edited March 2014 in Running Form
Many runners refer to their 'push-off' or 'toe-off' when describing their running or perhaps discussing shoes that are designed to help it.

I wrote an article recently as to why 'pushing-off' slows you down when you run and is something you should work at eliminating.



  • You clearly don't understand running mechanics.  Sadly being a Pose advocate requires you to leave understanding of actual physics behind and enter the world of make believe.

    Gravity absolutely doesn't pull you forward, it's a vertical force, it accelerates you downwards.  We even define up in our world based on the direction that aligns with the local gravitational vector, so by definition gravity *always* pulls you downwards. (FYI I have an engineering degree so I really do know about this stuff.)

    One of the subtle effects of being a biped is that we use imbalance to helps go faster or slower.  When we want to accelerate, such as when we start running, what we do is we lean forward.  If we lean forward and not push off at all, such as if you locked your knees and totally relaxed your ankle what would happen is that you'd fall forward an land on your face.  Here we've created an imbalance but never re-balanced it by pushing off.  

    Thankfully from our very first attempts at walking when we were babies we worked out that this initial imbalance created by leaning has to be balanced by pushing off using a combination of straightening our knees and pushing off with our calves through stiff ankle.  From a very early age we master actually how much push off to generate to get back into balance and also how we can use the amount of initial imbalance to generate different amounts of horizontal acceleration.  The extreme of imbalance can be seen by a sprinter at a starting block - their center of gravity is along way forward of their feet and requires massive amounts of force to keep balance once they lift their hands once the gun goes off.

    The imbalance is also used when we want to slow down - we lean backwards just enough to force us to generate the slowing force we want.  If we want to slow down quickly we have to lean along way back, if we want a gentle slow down then a small lean will be required.

    The imblance is also used when running around a bend - we first generate the lean then the forces follow, when we no longer want the lean we actually generate a little more horizontal force and the body then pops back up level.

    This interplay between imbalance and the horizontal forces we generate to accelerate/decelerate and go round bends becomes complete ingrained and automatic from a very early age - you really don't have to think about it, and you really shouldn't because conscious thought doesn't have anything like the rapid flow of control needed for running - instead we should trust in our very clever central nervous systems and the motor programming that it's built up over years of practice.

    Now that said, once running at a steady state, the motor programming that you've developed may not be achieve perfect efficiency and may even be causing a injury risk.   It's quite possible to retrain the motor programming using cue.  There may even be some cases where runners do indeed push off to strongly and using a cue to not push may improve their running, but also the opposite can be the case a runner doesn't push off enough and will benefit from pushing off more.  To know which you'd need to know the specifics of the runner you are working with and the type of event they are training for. 

    Making blanking claims that pushing off is bad is a really bad thing for a coach to do, even worse when you try to fool them with pseudo science.  I'm afraid you've fallen into both these areas, your understanding of mechanics is pitfall and as consequence advice largely misguided.  It's time to rip up your POSE coaching books and go learn from find a source that actually understand mechanics of running.

  • You haven't rwad what I said carefully enough.

    I didn't say Gravity pulls you forward. I also agree that gravity is a downward force. What I actually said was that gravity pulls your bodyweight over when we are unbalanced. We cannot move ourselves without first unbalancing. Until then we're balanced and don't move anywhere. Pushing when our COM is over our support point will only push you up.

    I also didn't say there was no pushing involved. There is, at mid-stance. I said that. My point is, there's no need to push again as we reach terminal stance because it's unnecessarry (because we're already rising) and doing so can only push us up.

    We push to gain height so we have room to regain support.

    If, for example, someone pushed you from behind by surprise, you would instinctively PULL your foot off the ground so you could regain support to stop you falling on your face, you wouldn't instinctively PUSH into the ground. Repositioning the foot under the hips is what stops us falling, not pushing into the ground.
  • edited March 2014
    idenby- Don't pay too much attention to the other poster on this thread. Most everyone here is quite polite and willing to engage in thoughtful discussion. Beginning a post with personal insults is likely not the best way to foster constructive discussion nor to persuade anyone of the strength of one's arguments.

    For what it's worth, I'm with you on the efficiency of the "push-off". Feels to me like I'm straining too much if I try to do so, and just asking for injuries based on awkward mechanics. But fortunately, much more qualified authorities than myself also agree. Steve Magness has a couple good posts on running mechanics on his Science of Running blog. Links are below, but here's the specific text regarding push-off:

    "Unlike what many suggest, do not try and get any extra propulsion out of
    pushing off with the toes. It is too late in the running cycle to net
    any forward propulsion and will instead result in simply making your
    stride flatter. Instead, the forward propulsion should come from the hip
    and the foot should be thought of as being along for the ride, which we
    will discuss shortly. Essentially, once the hip is extended, leave the
    foot alone."


