Why Force Plate Normative Data Fails Your Athletes (And 1 Strategy to Improve Interpretation).

If you use a force plate for athlete monitoring, you are almost certainly comparing your athletes against a normative database. The platform you use probably does this automatically — a percentile appears next to each metric, and you use it to decide whether an athlete’s result is good, average, or below par.

This is a reasonable approach. Raw force plate numbers are difficult to interpret in isolation. Normative comparisons provide context. The logic is sound — until you look at who was actually in the database.

Why Force Plate Normative Data Exists (and Why the Logic Is Sound).

A peak braking power of 23 W/kg, a jump height of 31 cm — without context, these numbers are hard to act on. Normative databases address this by collecting large athlete samples and building a distribution. Any practitioner can then see where their athlete sits relative to the broader field.

This is genuinely useful. A 40th percentile result may signal capacity to develop. An 80th percentile result may suggest strength is not the limiting variable. These are actionable inferences — when the reference group is relevant.

The problem is that “relevant” is doing a lot of work in that sentence.

Why Normative Data Is Often Misleading for Elite Athlete Benchmarking.

Most published normative databases were built from the populations easiest to recruit: university-aged recreational athletes, general sport samples, and in many cases, populations where men substantially outnumber women and elite athletes are outnumbered by their development-level counterparts.

Some databases are more specific — sport-separated, position-separated, competition-level separated. These are an improvement. But they are still built from the same underlying logic: a group average as the benchmark.

Here is the clearest way to see the problem.

Take 20 world-class weightlifters and run a max strength battery. One athlete lands in the 5th percentile. That athlete is almost certainly not weak — the floor of that distribution is already extraordinary. The percentile is technically accurate. The distribution is the problem.

Now take 20 young female ice hockey players who have never done a structured strength program. The strongest athlete in the room tops the chart — 99th percentile by every metric. She is, in all likelihood, still undertrained.

The percentile is accurate in both cases. The reference group is not. A group average tells you where an athlete ranks within a room. It says nothing about whether the room is right.

The Reference Group Problem: You Might Be Using the Wrong Population.

The reference population determines what a percentile means. When practitioners use published normative databases without interrogating who was in them, they are trusting a comparison that may have been built on athletes who do not resemble their own population in sport, sex, training history, or competition level.

This problem is most acute in three contexts:

Female athletes. Many widely used normative databases were developed primarily or exclusively from male samples. Applying male-derived norms to female athletes produces comparisons that are systematically off — either inflating how an underdeveloped female athlete appears, or creating unrealistic benchmarks for a trained one.

Youth and development athletes. Normative databases often pool athletes across wide age ranges or mix development and elite performers. The result is a reference group with a wide performance distribution that obscures meaningful differentiation within a development cohort.

Specialized sport populations. A weightlifter, a sprint hurdler, and a field hockey midfielder all load the force plate differently. Comparing any of them against a general athletic population or even a broad “team sport” norm produces comparisons with limited interpretive value.

The Equal Weighting Problem: 30 Metrics, 30 Equal Signals.

There is a second problem beyond the reference group that receives less attention.

When every metric in a force plate dataset is evaluated against a normative distribution independently, each one is implicitly assigned equal weight. A force plate assessment can generate 30 or more variables — jump height, peak braking force, relative concentric impulse, eccentric deceleration impulse, landing stiffness, and many more. Examining each one against a population norm treats every variable as equally important for every athlete.

But what is “good” for one athlete profile is not necessarily good for another of the same age, sex, and sport. What distinguishes high-performing rugby props from average ones is not the same as what distinguishes elite sprinters. And what matters at the beginning of a strength program is not what matters for an athlete with four years of progressive training history.

Treating all metrics equally obscures this. A practitioner reviewing 30 percentiles across 30 metrics does not have more information; they have more noise.

How to Build Better Benchmarks: Internal Norms and Athlete Profiling.

The response to both problems is the same: build a reference group that actually reflects your athletes.

Athlete Profiling.

At the more sophisticated level, profiling replaces normative comparison with pattern recognition.

Rather than asking “where does this athlete rank in a database,” profiling asks “how closely does this athlete match the neuromuscular pattern of high performers in our specific environment?” Using data-driven methods — including unsupervised machine learning approaches — it is possible to identify the combination of metrics that characterize your best performers across sport, sex, and training age.

This approach addresses both problems. It uses your own athlete data as the reference, so the population is relevant by construction. And it identifies which metrics actually distinguish high performers in your environment — solving the equal weighting problem by surfacing athlete-specific key performance indicators rather than treating all metrics equally.

The output is not a percentile. It is a pattern fit: how closely this athlete matches the profile of athletes who perform well. Combined with longitudinal trend data — is this athlete moving toward or away from the target profile over time? — it provides considerably more actionable information than a spreadsheet of 30 normative comparisons.

What Changes on Monday Morning.

For practitioners who cannot yet build robust internal norms or implement profiling, two practical shifts improve the current approach immediately.

Interrogate the reference group. Before relying on a normative comparison, identify which database the platform is using and whether the reference population is appropriate for your athletes. VALD Hub, for example, uses its own normative database. Understanding the composition of that sample (sport, sex, age, competition level) is a prerequisite for interpreting the percentile it produces.

Pair cross-sectional comparisons with longitudinal tracking. An athlete at the 40th percentile who has improved consistently over three consecutive testing blocks is telling a different story from an athlete at the 40th percentile who has not changed in eight months. The normative comparison tells you where; the trend tells you whether the direction is right.

For return-to-sport decisions, where the stakes are highest, the most defensible criterion is not a population percentile but an athlete-specific comparison: does this athlete’s braking profile, reactive strength, and limb symmetry match their own pre-injury baseline? Or match the pattern of healthy athletes who play the same sport at the same position?

The normative database may still serve as a directional tool. But the reference group question comes before the percentile. If you cannot name who was in the database, you cannot know what the number means.

Zach McClean is a sport scientist and PhD researcher at the University of Calgary. His research focuses on ML-driven athlete profiling and force plate data interpretation. He consults with professional sport organizations on performance monitoring and return-to-sport decision-making.

Sign up for McClean Performance Insights — a weekly newsletter for sport science practitioners. https://www.mccleanperformance.com/members