Stop Dragging Your Toes!

Stuart McMillan

Stuart McMillan


Over the last 15 years or so, there has been a debate in the track and field world around the recovery height of the heels during early acceleration in sprinting. 

We actually don’t obsess over ‘low recovery of heel’ in the initial steps, but we understand why many do.  

Many of the world’s top sprinters’ heel recovery is so low that they drag their toes through the recovery stage of their stride in the initial steps of an acceleration.  In fact, this has become so popular that the ‘toe-drag’ has, in many circles, morphed into the outcome objective, rather than a resultant byproduct of attempting to maintain low heel recovery.  

It is important to understand that the genesis of this particular technique was as an error-correction reaction for athletes who would accelerate with an exaggerated cyclical heel recovery.  A common technical error in the start is limb motion that is overly cyclical – especially during block clearance. 

This is typified by excessive backside motion with the foot immediately after clearing the blocks, and is often observed in developmental and elite sprinters, but for different reasons. In developmental sprinters this is often observed in the foot exiting the rear block – perhaps due to a lack of either the physical capability to generate force, or the skill to apply force effectively. In elite sprinters, on the other hand, this may be observed in the foot exiting the front block – often caused by an effort to over-push horizontally.  

However, what began as a well-intentioned example of ‘over-cueing’ a correction of a technical error, has seemingly mutated into a technical objective for all sprinters, all the time.  

In the following article, we will present our thoughts on the height of heel recovery, generally, and the ‘toe-drag’, more specifically.  Before we dive into the nitty gritty of this somewhat controversial topic, though, we need to step back a little and discuss how we look to analyze movement in the first place. So please bare with us quickly before we return to the topic at-hand.


In many cases, movement evaluation begins with a biomechanical analysis. This analysis may include an examination of the motions (kinematics), forces (kinetics), and muscle activation involved with the athlete’s movement pattern. Sprinting lends itself nicely to this type of assessment – either in the lab, or on the turf/track. However, while it is often straightforward to collect biomechanical data, using this data to elicit positive changes in the athlete’s movement pattern can be more challenging.

All athletes are unique, and will move in unique ways.  In coaching mechanics, one of our greatest challenges is to develop biomechanical models that respect the individual nature of the athletes we are working with, while appreciating the futility of truly individualized models of movement.

We all know athletes with limited ranges of motion (ROM), for example.  This lack of ROM can be considered an organismic constraint (sometimes called a ‘performer constraint’).  A constraint acts to guide an athlete’s technique. We can only move within the boundaries of what constrain us – either structurally or functionally.  A constraint regulates our movement, not by prescribing certain configurations, but by eliminating what is not possible (Kugler, Kelso & Turvey, 1980).

For example, if you are weightlifting, and you have the objective of catching your clean at the very bottom (i.e. a full squat position) and lack the required mobility to sit there, then it won’t matter how much technical instruction you receive, or how hard you try, you will not be able to achieve your objective: you are limited by the structural constraint of a relative lack of mobility.  We cannot separate the movement outcome from the system that produced it; the ultimate technique is governed by the existing organismic constraints.  

These constraints then interact with task and environmental constraints to govern the ultimate control and coordination of movement – a coaching and pedagogical method known as the Constraints-led Approach (Newell, 1985).

Newell’s model of interacting constraints adapted to illustrate the resulting effects on variability of physical performance (Davids, et., al, 2003)

There is a lot to unpack here, so let’s go through it in order:


Firstly, let’s consider the task – ‘accelerating the body from a static, starting position.’  

The sprint start has two basic biomechanical requirements: 

  1. Apply enough force down into the ground to support the body
  2. Apply force backwards so the ground-reaction forces will propel the body forwards (Clark & Weyand, 2015)

Recent research has demonstrated that, per Newton’s laws, sprinters who apply more horizontal force relative to body mass will attain the greatest initial acceleration (Rabita et al., 2015). However, an optimal balance of vertical and horizontal forces is required to achieve a desirable projection angle from the blocks, to ensure that the subsequent steps can be executed successfully. Since gravity is always pulling downwards on the center-of-mass, the sprinter needs to project both out and up, in order to allow for the appropriate space and time for the limbs to correctly reposition in preparation for the next ground contact. 

