Acceleration Speed Development
Faster legs or more force? What the research says actually makes athletes accelerate, and how to coach it.
The short version
- Force beats fast legs. Faster runners push harder into the ground. They don’t swing their legs any quicker.
- Direction matters. Horizontal force predicts sprint performance. Total force doesn’t.
- Hamstrings drive it. High hamstring activation before contact and eccentric strength go with more horizontal force.
- Not all steps are equal. Early steps rely on propulsion and step length. Later steps rely on frequency and reactive stiffness.
- Train it, then test it. Use resisted sprints, strength and plyometrics, and profile athletes so you train their actual gap.
In most field and court sports, the sprints that decide games are short. A typical sprint in team sport lasts 2 to 4 seconds and covers 10 to 30 metres. Even in the 100 m, an athlete can only accelerate at maximal effort for around 5 to 7 seconds, so the ability to accelerate shapes the whole race.
That makes acceleration one of the most valuable physical qualities a coach can develop. It also attracts plenty of myths, such as “quick feet” drills, cues to turn the legs over faster, and the idea that speed is purely genetic. The research below, from biomechanics labs to professional football clubs, points to a clearer picture. Acceleration depends on how much force an athlete can produce, in which direction, using which muscles, at each phase of the sprint.
Force beats fast legs
In a landmark study, Peter Weyand and colleagues at Harvard put 33 runners on a force-plate treadmill. Their top speeds ranged from a jog-like 6.2 m/s to an elite 11.1 m/s. The researchers measured two things: how hard each runner pushed into the ground, and how quickly they repositioned their leg in the air between steps (swing time).
If fast runners were fast because they moved their legs faster, swing times should have shrunk as top speed rose. They didn’t. Swing time showed no significant difference across the whole range of abilities. What did change was ground force.
As speed rises, step frequency increases because both contact time and swing time get shorter, but the minimum swing time a runner can reach is similar for fast and slow athletes alike. The difference between them is how much force they can put into the ground in the brief time their foot is in contact.
Both are greater in faster sprinters, but the two trade off against each other. Longer steps come from bigger impulses, which take ground time. Higher frequency comes from shorter contacts, because fast and slow sprinters reposition their limbs in about the same time (~0.35 s). Elite sprinters who rely on neither one alone tend to be the fastest (Salo et al., 2011, cited in Sims, 2016).
Push back, not just down
Weyand’s study looked at top speed, where the main job is supporting body weight. Acceleration is different. To speed up, the athlete has to push their mass forwards, and only the horizontal part of the ground reaction force does that. The vertical part keeps the athlete off the ground but doesn’t drive them forwards.
Jean-Benoît Morin and colleagues measured this directly on an instrumented sprint treadmill, testing sport science students, national-level sprinters and a world-class sprinter. They described force orientation with the ratio of force (RF): horizontal force divided by total force on each step. They also measured how quickly RF falls as speed increases, called DRF.
Across both parts of the project, the amount of horizontal force was significantly related to 100 m performance, while total force was not. Vertical force was linked to top speed but not to 100 m time. The world-class sprinter stood out because he kept his force horizontal for longer as he sped up.
The same principle holds in football. In 37 male soccer players sprinting 60 m, Murata and colleagues found that net horizontal (anteroposterior) impulse was related to the change in running speed in every section of the 50 m analysed, with correlations up to r = 0.68.
Hamstrings drive it
If horizontal force is the target, which muscles produce it? Modelling and clinical studies had pointed to the hip extensors. In 2015, Morin and colleagues tested this experimentally with 14 sprint-trained males.
Each athlete completed isokinetic strength testing of the knee and hip muscles, then performed maximal 6-second sprints on an instrumented treadmill. Muscle activity (EMG) of the hamstrings (biceps femoris), glutes and quadriceps was synchronised with the force data.
01 Activation timing
The athletes who produced the most horizontal force activated their hamstrings strongly at the end of the swing phase, just before the foot hits the ground.02 Eccentric strength
They also had high eccentric knee-flexor peak torque, meaning their hamstrings were strong while lengthening under load.Together, these two factors were significantly related to horizontal force production (P = 0.024). The authors concluded that both hamstrings and glutes should be considered essential components of any training programme for acceleration. They noted that the same strengthening work may also help prevent posterior thigh injuries.
Not all steps are equal
Acceleration is often coached as a single movement, but the mechanics change from step to step. Nagahara and colleagues tracked 12 sprinters through a 60 m sprint with 60 infrared cameras. Using the height of the centre of gravity, they found two clear transitions that split acceleration into initial, middle and final sections.
At the first transition, the foot started landing in front of the centre of gravity, the knee began to flex during ground contact, and step frequency stopped increasing. At the second, body posture stopped changing.
What changes across the phases
In soccer players, a bigger change in speed over the first steps went with longer steps (r = 0.70 over steps 1 to 4), driven by greater propulsive impulse (r = 0.74). From around the 9th step, faster players relied more on step frequency, achieved by producing less vertical and braking impulse.
