

However in my opinion the most interesting issue for debate is the concept of pulling from stance rather than pushing. The Pose mantra is ‘Pose, Fall, Pull’ implying that the crucial actions are:

1) adopting the classic pose with the point of support (BOF), hips, and shoulders aligned, in mid-stance;

2) falling under the influence of gravitational torque as the centre of mass moves in front of the point of support;

3) pulling the leg from stance by means of a contraction of the hamstrings.

In practice, I think that getting airborne is impossible without a push. Pulling from stance would amount to lifting oneself by one’s own boot straps. Of course pulling the leg forwards by hip flexor contraction once airborne is essential, but is largely automatic. While I think that pulling from stance is largely an illusion, I think it might be a useful illusion to cultivate because it encourages a short time of stance, high cadence and relatively short strides – possibly not the best style for record breaking performances, but nonetheless fairly safe. For most of us, avoiding injury is more important than maximizing mechanical efficiency.

So, while I think that the Pose concept of pull is not mechanically correct, I think that it is a useful mental image to cultivate, in the same way that some of the principles of Alexander technique are based on creating helpful illusions. Therefore, Pose drills such as Change of Stance that encourage the illusion of pulling are probably useful. But if one want to increase one’s speed of running, are Pose drills more important that developing the muscles responsible for the push?

Weyand: it’s the push that counts

Perhaps the most important evidence about the muscular action required for running fast comes from a study by Peter Weyand and colleagues from Harvard (J Appl Physiol Vol 89: pp 1991–1999, 2000). In a study of runners with various different maximum running speeds, using a force plate to assess ground reaction forces during treadmill running, they demonstrated that ‘human runners reach faster top speeds not by repositioning their limbs more rapidly in the air, but by applying greater support forces to the ground.’ The force plate data showed clearly that a larger vertical ground reaction force achieved by the foot pushing against the ground was the main factor distinguishing fast runners from slower runners.

The extensor paradox

However, Weyand’s findings present a problem: the extensor paradox. Studies employing electromyography to measure the electrical activity in muscles demonstrate the major muscle groups that would be expected to execute a push off the ground, especially the quadriceps which are responsible for knee extension, are almost silent in late stance (Elliot et al. Medicine & Science in Sports vol 11, pp 322-327, 1979). So where does the push come from? Much of it almost certainly comes from elastic recoil of quadriceps, acting in conjunction with the hamstrings, and soleus, all of which have been stretched on impact at foot fall. Thus the crucial strength that is required for running fast is the ability to sustain a strong eccentric contraction of the quads, hams and calf muscles.

Conclusions

In a program to rebuild the strength necessary to run fast, it is crucial to recover the capacity for strong eccentric contraction of quads, hams and calf muscles. Of course other muscles also play a role – especially the gluteals which help hold the pelvis in a horizontal position during swing, and the hip flexors, which drive the swinging leg forwards – but Weyand’s study indicates that it is the ability to push strongly that distinguishes fast runners from slow runners. Clearly any strengthening program should involve a reasonable balance between all relevant muscle groups as muscles do not work in isolation, but nonetheless, much of the focus should be on eccentric strengthening of quads, hams and calf muscles.

However there is a problem: as I have discussed several times previously on this blog, eccentric contraction damages muscle, possibly permanently, especially in older runners. So a program of plyometrics designed to increase the ability to withstand eccentric contraction must be undertaken cautiously. I am inclined to think that the first priority is to recover strength of the major muscle groups, especially quads and hams with a program of resistance training with free weights, including a gradual build up of eccentric loading, before moving on to a modest amount of plyometrics. Meanwhile, I will also continue with regular uphill running, as this invokes the required muscular actions in a way that is ‘natural’ for a runner.

As for drills, any drill that promotes a short time on stance is likely to be beneficial. I am still exploring the options, but despite my opinion that the pull is less important than the push in getting airborne, I think that the Pose ‘Change of Stance’ drill is beneficial, even if it is largely cultivating a beneficial illusion of pulling from stance.

Advertisements

57 Responses to “Weyand v. Pose: it’s the push that counts”

  1. Jeremy Huffman Says:

    Canute,

    Lana is Nicholas Romanov’s Daughter.

  2. canute1 Says:

    Jeremy, Thankyou for that correction.

  3. Jeremy Huffman Says:

    Your welcome

  4. Jeremy Huffman Says:

    Wish I had something more constructive to say. Unfortunately I don’t know where to begin 😦 I disagree with pretty much every word you say. Okay…okay…not every word but most of the points. Interesting perspectvie you have nonetheless. And you right with great thought 🙂 Good day

  5. canute1 Says:

    Jeremy, as the person who seconded the motion that I should be burned as a heretic (though of course only in jest, I am sure), I am not surpised that we disagree. However I appreciate the fact that you consider that my perspective is interesting

  6. RICK Says:

    Dear Canute,
    I spent the last two years trying different running methods, inc Pose, Chi, etc, my own feelings are pose and chi are not the answer to injury free or faster running, its a bit worrying to me that people get really hooked up [ brain washed] into these big money making bodies and believe everying their running guru’s tell them!
    i have a friend who was a pose instructor, and she gave it up after being sickened by their money making greed!
    a top running coach told me ‘ RUN AS FAST AS YOU CAN, WHEN YOU SEE A RUNNING GURU COMING TOWARDS YOU’ good advice as far as pose is concerned i think.

  7. RICK Says:

    P.s. on my long journey i have picked up on what works and disguarded what is dangerous, after reading ‘EXPLOSIVE RUNNING’ by Dr Yessis I would have to recommend that every runner who really wants to run faster and with less risk of injury read it. I doubt your find better advice!

  8. Jeremy Huffman Says:

    Rick,

    Be careful you don’t follow the “guru” Dr. Yessis 😦

  9. RICK Says:

    Dear Jeremy,
    I don’t believe any one coach or ‘guru’ has all the answers,I don’t believe any one method is right for everyone, but reading the Explosive running book is a very good place to start, get it from a library at little cost, better than paying for a $300 lesson with a ‘Guru’ :]
    RUN IN PEACE

    • Jeremy Huffman Says:

      Rick,

      I agree NO one person as all the answers.

      What are some of the points in the book “explosive running” that you think apply to a runner? and if they apply to everyone how come?

      I notice you are a huge fan of the “Paw back” as you call it. Is this something that Dr. Yessis advocates? Having spent many years running and much of it at an elite level, I am not an advocate for “pawing back”. As I see it as an empty motion void of providing any gains in efficiency. I am open to hereing your arguement on why it has some validated other than, you think you see some elite runners employ this action.

      P.s.- Making comments that suggest everyone who uses a specific technique is “brain washed” doesn’t really put you in a good light. It makes you sound as if you are not as open minded as you think 😦

  10. Peluko Says:

    Hello.

    I’ve been reading Weyand study (this is a boring afternoon…) and I want to ask some questions. Maybe my English skills are misleading me (I’m Spanish), but I think that I’ve understood the study but I’ve not undestood why they conclude that the greater ground forces are causing the speed. What they are measuring might be a consecuence of speed, not a cause, specially when they say
    “Although support forces differed roughly twice as much across this range of top speeds as did either step frequencies or contact lengths, we expected these force differences to be greater. Our regression relationship indicates that altering the support force applied by only one- tenth of one body weight is sufficient to alter top speed by one full meter per second. In contrast, Eq. 5 predicts the force necessary to affect this difference in top speed should be twice this large for a runner with average step frequencies and contact lengths.”

    I can’t understand all the wording that cames after this paragraph, concluding
    “Thus the sensitivity of top speeds to the forces applied to the running surface resulted from the positive effect support forces had on both the maximal stride lengths and frequencies that runners were able to attain.”

    Well, increased forces to the ground implies more aerial time (vertical height) and/or less contact time for the same aerial time (vertical height). I can’ see the cause-consequence relation with speed.

    I find that the conclusions aren’t clear, mainly because the difficulty to distinguish between vertical forces (which affect vertical displacement and directly the flying time) and horizontal forces, which are the ones that give speed.

    Am I missing something?

    And yes, I think that your conclusion that ‘it is crucial to recover the capacity for strong eccentric contraction of quads, hams and calf muscles’ is correct, because be it a cause of a consequence of speed, you will need it.

    Bye!!

  11. canute1 Says:

    Peluko,
    Thank you for your very interesting comments.

    You ask what Weyand and colleagues meant by the text leading to their conclusion “Thus the sensitivity of top speeds to the forces applied to the running surface resulted from the positive effect support forces had on both the maximal stride lengths and frequencies that runners were able to attain.”
    It seems to me that there are three items of information that lead them to that conclusion:
    1)The faster runners exerted a greater vertical force and so were air-borne for longer. As a result their steps were longer.
    2)Because the product of average vertical force x proportion of time on stance must be equal to weight, the faster runners spent less time on stance, but the data showed their bodies moved further while they were on contact (ie they had a longer contact length.
    3)Because time on stance was shorter, total time for the gait cycle was shorter in the faster runners. Therefore they had a higher step frequency.

    (Note that total time for each gait cycle = 2x airborne time + 2 x stance time; swing time = 2x airborne time + 1x time on stance; at top speed all runners had approximately the same swing time associated with shorter time on stance but longer airborne time)

    According to equation 5 in Weyand’s paper, speed is the product of step frequency, average vertical force per unit mass and contact length (the distance the body moves while on stance). For the faster runners, all three of these quantities were greater. Hence the faster runners were able to achieve a large increase in speed with only a moderate sized increase in force.

    What the authors observed was greater speed associated with greater vertical force. The authors attribute the benefits (faster speed, longer stride, higher step frequency) to the greater force. You raise the possibility that the greater speed causes the greater vertical force rather than the other way around. I do not think this is likely because if one starts with a greater force, one would predict greater stride length and higher step frequency. So there is a logical mechanism by which greater vertical force must lead to greater speed. In contrast, it is not obvious to me how greater speed would cause greater vertical force. If for example there was a strong wind from behind the runner, I think he/she would run faster but with no increase in vertical ground force.