  • Thanks TrailrunningDad.
    I don't actually agree with The quoted comments from Stece Magness.
    There is no forward propulsion from pushing-off. We know the quads turn off after mid-stance, so they can't straighten the knee to push us forward, and that leaves the calves, and they extend the ankle as my diagram shows, so only up. If it were possible, we'd be able to push ourselves forward from an upright balanced position.

    Also, pushing-off will make your stride bouncier not flatter.

    An ACTIVE push-off is a variable in running, just like bending at the waist and overstriding with a heel strike are. Some do it, some don't. They are not necessarry in order to run. However, unbalancing and falling are...and if you want to avoid a face plant, so is PULLING your foot off the ground to recover it.
  • Should have figured Ian would bring this one here.  It came up in another thread on Running Shoe Geeks a while ago.  Pretty sure he was the voice of the minority in that one.

  • Hi Idenby,

    "If, for example, someone pushed you from behind by surprise, you would instinctively PULL your foot off the ground so you could regain support to stop you falling on your face, you wouldn't instinctively PUSH into the ground. Repositioning the foot under the hips is what stops us falling, not pushing into the ground."

    I afraid you have the mechanics wrong.  When we are pushed off balance we instinctive push off to generate a horizontal force to rebalance ourselves.  It's so instinctive you don't even notice it - it;'s all part of how our central nervous system manages balance.  

    We don't pull our foot off the ground, if we didn't we'd fall downwards and stumble on the next step as our body weight will have fallen forcing us to collapse the opposing leg once it lands on the ground.

    As for push-off being a variable in running, yep absolutely agree on this.  What I don't agree with is that we all do, just to varying degrees.  Some people will push off too strongly, while others will benefit from pushing off more strongly. 

    A crucial element to good running form is synchronization of the various elements of the body, for instance in the second half of stance we are we not only doing the final push vertical off the ground, but we are also need to synchronize the active of the hip flexors with the push off from the leg.  If we get the synchronization right then we'll store energy in our hip flexors so the once we have left the ground passive mechanics can pull the knee forward and have the foot rise naturally and completely relaxed. If you get the timing right the legs will be relaxed, with the legs recovering into position automatically - the only thing you have to think about is keeping relaxed.

    You absolutely don't want to pull the foot off the ground, this will ruin the passive mechanics that you should be using when running.  The ham string and hip flexors should *not* be actively used to prematurely pull the foot through.  This is one of the key errors of POSE running that leads to it being less efficient than a normal stride (there is even a scientific study that found that POSE running makes runners less efficient.)

    I would recommend that you leave POSE behind, and find learn from those who have a proper understanding of running mechanics.  Steve Magness blog is great place to start.  I have the book he has just published and thoroughly recommend it.   Be aware that Steve has debunked POSE on his blog several times, so I'm not the only one that feels that POSE can be detrimental, not only does it harms a runners mechanics but it also fills runners heads with pseudo science where physically impossible mechanisms being peddled as fact.
  • Robert

    Can I just be clear that I'm not disputing a push per se, as I do understand the requirement to do so, as I've already said. I'm specifically taking about plantar flexing the foot at terminal stance in an attempt to propel ourselves further forward than we would if we didn't do this. It's this that I'm saying is a variable in running and unnecessarry. I can choose to let my foot leave the ground in a neutral position or I can push onto my toes, prolonging ground contact time and loading the calve more...all to gain *excessive* height -bounce.

    My understanding is that it's physically impossible to just push ourselves horizontally. Force plate data does show hGRF, but as I understand it, it's always below body weight -meaning it can't be coming from a push. Dr Romanov argues it's coming from Gravitational torque.

    I will read Steve Magness blog with interest, genuinely. May I also suggest you read Christopher Drozd articles on running too -yes, he is a Pose guy, so you might have to get over yourself on that one, but his 'Anatomy of a Stride' and similar articles on running form are terrific. There are points of difference of course, but I'm also sure you'll find some common ground.

  • Thanks for links.  Christopher's articles are riddled with pseudo science.  For those who don't understand the laws of physics no doubt it sounds compelling, but for those of use who do it's very misguided and confused.

    Gravitational toque is an effect that works on a planetary scale - it's what locked the moon in sync with the earth, it's not something that works on the scale a runner.  Gravity doesn't any measurable torque on humans, we are way too small in scale for the effect for that.

    The "Gravitational torque" made up by Romanov and perpetuated by supporter is a school boy error in mechanics where you one attempts to generalize a simplification that can be made in static systems to dynamic systems.  I know I made the mistake in an applied maths lesson why I was 16 and my teacher explain how posing problems in this way was a mistake and caused all manner of errors down the line - this single lesson made a huge difference to may understanding of mechanics and helped my progress (I went of do Engineering at Oxford.)  On a human scale Gravity actually works through our centre of mass and doesn't generate any moments, instead it simply applies a vertical force that accelerates us downwards.

