Sports science and strength & conditioning practice is built on a foundation of identifying a problem, testing the problem, applying an intervention and then re-testing to ensure progression. Athletes will buy into fitness testing, injury prevention and subsequent high performance behaviours if they are given the impression that their coach and medical team know what they are doing and things are done with a purpose (Kristiansen and Larsson, 2017). This begs the question whether coaches can justify and clinically reason their battery of performance tests.
When applying a performance measure, understanding of the underlying kinematics is essential to understand the validity of the test to the desired outcome. The OptoJumptm is a valid tool in assessing a reactive strength via drop jump (Healy et al., 2016) however what components of the jump is the coach wishing to address? The validity of the tool is the not the same as the validity of the test. For example, reactive strength index (RSI) can be influenced by a reduced contact time (stretch shortening cycle via the musculotendinous unit) or via total jump height (power output throughout the lower limb and nervous system) or a combination of both (Healy et al., 2017). Understanding these mechanisms may influence the instructional bias of technique given by the coach in order to test what is desired.
With complexities over a test like an RSI to something seemingly obvious like a jump, testing for broader components of fitness and multiple movement patterns is much more difficult.
The Yo-Yo intermittent recovery test (IRT) is reported to be a valid measure of fitness and correlates to match performance in football (Krustrup et al., 2003). However, this is an example of a fitness capacity test and in fact correlates to fitness capacity in a match scenario. In field based team sports, there are a large number of variables and complex interactions that all contribute towards “performance” as an outcome (Currell and Jeukendrup, 2008). Krustrup’s conclusion was based on correlated Yo-Yo IRT results to high speed running in a game (>15km.h-1) with a strong correlation (r=0.58). Overlooking the methodological accuracy of this (pre-GPS, using VHS locomotive assessment retrospectively), the correlation is between two differing metrics. Where the high speed running was recorded over 90 minutes of varying intensities and periods of effort (12 players across 18 different games), the Yo-Yo IRT covered 1.7km in a mean time of 14.7mins with incremental increases in pace dictated externally. For a test to be considered a valid indicator of performance, it should meet the same metabolic demands as the sporting activity (Currell and Jeukendrup, 2008). The Krustrup paper does not make this comparison, instead analysing physiological markers from rest to exhaustion during the Yo-Yo IRT, not exhaustion markers in comparison to game data.
Perhaps semantics, but in fact there should be differential terminology to distinguish “fitness performance” from “athletic” or “sporting performance.” It should be considered that sporting performance is influenced by a large number of uncontrollable and non-modifiable factors that would make any comparison of validity and reliability to outcome measures unfair. Essentially, recreating a competitive environment is near impossible. This raises the question whether we are exercising just to improve test scores or, closing the loop and relating exercises to performance? Does raising the envelope of one, consequently improve the other? Something that we should not only be asking ourselves, but a question we could come to expect from coaches and athletes a like.
Does the research answer this?
It has been suggested that stronger athletes produce faster sprint time, quicker change of direction speeds and higher vertical jump scores when compared to weaker athletes of the same sport (Thomas et al., 2016). Squat jump (r = -0.70 to -0.71) and counter movement jump (r = -0.60 to -0.71) demonstrate strong correlations to change of direction speed (Thomas et al., 2016). Peak force during isometric mid thigh pull was significantly correlated to 5m sprint time (p <0.05) however this correlation was only moderate (r = -0.49). But again, does this correlation transfer into performance if the testing protocol doesn’t accurately mirror sporting performance?
Sprint times over 40m have been shown to decrease following an acute bout of heavy loaded squats, hypothesised to be due to post activation potentiation (Mcbride et al., 2005). Higher squat strength scores also correlate with sprint times over 0-30m (r= 0.94, p=0.001) and jump height (r = 0.78, p=0.02) (Wisløff et al., 2004). Importantly, we know sprint performance tests have demonstrated construct validity to the physiological requirements of a competitive field based game (soccer) (Rampinini et al., 2007), which is ultimately what we are aiming to do; relating performance testing to physiological and metabolic markers from a given sport.
The addition of a jump squat exercise into a training program may help improve 1RM squat and 1RM power cleans (Hoffman et al., 2005). So perhaps yes, there is a perpetuating loop between exercise, tests and performance but the link between them all may not be tangible or direct.
But how do we translate all of these statistics and data sets this to a non-scientific population, as a lot of our athletes are? I’ve developed the following analogy to try and help with this.
Solar system analogy:
If we consider that “athletic performance” is the main focus of any intervention, much like the sun at the centre of the solar system. This is the bright light that everything revolves around; media, finance, fan base and support and so on. It could be argued that any intervention we have as coaches will never truly replicate “athletic performance” but should be influenced by it. This influence works both ways, positively and negatively. For example, if we maximally test an athlete before a competition, this will likely have a negative impact on “athletic performance”. Conversely, if we were able to collect data that informed a training program to improve athletic performance, despite not actually replicating “athletic performance” it would (hopefully) have a positive impact. For example, a football game is determined by so many uncontrollable variables that can not be replicated in a gym, but we might identify that a player needs to improve their 5m sprint time which in turn, will benefit performance.

Let’s consider our potential interventions to be orbiting the sun (Figure 1). There is an interaction between the planets and the sun via gravity but they do not have a direct overlap, where the planets do not collide with the sun just as an outcome measure does not truly match sporting performance. We know that larger planets have a greater influence, so as coaches, we are trying to affect the level of positive interaction with “athletic performance”, the gravitational interaction. By influencing links between exercise intervention and outcome measures, we can affect the size of these planets. In turn, this will have a greater interaction with the centre of our solar system, “athletic performance” (Figure 2). Much like the universe, there will be many different solar systems just as there are different sporting codes and contexts, so the skill lies in identifying the most influential planets in your solar system.

A clinical reflection:
For long term injuries, I utilise a continuum to guide return to play (train / play / perform), often these stages are guided by outcome measures linked to goals and aims for stages of rehab. Typically these tests are scheduled in advanced and often follow a planned “de-loading” micro-cycle. This helps with continuity and, as much as you can in sport, standardisation of the test.
A recent case study found me questioning my judgement and to a degree, wondering if my intrigue and curiosity about my rehab plan drove me to test out of sync with the schedule, instead of doing the test for the athletes benefit.
Following a good period of return to train, the proposed testing date previously scheduled clashed with a squad training session. Observational assessment suggested the athlete was coping well with the demands of training and it seemed counter-intuitive to pull them out of training to undertake some tests. A few weeks later, a gap in the daily schedule presented an opportunity to re-test. The test scores were down compared to the previous month, most likely because the athlete had trained in the morning and trained the 4 out of the last 5 days in some capacity. In previous tests, the athlete had come off of a de-load week and tested the day after a day off.
The result:
The athlete began to question their ability and availability to train. They were visibly knocked in their confidence given a drop in scores, despite me being able to rationalise why this could be. Having had the opportunity to feed my own interest and try to prove to myself that a rehab program had worked, the outcome was much worse. I threatened the confidence of a long term injury returning to training, potentially adding doubt and hesitation to their game and I did not get the results I was expecting.
On reflection, given their time out through the season so far, I should have stuck to protocol and tested on the scheduled day (one training session was not going to increase their chances of availability).. or, not tested at all. Instead, i shoe-horned some testing into an already busy schedule. What did I expect given the current level of fatigue?!
Previous results had reached a satisfactory level to return to train and I was now chasing the final few percentages available. To give them confidence? Probably not, as they were training and enjoying the return to train. So perhaps it was just to give myself confidence. An interesting lesson learnt, mostly about myself.
Yours in sport,
Sam