    Finally you state that horizontal forces are the ones that give the speed. If we ignore air drag, the braking forces must always balance the forwards forces according to the law of conservation of momentum. Therefore the horizontal forces do not produce speed.

    I am pleased that we both agree with my final conclusion about the need to increase the strength of eccentric contraction of quads, hams and calf muscles

  12. Peluko Says:

    Hello again.

    Sometimes, when I’m writing English, I lose the track of my thoughts trying to find the correct words and I miss something in what I say.

    My thoughts were in the line that the impulse (force by time, translatable to work and energy) where the same for slow and fast runners.

    Your statetement:

    “1)The faster runners exerted a greater vertical force and so were air-borne for longer. As a result their steps were longer.”

    contradicts this Weyand’s statement:

    “The time spent in the air (mean aerial time 5
    0.128 6 0.004 s, Fig. 5C) at top speed did not vary as a
    function of the top speeds for our 33 subjects during
    level running. This was due to the equivalence of the
    vertical impulses determining aerial times among fast
    and slow runners.”

    The faster runners exerted a greater force during shorter time, so the impulse, and finally vertical displacement and time airborne, was the same. The longer steps were consequence of their greater speed applied during the same time.

    So the forces are greater because the work for vertical support is more ‘explosive’, but the total work is the same. Making the work more explosive lets faster runners to spend less time in contact with the ground, and thus lets them to get greater step frequency without shortening the airborne time. This is what I see.

    PS: Oh, and yesterday I was looking this video:

    and I was thinking “eccentric contraction or elastic recoil?”

  13. Peluko Says:

    Oh, well, again I forgot something…

    Well, what I miss in this study is that it doesn’t explain from where the speed comes from. It explains that faster runners applies greater forces during shorter times, increasing stride frequency, but this frequency alone doesn’t provide the speed. Someone jumping without moving forward also could apply great forces and shorter contact times, but from where comes the forces that first accelerates forward and second maintain the speed against braking forces?

    Maybe the gravity Dr. Romanov states for? Propulsive conscious or unconscious muscle effort? Both?

  14. Ewen Says:

    Hi Canute. I had a good laugh at your detractors wanting to ‘burn the heretic’.

    All I have to say about Pose, is if it works for some people and they enjoy running with that method, then fine. However I side with Weyland if he believes ‘pushing off’ with force (even if that comes from ‘recoil’) produces a longer stride, and hence, faster running. In the end, the stride needs to be long, because the ‘steps per minute’ has a natural limit.

    Also, I’d just say that people ran well, won medals and enjoyed their running for a long time prior to the development of Pose.

  15. Aliesa George Says:

    Hi! I am only a recreational runner, but also Pilates teacher with great personal and professional interest in foot fitness! Haven’t studied different running techniques, but spend quite alot of time watching bodies in motion and helping clients improve strength, mobility, and mechanics. Seems to me that pulling the leg forward with the hip flexors will shorten leg swing and stride. I know that for my walking and running technique, that when I actively push with my hip extensors and carry the push all the way through the foot both my speed and stride length increase with less effort. The challenge then becomes my heart rate being able to keep up! Another observation – due to the high use of flip-flops and other mechanically limiting things in life…I am noticing that more and more people walk and run with their legs in front of them basically hanging off their hip flexors, and are completely missing the balance of hip extension that should be a part of stride. (no hip extension means no chance of a push through the ankle and foot!) Hope that perhaps my comments and insight will be helpful to someone.

  16. canute1 Says:

    Peluko,
    I agree with you that that it is the vertical impulse (vertical force x time) that determines the amount of upwards momentum imparted to the body. I had not looked closely enough at figure 5 of Weyands paper, and on closer inspection I note that the tendency for greater airborne time at faster speed is not statistically significant, and that the authors make this point in the text. One the other hand, the observation of similar swing time despite shorter time on stance does suggest a tendency for longer airborne time in the faster runners, because swing time is the sum of 2xaerial time + time on stance. However, the issue of any difference in airborne time is not the crucial issue. The crucial points I was drawing attention to were that both stride length and step frequency (cadence) were greater for the faster runners who exerted a greater push on the ground. However these observations alone do not prove the direction of causality. As you point out, the longer stride length could be because they are running faster, or it could be that they are running faster because they have a greater stride length.

    You also ask about the source of the energy for the initial acceleration and for overcoming breaking forces. The source of the energy for initial acceleration almost certainly is muscular contraction that produces a push backwards that in turn generates a forward ground reaction force. I accept that gravitational torque assists us to accelerate more rapidly, but I do not think that this is important for distance runners. However I think it is important to note that if we ignore wind resistance, when running at a constant speed on level ground, we do not need to expend any energy simply to maintain speed. That is what is predicted by Newton’s first law of motion. However we do need to do work to overcome braking forces. I believe that the energy for overcoming braking is largely provided by the muscles that produce a backwards push in late stance thereby generating a forward ground reaction force.

    I question I was wishing to address in my blog was what a runner needs to do in order to be able to run faster. The evidence from Weyand’s study suggests that decreasing time on stance is likely to be more useful than decreasing swing time. A crucial requirement for decreasing time stance is increasing the ability to exert a forceful downwards push. However the evidence that quads are silent in late stance (the extensor paradox) indicates that this push is not achieved by a strong concentric contraction in late stance. It appears that the best way to prepare for exerting the required force is to strengthen eccentric contraction of quads, hams and calf muscles. I think we agree on this.
    In addition, it is necessary to develop the neuro-muscular coordination to achieve the required short time on stance. As I stated in my post, I believe drills such as the Pose ‘Change of Stance’ drill might be helpful for this – though this hypothesis is only based on observing a few runners who have employed the COS drill. I also think that cultivating a mental image of pulling rapidly from stance can help, though what actually happens should probably be described as a short sharp push rather than a pull.
    On account of the consequence of the law of conservation of angular momentum, I do not believe that gravitational torque plays a substantial role in running faster, but that is a debate that I am still having with Simbil – I hope to get time in the near future to do some further calculations in response to his most recent comment on my February post on this topic.

  17. canute1 Says:

    Aliesa
    Thank you for your comment. I agree that too muuch focus on swinging the leg forwards it not helpful, and can lead to over-striding (landing too far in front of the centre of mass), which produces unnecessary braking. I was interested in your comment that you fnd it helpful to conciously push using the hip extensors. Do you mean the hams or the glutes? Contraction of the hams will proudce knee flexion as well as hip extension unless there is simulataneous quad contraction to stablise the knee, so I am not quite sure that I understand what you mean. However, it is true that a strong push is essential to get airborne efficiently, and the Weyand study indicates that this results in a long stride, though I am not sure what is the best mental image to invoke in order to produce the required push.

  18. RICK Says:

    sorry canute I think jack nirenstein seems to have a far better understandimg of how we run!
    I believe he is right to say its not possible to push backwards because once the foot is grounded the muscles in the front of the leg are activated so its not possible to use the oppossing muscles to push back unless you bend your knees like at the start of a sprint and drive with the quads.
    i think it is a upwards hop plus gravity and momentum along with driving the knee forwards so that the body moves forwards over the grounded leg, that moves us forwards with a longer faster stride.
    also you will not over stride along as you pull your foot back [paw back] to land just forward of your centre of gravity.

  19. RICK Says:

  20. RICK Says:

    above you will see how to run very fast, this is not pose, its almost the same style as Percy Cerutty taught over 50years ago, [ and runners used for 100’s of years before that]why because its the best way to run, most kenyan runners do lots and lots of drills, high knees, skipping, paw back etc.

  21. Peluko Says:

    Hello again… I continue thinking in the subject (I’m interested in it), and I want to share with you some of my thoughts.

    Yesterday I did some speed work (200m and 400m intervals), and I was paying attention trying to find out if the ‘vertical impulse’ is what I need to train to get more speed. Well, I found that I had enough strength in my legs to sustain a good and fast vertical impulse, but I found two other things that really blocks my ability to sustain greater speeds:

    – First: leg swing. I found that my grounded foot gets far backward due to greater contact length and, most important, due to greater relative speed of my foot to my body, increasing the kinetic energy of my foots (my body is moving forward while my foot is stopped in the ground). So the range of motion of my feets is increased, and also I need to put more energy to first decelerate my backward moving foot and second accelerate it again to reach my body. And as Weyand concludes, airborne time is not increased, so I have to do this movement in the same time, applying greater forces and accelerations.

    – Second: balance. Range of motion increases, making more difficult to keep balance. During ground contact I recover the lost balance. But shortening the contact time makes this balance recovering more difficult.

    So to improve speed, at least for me, I need to train this two skills. For now my eccentric contraction skills are sufficient. I think that for balance, some Pose drills will work fine (pose stance, cos, forward cos), but I don’t know which drills to do to improve leg swing.

    And for the propulsive question, I think that wind resistance is not significant compared to other forces that put braking on the motion. For example, our legs are a lot more complex than wheels and leg movement can generate unbalancing inertial forces which summed up could be possitive or negative. But this is another subject. I’ll keep reading on your debate with Simbil.

  22. canute1 Says:

    Rick,
    I agree with you and Jack Nirenstein that the most important forces when running at constant speed are the vertical forces that get us airborne. However force plate data (eg Cavagh and Lafortune, Journal of Biomechanics, 1980; see also the ullustrative diagram in Dancing with the Devil, Part 1, listed in the side panel of my blog) shows clear evidence of a backward directed ground reaction from in early stance (braking forces) and a forward directed ground reaction force in late stance that approximately balances the braking forces. The forward directed ground reaction force must be generated by a backwards push of the foot (Newtons’ third law: action and reaction are equal and opposite). I believe that the backwards push is mainly generated by muscle action. though when running into a head wind, perhaps the wind provides part of the backwards push.

    I note that ‘paw back’ might reduce braking forces and it would be interesting to see force plate data from an elite runner. However, as far as I can tell from video images taking from the side, even elite runners appear to have some braking force and I think this will be overcome by an oppositley directed push back against the ground.