In the initial steps after block clearance, ground contact times will get progressively shorter and flight times will get progressively longer as the sprinter gradually becomes upright. The body can only accelerate during the ground contact phase because force can only be applied when the foot is in contact with the ground. Therefore, flight time during initial acceleration should be just long enough to allow the limbs to properly reposition prior to the next ground contact, but without any extra time in the air. During the ground contact phase, the sprinter is challenged with applying sufficient vertical support force and maximal horizontal propulsive force in increasingly brief contact times. 

With this in mind, the mechanical goals of initial acceleration include: 

  1. Prior to ground contact, limb in front of body attacks down and back toward ground
  2. Initial ground contact occurs underneath (or behind) the center-of-mass
  3. During ground contact, forceful extension of stance-leg and forceful flexion of swing-leg
  4. At toe-off, a projection angle that allows for maximal horizontal displacement with sufficient vertical displacement to reposition limbs in preparation for next ground contact


Now that we understand the dynamics of the task, let’s consider the athlete’s interpretation of it.  

We have already briefly discussed that the control and coordination of movement is a result of the interaction of the task, the athlete, and the environment. Especially with young coaches, however, we tend to apply a one-size-fits all approach to the training of an athlete’s technique; rather than appreciating the individuality of the athletes we work with, we attempt to shoe-horn them into some ‘ideal’ biomechanical model.  

Often, this ideal is based on the average of a group, and occasionally, it is based on those in the sport who are the best performers. 

We see this in sprinting all the time, for example.  The following is a typical sequence of events:

  1. Observe the best in the world
  2. Pick an outlying mechanical aberration (something that is somewhat unique to that performer)
  3. Assume that this aberration is the key to their success
  4. Copy and apply to everyone 

Think back to Ben Johnson – wide hands on the track in ‘set’, and jump out of the blocks to an upright position as soon as possible.  It took perhaps a generation of young sprinters falling on their face before we figured out that this ‘jump’ start just wasn’t for everyone.  

Many went through this same process in reaction to Michael Johnson’s truncated arm swing, as well as John Smith’s late 1990s group’s exaggerated ‘drive phase’.  

It is important to remember that the best athletes are outliers; by definition, they can do things that a vast majority of other athletes cannot.  In addition, when we base our models on averages, we tend to end up with biomechanical models that match no single athlete.  

Let’s consider the ‘toe-drag’ as it relates to the above:

What started as a mechanical adaptation by Jamaican sprinter Michael Frater was adopted by his more-famous, and faster, teammate Asafa Powell, and quickly became the ‘secret’ to a world-class start.  

However, what might be appropriate for Frater, and what might be appropriate for Powell, is not necessarily what is appropriate for every other sprinter.  For every sprinter that can successfully navigate the ‘toe-drag’ start, there are perhaps dozens who are doing little more than shuffling for 5m, and ruining their shoes in the process.  

Younger, slower athletes simply do not have the force-producing abilities, nor the technical competence, to effectively do what Frater and Powell did, and as many elite sprinters do now. In fact, we would go so far as to state that no sprinter – elite or not – should be dragging their toes – and the concept of low heel recovery in the first place is one we simply do not place great emphasis on. There are simply far too many truly important factors that determine a successful start than to spend significant time on recovery height metrics (unless, as mentioned earlier, the recovery height is excessive, and performance-limiting). We will explore these factors in greater detail in part II of this article.

For now, let’s take a step back – as it is important to point out the reason(s) why many coaches recommend low heel recovery, generally, and the ‘toe-drag’, more specifically.  

  1. It was originally thought that encouraging the ‘toe-drag’ would eliminate excessive flight time during initial acceleration. 
  2. Some coaches have cued the ‘toe-drag’ with the goal of eliciting a better flexed position of the swing limb prior to the next ground contact. 

Although these two outcomes may occur for sprinters that can effectively execute a low heel recovery start, for sprinters who struggle with learning this motion, attempting to artificially recover a low heel may, in fact, delay the forward recovery of the swing leg and disrupt the appropriate ratios of ground contact and flight time.

Coaching cues should remain true to basic biomechanical principles of effective acceleration. Specifically, the sprinter needs to apply sufficient vertical force and large amounts of horizontal propulsive force in relation to body mass. Successfully accomplishing this requires precise recovery mechanics of the swing-leg in preparation for the next ground contact.  After reaching peak flexion in front of the body, the swing-leg thigh must rapidly reverse to attack the ground and apply force.

