Or did it?

Before we start attributing that improvement to the new strength programme, plyometric block, recovery strategy or anything else we’ve changed, there is a more fundamental question we need to answer: Is that 5% change actually larger than the amount of variation we’d normally expect?

It’s really important to state from the outset here if it’s not already clear, not every change in a number represents a meaningful change in the athlete.

Different doesn’t automatically mean anything has changed

This is one of the most important concepts in objective testing. If I test an athlete twice, I don’t expect to get exactly the same number. Even under very well-controlled conditions. An athlete might jump 40.2 cm on Monday and 39.6 cm on Wednesday. That doesn’t necessarily mean they became worse. Likewise, they might jump 41.0 cm the following week. That doesn’t necessarily mean they became better. There is always some amount of variability around a measurement. Our job is to understand enough about that variability to distinguish:

Signal from Noise.

That’s really what this entire conversation is about.

Where does the noise come from?

When a measurement changes, there are multiple possible sources. The first is the athletes themselves. Human performance naturally varies, that much is commonly accepted (I hope!). Factors such as sleep, fatigue, motivation, soreness, nutrition, hydration, training load, intent, coaching cues and much more can influence the results between reps and between testing sessions. No matter how well we control the environment, we can never fully control the athlete, they’re not robots.

The second source of potential change is the testing process itself. Did they use exactly the same technique? Did the warm-up, testing instructions and number of trials change between testing sessions or remain the same? Did the same coach administer the test or was it a different coach with slightly different cues? There are a bunch of key questions around the process that need to be considered when looking at comparison data.

The final source can be the measurement system itself. For example, if you’re using force plates, it’s important to understand the typical variation of metrics collected with that system. I’d always recommend utilising technology that has undergone independent, peer reviewed validation and reliability testing. For example, I use ForceDecks, and below is a section from the Collings et al (2024) study where they investigated validity and reliability of all key tests utilised within the system..

The goal of highlighting these potential sources of noise isn’t to have zero variability in testing, it’s to be aware of what’s normal and anticipate.

Reliability matters before change matters

Let’s look at 2 imaginary metrics for a moment. “Metric A” typically varies by around 2% between repeated tests. “Metric B” typically varies by around 12%. After completing a re-test we observe that both metrics have improved by 5%, should we interpret those changes the same way? I’d say probably not.

In metric A, an improvement of 5% is relatively large, keeping in mind it normally deviates by 2%, so 5% is above the point of normal variation. In metric B however, 5% could easily sit within the amount of variation we’d naturally expect anyways. While with metric B, which normally shows variation of around 12%, a 5% improvement may sit within the normal degrees of variation anyways. This can turn you into a bit of a buzzkill when an athlete initially sees the green % increase on the screen, but unfortunately, that’s your job.

This is why reliability should be considered before we become excited about change. If a metric can’t be measured consistently, it’s difficult to use small changes in that metric to guide decisions.

What does reliability actually mean?

Read more

If jump height is the only thing we were interested in, that’s probably a reasonable conclusion to come to, but as this series has constantly advocated for, going beyond just the outcome metric is where the real value lies for understanding neuromuscular changes in our athletes. The athlete may have achieved the same jump height, but taken significantly longer to perform the jump or increased their countermovement depth to be able to reach the same jump height.

The distinction between looking at an athlete’s performance and their movement strategy is an important one and for me, it’s just as important to understand when you should be looking at each. Performance based feedback can be crucial in profiling / re-testing sessions to quantify progress as both the athlete and coach are looking for some validation in the work they’ve put in. Insights around their strategy are important descriptors for the same profiling and re testing scenarios, as we need to understand their basic movement signature. However, tracking the change in that strategy over time and mapping it with relevant context, training stimulus and other factors are crucial.

Start with the outcome

Imagine these two jumps:

But it’s difficult to argue that the performances are identical in every respect. They’ve produced the same external outcome under a different time constraint and with different force-velocity characteristics. Whether that matters depends on what we’re trying to understand.

Movement Problems Can Have Multiple Solutions

This is an important principle. Due to the variability of humans in general, there is no ‘textbook’ approach to jumping. Depending on strength levels, task proficiency / familiarization, anthropometrics and a range of other factors different athletes will lean into different strategies to produce the same 40 cm jump height. One may be a deeper, longer countermovement while the other may be a shallower, more explosive effort. Depending on the athlete, their sporting requirements and training emphasis, neither of these are right or wrong.