  23. canute1 Says:

    Peluko,
    Thank you. I too find it interesting to discuss questions of technique with runners who examine the relationship between their own experience of running and the findings of research studies. Because of the complexity of the human body it is not easy to obtain definite answers, but I do believe that we can improve our performance by improving our running efficiency. Physiologist Andrew Jones reports that Paula Radcliffe improved her performance over a period of 15 years mainly by improving running efficiency rather than aerobic capacity.

    In response to your comment: Although Weyand’s data indicates that greater vertical force is not associated with significantly greater airborne time (when comparing faster with slower runners at their top speed), the data does confirm that greater force is associated with shorter time on stance. The problem you are describing sounds to be related to spending too long on stance. As I have discussed previously, I think the two things that are necessary to decrease time on stance are increasing the eccentric strength of hams, quads and calf muscles and developing go neuromuscular coordination. I think it is likely that Pose drills will improve neuromuscular coordination. Maybe it is also useful to focus on the direction your foot travels after leaving the ground. If you are focusing on pulling the foot towards your buttocks, maybe this will result in a greater proportion of the push being directed in vertical direction, and also result in a smaller range of motion with your foot landing only a small distance in front of the centre of mass. Then there will be less braking force and less need to push backwards in late stance, so the final result will be a shorter time on stance. There will also be a smaller amount of unbalancing.
    Maybe it would be useful to get a video recording of yourself taken from the side so that you can measure the time on stance, and also examine the direction your foot travels when it leaves the ground.

  24. Jon Port Says:

    Hi Canute,
    If you have time, may you give some thoughts…

    Canute wrote: “I believe that the energy for overcoming braking is largely provided by the muscles that produce a backwards push in late stance thereby generating a forward ground reaction force.”

    Thanks for that statement, I’ve been trying to pin you down on the thoughts of what overcomes braking in early stance for a while.

    Once our centre of mass passes forward of our point of support out bodyweight is in rapid decline. If that is the case, how can you push something that isn’t there to push. You can push (and in fact I believe the impulse to do so) only really happens at mid-stance when COM of the body is directly over the grounded foot. Obviously this will only take your body upwards but not accelerate it forwards.

    Once your bodyweight is gone, there is nothing to push!

    More to the point you would expect (from your statement) that in order to increase your speed during a run that you would have to push harder. But momentum is taking your COG past your support foot even more quickly as you accelerate up to a faster speed. How can we push something harder that is gone more quickly and yet also accelerate.

    Of course it is feasible to push while bodyweight is fully loaded, but that will mainly take your body up not forward.

    So the paradox of the push is that you have to push harder to move faster, but you can only push something that is there (bodyweight in this case), but that can only really happen when the body COM is over the support point which would show as a large vertical oscillation of the COM. But then the worlds best runners show the opposite – a minimal vertical oscillation of the COM. So how do you push that that is not there and minimise vertical oscillation?

    I will keep the extensor paradox out of the discussion, therefore leaving only the calves and ankles to provide push. I guess those with longer feet are the fast runners, right? (that’s just a joke)

    Thanks for your time, and hope you are fit and well and enjoying your running. If you are ever over in Manchester give me a shout and we could go for an easy run.

  25. canute1 Says:

    Jon,
    Thanks for your comment. In my response to Peluko, I stated I consider that push backwards is the main compensation for braking forces when running at constant speed. However, when at constant speed, the work done against braking forces is only a minor part of the work that is required.

    The major task is getting airborne. This requires a strong vertical push against gravity. The force plate data confirms that the vertical forces involved in getting airborne are several times greater than the horizontal forces. Nonetheless, the average value of the backwards directed braking forces is approximately equal to the forward directed horizontal forces due to push backwards in the second half of stance, consistent with my claim that backwards push balances braking .

    I agree with you that the horizontal pushing forces are relatively weak. For a runner who avoids over-striding and maintains a short time on stance, the braking forces and the compensatory backwards push forces should be very small.

    If we want to run very fast, we need to put our efforts into ensuring a strong vertical push. This requires eccentric contraction strength to ensure we capture as much gravitational energy as possible as elastic energy, and good neuromuscular coordination. A well coordinated strong push will achieve a short time on stance, a longer stride and a higher cadence (as illustrated by the data from Weyand). It is of course important to note that short time on stance necessary demands a large vertical push and hence creates risk of repetitive strain injury to the foot when running long distances. That is why I think that when running long distances it is important to relax the ankle even if this does result in a slightly weaker vertical push.

    All of the above discussion refers to maintaining a constant speed (in the absence of wind resistance). When we accelerate, I think the horizontal push plays a somewhat bigger role. A sprinter uses blocks to maximize the push. I suspect that a long distance runner also pushes back against the ground, but the question of how a long distance runner accelerates is much less important than how he/she maintains a high speed. The recently reported study by Fletcher, Dunn and Romanov suggests that gravitational torque is important in the acceleration phase. I suspect that gravitational torque does promote a more rapid acceleration by promoting reflex muscle contraction, but ultimately, the net work is done by the muscles, because the centre of mass actually rises and work must be done against gravity (at least when we start from a crouched start)

    Thanks for the invitation to contact you when I am next in Manchester. It would be good to go for a run together.

  26. RICK Says:

    http://runwitharthurlydiard.blogspot.com/2009/10/ryan-hall-in-best-shape-ever-for-new.html
    Dear Canute, check out the photo of ryan hall at high speed { about 13 mph]
    He is is moving at 13 mph yet his foot is at zero mph, i think its only possible for a upward jump but due to momentum of his body moving over his cog this will show up as a backwards push on a pressure plate!

  27. canute1 Says:

    Rick
    Thanks for your comment. I did not mean to imply that there has to be a specific muscle contraction at the time of the back wards push. When the leg is angled backwards and down, as it is for the brief period after mid-stance, there will be a backwards push due to forces exerted along the length of the leg. Elastic energy stored at footfall could produce the required backwards push without requiring an active muscle contraction at that instant.

  28. Jon Port Says:

    Thank your for your reply Canute. I think we can agree on somethings in there for sure. I thought your last comment to Rick was bang on the money. Pull (aka Pose) and push co-exist due to Newton’s 3rd.

    I use lots of plyometric type hop and jump exercises with my clients. Here are some examples of me performing them:

    The most important thing is that I am not focussed on any form of push. I am focussed on lifting with the pelvis with a rapid pull of the foot. Whatever opposite push happens – just happens, but my conscious hierarchy of thought is on movement of the pelvis and feet upwards rather than conscious muscle contraction. If you focus on pushing to do this you lose the elastic return energy which requires should ground time. Also if you focus on push with the foot/calves you will plantarflex the foot which cause achilles tendon strain very quickly. The key is to remain like a coiled spring but be flexible.

  29. RICK Says:

    thanks canute for reply,
    one thing you have not talked about is the use of the upper body in running.
    i have found that using the arms adds no extra speed to my running BUT using shoulder rotation does!
    watch the video of Wanjiru [above] and you can see his shoulders twist which help generate more forward drive in the opposing knee.
    when running to college with a ruksack on my back i could not use my arms but by twisting [ rotating my shoulder a found i could increase speed!
    jack nirenstein http://books.google.co.uk/books?id=t8NoK2cxQioC&printsec=frontcover&dq=just+undo+it&hl=en#v=onepage&q=&f=false recommends using a shoulder twist, as well as most sprint coaches, also look at any fast kenyan runnerhttp://runwitharthurlydiard.blogspot.com/2009/11/running-in-kenya.html they all use use the twist [it works] :]

  30. Peluko Says:

    Hello again. My last (wrong) thoughs in your gravity post, served me to review and refresh my physics knowledges, and I’ve applied them to my last thoughts expressed above about where goes my energy during fast runs and what I have to do to not get exhausted too quickly.

    ‘The major task is getting airborne. This requires a strong vertical push against gravity. The force plate data confirms that the vertical forces involved in getting airborne are several times greater than the horizontal forces.’

    Lets do some math. As Weyand’s study measures, airborne time is almost the same for fast and slow runners, around Ta = 0.128s. If the flying time is the same, the height of the flight must be the same. This height of the flight and the initial vertical speed can be calculated using g = 9.8m/s ant time = Ta/2 = 0.64s. This height is 0.04m, and the initial speed in which the runner gets airborne is S0 = 0.6272 m/s. The work done is the change of kinetic energy in the body, related to vertical movement. The initial kinetic energy is 0, and the final kinetic energy is (m*S0*S0)/2 = m*0.1967 joules, where ‘m’ is the mass of the body. This work is constant for fast and slow runners, and the energy implied in this work is also m*0.1967 joules.

    This work or energy is the same for slow and fast runners. Over time, the variable here is the stride frequency. Slow runners, at 1.8 strides per second (or 3.6 steps per second), spent m*0.1967*3.6 joules per second or power watts. Faster runners run at 2.4 strides per second or 4.8 steps per second, so the energy implied is m*0.1967*4.8. So the variation in energy consumed comes from stride frequency, and varies 1.33x times in the 1.79x times speed variation, but the 1.33x variation comes from the stride frequency, not from the increasing forces.

    As I’ve said before, I don’t perceive that the quick push or pull to get airborne is where I spend most of my energies. As you say, maybe I spend too much time in stance, but for sure, when I’m running fast I spent much less time than when I’m running slow. An when I run fast I get out of fuel quickly. And I can get to run slow with short contact time (is something like jumping a rope). Of course, if I have a short contact time, I have to do more force to get airborne, but the impulse (force by time) remains the same, so I get airborne for the same height and for the same time. When I do this, without increasing my speed, I’m also increasing my stride frequency, just as when I run fast. But when I do this, moving slow, I don’t perceive the same level of fatigue, just a bit more because I’m increasing my stride frequency and jumping more ‘intensely’ (more force during less time, I need to put more power).

    Of course, I’ve not measured my ground contact times, are just impressions, but I think they are representative. If I have the time, I’ll try to measure it by filming.