    When we lean forward it kinda feels like it's Gravity "pulling" us forward, but in reality it's Gravity pulling us downwards and the compressive forces in our legs that were present when were standing now angled forward and it's this force that provides some horizontal force. However, if our feet are only a slippery surface what will happen is that our feet will slip backwards out from underneath us and will not be able to impart any horizontal force to push us forward, instead our center of gravity will fall directly down wards.  If it were true that "Gravitational Torque" were a real affect as described by Romonov then we'd move forward regardless of how slippery the surface were were on.  This isn't the case though.  For all Romanov and Drozd quoting of laws of physics neither grasps this subtle but absolutely fundamental point.  "Gravitational toque" as they propose is complete fiction and directly contravenes the laws of physics.

    What actually generates both the vertical and horizontal forces during stance is not gravity but our bodies.  Gravity provides a constant vertical acceleration upon every atom in our bodies, but do so equally, there isn't ever any moment generated.  The only moments (Torques) generated on our bodies come from the forces we generate that are off centrer from our center of mass.  

    A skilful runner will actual generate a the right combination of horizontal and vertical forces on stance so that the resultant force vector stays close to their center of mass.  Such runners are able to run with smooth motion with little pitch forward/backwards or side to side when moving through the gait cycle.

    By contrast a less skilful runners who don't manage this to generate the right combination of horizontal and vertical forces will have a resultant force vector that periodically moves either side of their center of mass.  Such runners will exhibit a awkward pitching of their body forward/back and/or side to side.  Such movements patterns are a sign of wasted energy.

    Now if bring this understanding of optimium balance of forces back to the what happens on stance you'll see that on landing we actual want the breaking and vertical forces to align right through the leg to just above the hips where our center of mass is, at mid stance we want the force to vertical and then towards to off we want to keep the overall force vector through the center of mass. 

    If we push off too much with our calves then the ankle will generate too much rotation and horizontal force compared to vertical force and the resultant force vector will pitch our bodies head back.  This isn't good.  Too little push off at the ankle and we will have insufficient horizontal force generated and the force resultant too vertical and we will pitch forward.  

    If we pitch forward then we end up having to jam our foot down on landing to generate enough vertical force to ensure the overall resultant is infront of the center of mass to arrest the forward pitch.  We know this to be the case as the overall cycle has to be in balance, even if the parts within it aren't well balanced - otherwise we'd fall over after a few steps.

    Some runners will be pushing off too much, you'll see it if they are pitching backward/forwards through the gait cycle.  If they do correct them with cues such as pushing off through the heel/relaxing the ankle/"leaning from the ankle" more.  Cues are cues though, don't confuse them with actual mechanics, they are just hints to help runners change their ingrained motor programming.
  • Robert

    Just to be clear. I never said gravity pulls us *forward*.

    Running (and walking for that matter) is controlled falling. Falling occurs when we're unbalanced -our center of mass is out of line with the direction of gravity's force. Falling is only relevant because of the existence of gravity. Unsupported objects fall to the ground, but supported (rigid) objects including human beings rotate via their support point until they become unsupported. A felled tree or the leaning tower of Pisa haven't pushed themselves over. I don't think iether of those generate their own horizontal and vertical forces.

    Bottom line. We cannot move ourselves from the spot we're standing on horizontally until we fall.

    Robert -thanks for the debate

    PS. Pushing-off through the heel? That would be impossible.
  • Neither running or walking are controlled falling.  I don't know where this silly idea came from but it's total hogwash.  If you ever see someone suggesting that running or walking are a controlled fall then it's a red flag that they don't understand mechanics.

    We don't don't rotate over the pivot like a falling tree, we move over our foot with our upper body near vertical and at a near constant angle to the horizon, our support leg rotates forward but only because it connects a fixed point on the ground where our foot is, and the hip which is translating horizontally and moving vertical (down prior to mid-stance and up after.) During stance our opposing leg is actually being rotated in the opposite direction to the supporting leg.  

    So how are we ever rotating if our torso is near fixed, one leg rotating in one direction but the opposite rotating in the other....  Net rotation is actually zero if you are running correctly.

    The only time we are doing anything that could be called falling when running is in the second half of time you are airborne to when your foot touches down.  It isn't really helpful to think in these terms though, we aren't falling when we hop, we aren't falling when we running, we are running when we run. The only time you really fall when running is if you trip up or accidentally run over the edge of cliff.

    The most daft thing about "running is controlled falling" is that you try and make it out that after mid-stance you are actually falling forward.  After mid-stance the body actually is accelerating upwards, we are doing the *exact* opposite of falling, we are *rising* up and launching ourselves. 