Sport is full of examples where athletes achieve similar outcomes using different movement strategies, think 5 foot 7 inch Lionel Messi jumping against 7 foot 4 inch Victor Wembanyama. While the outcome in heights achieved may be different, visually the strategy would likely be massively different too.

Force plates simply give us a way of quantifying some of those differences with hard numbers. At the end of the day both athletes are jumping, they’ve just solved the task with different answers,

The Same Athlete Can Change Strategy

Tracking athletes longitudinally provide a real opportunity for appreciating how strategy fluctuates and changes in line with training adaptations and response to load. Body mass changes, strength developments, movement skill increases, injury occurrence can all change across the phase of a season and as such, we can expect to see knock on effects in movement strategy. The key is understanding when a change is big enough to warrant an intervention. That’s our job as practitioners. Lucky us right!

At the end of a period of training, when sharing results from re-testing with the relevant stakeholders, imagine this:

Read more

Perform three countermovement jumps and within seconds you can have jump height, peak force, mean force, peak power, peak power relative to body mass, eccentric duration, concentric duration, braking impulse, concentric impulse, concentric peak velocity, eccentric peak velocity, rate of force development, contraction time, RSI-Mod and dozens of other variables sitting in front of you. Seriously impressive, but it does raise the question of which ones she should care about.

Access to metrics alone isn’t what gives us more useful information. In fact, I think one of the biggest progressions a practitioner can make with force plate testing is moving from what can I measure to what do I need to measure? This article dives into ‘some’ of the key metrics I’d recommend practitioners look at with their testing. It’s important to note that depending on the specific circumstances, certain metrics become more valuable than others.

Start with the outcome

If I’ve asked an athlete to perform a maximal countermovement jump, one of the first things I want to know is how well did they actually perform the task of jumping maximally. A maximal jump to me, at least, would showcase their maximum jump height. For that reason, I’d automatically start by looking at jump height. Jump height as a concept itself can cause debate depending on which calculation you choose to use. As mentioned early in this series, when using force plates and I’m fully confident in protocol execution, I’ll use the impulse-momentum (IMP-MOM) calculation for jump height. You might be reading this feeling like I should be diving deeper, but as an outcome, jump height is literally it. So many metrics feed into it, which we’ll look at shortly. Ultimately, if jump height has decreased from the norm, you can bet it’s not an isolated change to only that metric.

Then ask how they produced it

Once we have a decent feel for the jump itself from an outcome metric point of view. The next layer of the onion is to examine the strategy used through the movement to achieve that jump height. This for me, is where force plates become to really show their value in comparison to devices that just give you an outcome metric. A simple way to remember it is if jump height shows us ‘what happened’, then strategy metrics show ‘how it happened’.

At this stage, it can become overwhelming. We’ve crossed jump height off our list of 200+ metrics, now across the remaining 199 how do we decide which to invest our time in? I feel at this point it becomes necessary to divide metrics into categories. The below graphic details those metric categories and an example of a metric that may fall into it. It’s important to note that some metrics may overlap into different categories and this isn’t meant to be exclusive. For me at least, it’s been a good way to understand what I’m looking at.

Read more

This week, I want to take that a step further and discuss what I consider to be one of the most important concepts in understanding jump performance: Impulse.

When coaches first start exploring force plate data, there’s a natural tendency to pay a lot of attention to force-related metrics. Peak force, mean force, eccentric force, concentric force, rate of force development and so on. It’s understandable, particularly when we consider how much emphasis we place on developing force-producing capabilities through strength and conditioning. If we’re trying to jump higher, surely producing more force is a good thing?

Generally speaking, yes. However, the important distinction here is that force alone doesn’t determine how high an athlete jumps. We also need to consider how long that force is applied for. This is where impulse becomes particularly interesting. Once we understand the relationship between force and time, we can begin to explain why athletes with very different force profiles can achieve similar jump heights, why the highest peak force doesn’t necessarily belong to the best jumper, and why changes in jump performance aren’t always as straightforward as we might think.

What Actually Is Impulse?