    So, as I’ve said before, I’m thinking that maybe what I need to train is not the quick jump. So I’ve been doing some simple math. And I’ve concluded that the energy required to move your feet on the air, to bring the back foot to contact position in front, is almost directly proportional to speed, independtly of short or long contact times. When the foot is grounded, the foot speed is 0 m/s. But the body is moving at an speed (S). To bring the foot to the front contact position, you need to accelerate it to catch the body. When the foot takes contact at 0 speed, the body continues moving. The next time this foot takes contact is determined by stride frequency or stride duration. Lets call Ts to the aerial time a foot is on the air (swing time), and Tc to the time a foot is on the ground (contact time), and T0 to the moment in which a foot touches the ground, T1 the next moment the foot touches the ground, etc. Lets also call D0 to the linear position of the foot at T0, D1 to the linear position of the foot at T1, etc. The foot touches the ground at T0, T1, T2… with 0 speed. So T1-T0 = Ts+Tc. Between T0 and T1 the body has moved (S*(Ts+Tc)) meters, D0 = 0, D1 = (S*(Ts+Tc)), D2 = 2*D1. After contact time Tc, the grounded foot has to reach the next contact position in Ts seconds. To do this it has to move (S*(Ts+Tc)) meters in Ts time. To do so with an initial speed of 0 m/s, it has to accelerate. The distance formula respect to acceleration is distance = acceleration*time*time. Lets call A to the acceleration. The we have (S*(Ts+Tc)) = A*Ts*Ts, so A = S*(Ts+Tc)/(Ts*Ts).

    Applying all of this to the data in the Weyand’s study, we found that Ts is 0.350s, constant to all runners at top speed, fast and slow runners. S (speed) vary from 6.2 to 11.1 m/s, and Tc moves between 0.149s and 0.080s. We have:

    A = S*(Tc+0.350)/0.1225

    So for 6.2 m/s with a contact time of 0.149s we need a foot acceleration of 25.255 m/(s*s). For 11.1 m/s with a contact time of 0.080s, foot acceleration is 38,963 m/(s*s). The speed increment is 1.79x. The acceleration increment is an 1.54x times increment. And this increment doesn’t only apply to feet. As it is proportional, it applies to the whole leg. For example, the knee acceleration will be smaller, but the knee acceleration at 11.1 m/s will be 1.54x times the knee acceleration at 6.2 m/s. This will be in the ideal case of a leg like a perfect pendulum, but I think this will be a good approximation. I think that some of that proportion also applies to the arms and some part of the torso.

    Force increment is proportional to acceleration increment, and as the times doesn’t vary (Ts is constant) also work and energy spent are also proportional to this acceleration increment. So we find that the energy spent in getting airborne is the same for fast and slow runners, but the energy spent in moving our limbs is proportional to the formula S*(Tc+Ts)/Ts, where Ts is constant. The 1.33x stride frequency variation applies to both, airborne and limb movement, so it can be left out of the comparison.

    Other thing is the peak power, which depends on the time during which forces are active. For getting airborne the time is shorter for fast runners, this time is contact time, which is 0.080s for runners at 11.1 m/s and 0.149s for runners at 6.2m/s. For fast runners the power our muscles have to generate is m*0.1967/0.080 = m*2.459 watts, and for slow runners is m*0.1967/0.149 = m*1.32 watts. For moving our limbs, the peak power also increases, but as time is constant, it only varies proportional to the increasing in forces. And I think that at a long run, finally what accounts the more is the total energy, not peak powers. Also this increasing peak powers for getting airborne take much energy from the elastic recoil, so maybe aren’t very dependent on muscular effort.

    I’m not completely sure that all this calculations are accurate, but for me and my actual knowledge about this item, seems ok.

  31. Peluko Says:

    Hello again.
    I’ve found another conclusion: for fast running, short contact time is mandatory, long contact time is impossible, even if the runner tries to force it. Let me explain.
    Weyand’s study data shows that for runners at 11.1 m/s, contact time is 0.080s and for running at 6.2 m/s, contact time is 0.149s. During this contact time, the body travels 0.888m and 0.924m (contact length). This is a 1.04x times variation for a 1.79x variation in speed. Virtually it’s no variation.
    This contact length is a physiological limitation of our body. If at a faster speed we try to keep a longer contact time, as our body moves forward the angle between our leg and the ground must increase, and the contact only could be maintained if we let our body to fall more to the ground, which implies more vertical movement, which is a waste of energy.
    So for any given speed, the maximal contact time is fixed, only can vary with variations in vertical displacement. On the other side, the minimal contact time is not fixed, so runnig slow with a short contact time is possible, but running fast with a long contact time is not possible.

    There must be an optimal relation between the vertical displacement, contact time, elastic recoil, muscle effort, etc.

  32. Peluko Says:

    Hello again.

    Well, it’s evident that it’s a long time since I did my last physics exercises…

    I’ve made a mistake in the distance formula respect acceleration. I’ve written that the formula is distance = acceleration*time*time. Correct formula is distance = acceleration*time*time/2. This changes some of the numbers, but the conclusions are the same, as the proportions are identical. So the actualized data is:

    The height of the vertical displacement is 0.02m instead of 0.04m.
    The formula to calculate foot acceleration is A = 2S*(Ts+Tc)/(Ts*Ts).
    The formula with the Weyand’s data values is A = 2*S*(Tc+0.350)/0.1225, so the accelerations of the foot are double of the calculated.

    But this doesn’t change the proportions between fast and slow runners, so the conclusions are exactly the same.

  33. Simon (simbil) Says:

    Canute,
    Thanks for the mention – looking forward to your next article addressing angular motion in running. By the way I am based in Guildford rather than Sheffield, but no problem.

    Peluko,
    I was interested to read your comments about the energy cost of running as it increases in speed.
    I think one key point that has not been mentioned so far is the efficiency of muscular elastic return. In some studies of ‘blade runner’ (the guy with the carbon fibre lower legs) vs elite runners, they found that an elite runner can store and reuse 50-60% of the energy from footfall to get airborne again.
    Form other studies it seems that the storage of energy is time sensitive and their is a particular frequency or running cadence that gets the best return – this is thought to be around 180 steps per minute, but you may want to do your own research on that figure to check.
    When a sprinter goes into an all out sprint, the steps per minute typically goes above 240. I suspect that the return from the muscular elastic store will decrease at these high frequencies and so the cost of getting airborne in sprinting will be significantly higher than the cost when distance running.
    However, I agree with you that the repositioning of limbs is a significant cost. When I watch Bolt run, it seems to be what separates him from his competitors – he recovers his foot under his hips whilst his competitors struggle with their foot hooking up behind their hips.
    In summary, to sprint well you need to recover the foot well and be able to handle short sharp pushing off the ground via strength, technique and plyometric conditioning.

  34. canute1 Says:

    Peluko, Thank you for those comments.

    If I understand you correctly you conclude that ‘maybe we aren’t very dependent on muscular effort’ exerted in getting airborne.
    I have not yet had a chance to go through all the steps of your calculations in detail and therefore do not want to draw definite conclusions about your argument until I have had a chance to examine both the underlying hypotheses implied in your calculations and the also the mathematical steps in detail. However after first inspection of the three most recent comments, I have several initial remarks:

    1) you argue that short time on stance is essential when running fast, in order to avoid the inefficient loss of height associated with a large rotation. I agree entirely. I would note that if we have a short time on stance, the average vertical ground reaction force must be large, so that the upwards impulse during each gait cycle matches the downward impulse due to the weight of the body. Therefore large push against the ground is essential if we are to run fast. Of course some of this push can be obtained by elastic recoil. Because muscles and tendons are viscoelastic, it is not easy to compute what fraction of energy can be recovered by elastic recoil, but the observational data indicate that not much more than half of the gravitational potential energy can be captured and recovered, so running fast is hard work, and I believe most of that work is done in getting airborne.

    2)You state that swing time is the same for fast and slow runners. Swing time is twice aerial time + stance time. Aerial time is approximately the same and stance time is less, so swing time would be expected to be less in the faster runners, though Weyand reports that it is not significantly different. If it is true that time on stance is less (which is strongly supported by the evidence) the fact that Weyand reported no significant difference in either aerial time or swing time must mean that the study did not have sufficient statistical power to detect a real difference in swing time and/or aerial time. Therefore we should be careful about drawing conclusion from the measurements of aerial time and swing time. The measurement of ground reaction force and step frequency are probably more reliable.

    3)When describing your own subjective effort when running slowly compared with running faster, you repeat the claim that airborne time is constant at different speeds. That is not true for different speeds in the same runner. While Weyand;’s data show that at maximum speed different runners have very similar airborne times, the data also show that as speed increases from sub-maximal to maximal speed there is a substantial increase in airborne time. As speed increases from 3 to 5 m/s, aerial time increases from 110 milliseconds to 150 milliseconds. The work done in raising a body of mass M to height h against gravity is given by M*g*h where g is gravitational acceleration. The height h is given by ½*g *t*t where t is half of aerial time. Therefore the work done per step is ½*M* g*g*t*t. M and g are constant, so an almost 40% increase in t will be associated with nearly twice as much work done against gravity per step (while speed increases from 3 to 5 m/s). If cadence also increases there will be an even greater increase in work against gravity per unit time.

    4) If we accept Weyand’s data showing that aerial time at maximum speed does not differ appreciably between fast and slow runners, but that step frequency is 33 percent greater in the faster runners, the amount of work done against gravity per step will be similar, but the amount of work per unit time is 33 percent greater in the faster runners. However, stride length is almost 70 per cent longer in faster runners so the energy cost of getting airborne will actually be less per Km in the faster runners. Thus Weyand’s data indicate that is more efficient to exert a greater push

    When I have more time I will go through your calculations in greater detail, but at this stage, I continue to believe that the evidence indicates that it is necessary to exert strong vertical forces when running fast and that a major portion of the energy cost of running is spent on getting airborne.

  35. Peluko Says:

    Hello Canute.

    Please, I think that is necessary that someone review my calculations and assumptions because I’m disappointed about the little that I remember about my other time loved physics lessons.

    I want to comment on some of your appointments.