Jump tests line up with these phases. In 19 sprinters, squat jump height was related to acceleration at steps 6 to 10, and countermovement jump height at steps 5 to 11 (r ≈ 0.46 to 0.54). A rebound ankle-jump index, which reflects reactive ankle stiffness, was related to acceleration later on, at steps 14 to 19.
| Jump test | Steps where it related to acceleration | Correlation (r) |
|---|---|---|
| Squat jump height | Steps 6–10 (early to middle) | 0.48–0.51 |
| Countermovement jump height | Steps 5–11 (early to middle) | 0.46–0.54 |
| Rebound ankle-jump index | Steps 14–19 (later acceleration) | 0.48–0.54 |
| Rebound continuous jump index | No step | Not significant |
Table 1. Jump tests and step-by-step acceleration in 19 male sprinters (Nagahara et al., 2014, J Sports Med Phys Fitness).
Train it. Test it.
Training methods
In a review of sprint training methods, Rumpf and colleagues (2016) found that resisted sprint training was the most effective method for improving 10 m and 20 m sprint performance. Loading the sprint itself overloads force application in a task-specific way, and horizontally resisted sprinting encourages a greater forward lean.
| Goal | Load | Reps and volume |
|---|---|---|
| Acceleration | Resistance causing a 10–30% drop in max velocity (horizontal-force bias) | 10–30 m reps, 250–350 m total per session |
| Max velocity | About 10% of body weight (vertical-force bias) | Limited evidence for specific recommendations |
Table 2. Resisted sprint recommendations, from the speed development critical review (2016).
Strength and plyometric training complete the picture. Stronger athletes generally produce more force, and both methods raise rate of force development (RFD). In sprinters who already have adequate strength, the review suggests a ceiling where RFD becomes more decisive than maximum strength, so training should shift towards applying force in shorter times. All speed work should be done at maximal intensity, with minimal fatigue and full focus on movement quality.
Three tiers of speed training
The critical review groups speed training into three tiers. Free sprinting forms the base. Resisted and assisted sprinting overload it. Everything else supports it.
| Tier | Methods | What it develops |
|---|---|---|
| Primary | Free sprinting with no equipment | Longer steps, shorter contacts, technical force application. Specific training gives the greatest transfer. |
| Secondary | Resisted (sleds, hills) and assisted (towing, treadmill, downhill) sprinting | Task-specific overload. Resisted work targets acceleration; assisted work targets max velocity and step frequency. |
| Tertiary | Strength, power, plyometrics, Olympic lifts, mobility and trunk stability | Force capacity, stiffness and RFD, plus the joint ranges and postural control to apply force. |
Table 3. Primary, secondary and tertiary speed training methods (Sims, 2016, after Behrens & Simonson, 2011 and Rumpf et al., 2016).
Assisted sprinting for top speed
| Method | Guideline |
|---|---|
| Towing | No faster than 106–110% of max speed. Reps no longer than 40 m, as mechanics are hard to control at supramaximal speeds. |
| High-speed treadmill | No established guidelines. By analogy with towing, ~106–110% of max speed for 3–6 s reps, as long as mechanics hold. |
| Downhill sprinting | Slope no steeper than 7°. Angles of 2–5.8° may be most effective. Rep distance set by maintaining good form. |
Table 4. Assisted sprinting recommendations (Sims, 2016). Evidence is limited, so treat these as starting points.
2. Optimise acceleration mechanics so athletes can accelerate for longer, using resisted and unresisted sprints.
3. Expose athletes to max-velocity running, for performance and to reduce injury risk.
Force-velocity profiling
Two athletes can post identical sprint times for very different reasons. A sprint force-velocity profile shows why by describing the athlete’s maximal horizontal force (F0), maximal velocity (V0), peak power (Pmax) and force orientation (RF and DRF).
Player C Force-oriented
Higher force effectiveness in the first pushes (higher RF max), but RF drops quickly as speed rises. He can’t hold his forward lean.Player D Maintains RF
Same power and 20 m time, but keeps force horizontal for longer, so he accelerates for longer and reaches a higher top speed.These two elite rugby union players (Morin and Samozino, 2016) show why profiling matters. The same programme would suit neither of them well.
Testing in real play
For football, Morin and colleagues (2021) took profiling a step further with an in-situ acceleration-speed profile built only from GPS data recorded during normal training. For 16 professional players at an Italian top-flight club, they kept the highest accelerations reached at each speed from 3 m/s up to top speed (around 50 points per player) and fitted a straight line.
Typical values were a theoretical maximal speed (S0) of 9.2 to 9.5 m/s and maximal acceleration (A0) of 7.2 to 7.7 m/s², close to force-velocity values reported in professional footballers. A follow-up pilot study (Komino, 2022) found that top speed agreed closely with classic sprint testing. Estimating maximal acceleration accurately needed more maximal efforts than a single 45-minute small-sided game provided.
| Variable | Weeks 1–2 | Weeks 4–5 | Typical error |
|---|---|---|---|
| S0 (m/s) | 9.21 ± 0.43 | 9.47 ± 0.52 | 3.3% |
| A0 (m/s²) | 7.70 ± 0.53 | 7.20 ± 0.40 | 5.4% |
| AS slope (1/s) | −0.84 ± 0.09 | −0.76 ± 0.07 | 7.6% |
Table 5. In-situ profile results for professional footballers across two 2-week training blocks (Morin et al., 2021). Typical error is the standard error of measurement.