At its most basic level, impulse represents the application of force over a period of time. The simplest way to demonstrate this is through the following equation:

Impulse = Force × Time

Example 1: we apply a constant force of 2,000 N over a period of 0.10 seconds.

2,000 N × 0.10 s = 200 N·s

Example 2: we apply a considerably lower force of 1,000 N, but over twice the duration.

1,000 N × 0.20 s = 200 N·s

Despite having very different force magnitudes and durations, both examples result in exactly the same impulse. Of course, when we look at a countermovement jump, force isn’t constant throughout the movement. As we established in Part 3, the force being applied changes continuously as the athlete moves through the different phases of the jump.

Rather than using one force value multiplied by one duration, we calculate impulse by determining the area underneath the force-time curve over the period we’re interested in. Think about that force-time curve again. The vertical axis represents force, while the horizontal axis represents time. The area underneath the curve combines both of these variables into a single value.

This is an important concept to understand because it immediately demonstrates a limitation of looking at force values in isolation. A very high peak force might look impressive on a report, but if it is produced for an extremely short duration, it doesn’t necessarily tell us a great deal about the overall impulse generated during the movement. Importantly, the specific period of the movement we’re examining, and whether we’re looking at total ground reaction force or net force, will determine what that impulse value actually represents.

Why Does Impulse Matter for Jumping?

Now that we’ve established what impulse actually represents, the next question is why we should care about it.

To answer that, we need to return to some relatively basic physics.

Impulse for as long as I’m aware has been explained as a change in momentum.

Momentum is calculated as:

Momentum = Mass × Velocity

When an athlete begins a countermovement jump from a stationary standing position, their vertical velocity is initially zero. Throughout the movement, their centre of mass accelerates downward, changes direction, and eventually develops an upward velocity at take-off.

Read more

The action of jumping looks relatively straight forward, the athlete stands still, they dip down, drive up and stick the landing. Then the software (usually instantly) gives you a graph with force on one axis, time on the other, multiple phases of movement, and potentially dozens of metrics describing what happened within a movement that lasted little more than a second.

This is the part, and was the part for me starting out, where jump testing starts to feel intimidating for a lot of coaches. Before getting overwhelmed by the length and breadth of available metrics and the need to understand them all, it’s important to keep in mind that everything is derived from 2 measurements, force and time.

The force-time curve (aka force trace) allows us to see simply how much force the athlete applied and when they applied it. Once you understand the basic shape of the curve and potentially how it deviates from what you’d expect, it can point you to the metrics worth looking at pretty easily.

Start With What The Force Plate Actually Measures

This is an important place to start. As I mentioned in the previous section, jump height, impulse, power or velocity aren’t directly measured by the force plate, they’re derived from Force and Time. More relevant to this topic, in the context of a vertical jump test, we’re primarily interested in the vertical force applied to the ground.

The force plate measures the interaction between the athlete and the ground and from that measurement, with some additional information such as the athlete’s body mass, we calculate the multitude of variables that are available instantly as metrics after each jump. I think it’s really important to emphasize that there are 2 categories for these variables. Measured and derived.

Measured variables are force and time while derived variables are everything else; velocity, displacement, power, impulse, jump height and so on. When you look at a force time curve, it’s foundational to the entire analysis. Being able to spot discrepancies in a force time curve, can lead you to understand knock on effects that may be displayed in the numbers.

What Is a Force-Time Curve?

The clue is in the name. The X axis is typically time while the Y axis is typically force. As an athlete moves, time will continue to move at the same rate, while the rate of force being applied varies based on movement demand and physical ability. Plot those changes in force over time on a chart and we are presented with a force-time curve. Importantly, if an athlete is standing still on the plates, the force will be constant at a number equal to their body weight.
For a 100kg athlete, we’d expect that force to be 981 N as a reference point. The logic for that calculation is pretty concrete;

Force = Mass x Acceleration

Acceleration due to gravity is approximately 9.81 m/s².

So for the 100Kg athlete; 100kg x 9.81 m/s² = 981 N

This isn’t exactly exciting as the athlete is just standing still, but fundamental to be aware of before moving forward. Once movement begins, then we see a more dynamic force-time curve emerge.