    As for trees falling over, I'm afraid you mis-understand the mechanics, please re-read my previous statements on this, or just go and enrol in applied maths course so you can actually learn properly what you are attempting to talk about. Please don't go to Romanov's sources for how to understand gravity, he doesn't understand basics mechanics.

    And the leaning tower of Pisa is it's barely moving so we know it's in almost in equilibrium, this means all the forces and movements on it balance out. It's essentially a static system.

    The bizarre thing about Pose and apologists for it is that we get thrown these examples that have absolutely no relevance to the the dynamics of a runner.  A runner doesn't fall after mid-stance but does the exactly opposite they rise, unless of course we want to redefine falling as something that accelerates upwards.

    If we are to take this to it's logical conclusion then we could say that when a space rocket launches it's falling upwards into space.  Yep I'm sure this sounds pretty dumb a statement, but that's basically how explanation of Pose comes across to those of who educated in mechanics.

    > PS. Pushing-off through the heel? That would be impossible.

    Did you not read the bit about a cue not being actual mechanics?  Please go re-read the last paragraph of my previous post.

    A cue to push off through your heels is intended to encourage a runner to push less off with their toes, it doesn't actually mean they will dorsa flex their foot on late stance and push off with their heel.  

    "Lean from your ankles" is also a similar nonsensical cue if you try to think about the actual mechanics, but it still can help runners trying to push off so strongly late on stance.

    There are other cues that can help fine tune form that don't map exactly to actual movement pattern you are after. I would encourage you to learn properly about running mechanics and then start viewing coaching cues as hints for achieving better form.  In some cases this better form might be cues to reduce the push off, other case it'll be to encourage more push off, this applies the whole spectrum of variables that we have to work with.
  • The hips rotate from the point of support (the foot), and occurs from when we land to when we reach 1body weight at terminal stance. Our centre of mass is between the hips, so the torso can remain vertical (ish) during the rotation, as Michael Johnson's posture displayed. Nevertheless, his hips still rotated over. Imagine 4 minutes past 12 on a clock face. But, the hips are rising as the are rotating, due to an involuntary push at mid-stance as the body comes out of a compressed position. The two motions together cause the slightly upwards arc of the hips we see in everyone's running stride. I said this in my article. I'm not and never have been disputing this push up. The body is already rising -there's no need to add to it by plantar flexing the foot to push into your toes. The fact that not every runner does this demonstrated is unnesesarry.

    If we stand on one leg and fall over, we have two options. 1. Do nothing and faceplant! (uncontrolled fall), or 2. Pull our support foot off the ground to regain support with the other foot or stick the other foot out (controlled fall).

    Falling OVER and falling DOWN are different and both are the affects of gravity's force.

    Thanks for your insight Robert. We'll move on... in different directions.
  • Hips don't rotate from the foot, hips rotate from the.. wait for it... HIP, that's the job of hip joint.

    You can't define a rotation where something doesn't rotate, just how stupid this you even point out when looking real-world runners and yes Michael Johnson's upper body posture when sprinting is great example of we aren't rotating when we run.  Our legs rotate from the hips, and arms rotate from our shoulders, and our spins twist, all these rotations work in nice cycles that balance and work with each other in a well balanced and coordinate runner.

    To qualify just how stupid it is to claim that we are rotating about our support point draw the a the motion of the center of mass relative to the ground and the points we land at.  What you'll see is wave pattern, with the bottom at mid-stance and the top mid way between landing points.  The point of rotation if you were to try and map the two opposing bits of the curves would be underneath the center of mass whilst we are in the air, and above the center of mass when we on stance.  Of course these two center of rotation points are really just totally meaningless for the actual physics that's going on, but it's nicely illustration how you can choose various different but in the end unhelpful ways to measure rotation.

    For engineers/physists we set up the rotations that are helpful for formulating the maths that can model what is going on.  When tackling the mechanics of runner this means using the center of mass of the runner and local axis relative to the center of mass.

    What you don't do, because it's a error is start working out the system from a point on the ground and then start calling points of force a support point.  Right away such formulations lock you in models that only work for static systems.  It's a subtle point but crucial, you can't use the same simplications that you use when analysis static objects as you can with dynamic systems.  If we were to attempt to use static model for a dynamic system you'd end up lots of errors and complete unrepresentative model that has no baring to the real-world.  This is *exactly* what has happened with POSE.  So far you have lacked the ability to pick up these errors and keep repeating the same mis-understanding of laws of physics.

    The difference between us is that I have degree in Engineering from a repeatable University and your understanding of physics has been learnt form a widely dunked model developed by someone who doesn't understand the laws of physics.  It's time you walked away from the pseudo science and actually started to learn and think.

    You *really* don't understand what you are trying to talk about, sadly it doesn't seem like you've twigged that this is the case yet.  The minute you do, is the minute you can begin the journey of learning that will result in your becoming a better coach.
Sign In or Register to comment.