It is important to point out that there is a potential trade-off between recovery height of the swing-leg and the amount of force an athlete can apply into the ground.  In addition, due to the relative lack of height of the swing-leg foot, there is less time to reverse the foot so as to effectively initiate ground contact in the appropriate position relative to the center of mass (i.e. behind, or under).  

Elite sprinters have greater step frequencies and faster reversal of the thighs at the point of maximal flexion/extension. The ability to rapidly reverse the thighs may allow elite sprinters to require less distance from which to apply effective ground reaction forces, so the recovery height of the swing-leg can be lower, relative to their less-elite counterparts.

Furthermore, rapid thigh reversal allows elite sprinters to initiate ground contact in the appropriate position relative to the center of mass (i.e. behind, or under), even with a relative lack of height of the swing-leg foot. Developmental sprinters, who lack the same step frequency or ability to rapidly reverse the thighs, may not be able to effectively apply force, or appropriately contact the ground given a similar recovery path of the swing-leg foot.

The bottom line is that developmental sprinters (and in fact, many elite sprinters) often simply lack the specific force-producing abilities to accelerate with exaggerated low heel recoveries in an effective and efficient manner.  We must remember that there is a direct relationship between kinetics and kinematics (force and motion), and limitations in force production may constrain the movement positions that developing sprinters can achieve.


A mistake that many athletes make is attempting to accelerate the same way, regardless of the environment.  It perhaps goes without saying that a competition block start will look significantly different from a 3 point start on turf.  The two tasks are not too dissimilar, but the resultant mechanics will be.  

Differences in footwear and running surface are the two most-significant environmental constraints that will affect early acceleration technique. Quite simply, the grippier the footwear and surface, the greater the possibility of a more horizontally-oriented projection.  Athletes will have to make adjustments for less grippy environments by projecting relatively more vertically. For example, a projection angle of an elite male starter in an international competition may be less than 35 degrees – and this angle of projection may lead to a very low heel recovery.  However, if this same sprinter attempted this same angle of projection in running flats on a wet or very dry grass surface, for example, chances are high that he would slip, stumble, or otherwise accelerate ineffectively. 

Dragging the toes (or for that matter, attempting to produce low heel recoveries)  in an environment that does not allow for low angles of projection just leads to inappropriate mechanics, often manifesting in the ‘casting’ of the swing-leg, and an initial ground contact that is too far in front of the center-of-mass.  

We have to remember that athletes should move in ways that are most-appropriate to the task and the environment – rather than using the goal of ‘dragging their toes’ (or, accelerating with ‘low heel recovery’) as the initial starting point. 

Low heel recovery is often the effect of a successful acceleration; it is not the cause of one.  Christian Coleman is not the world’s fastest starter because he has low heel recovery. He has low heel recovery because he is the world’s fastest starter.  

In part II of this article, we will present what we feel is a more effective means of coaching initial acceleration, based on biomechanical first principles, while still respecting the unique abilities of each individual athlete.  

Thanks for reading,


Thanks to Dr. Ken Clark for his input on the biomechanical requirements of the sprint start, and to Coach Kebba Tolbert for the helpful feedback“ – Stu

Clark, Kenneth & Weyand, P.. (2015). Sprint running research speeds up: A first look at the mechanics of elite acceleration. Scandinavian Journal of Medicine & Science in Sports. 25.

Davids, Keith & Glazier, Paul & Araujo, Duarte & Bartlett, Roger. (2003). Movement systems as dynamical systems: the functional role of variability and its implications for sports medicine. Sports medicine (Auckland, N.Z.). 33. 245-60.

Kelso, Scott & Holt, Kenneth & Kugler, Peter & Turvey, M.T.. (1980). 2 On the Concept of Coordinative Structures as Dissipative Structures: II. Empirical Lines of Convergence.

Rabita, G., Dorel, S., Slawinski, J., Sàez‐de‐Villarreal, E., Couturier, A., Samozino, P., & Morin, J. B. (2015). Sprint mechanics in world‐class athletes: a new insight into the limits of human locomotion. Scandinavian journal of medicine & science in sports, 25(5), 583-594.


Stop Trying to ’Fix’ Athletes!

I’ve never been comfortable with the term ”corrective exercise”. In this short post, I’ll tell you why.

Coaching Chat #3: Coach to Olympic Gold Medalist – Sue Humphrey

Let’s Connect

ALTIS 360 Logo


Only $9.99/mo

Gain exclusive access to 100s of ALTIS education videos and join a global community of coaches and athletes.