The First Thing That Surprises People

When an athlete begins a countermovement jump, the force doesn’t immediately increase. It initially decreases. Particularly for those new to interpreting force traces, it can be easy to associate the decrease on the graph as movement downward and the increase as movement upward. I made that mistake at the very start when I was using my Pasco plates. The actual expected behaviour on the force trace makes more sense when we tie it back to biomechanics and basic principles of physics.

Read more

An athlete could perform the same countermovement jump and receive different jump height values depending on the technology being used and, importantly, how that technology calculates jump height. I think this is a crucial knowledge gap regarding the utilisation of technology in our industry currently. People need to understand exactly what goes into the number you receive immediately after a jump.

The differences in results aren’t because one device is good and another is bad; it’s often down to the fact that the numbers have been calculated differently. The most important consideration from this for me is if we’re going to track athletes over time, compare athletes against normative data, or compare results between systems, understanding that difference matters.

So before getting anywhere near the more complex metrics available from jump testing, I think we need to understand the metric almost everybody starts with, jump height

You Don’t ‘Measure’ Jump Height

This is the first important concept to wrap your head around. Most technologies don’t literally measure how high the athlete’s centre of mass travels in the air. Instead, they measure other variables and use that information to calculate jump height. Depending on the technology, those variables might be flight time, force, velocity or displacement. The distinction is important because the calculation determines what assumptions we’re making.

The two most common calculations for jump height practitioners will encounter in the industry are either flight time or impulse-momentum derived jump heights. Both give a value in the same unit (centimetres, inches, metres), but both are calculated very differently.

Flight Time

Conceptually, the flight time method is very simple. When an athlete leaves the ground, gravity acts as a constant force to pull them back to the ground. If we can accurately measure how long the athlete spends in the air, we can estimate how high they must have traveled.

Spend more time in the air, get a higher flight time number and in turn, a higher calculated jump height. Very easy to understand at a high level in my opinion. This is why we see technologies like jump mats and optical systems being able to estimate jump height without actually measuring force. They detect three key phases; Takeoff, flight and landing. Then they ultimately calculate jump height from the duration between takeoff and landing (flight).

This all sounds great and almost makes you think, why do we need any more complexity with jump height if this works so robustly. The key thing to highlight at this point is the assumption hiding underneath the flight time calculation.

The Landing Matters

It’s assumed that the athlete’s centre of mass is at the same height during takeoff and landing. Athlete’s usually do everything in their power to challenge this assumption, such as throwing their legs forward in the air to generate a longer flight time or landing in a deep squat position before their feet contact the ground. All of which ultimately inflates the flight time and results in overestimated jump heights. In these examples, the athlete changes their strategy while in the air, but the takeoff effort doesn’t necessarily propel their centre of mass any higher.

For these reasons alone, standardising execution of testing with flight time calculations is the most important factor in accurately quantifying if you’re actually making progress with your athletes. This is why something as potentially trivial as landing from a jump is so important. When we see a 4cm increase in jump height from 40cm to 44cm, we need to make sure we’re confident in how the movement was performed to have true confidence in the progress we’re making with our athletes.

Impulse-Momentum Takes a Different Route

When we have force plates, we have another option. Rather than waiting until the athlete lands and calculating height from how long they were airborne, we can use what happened before they left the ground.

Read more

The addition of technologies (such as ForceDecks) has made this considerable easier. What once required a million dollar biomechanics lab can now be performed relatively quickly on the gym floor, training facility or even in the locker room. Force plates can capture thousands of datapoints during one jump and immediately spit out hundreds of metrics. This is awesome, but, it can also create a potential problem.

As an industry we’ve gotten very good at collecting jump data. Being able to collect proper data with high level protocols is one thing, being able to interpret said data is a whole different kettle of fish! Before we dive into metrics, force-time curves, impulse, eccentric braking, or any other topics this mini series dives into, the main question to ask right now is why are we testing jumps in the first place?

Read more

The Study

Xu and colleagues used a countermovement rebound jump (CMRJ), where the athlete performs a countermovement jump, lands, and immediately performs a second rebound jump. This gives us two distinct actions within the same task: an initial countermovement jump followed by a rebound action.