    —-
    1) you argue that short time on stance is essential when running fast, in order to avoid the inefficient loss of height associated with a large rotation. I agree entirely. I would note that if we have a short time on stance, the average vertical ground reaction force must be large, so that the upwards impulse during each gait cycle matches the downward impulse due to the weight of the body. Therefore large push against the ground is essential if we are to run fast. Of course some of this push can be obtained by elastic recoil. Because muscles and tendons are viscoelastic, it is not easy to compute what fraction of energy can be recovered by elastic recoil, but the observational data indicate that not much more than half of the gravitational potential energy can be captured and recovered, so running fast is hard work, and I believe most of that work is done in getting airborne.

    Here we have to clarify what is hard work. It’s energy (work) or peak power? For the total energy consumed in getting airborne, what accounts is the height of the vertical movement, not the contact time. For peak power, the contact time must be taken into account. But, as a long distance runner… which is the more important factor? Energy or peak power?

    —-
    2)You state that swing time is the same for fast and slow runners. Swing time is twice aerial time + stance time. Aerial time is approximately the same and stance time is less, so swing time would be expected to be less in the faster runners, though Weyand reports that it is not significantly different. If it is true that time on stance is less (which is strongly supported by the evidence) the fact that Weyand reported no significant difference in either aerial time or swing time must mean that the study did not have sufficient statistical power to detect a real difference in swing time and/or aerial time. Therefore we should be careful about drawing conclusion from the measurements of aerial time and swing time. The measurement of ground reaction force and step frequency are probably more reliable.
    —-
    I think I don’t understand your arguments here. Weyand’s data concludes clearly that at top speed swing times are almost the same for all the runners. If we can’t trust on aerial and contact times measurements, I think that we can throw this study to the recycle bin, because it only concludes that fast runners must do more force to the ground… but does not clarify anything about how to do this force. For example, the most important is the time of the force. If we apply the greater force during more time, this indicates more vertical movement, which is a bad thing which involves longer aerial time and lower cadence.

    —-
    3)When describing your own subjective effort when running slowly compared with running faster, you repeat the claim that airborne time is constant at different speeds. That is not true for different speeds in the same runner. While Weyand;’s data show that at maximum speed different runners have very similar airborne times, the data also show that as speed increases from sub-maximal to maximal speed there is a substantial increase in airborne time. As speed increases from 3 to 5 m/s, aerial time increases from 110 milliseconds to 150 milliseconds. The work done in raising a body of mass M to height h against gravity is given by Mº*g*h where g is gravitational acceleration. The height h is given by ½*g *t*t where t is half of aerial time. Therefore the work done per step is ½*M* g*g*t*t. M and g are constant, so an almost 40% increase in t will be associated with nearly twice as much work done against gravity per step (while speed increases from 3 to 5 m/s). If cadence also increases there will be an even greater increase in work against gravity per unit time.
    —-

    First here, if cadence increases, work increases per unit of time, but this increment is exactly the same for getting airborne and for leg swing, so cadence can be left out of the comparison.
    I can’t find the data you show about the speed increment from submaximal to maximal speed (now I don’t have the time to reread the whole study). I think this could be interesting data, because if you increase your aerial time you are increasing your energy waste on vertical displacement, and also you are competing against cadence. More aerial time is directly related to less cadence, so for cadence incrementation your contact times must compensate greatly for the aerial times. So this would be valuable data to find some evidence about how is ‘efficient running’ from the mechanical point of view.

    —-
    4) If we accept Weyand’s data showing that aerial time at maximum speed does not differ appreciably between fast and slow runners, but that step frequency is 33 percent greater in the faster runners, the amount of work done against gravity per step will be similar, but the amount of work per unit time is 33 percent greater in the faster runners. However, stride length is almost 70 per cent longer in faster runners so the energy cost of getting airborne will actually be less per Km in the faster runners. Thus Weyand’s data indicate that is more efficient to exert a greater push.
    —-
    Well, as I’ve said before, step frequency applies to the work done against gravity and also to the work done to move our legs, so it could be left out of the comparison. And I think that the efficiency is not to exert a greater push, efficiency is to expend the minimum energy to travel de distance.

    I want to do one more calculation on this data, but now I have to review some other lessons I forgot to be able to do it. Let me explain.

    The energy to get airborne is M*g*h, where M is the mass of the body. As we calculated before, we can calculate that h = 0.02m, and g is 9.8m/(s*s). So the energy to get airborne is M*0.196 joules.
    If we simplify our model of a runner, we could consider that the foot has a mass m, and we reject the leg mass, as if the foot were connected to the body only by a thread. Let’s simplify a lot more and assume that the foot does not fly. It simply stop movement (speed = 0) at T0 at point D0, and after contact time (Tc) it moves to the next point D1 in the swing time interval (Ts). The difference between D1 and D0 is dependent on speed, and for the data we have from Weyand, for 11.1 m/s, the distance is 4.625m, and for 6.2 m/s the distance is 3.44. Distance travelled by an object is directly implicated in the work done and in the energy spent, and the distance travelled by our simplified foot is hundreds of times greater than the distance travelled vertically by the body (231.25x for 11.1 m/s and 172x for 6.2 m/s). I have to do this calculation to find how this distance is affected and confirm and correct this multiplicative factor… any volunteer?

  36. Peluko Says:

    Hello again. I’ve completed some calculations, based on the simplification of the model I’ve explained before.

    The simplified model is simply to consider the foot as a mass that has to move linearly from contact point 0 to contact point 1 to contact point 2, etc. In this movement, the initial speed is 0, the foot accelerates, and then decelerates to reach next point with a final speed which is also 0. We know the distance that need to be travelled (distance the body travels during stride duration) and the time in which it has to travel (swing time, or stride duration minus contact duration).
    We don’t know how are the accelerations and decelerations of the feet or the forces involved. But the sum of all produces a work and this work is energy consumed. By doing some calculations (too tedious to express here in ascii… and in english!!), I’ve found that the work done to move an object a distance d during a time t, with an initial and final speed of 0m/s is W=k*m*d*d/t*t, where m is the mass of the object. ‘k’ is a number dependent on the intensity of the forces and the duration of the application of the forces. When the forces tend to infinitum and the time tends to 0, the limit of k is 1. So the minimum work is done by hipotetical forces of infinitum and hipotetical duration of 0. All other realistic applications results in k > 1.

    To summarize:

    We have:
    Ta = aerial time
    Tc = contact time
    Ts = swing time = 2*Ta + Tc
    Tt = stride time = 2*Ta + 2*Tc = Ts + Tc
    Mb = mass of the body
    Wa = work of getting airborne
    g = gravity = 9.8 m/(s*s)

    Wa = (1/2)*Mb*g^2*(Ta/2)^2 = 12*Mb*Ta^2

    For the leg swing we have:

    Ws = work on one idealized leg swing
    Mf = mass of the idealized foot
    V = speed of the runner
    D = distance traveled by the runner during one stride = V*Tt

    Ws = Mf*D^2/Ts^2 = Mf*(V*Tt)^2/Ts^2 = Mf*V^2*Tt^2/Ts^2 = Mf*V^2*(Tt/Ts)^2

    So the work done in swinging legs is directly proportional to speed squared and to the relation between stride duration and leg swing duration, so that when contact time decreases, the relation tends to 1, and when contact time increases, the energy consumed also increases.

    Well, for one stride, we have one leg swing and one leg push for each leg. Now lets put some Weyand’s number in it.

    For the fast runner we have:
    Ta = 0.128s
    Tc = 0.080s
    Ts = 0.336s
    Tt = 0.416s
    V = 11.1m/s

    The energy spent for push is Waf = 12*Mb*(0.128)^2 = 0.1967*Mb joules
    The energy spent for leg swing is Wsf = Mv*11.1^2*(0.416/0.336)^2 = 188.85*Mf joules

    For the slow runner we have:

    Ta = 0.128s
    Tc = 0.140s
    Ts = 0.405s
    Tt = 0.554s
    V = 6.2m/s

    The energy spent for push is Was = 12*Mb*(0.128)^2 = 0.1967*Mb joules
    The energy spent for leg swing is Wss = Mv*6.2^2*(0.554/0.405)^2 = 71.92*Mf joules

    For example, if we assume that our runner has a mass of 80Kg and our idealized foot has a mass of 1Kg (which in reality should be a lot more, due to the mass of the leg and other implications), we have:
    Waf = 15.736 joules
    Wsf = 188.85 joules
    Was = 15.736 joules
    Wss = 71.92 joules

    The energy implied in moving the foot is several times greater than the needed to get airborne. And this energy increases more rapidly than speed, 2.62x times within the 1.79x times the speed.

  37. canute1 Says:

    Peluko,Thanks for your further comments. I will look at your detailed calculations when I have time. Meanwhile, here are a few comments on your comments on my comments on your comments on my comments on your comments on my posting. (This discussion is is like a good game of tennis.)
    1) You ask whether it is energy or power that is important for a long distance runner. Clearly for a long distance runner, it is important to adopt the style that is fairly efficient (that is, consumes fairly little energy per Km). However, if you want to win a race, you must also run fairly fast – so you need to balance economy with speed. I believe that the most efficient way to do this is to employ a fairly short time on stance. If I simply want to cover the maximum distance possible without concern about speed, I would walk (with no aerial time at all). When running a marathon, I would aim for a time on stance that that is around 150millisec but when sprinting, I try to make time on stance as short as possible, preferably less than 100 millisec.
    2) You say you do not understand why I stated that if stance time decreases it is impossible for both aerial time and swing time to stay the same. The reason is that swing time consists of two aerial times + stance time. So if stance is shorter (which Weyand reports to be the case) then either aerial time and/or swing time must change. The fact that Weyand did not observe statistically significant differences between fast and slow runners at maximum speed for either swing time or aerial time indicate that he did not have the statistical power to detect the effect. The data do show very small increases in both aerial time and swing time but these are trivial. However, as I mentioned in my previous response to you, the findings about aerial time and swing time are not the most important observations in this paper. The really important observations are greater ground reaction force, shorter time on stance and higher step frequency in the faster runners. These finding appear to be statistically robust.
    3) the increase in aerial time from 110 milliseconds to 150 milliseconds as speed increases from 3 to 5 m/s is shown in figure 2 of Weyand’s paper.
    4) I do not agree that the goal is to expend the minimum energy to travel the distance. As discussed in point 1) , a competitive runner needs to achieve both speed and economy. If economy was the only goal, it would be better to walk.

