Why strength drives speed

Force, not leg speed. The classic study by Weyand et al. (2000) compared runners of different abilities on an instrumented treadmill. Faster runners did not reposition their legs any faster: swing times at top speed were similar across groups. What separated them was the mass-specific vertical force they applied during shorter ground contacts. Later work confirmed that the limit to running speed is set “from the ground up”, by how much force the limb can apply in the brief contact window (Weyand et al., 2010).

Direction matters. Maximal velocity rewards vertical force, but acceleration depends on how much of the total ground reaction force is directed horizontally. Morin et al. (2011) called this the ratio of forces (RF): the share of horizontal force in the resultant force at each step. Across sprinters, the rate at which RF falls as speed increases was a stronger predictor of 100 m performance than total force alone (Morin et al., 2011; Morin et al., 2012). In practical terms, the best accelerators keep pushing horizontally for longer.

Strength transfers. The link between the weight room and the track is not theoretical. A meta-analysis by Seitz et al. (2014) found that increases in lower-body strength transferred positively to sprint performance, with an overall effect size of about 0.87 and greater transfer in younger and less trained athletes. Stronger athletes also tend to sprint, jump and change direction better, and they respond better to power-type training (Suchomel et al., 2016).

The takeaway for coaches is simple. If the athlete cannot produce high forces, technical sprint work has little raw material to organise.

The force–velocity curve

Muscle produces less force as its contraction velocity rises. A. V. Hill described this hyperbolic relationship in isolated muscle in 1938, and it holds for multi-joint movements too, where the relationship is close to linear (Cormie et al., 2011a). Heavy loads move slowly; light loads move fast. Power, the product of force and velocity, peaks somewhere in between.

For programming, the curve is most useful when it is divided into zones by the mean concentric velocity of the bar. Each zone emphasises a different strength quality, and each is trainable. The velocity bands below are adapted from Mann (2016) and are guidelines rather than hard cut-offs; exact values vary by exercise and athlete.

Develop the entire curve. Training only one end of the curve tends to produce adaptations specific to that end. Heavy strength training mainly raises the high-force portion, while light ballistic training mainly improves the high-velocity portion (Cormie et al., 2011b). Athletes need both. The goal is to train the whole curve so it shifts up and to the right: more force at every velocity, and more velocity at every force.

Where the athlete sits on the curve should guide emphasis. A weak athlete gains speed from almost any strength stimulus. A strong athlete with a force-dominant profile is likely to benefit more from work toward the velocity end. Force–velocity profiling, for jumping or sprinting, can help identify which side of the curve is the limiting factor (Morin & Samozino, 2016).

Rate of force development

Maximal strength sets the ceiling, but sprinting rarely gives the athlete enough time to reach it. Reaching peak force in a maximal voluntary contraction typically takes more than 300 ms (Aagaard et al., 2002; Maffiuletti et al., 2016). A ground contact at top speed lasts roughly 0.08–0.10 s in well-trained sprinters (Weyand et al., 2000), and even early acceleration contacts are well under the time needed to reach peak force.

This is why rate of force development (RFD), the slope of the force–time curve, matters so much. Two athletes can share the same peak force. The one with a steeper early rise will have produced more force by the end of a 100 ms contact.

The physics comes from the impulse–momentum relationship:

Net impulse (J), the average net force multiplied by the time it acts, equals the change in the athlete’s momentum. When contact time is fixed by the speed of running, the only way to increase impulse, and therefore change in velocity, is to apply more force within that window. High RFD means more impulse, sooner.

RFD is highly trainable. Heavy resistance training increases early RFD and neural drive, as shown by greater EMG amplitude and rate of rise in the first 50–200 ms of contraction (Aagaard et al., 2002). Ballistic and plyometric training, performed with maximal intent to move fast, add to this by targeting the earliest phase of force production (Maffiuletti et al., 2016).

Specificity & force vectors

General strength raises the ceiling, but transfer to sprinting improves when exercises resemble the task. Verkhoshansky’s principle of dynamic correspondence gives coaches five criteria for judging that resemblance (Verkhoshansky & Siff, 2009):

1. Amplitude and direction of movement. Do the range of motion and the direction of force match the sport action?
2. Accentuated region of force production. Is peak force produced at the joint angles that matter in sprinting?
3. Dynamics of the effort. Is the magnitude of force similar to what the sprint demands?
4. Rate and time of maximum force production. Does the exercise require force to be produced as quickly as sprint contacts do?
5. Regime of muscular work. Does it use the same type of action, such as a fast stretch–shortening cycle?

No single exercise meets all five. A back squat scores well on force magnitude but poorly on rate and time; a bounding drill is the reverse. That is why a good program combines exercises from across the spectrum rather than searching for one perfect lift.

Cover both vectors. Acceleration and maximal velocity load the athlete in different directions. Acceleration needs horizontal force production, so horizontally directed work such as sled pushes and pulls and hip thrusts is a logical fit. Maximal velocity needs high vertical force in very short contacts, so squats, jumps and vertical plyometrics fit better. The evidence for strict vector specificity is mixed. Contreras et al. (2017) found hip thrusts and front squats produced broadly similar transfer in adolescents, with only small differences favouring the hip thrust for 10 m and 20 m sprint times. We therefore see vector-based selection as a way to cover both qualities, not a rule that one exercise transfers and another does not. Sprint force–velocity profiling can show which vector an athlete is lacking (Morin & Samozino, 2016).