Twenty-five recreational team-sport athletes performed the test under three cueing conditions. The height cue (HC) asked athletes to reach …

Read more

It’s what VBT allows us to understand when we combine it with other information. A GymAware measurement tells us something about how an athlete moved an external load and that’s useful. But athletes don’t compete in a GymAware assessment. The goal is to understand whether the qualities we’re developing in training transfer toward the physical capabilities required by the sport. That’s where integration becomes important.

Think in Physical Qualities, Not Technologies

One mistake I see in sports technology is allowing the device to define the assessment system.

Where VBT Fits

VBT provides a particularly useful window into strength training performance. A standardized external load can become a repeatable stimulus. Velocity tells us how the athlete performed against that stimulus. Over time we can ask: Is the same load moving faster? Is the athlete capable of moving h…

Read more

Whenever a new measurement technology enters our workflow, there’s a dangerous period. We can collect the data before we fully understand it.

VBT is no exception. A device can tell me that an athlete’s squat velocity changed from 0.61 m/s to 0.64 m/s. That’s a 4.9% improvement.

But the important question isn’t: Did the number change? It’s: Did the athlete change?

Those are not necessarily the same thing.

Reliability Comes First

Every measurement contains noise. Athlete variability. Technical variability. Device variability. Protocol variability. If the typical variation in a measurement is ±5%, then celebrating a 2% improvement probably isn’t justified. Some more cost effective LPT (linear position transducer) solutions measure movement of a tether assuming it’s vertical, when we know that bar path is rarely completely vertical. GymAware accounts for bar path by correcting the angle of the x-axis…

Read more

Understanding VBT theory is relatively easy. Implementing it without making training unnecessarily complicated is harder. If every exercise has a velocity target, every set has a velocity-loss threshold and every repetition requires the athlete to stare at a screen, we’ve probably gone too far. My preference would be to identify where velocity improves the program.

Start With the Training Goal

Before attaching a device to anything, ask: “What adaptation/environment are we trying to create?” Maximum strength? Power? High-velocity intent? Hypertrophy? Fatigue resistance?

The application of VBT should follow that answer.

Heavy Strength Work

During heavy compound exercises, velocity can provide context around relative intensity. Instead of prescribing only: 4 × 3 @ 85% 1RM, we could incorporate an expected velocity range. The athlete works toward the intended loading zone while the coach observes whe…

Read more

“Autoregulation” sounds sophisticated. In reality though, good coaches have been doing it forever. An athlete walks into the gym looking exhausted. Warm-ups look poor. The first working set is unusually difficult. The coach adjusts the session.

VBT didn’t invent autoregulation. It potentially gives us another objective piece of information to support it.

A Simple Example

Imagine an athlete regularly squats 100 kg during their warm-up. Across previous sessions, their mean velocity is typically around 0.72 m/s. Today its hitting 0.63 m/s.

Something is different, but we don’t definitively know what. Perhaps they’re fatigued, they travelled yesterday, they played two nights ago or their technique just changed. We need to get in the frame of mind to critically evaluate if the measurement itself is noisy.

The velocity doesn’t tell us why performance changed. It tells us that performance changed. That d…

Read more

Part 4 of an 8-part series exploring Velocity-Based Training from a practitioner’s perspective.

Most resistance training programs prescribe volume using repetitions. Three sets of five. Four sets of eight. Five sets of three. It’s simple, practical and easy to communicate. But there’s an assumption hidden inside that prescription that the same number of repetitions represents a similar stimulus for everyone.

We know that’s not necessarily true. VBT gives us another way to look at the problem. Instead of only counting repetitions, we can measure how much performance declines during the set.

What Is Velocity Loss?

Using the example of an athlete squatting 100 kg:

Rep 1 — 0.70 m/s

Rep 2 — 0.69 m/s

Rep 3 — 0.66 m/s

Rep 4 — 0.63 m/s

Rep 5 — 0.59 m/s

Rep 6 — 0.56 m/s

Compared with the fastest repetition, velocity has now fallen by approximately 20%.

That decline can be described as velocity loss. Rather than prescribing six repetitions, we could theoretically prescribe to simply continue the set until …

Read more

Testing maximal strength presents an interesting problem. A true 1RM can provide useful information, but repeatedly exposing athletes to maximal attempts isn’t always practical or necessary, particularly in environments where strength training supports sporting performance rather than being the sport itself.