Intervention strategies for skill errors
Knowing the science of acceleration is one thing. Seeing a fault in an athlete and fixing it is another. The Kinetics Manual makes a clear point: most technical errors have a physical cause, such as posture, strength, stiffness or tissue tightness. Shouting a cue rarely fixes them.
Before stepping in, the manual sets two conditions. The athlete should be consistently showing the fault, not just once. And the coach should check whether the athlete actually has the strength, joint positions and anatomy to move differently. Without those prerequisites, many verbal cues are given in vain.
| What better sprinters show | Source |
|---|---|
| Higher RF and a smaller DRF in acceleration | Morin et al., 2011, 2012, 2015 |
| Short ground contact times | Weyand et al., 2000; Clark & Weyand, 2014 |
| Greater step frequency | Weyand et al., 2000; Coh et al., 2001 |
| Greater hip, knee and ankle extension at take-off | Hay, 1994; Wild et al., 2011 |
| Higher angular velocity of the stance hip | Mann & Sprague, 1983; Yu et al., 2016 |
| Faster limb segments in hip extension | Mann et al., 1984; Ae et al., 1992 |
| Better developed motor programming | Mero & Komi, 1985 |
Table 6. Technical characteristics that distinguish sprint skill levels (Sims, 2016). Use these as the “what good looks like” picture before diagnosing a fault.
Adapted from Figure 5, Intervention Strategies of Skill Errors, Kinetics Manual (2013).
Less talking, better tasks
The manual argues that over-reliance on verbal instruction is one of the weakest parts of coaching. Much of a sprinter’s stride comes from unconscious sources: central pattern generators, stretch reflexes and the interaction of ground reaction forces with each athlete’s anatomy. Too much cueing can interfere with that.
Instead, it recommends shaping learning through task design, physical preparation, appropriate stress and an emotionally healthy environment. The coach’s job is to remove what interferes with natural running and use the minimal amount of guidance. Athletes given the right task tend to learn more deeply, and faster in the long run, than athletes who are constantly corrected.
Take home messages
- Build force first. Speed comes from force applied to the ground, not faster leg turnover.
- Coach direction. Use body lean and a “push back” intent to keep force horizontal.
- Strengthen hamstrings and glutes, including eccentric work, for performance and resilience.
- Match the method to the phase. Use heavy, horizontal pushing early and stiff, reactive contacts later.
- Use resisted sprints causing a 10–30% velocity drop, over 10–30 m and 250–350 m per session.
- Keep speed work fresh, early in sessions and at maximal intensity.
- Fix technique faults at their physical cause. Only intervene on consistent errors, and use tasks before cues.
- Profile athletes and retest, so each athlete trains their own gap.
References
- Weyand PG, Sternlight DB, Bellizzi MJ, Wright S. (2000). Faster top running speeds are achieved with greater ground forces not more rapid leg movements. J Appl Physiol 89: 1991–1999.
- Morin JB, Edouard P, Samozino P. (2013). New insights into sprint biomechanics and determinants of elite 100m performance. New Studies in Athletics 28(3/4): 87–103.
- Morin JB, Gimenez P, Edouard P, et al. (2015). Sprint acceleration mechanics: the major role of hamstrings in horizontal force production. Front Physiol 6: 404.
- Nagahara R, Matsubayashi T, Matsuo A, Zushi K. (2014). Kinematics of transition during human accelerated sprinting. Biology Open 3: 689–699.
- Nagahara R, Naito H, Miyashiro K, Morin JB, Zushi K. (2014). Traditional and ankle-specific vertical jumps as strength-power indicators for maximal sprint acceleration. J Sports Med Phys Fitness 54: 691–699.
- Murata M, Takai Y, Kanehisa H, Fukunaga T, Nagahara R. (2018). Spatiotemporal and kinetic determinants of sprint acceleration performance in soccer players. Sports 6: 169.
- Sims L. (2016). Speed development: a critical review. MSc Strength and Conditioning assignment, Cardiff School of Sport. Includes Rumpf et al. (2016), Morin & Samozino (2016), Hunter et al. (2005), Faude et al. (2012) and Alcaraz et al. (2009).
- Morin JB, Le Mat Y, Osgnach C, et al. (2021). Individual acceleration-speed profile in-situ: a proof of concept in professional football players. J Biomech 123: 110524.
- Komino P. (2022). Acceleration-speed profile in-situ in soccer: direct comparison with the classic sprint force-velocity profile. Master’s thesis, Université Jean Monnet, Saint-Etienne.
- Kinetics Manual (2013). Speed Development: Instructing and Training Power Athletes, Chapter 1, Figure 5: Intervention Strategies of Skill Errors.