  38. Peluko Says:

    Hi Canute. Communication is this, a fluid interchange of ideas.

    I agree with your view of efficiency and racing.

    I didn’t pay attention to the figure with the representative runner’s data. But I’ve found that if I throw the numbers stated there into my calculations, the energy needed to swing legs continues being several times the energy needed to get airborne. With this calculations, there is a wide range of aerial and contact times into which the swing leg energy is greater than the pushing energy. And respect to speed the swing energy increases faster than airborne needed energy.

    I would appreciate if you review my calculations. To me it looks correct, but I’ve noticed that I’ve forgotten much more physics and math than I would like… maybe that’s the cause that I’ve been enjoying those calculations too much.

  39. canute1 Says:

    Peluko, I have started to check your calculations, but this is a time consuming process. Here is a list of points on which I disagree with your post on 28th November:

    Peak height = 0.02m (You yourself discovered that you had omitted the factor ½) .

    Acceleration of the foot is approximately twice the value you calculated because the foot must accelerate for approximately half of swing time and decelerate for the other half. However these errors do not affect you conclusion that foot acceleration will be 1.54 times greater in the faster runners than the slow runners.

    However I think your calculation of the relative power required to accelerate the foot forwards is wrong because the distance covered during swing is 1.79 times greater in the faster runner (assuming same swing time). Assuming uniform acceleration in the first half of swing, the work done accelerating the foot is 1.54* 1.79 = 2.76 times greater for the faster than for slow runner. Because swing time is same for fast and slow runners, the muscle power required to swing the foot forwards is 2.76 times greater in the faster runner.

    However, this calculation tells us nothing about the ratio of power required to lift the body compared with power required to accelerate because we have made no attempt to calculate the mass that is being accelerated.

    The average energy consumed per second will be greater in the faster runner. However what matters when we calculate efficiency is the energy consumed per Km. The energy consumed per Km in lifting the body against gravity will be less by a factor of 1/1.69 (=.59) for the faster runner because the work against gravity per step is the same but number of steps per Km is less by 1/1.69. However the energy consumed in accelerating the leg and foot forwards will be greater in the faster runner because the amount of work per stride is 2.756 times greater yet the number of strides is less by a factor of 0.59. Thus the amount of work done accelerating the foot forwards is 1.63 times greater per Km for the faster runner.

    In summary, for each Km, the faster runner uses only 0.59 time as much energy lifting the body against gravity as the slow runner, but 1.63 times as much energy swinging the foot forwards compared with the slower runner. However because we do not know the absolute value of the energy required to swing the foot forwards we cannot calculate whether the total energy cost per Km is different between the two runners.

    Other studies have shown that the energy cost /km is similar for all running speeds. If we assume that the energy cost per Km is the same for fast and slow runners, you can easily show that the energy cost of lifting the body is approximately 1.5 times as great as the cost of swinging the leg. However, this is only a very rough approximation.

    I think that the important conclusions from the Weyand study are that faster runners exert a stronger push against the ground, have a shorter time on stance and a longer stride length. They also have a faster cadence. The energy cost per Km spent in getting airborne is less than for slower runner because the faster runner takes fewer steps per Km. However, the energy saving in getting airborne is offset by greater cost of accelerating the foot forwards, because the foot must accelerate a lot faster in the faster runner. The faster runner needs to be stronger for two reasons: he/she must exerts a greater push against the ground and also must be able to accelerate the swinging leg much faster.

    I disagree with your conclusion that ‘we are not dependent on muscular effort.’

  40. Peluko Says:

    Hello Canute.

    My post on 28th November is only the beginning of the calculations and corrections that ends with the post on 30th November, which I think is more accurate and revealing. I’ve done it this way because as I’ve been remembering things I’ve been adding to the calculations.

    Canute: “this calculation tells us nothing about the ratio of power required to lift the body compared with power required to accelerate because we have made no attempt to calculate the mass that is being accelerated.”

    The last calculations on the 30th Nov post shows that making a very conservative assumption about the mass of the feet, the energy needed for leg swing is several times the energy needed to lift the body, and for sure that real feet mass numbers must be greater. Maybe with real numbers the ratio easily could be ten to one for moderate speeds.
    If you put in the distance factor (energy consumed per kilometer), the total number of steps needed for the fast runner is less than the number of steps needed for the slow runner, but the ratio between body lift and swing leg energy remains the same as if calculated for only one step.

    The total energy per km could be calculated based on the total energy per step. When we increase the speed, we decrease the number of steps per km, but increase the energy needed for every speed. This could be a relation that gives the same energy per km independently of speed, but this doesn’t affect to the ratio between body lift and swing leg energy.

    Canute: “I think that the important conclusions from the Weyand study are that faster runners exert a stronger push against the ground, have a shorter time on stance and a longer stride length. They also have a faster cadence. The energy cost per Km spent in getting airborne is less than for slower runner because the faster runner takes fewer steps per Km. However, the energy saving in getting airborne is offset by greater cost of accelerating the foot forwards, because the foot must accelerate a lot faster in the faster runner. The faster runner needs to be stronger for two reasons: he/she must exerts a greater push against the ground and also must be able to accelerate the swinging leg much faster.”

    Well, it’s ok. But all of this discussion started because my impressions that were I spent more work is on swinging legs, so instead of doing plyometric work on the muscles involved in lifting the body, I’ll prefer to spend more training in the leg swing muscles. The ratio of work done with this muscles seems to be several times the work done with the others. When I run fast, I feel that I can do easily a short contact time, but to maintain speed I need to be able to get support points in place at time, and that’s the hard part. If my leg swing is not fast enough, I never could get the step length needed to get the speed. If my leg swing’s muscles aren’t trained enough, they will exhaust soon and I couln’t maintain the speed much time. This calculations confirm this. To translate it into Pose push/pull terminology, the muscles involved in the push do only a small part of the work compared to the muscles involved in the pull.

    I don’t say that Weyand’s study is wrong. I only say that it only shows a little part of the whole picture. Only shows the part that refers to contact with the ground, but don’t show anything on the airborne part, where the body continues working. If we take into account the whole picture, the conclusions could change.

  41. Stefan IJmker Says:

    Some suggestions:

    -experimental data shows that swinging the leg costs up to 25% of the energy spended during a running cycle (Marsh et al. Science 303,80-83, 2004)
    -most energy is needed for vertical oscillation. The muscles need to support the falling body in order not to collapse (load tendons for elastic recoil, and then some form of push). More oscillation (10 cm vs 6 cm) coincided with poorer performance: (14:50 vs 16:30 average 5k time) VO2 max was te same in both groups 68 ml.kg(-1).min(-1) (Zatsiorsky.Kinetics of human movement, 2002, p.527). Cavagna and Kaneko Journal of Physiology 268:467-481) show that energy needed to swing limbs increases with speed, but remains far less than 50% of total work cone.
    – more energy cost at higher speeds: less time to do work, shorter ground contacts
    – trying to run with artificial or higher cadence costs more energy (Cavagna research ==> your preferred cadence costs less energy, higher or lower cadence costs more energy, ). the Dallam et al study points to this ==. more energy cost after pose training.
    – Pawback might simply be swing leg retraction, something that happens automatically. Leg comes forward due to elastic recoil ==> posterior muscles are stretched (much more so during higher speeds) ==> recoil ==> deceleration of feed in relation to trunk
    – on force plate the braking force is due to the fact the body is moving forward, still, relatively to the ground. If pawback would work we should be able to measure a negative speed relative to the ground (which I have not seen across). If that is present than action = – reaction ==> more braking force……
    – muscle action during running might well be isometric instead of eccenctric-concentric / very slow v/vmax (see The integrated function of muscles and tendons during locomotion. Roberts TJ.Comp Biochem Physiol A Mol Integr Physiol. 2002 Dec;133(4):1087-99. And his more recent publications of in vivo measurements)

  42. Peluko Says:

    Interesting.

    The Cavagna and Kaneko study indicates that the relation between swing leg energy and vertical oscillation energy is less than in the simplified model explained above. Maybe the key is that the Cavagna’s study includes full vertical movement and here we only dealed with airborne time. But still it shows that the need in vertical energy decreases with increasing speed, and swing leg energy increases.

  43. canute1 Says:

    Peluko,
    The main issue I would raise about your calculation of the work done during the swing is that you have calculated work done to accelerate and stop an isolated 1 Kg object. This is not the situation with the foot when running.

    The torso and head move forwards with almost constant velocity throughout the gait cycle. During the first half of swing, muscle contraction pulls the foot towards the torso. This accelerates the foot forwards while slowing the torso and head, but because torso and head are more massive than the foot, the amount of slowing is quite small. In the second half of swing, the muscles (mainly hip extensors) arrest the forward swing of the foot. This action decelerates the foot, but accelerates the torso and head, thereby recovering much of the momentum that had been lost in the first half of swing. In other words, in the first half of the swing the foot is accelerated by extracting energy from the moving torso and head, but much of this energy debt is repaid in the second half of swing. So the main costs are the metabolic costs in the muscles.
    Therefore I believe your calculation seriously over-estimates the net work done during swing phase.
    The experimental data of Cavanagh and Kaneko (mentioned by Stefan) confirms that the work done in driving the swing is less than the work done in getting airborne.

    Stefan, thank you for your interesting comments.

  44. Peluko Says:

    Hello again.