This is where the load–velocity relationship becomes interesting. If heavier loads consistently move slower, can we use submaximal velocities to estimate where an athlete’s maximum might be?

Building the Profile

The example below shows that an athlete performs a squat at several progressively heavier loads:

80 kg — 0.85 m/s

100 kg — 0.70 m/s

120 kg — 0.54 m/s

140 kg — 0.38 m/s

Plot those values and we can show a visual describing the relationship between load and velocity. If we have an appropriate estimate of the velocity associated with maximal loading for that athlete and exercise, the relati…

Read more

One of the easiest ways to make sports technology unnecessarily complicated is to collect every metric a device gives us.

VBT is no different.

Attach a linear position transducer to a bar and suddenly we can have velocity, peak velocity, power, force, displacement, rep duration and numerous other variables. The temptation is to assume that more information automatically means better decision-making.

It doesn’t. For me, the better starting point is always: What question am I trying to answer?

Mean Velocity

Mean velocity represents the average velocity across the concentric phase of the repetition. For traditional strength exercises such as the squat or bench press, this is often one of the most useful variables.

Why? Because it provides a relatively stable representation of how the external load moved throughout the repetition. If I’m trying to establish a load–velocity relationship or monitor how …

Read more

Velocity-Based Training (VBT) is another piece of that puzzle.

At its simplest, VBT involves measuring the velocity of an external load during resistance training and using that information to help inform training decisions. Before getting into velocity zones, autoregulation, velocity loss or estimated 1RM (one-repetition maximum), however, there is one relationship that underpins almost everything in VBT:

As load increases, movement velocity decreases.

The Load–Velocity Relationship

Imagine an athlete performing progressively heavier back squats.

At 60 kg they might move the bar at 1.00 m/s.

At 80 kg, 0.82 m/s.

At 100 kg, 0.65 m/s.

At 120 kg, 0.48 m/s.

At 140 kg, 0…

Read more

Recently, it was announced that GymAware is becoming part of VALD, which is an exciting addition to the broader VALD ecosystem and, for me personally, a great opportunity to talk a little more about velocity-based training (VBT).

At its simplest, VBT gives us another way of understanding what’s happening in the gym. Traditionally, resistance training is largely prescribed around load; 100 kg, 80% of 1RM, three sets of five, and so on. All perfectly useful information.

Velocity simply adds another layer. Instead of only knowing what an athlete lifted, we can understand how they lifted it.

How quickly did they move the bar? How does that compare to previous sessions? Are they getting stronger? Is today’s performance different from what we’d normally expect? Are they actually producing the intent we want from the exercise?

But like any technology, collecting the number is the easy part. Understanding …

Read more

So that’s exactly what this article is.

Rather than simply posting the graphic, I wanted to walk through each section and explain not just what I’d do, but why I think it’s important.

One point I’d make before diving in is that this isn’t intended to be the definitive framework for athlete profiling. Every coach will have their own preferences, testing battery and philosophy. This is simply the process that I’ve found myself gravitating towards over the past few years. It helps me organise my thinking, make better decisions and, ultimately, keep the focus where it belongs, on the athlete rather than the numbers.

1. Assess the Key Physical Qualities

Read more

Like most practitioners, I was introduced to the gold standard warm up structure, alsa known as the RAMP framework: Raise, Activate, Mobilize and Potentiate. It’s a framework that’s stood the test of time and, in my opinion, it’s still a fantastic way of preparing athletes. This article isn’t me saying RAMP is wrong or that I’ve somehow found a better way. It’s simply a small adjustment that I’ve naturally gravitated towards over the last few years because I’ve seen it work well with the athletes I’ve been fortunate enough…

Read more

Over the last year or two, we’ve seen a lot more force plate options come onto the market. This has probably accelerated in the last few weeks with more options being released by more companies. Many of them are significantly cheaper than the established systems, and on paper they can look like a great deal. Accurate hardware, no subscription fees, one-off payment... it’s easy to see why they’re catching people’s attention.

And honestly, for some practitioners, they might be exactly the right choice. But I think there’s a question that often gets overlooked when people are comparing spec sheets and prices.

Where do you see yourself as a practitioner in three years’ time?

Because in my experience, buying …

Read more