Five tools for faster athletes

The five tools below cover the force–velocity curve from the heavy end to sprint-specific loading. Used together, they build force, teach the athlete to express it quickly, and transfer it to the sprint.

Maximal strength. Heavy loads of 80–95% 1RM recruit high-threshold motor units and raise peak force and early RFD (Aagaard et al., 2002). For less experienced athletes this is the most productive single tool, because strength underpins every other quality (Suchomel et al., 2016; Suchomel et al., 2018).

Ballistic lifts. In a ballistic lift the athlete accelerates through the whole range and leaves the ground or releases the load, so there is no deceleration phase. Light loads of 0–30% 1RM, moved with maximal intent, target the velocity end of the curve and peak power output (Cormie et al., 2011b).

Weightlifting derivatives. Pulling derivatives such as the hang clean, clean pull and jump shrug train fast triple extension of the hip, knee and ankle with high RFD. Pulls from the knee or mid-thigh remove the technically demanding catch and allow loads to be biased toward force or velocity (Suchomel et al., 2015).

Plyometrics. Plyometrics train the stretch–shortening cycle (SSC). Fast SSC actions, with ground contacts under about 250 ms, best match sprinting; slower SSC actions such as countermovement jumps take longer (Schmidtbleicher, 1992). Bounds, hurdle hops and drop jumps progress from lower to higher intensity and should be dosed by ground contacts, not just sets.

Resisted sprints. Sled pushes and pulls overload the horizontal force demands of acceleration while keeping the movement specific. Loads are best prescribed by velocity decrement (Vdec), the percentage reduction in sprint velocity compared with unresisted sprinting, because body mass alone does not account for surface friction or athlete strength. Light loads (~10% Vdec) preserve sprint mechanics. Heavier loads near 50% Vdec put the athlete close to the load that maximises horizontal power (Cross et al., 2017). A systematic review supports sled training for improving early acceleration (Petrakos et al., 2016).

Putting it into practice

Knowing the tools is half the job. The other half is ordering and dosing them so that strength turns into speed rather than fatigue.

1. Sequence the qualities. Build maximal strength first, then convert it into strength–speed and speed–strength. Phase potentiation models hold that a block of strength work raises the potential that later power and speed blocks can express (Haff & Nimphius, 2012). Cormie et al. (2010) showed that weaker individuals improved jump and sprint performance with either heavy strength or ballistic training, which supports building a strength base before shifting emphasis. Relatively strong athletes, by contrast, often need more velocity-focused work to keep improving.

2. Complex training (PAP). Complex training pairs a heavy conditioning set with a biomechanically similar explosive task, such as a heavy back squat followed by jumps or short sprints. The heavy set can acutely enhance the explosive effort that follows. Seitz and Haff (2016) found the effect is larger in stronger athletes, after multiple-set conditioning activities, and after rest periods of at least 5 minutes; we use about 5–7 minutes as a working range. Researchers now distinguish classic post-activation potentiation, which is short-lived, from post-activation performance enhancement (PAPE), which develops over several minutes and likely explains most of the effect seen in practice (Blazevich & Babault, 2019). Individual responses vary, so test the pairing with each athlete.

3. Autoregulate with VBT. Velocity-based training (VBT) uses bar speed to guide load and volume. Ending each set at a velocity loss of 10–20% from the fastest rep limits fatigue while still building strength. Pareja-Blanco et al. (2017) found that training to 20% velocity loss gave similar squat strength gains to 40% velocity loss, but with better jump outcomes and fewer shifts toward slower muscle fibre types. Higher velocity losses drove greater hypertrophy but less benefit for speed qualities.

4. Sprint fresh, then lift. Maximal sprinting is a high-quality skill that suffers under fatigue. Schedule it early in the session, or on separate days, before heavy lifting. In season, strength can be maintained with as little as one heavy session a week. Professional soccer players who trained once a week maintained strength and sprint performance across a season, while those training once every two weeks lost ground (Rønnestad et al., 2011). We recommend one or two heavy sessions a week depending on the competition schedule.

Key takeaways

• Sprinting is a force problem. Faster athletes apply more force into the ground, in the right direction, during shorter contacts.
• Strength raises the ceiling; RFD uses it in under 0.1 s. Build peak force, then train the athlete to express it quickly.
• Train the full force–velocity curve. Use all five tools to shift the curve up and to the right, and let profiling guide emphasis.
• Match the force vector to the sprint phase. Horizontal work supports acceleration; vertical work supports maximal velocity.
• Use VBT to keep speed in the weight room. Stopping sets at 10–20% velocity loss protects the qualities you are trying to build.

Strength training will not replace sprinting. Athletes still need to sprint at maximal speed, often. But a well-planned strength program gives every sprint session more force to work with.