    ‘The experimental data of Cavanagh and Kaneko (mentioned by Stefan) confirms that the work done in driving the swing is less than the work done in getting airborne’

    The relation between works is more complex. In the model applied to my calculations I only taken into account the airborne energy, but full vertical movement is more than that, the center of mass moves more than that due to leg flexion and body elasticity. And this is not the only work implied in this study. The Cavagna and Kaneko study talks about Wext and Wint. Wint is limb movement, and they conclude that is proportional to the square of the speed, as we calculated here. Intrigued by Wext, I’ve tracked it to the source, this other study referenced by the Cavagna and Kaneko:

    http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1307665/

    (The sources of external work in level walking and running. G A Cavagna, H Thys, and A Zamboni)

    This last study explains how they calculated Wext. Wext is made up from Wv, which is the energy implied in vertical motion, and Wf which is the energy spent due to horizontal speed changes (interesting thing, more on that later). This study concludes that Wv is nearly constant with speed (even for faster speed it decreases), but Wf increases linearly with speed. At 15Km/h we get Wv = Wf.
    This makes clear that the energy spent in vertical movement changes little with speed, the energy ‘Wf’ increases linearly with speed, and the energy spent in limb movement increases with the square of speed. But my assumptions about proportions where wrong.

    And now this Wf. If I’ve understood the study (I want to review it slowly and with an English dictionary), the most energy goes in getting again the speed we loose on each step!! So finally a runner brakes and accelerates with each step, and this is an important part of the energy waste??

  45. canute1 Says:

    I agree entirely that braking leads to slowing and then accelerating on each step and that this is wasteful of energy. (It should be noted that consumption of energy due to braking is separate from the consumption of energy by the muscle actions during the airborne period that promote swing, which is what you were previously arguing was much greater than the energy cost of getting airborne.) I strongly support the proposal that we should minimize wasteful braking.

    Braking can be minimized by spending a short time on stance because when time on stance is short, the foot lands only a short distance in front of the centre of mass, and the horizontal component of the force acting along the length of the leg is relatively small. In order to spend a short time on stance we need to exert a strong vertical push. That is predicted by theory and confirmed by Weyand’s data. So I continue to maintain that the important conclusion from Weyand’s data (and from theory) is that if we want to run fast we need to exert a strong vertical push on the ground. However as I have pointed out in several of my blog postings, a strong vertical push produces large stresses on the foot. Therefore faster speed is purchased at the price of greater risk of injury. The risk can be reduced by relaxing the ankle so that the heel touches the ground.

    An important feature on which we agree is that a strong vertical push does not necessarily mean that greater work is done against gravity. As Weyand’s data shows, the energy cost per Km of getting airborne is actually less at high speed when the vertical push is greatest, though the cost of getting airborne are likely to be a fairly large proportion of the energy cost of running at all speeds

  46. canute1 Says:

    Two small caveats about my comments posted last night.
    1) With regard to minimization of braking, although it helps to have the leg nearer vertical at footfall, if the distribution of forces throughout the stance is constant, the advantage will be approximately canceled out by the greater magnitude of the force directed along the length of the leg. To capture the potential reduction in braking associated with shorter time on stance, it is necessary to distribute the impact so that it rises gradually in early stance and then reaches a fairly sharp peak at mid-stance. This is one of the reasons why I favour a forefoot landing with the ankle sufficiently relaxed to allow the heel to brush the ground at mid-stance (though I do not think this is practical when sprinting)
    2) The statement that the energy cost per Km of getting airborne tends to decrease with increasing speed is true in the upper part of the range of speeds, because aerial time is almost constant and cost per step of getting airborne is approximately constant, but stride length increases, so energy cost per Km of getting airborne will decrease. At slow speeds we need to consider changes in both aerial time and stride length. The airborne time increases appreciably with increasing speed, which will lead to increasing cost of getting airborne on each step. But stride length usually increases even more markedly as speed increases, so I would expect the energy cost per Km of getting airborne to decrease with increasing speed in the low speed range as well as at higher speeds.

  47. Ivan Rivera Says:

    Canute:

    I think that you bash the Pose Method’s theoretical underpinnings a bit unfairly.

    I disagree that you can’t pull yourself up by your own bootstraps. A jump is basically full extension of the body during stance with subsequent flexion of the lower body during flight. Extending the body during stance would bring the GCM (center of mass) as high as possible given the dimensions of the particular person.

    Flexing the lower body during flight means that the GCM is now higher than it was before: the GCM gets averaged out at a higher point, given that postural reconfiguration has brought more of the body’s mass higher.

    This is what we see a lot of in Parkour: the reason a backflip gets you higher than a frontflip is because postural reconfiguration (in particular, body extension at the height of the flight phase) which means that you can raise the GCM substantially more, relative to a frontflip (where there is full flexion at the height of the flight phase).

    The force exerted on the ground isn’t that much bigger–if at all–for the backflip than for the frontflip.

    Internal (postural) reconfiguration to raise the GCM has always been a useful way of getting objects on the ground to fly (also the reason why when cars slide going too fast, they typically come off their wheels on both sides before any other part of the car has hit the ground yet).

    Basically, centrifugal force beats gravity+friction, and when that happens, the car tilts, raising the GCM.

    In other words, I don’t think that a “push” only way to generate substantial–I’ll say again–substantial vertical motion.

    • canute1 Says:

      Ivan,
      Thanks for your comment.

      Internal re-configuration does not elevate the COG, though internal re-configuration might allow the body to clear a barrier without lifting the COG as high as that barrier. Perhaps the best illustration of this is the Fosbury flop. In a well executed Fosbury flop, at all stages the COG is below the bar. As the waist passes over the bar, the head, shoulders, hips and legs are below the bar, yet the trunk at waist level is above the bar.

      In similar manner, a rolling car might get all four wheels off the ground without raising its COG if the superstructure loses height as the wheels leave the ground.

      I disagree with your description of a jump. A jump requires a push of the ground. Merely extending or flexing the body without a push against the ground does not elevate the COG. Flexing the body once airborne brings the feet (and also the head, near to the COG but does not elevate the COG. According to Newtons laws, elevation of the body demands that work is done on the body by an external force. During running and jumping the only external force (apart for air-resistance) is ground reaction force.

      During the second half of the airborne air-borne phase of running the COG falls freely. To allow this,fall work must be done by pushing against the ground during stance to provide upwards acceleration. No internal re-arrangement of the body can achieve this.

  48. Ivan Rivera Says:

    *My comment about the backflip was bad. There isn’t full body extension at the height of the flight phase. But the postural reconfiguration allowed by jumping backwards allows you to put the GCM substantially higher for the backflip than for the frontflip.

    • canute1 Says:

      Ivan,
      As described in my account of the Fosbury flop, which is a very effective form of back-flip, the key challenge is getting the body over the barrier without needing to lift the COG (i.e. GCM) above the height of the barrier. It is easier to achieve this with a back-flip than a front flip.

      The fact that the spine is more rigid when the body arches backwards than forwards makes it easier to draw the hips and legs up and over the barrier in the final stages of the passage over the barrier. In the final stages, the problem is not keeping the COG high because by this time the chest, shoulders and head are already well below the barrier. The main challenge is ensuring that the legs continue to rise in order to clear the barrier. This is very difficult to achieve during a forward flip because of the lack of rigidity of the spine.

  49. Ivan Rivera Says:

    Canute:

    Merely flexing the body doesn’t elevate it, as you said. But if I first extend my body completely (full hip, knee, ankle extension, plantarflexion, and raising of the shoulder girdle), I’ll bring my GCM (let’s call it the hips) higher than it was during neutral (standing position).

    If at this moment in time, I flex my lower extremities, I will in fact become airborne (my lowest point, call it my foot, will clear the ground). Immediately, I will begin to lose altitude and very quickly my GCM will return to the height it was at neutral.

    I don’t believe I’m splitting hairs here—It is possible to raise the center of mass a few centimeters due to full extension, and then simply flex the lower body for an instant, get airborne (without gaining further altitude) and then extend the lower body reflexively to catch yourself when you’re back at neutral.

    I see what you mean about the Fosbury Flop, though. My analysis there was wrong.

    I’d still like to mention this: when I extend my body fully, as mentioned above, and then flex my lower extremities so that I am balanced in the sagittal plane (with the same amount of lower extremity mass behind my COG as there is forward) I can maintain full extension of the upper body (including spine and scapula).

    In my opinion, the raising of the center of gravity by postural reconfiguration is well-illustrated by single-leg plyometric butt kicks. One leg has difficulty pushing off (and yes, it is a pushoff). When the lats are engaged and the scapular girdle is kept in place, it’s almost impossible to substantially raise the foot off the ground before reflexive leg extension kicks in. However, when we raise the scapular girdle ( a maneuver similar to the one that initiates a snatch) prior to pushoff, a much greater amount of knee/hip flexion can be produced.

    However, I do not contest that GRF is–and must be, as you mention–the major driver of any propulsion. I am a pretty enthusiastic advocate of the Pose Method. However, it is extremely important to me, as it is to you, to reconcile the theoretical underpinnings of any theory, whether or not its coaching cues work well.

    Many of the concepts in Pose, whether they are theoretically sound or not, are often presented obscurely, and so inaccessible to direct critique. (I am thinking of the “fall” here). In principle, I find it odd that you think that several Pose coaching cues are effective but that Pose is unscientific (or worse, ascientific). Now, I don’t mean this to say that you are wrong and that Pose is right, or vice versa, but I do believe that there is much to be reconciled–and much that can be reconciled here.

    • canute1 Says:

      Ivan
      Thanks for your continuing comments.. Although I would not describe myself as a Pose runner, I learned a lot by experimenting with Pose. Like you I wanted to reconcile what I experienced as the practical benefits of Pose as far as possible with theory of Pose, even though this theory is clearly misleading, and in some instances, plainly wrong. In seeking a plausible theory of running mechanics one absolute principle is that the action of running obeys Newton’s laws of motion (and the associated laws of conservation of energy and momentum). This is an absolute principle because a massive amount of evidence confirms that objects of human size moving at running speed obey these laws to an exquisite degree of accuracy.

