Measured, Not Inferred: Why Horizontal FVP Needs a Loadcell
Summary
Force-velocity profiling has become one of the most useful diagnostic tools in applied sport science. By mapping how much force an athlete can produce across a range of movement velocities, practitioners can distinguish...
Measured, Not Inferred: Why Horizontal Force-Velocity Profiling Needs a Loadcell
Force-velocity profiling has become one of the most useful diagnostic tools in applied sport science. By mapping how much force an athlete can produce across a range of movement velocities, practitioners can distinguish a force-deficient athlete from a velocity-deficient one and direct training accordingly. The principle is well established in the literature, from the sprint mechanics work of Samozino and colleagues to the individualised training framework described by Morin and Samozino (2016). What receives less scrutiny is a question that sits underneath every profile: where does the force number actually come from?
Horizontal is where sport happens
Most field sport actions are horizontal. Acceleration, maximal sprinting, change of direction, and contact events are dominated by horizontally oriented ground reaction force. This is why horizontal force-velocity profiling (HFVP) has attracted so much interest. Cross and colleagues (2017) formalised methods for power-force-velocity profiling during sprinting, and subsequent work on resisted sled sprinting examined the loads that maximise horizontal power output. Vertical assessments such as the loaded countermovement jump and barbell-based testing remain valuable, but they characterise a different plane to the one in which most athletes compete.
The inference problem
Here is the part worth examining. Almost every commonly used method does not measure force at all. It infers it.
The sprint profiling methods derived from Samozino's simple field approach reconstruct force from a velocity-time curve, the athlete's body mass, and a modelled estimate of air resistance. Radar and timing-gate systems that report force are doing the same thing: they measure velocity well and then calculate force from how that velocity changes. Loaded vertical jump profiling estimates force by working backwards from jump height, which is itself derived from flight time, across a series of added loads. Barbell velocity-based training devices measure bar speed accurately, but the force in that system is simply the known mass on the bar.
None of these are flawed. They are elegant, accessible, and supported by good validation work. But each arrives at force through a chain of assumptions rather than a direct reading. When the model assumptions hold, the estimate is good. When they do not, the error propagates silently into the profile.
Measuring force from first principles
Direct measurement removes the chain. A loadcell placed in the line of resistance reads the actual force being applied, continuously, as it happens. There is no body-mass term, no aerodynamic model, and no flight-time reconstruction between the athlete and the number.
This is the basis of the DynaSled. As a friction-resistance sled instrumented with a loadcell, it measures the horizontal push and pull force the athlete produces directly, rather than estimating it from velocity. The distinction is not academic. Direct measurement allows force to be resolved at a resolution and in conditions where modelled approaches struggle, including on natural surfaces where friction is variable and difficult to assume.
Why this matters for profiling and screening
Two applications benefit most. The first is multi-load horizontal profiling. Building a force-velocity relationship across several sled loads is far more defensible when the force axis is measured rather than reconstructed, because measurement error does not compound with modelling error. The second is asymmetry. Identifying a left-right difference in horizontal force production requires force resolved per side. That information cannot be recovered from a single velocity trace, which by definition describes the system as a whole.
For return-to-sport and injury screening, where a meaningful between-limb difference may be small, the difference between a measured value and an inferred one is the difference between a usable flag and noise.
The practical takeaway
Force-velocity profiling is only as trustworthy as the force that goes into it. Inferred force is appropriate in many settings and will remain widely used. But where the priority is accuracy, horizontal specificity, and per-limb detail, measuring force directly is the higher standard. For practitioners building horizontal profiles or screening for asymmetry, it is worth asking of any tool a simple question: is this measuring force, or calculating it?
References: Samozino et al. (2016); Morin and Samozino (2016); Cross, Brughelli, Samozino and Morin (2017); Cross et al. (2018), resisted sprint optimal loading.
Mark Fisher
More from Sports Science