      With regard to lifting the GCM by a sequential extension and flexion of the body as you describe, the GCM only rises because of a push against the ground. Newton’s laws allow me to by completely confident that if I were to do a sequential extension and flexion of the body as you describe while weightless in the International Space Station as it orbits the earth, the distance of my GCM from the centre of the earth would not increase. My feet would move closer to earth while my head moved away. While I have not actually done this experiment, I have done the experiment of sequential extension and flexion of the body while standing on bathroom scales. During the brief period when my body is extending and hips are moving away from the ground, the pointer on the scales demonstrates that I am exerting a downward force on the ground, in accord with Newton’s laws. The ground reaction force is what causes my GCM to rise.

      I do not see any lack of logic in accepting that Pose cues can be helpful even if the theory of Pose is incorrect. You might find it interesting to watch the video posted by Pose coach Jon Port at 9:48 am on Nov 17, 2009 (in the above stream of comments). Jon executes both two footed and one footed hopping. If you observe the height of his hips relative to the door knob behind him, you will see that his GCM rises during each hop (just as GCM rises during the one-footed butt kicks that you discuss). Jon emphasizes that it is unhelpful to think of exerting a push against the ground. He states:

      ‘The most important thing is that I am not focussed on any form of push. I am focussed on lifting with the pelvis with a rapid pull of the foot. Whatever opposite push happens – just happens, but my conscious hierarchy of thought is on movement of the pelvis and feet upwards rather than conscious muscle contraction. If you focus on pushing to do this you lose the elastic return energy which requires should ground time. Also if you focus on push with the foot/calves you will plantarflex the foot which cause achilles tendon strain very quickly. The key is to remain like a coiled spring but be flexible.’

      Newton’s laws demand that he is pushing against the ground. Indeed it is standard proactive to use a force plate to measure the height of a counter=move jump . However Jon achieves an efficient upwards movement of the GCM by thinking about lifting the pelvis and rapidly pulling the foot up. He states that ‘whatever opposite push happens just happens.’ I agree with him. He uses a cue that neglects the push but the push nonetheless happens.

      When you do butt kicks, I accept that you use core muscles to stabilize the body. This stability allows you to exert a push against the ground with the grounded foot, thereby lifting the GCM.

  50. Ivan Rivera Says:

    No argument with you, or Newton.

    Moving the GCM absolutely must create a GRF. (I restate this a couple of times below—I think our disagreement is a lot more fine-grained than I previously thought).

    Sorry if you feel like I’m moving the goalposts on you. I promise, I only just grasped the arguments against Pose (if I did at all).

    If raising my scapula makes it easier for me to push off and allows me to increase my knee flexion, then I was probably creating a GRF by moving my mass around (upwards, in this case). The only way I got off the ground was by producing a GRF. (Please note the ifs—I haven’t really tested this).

    Where we’re probably getting lost is that this movement (shoulder shrug) feels very different to the kind of “push” that we think about when exerting force against an object, or kicking it. In any case, I would never deny or argue that moving my weight around doesn’t create a GRF. It would. Period. Although raising my scapula may have been de facto a push, in Newtonian terms, it sure as hell doesn’t feel like one. In essence, it feels like I’m lifting myself off.

    I think that there’s a very problematic semantic issue surrounding Pose. I think that this is what’s getting in the way of a lot of conversations. Until you laid your argument out in the terms that you did in your previous answer, I wasn’t really sure why we didn’t agree.

    (By the way, I don’t find a lack of logic in your reasoning. I’m just of the general opinion that if you agree with the coaching cues but disagree with the science, I can look at the line of communication and can reasonably expect something to be miscommunicated—which of course, doesn’t reflect on you).

    In light of this, I would like to discuss Pose principles further with you, because I do think that they can be reconciled—call it reinterpreted—to fit Newton’s idea of how the world works.

    • canute1 Says:

      Ivan
      Thanks for your comments. It is clear that we are largely in agreement. In particular, we appear to agree that it is important to understand why many of the cues used in Pose are effective, even though when analyzed using the terminology of classical physics, they appear to violate Newton’s laws. In fact , neither Nicholas Romanov nor other Pose coaches actually violate Newton’ laws when they run, so the issue is the language used to describe what they do, and the arguments they present to justify what they do.

      I agree that because Pose principles are helpful for many runners, it is potentially worthwhile to develop a language for describing what Pose technique entails that is consistent with Newton’s laws, because continued use of an erroneous description is likely to produce harmful results in at least some instances. For example, I believe the under-estimation of the importance of the push against the ground is almost certainly harmful for a sprinter aspiring to elite level. On the other hand, it appears that for many distance runners, focusing on pushing against the ground creates a risk of injury through misdirected force and a risk of decreased efficiency through remaining on stance too long.

      I believe that we need to sort out which Pose principles are potentially useful and under what circumstances this is the case, and provide a language that describes these principles in a way that does not violate the laws of physics.

      However, it is not clear that this can be done under the auspices of the Pose organization while that organization continues to obfuscate the evidence. For example, the Capetown study (Arendse et al) clearly showed that Pose decreased potentially harmful stress at the knee but increased that at the ankle. In the most recent update of Pose technique (presented in Dr Romanov’s new book and on the up-dated Pose Method website, there is at least small acknowledgment of the fact that Pose can be associated with Achilles tendon problems, and it is accepted that these problems might be due to undue emphasis on ball-of-foot landing, but the discussion of this issue is still rather misleading and in set in a framework based on the very misleading ‘Pose, Fall, Pull’ mantra ,

      I am not sure that it is possible to re-write the principles of Pose while still using the term Pose, unless Dr Romanov agrees to such a re-writing. Several other coaches have formulated their own versions of running mechanics, incorporating some of the good features of Pose. Nonetheless, as far as I can see, none of these alternative formulations properly address the issue of how to describe helpful Pose cues related to getting airborne, in a way that is consistent with Newton’s laws. Therefore, it is perhaps worthwhile to attempt to do this.

      • Ivan Rivera Says:

        Canute,

        I agree. As a running coach who follows the mantra of: first run pain-free, then fast, then long, it’s very important to me to find a set of principles that create this progression in my clients. There’s far too much misinformation and disagreement out there, and I think that just about the only thing that people can agree on is “don’t load your joints ahead of your center of mass.”

        I myself am an agnostic as to footstrike—I’m a midfoot-striker (or a posterior forefoot striker, if you will—but generally I just go by what will get my clients through the abovementioned progression. That said, I do think that generally speaking, the faster you go, you’ll be spending a greater percentage of your stance phase with your heel off the ground.

        What’s interesting to me is that Dr. Romanov emphasizes that the forefoot-strike is and should be a result of everything else that’s happening, rather than something you should focus on. For example, I find that when you tell people to forefoot-strike, what they really just do is clench their gastroc-soleus complex to point their foot and end up shredding their achilles (or their peroneus longus, or both). That’s why I don’t really mention it when I train people. It’s sad that, at least anecdotally, many people that try out the pose method get lower leg problems. Clearly something is not being communicated right (or theorized incorrectly), which is part of the problem.

        Do you have a post (or a source that you can lead me to) that talks about why anterior rotation of the GCM around the foot structure is not gravity aided (and furthermore, why it does not constitute forward travel)? I’m not a physics buff—yet—but that part of Dr. Romanov’s theory really made sense to me.

        I can’t really grasp why, if you take an object inside a gravitational field, which would accelerate in a straight vector (as the body does in freefall) and then you put a rigid lever connected to an axis of rotation, that object would not accelerate around the axis of rotation instead. If that lever extends appropriately as the body rotates from .001 degrees relative to the direction of force to say, 30 degrees, wouldn’t that rotational motion be translated mostly horizontally?

        (Note that this is no argument for either the idea that gravity is the sole or principal driver, or that GRF does not occur).

        Take my questions as me trying to pick apart the details of your disagreement with Pose.

        Ivan

    • canute1 Says:

      Ivan,

      Thanks

      We clearly agree in our approach to the subject but have a different perspective on some of the issues. With regard to Achilles problems, I do not think this is merely a matter of communication. In the Capetown study, Dr. Romanov was the principle Pose coach so we can reasonably assume that what was taught was what he intended. The published manuscript (Arendse et al) did report greater stress around the ankle joint, and the subsequent description in Science of Sport by Ross Tucker (who was an assistant coach in the study) reported a high incidence of ankle and related injuries in the runners trained in Pose technique. Therefore I believe that Pose coaches should be aware of this and explicitly advise their athletes about this. To give Dr. Romanic credit, one sensible approach is to build up distance slowly when learning Pose and another is to do drills that strengthen the calf muscles and tendons. However, the tendency for some Pose coaches to imply that Achilles problems are simply due to people not doing Pose properly creates a risk that athletes will not prepare the calf muscle sand Achilles adequately through no fault of their own.

      With regard to head forward and down rotation, that is largely a consequence of the trunk continuing to move forwards at approximately constant speed while the foots is grounded. After mid-stance, gravity will exert a torque that adds a trivial amount to this rotation but the major role of gravity in late stance is opposing the upwards acceleration of the body, thereby demanding a substantial expenditure of energy at this stage Some of this energy comes from elastic recoil of leg muscles and the reminder comes from active contraction of muscles. . I discussed these issues in detail in my extensive debate with Robert Osfield in the comment section of my page’ Running: a dance with the devil.’ The comment most relevant to head-forward rotation is my comment posted at 9:32 pm on Jan 11, 2012.

      Because efficient running requires a strong push after mid-stance, a coach should ensure that his athletes develop adequate power, though the tricky issue is that consciously pushing in late stance can result in poorly timed or misdirected pushing. I believe this is why Pose cues can be helpful, though there is a danger that failure to acknowledge the need to accelerate against the pull of gravity can lead to failure to develop the required power.

      Dr. Romanov’s claim that gravitational torque assists after mid-stance is seductive but misleading. Part of the seduction comes from the fact that when we start to run from a standing position, leaning forwards allows us to utilize gravitational energy at the cost of a fall in the GCM, though the main the effect of leaning forward is to encourage a reflex forward swing of one leg to prevent a face-plant. Within a few strides we must use muscle power to lift the GCM back to its usual height.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s


%d bloggers like this: