2 Literature review
2.1 Defining parkour techniques
Despite the decentralised growth of parkour, traceurs have adopted an emergent set of “core movements and techniques” (Aggerholm and Højbjerre, 2017). These techniques have names, are grouped according to related features, and are taught in parkour classes. Disputes over the attributes of these movements do not appear common amongst the parkour community in the same way traceurs often debate the broader definition of parkour. They have matured from the experimental movements performed by early pioneers, refined over time by traceurs across the globe, and firmly established themselves as identifiable parkour techniques, meeting the definition of distinct skills in their use to reliably achieve a desired result (Schmidt and Lee, 2019).
Yet parkour’s open-skilled nature (Grosprêtre and Gabriel, 2020), where techniques are performed in dynamic, varied, and continually changing environments, makes it challenging to epitomise any parkour technique for study. Other sports take place in fixed environments with clearly defined objective success markers that can provide structure to analysis, whether that be an immediate binary outcome (a goal was scored, the top of a climbing wall was reached) or a more extensive, nuanced expression of success (a judging panel scoring a routine). These definitions and markers are often codified in rulebooks or judging guidelines which, in turn, set a quantifiable outcome measure to evaluate a movement against, such as investigating the correlations between biomechanical variables in gymnastics and scores given by judges (Schärer et al., 2019). Parkour currently has no widely-accepted rulebook or judging guidelines, despite an increase in parkour competitions (O’Loughlin, 2012). Instead, parkour competitions remain provincial, with any adopted rulesets usually pertaining only to that singular competition.
Such competition frameworks can influence the scientific study of parkour despite a lack of widespread acceptance. Padulo et al. (2019) proposed a parkour-specific fitness test to assess traceurs competitive skills but drew from only one source of competition guidance (International Parkour Federation, 2019) to justify their choice of included skills. Other forms of parkour competition, such as speed run events where a traceur is given a fixed beginning and endpoint but are free to determine their method of traversing the environment, may be a more ecologically valid method of measuring parkour performance (Strafford et al., 2021). However the problem remains that, due to their freeform nature, parkour speed run events do not provide any further guidance on individual techniques as the decision to include, or not include, any movement in a speed run is entirely down to the traceur. There are also criticisms of competitive parkour from within the traceur community that may make it difficult to justify as a source of technique definition. Despite increasing acceptance, competition (notably commercialised, sponsored competition) is often viewed as deviating from the core principles of parkour (Angel, 2016).
Parkour coaching qualifications may provide an alternative source of technique definition that could be usefully applied to research studies. However, within such qualifications, coaching skills—such as class management, or observation and feedback techniques—often take precedence over teaching parkour movement itself. Parkour UK—the national governing body for parkour in the United Kingdom—uses a Competency Assessment Checklist in their coaching qualification, which contains several parkour techniques to be demonstrated or taught by the coach (Parkour UK, 2018). Technique names are listed without further description, and the assessment scoring consists of only ‘competent’ or ‘not yet competent’ marks without any specific guidance on how this competency should be assessed. Instead, the personal experiences and judgement of the tutors delivering the course become the primary criterion against which competency is assessed. This reliance on experience reflects the findings of Greenberg and Culver (2020) that describes a coaches’ prior experience as a traceur as their primary source of technique acquisition and development, developing their skills over time using anecdotal and experiential evidence. The overall experiences of parkour coaches that Greenberg and Culver (2020) report mirrors those of coaches across numerous sports found by Stoszkowski and Collins (2016), with knowledge transfer mainly occurring in the “social milieu” of peer-to-peer conversation and observation.
With no restricted environment and no formal technique definition, biomechanics researchers must therefore impose their own structure upon research into parkour movement, with delimits derived from prior studies or their own observations. It may be easier to justify these delimits for some techniques than others, and this could be the reason biomechanics-focused parkour research to date has predominantly investigated the ability of traceurs to mitigate ground reaction forces in jumping and landing tasks (Puddle and Maulder, 2013; Standing and Maulder, 2015; Grosprêtre, Gimenez and Martin, 2018). Such research can draw on other jumping sports to justify a quantifiable measure—for instance, measuring the height an athlete can jump to indicate their lower-limb power or how their landing style will affect their risk of injury. Only recently have studies begun exploring other aspects of parkour with less apparent parallels to other sports, such as Croft, Schroeder and Bertram (2019) investigating work optimisation in the parkour wall-run technique. Even so, the binary success outcome of either reaching or failing to reach the top of a wall still presents a clear objective marker that can be quantitatively measured - vertical displacement. If an adjustment to a technique increases the vertical distance travelled, allowing a traceur to reach the top of a wall they previously could not, that adjustment can be regarded as having improved their wall-run performance. However, some techniques are more nuanced in their goals and require equally nuanced understanding before study.
2.2 The kong vault
Vaulting an obstacle is a standard action in parkour, and within this sub-group of movement, the kong vault has become a frequently utilised parkour technique. The kong vault (alternatively known as the catpass, saut de chat, or monkey vault (Gerling, Pach and Witfeld, 2013)) consists of the traceur diving onto an obstacle with their hands, tucking their knees into their chest, and bringing their feet through their hands to pass over the obstacle. As reported in a recent narrative review by Pagnon et al. (2021), the commonality of the kong vault has led to it being the only parkour vaulting technique thus far reported on in the scientific literature, albeit in only a single paper. Participants in this single study of the kong vault by Maldonado et al. (2019) performed the vault from a standing position and vaulted over a bar, but no justification was given as to why that particular version of the vault was chosen. The kong vault moniker is also applied to similar techniques performed with a run-up, or from a plyometric movement, or on a flat-top obstacle (Henry, 2017). Consequently, the execution of a kong vault may alter to meet a particular result or to tackle disparate environmental factors.
While judging guidelines exist for vaults performed in a competitive gymnastics setting (Fédération Internationale de Gymnastique, 2017), the measure of a successful kong vault is less clear. Success could be a simple binary outcome—the traceur passed the obstacle or did not—or be broken down into a variety of outcomes based on the current environment or the personal subjective goals of the traceur at that moment. It is not clear how these numerous variables should subsequently be fixed for study in a lab, or why. As a result, the specific findings of Maldonado et al. (2019) may not be generalised to every movement labelled as a kong vault, which is emphasised within the paper when discussing the static nature of the take-off but incorrectly widened to a blanket statement on all kong vaults in the concluding remarks. Before any further quantitative assessment of the kong vault can take place, some attempt must be made to understand, define, and validate the use of any particular version of the technique, both in everyday parkour practise and when considered for study.
2.3 Deterministic models
Chow and Knudson (2011) propose using deterministic models in biomechanics studies to bridge the gap between theoretical and empirical understanding of sports research. Deterministic models are often depicted as block diagrams, showing the relationships between a desired result and the biomechanical factors that produce that result, as outlined in Figure 2.1. The top-most level of a model can be either an objective measure of performance, such as time, or an external subjective measure, such as a judges score. In cases where an individual movement may only contribute partially to an overall measure, it can also be assessed in terms of how successfully it contributes to the greater whole, often referred to as advantage (Hay and Reid, 1988).
Deterministic models are built using expert observation and analysis following the identification of the desired result (Ham, Knez and Young, 2007), but given the subjective and dynamic goal-setting of parkour movement, selecting a result for the top-most level of the model is difficult for the kong vault. Building the remaining levels of the model through observation alone could also be difficult due to the wide array of technique adjustments available to a traceur, influenced by the dynamic environments parkour is performed within. A researcher could never definitively say they have observed every possible permutation of a parkour technique.
Rather than observing a limited example, building a model by compiling the expertise of parkour coaches in using and teaching the kong vault would draw from a much wider variety of circumstances. Strafford et al. (2020) used this method in sampling the experiential knowledge of parkour coaches to catalogue the skills developed by parkour training and organise them into associated groupings, such as “physical” and “perceptual” skills. A similar approach in sampling expert understanding to identify emergent patterns can be used to construct a deterministic model of the kong vault that could summarise the most common aspects of the technique. Permutations within deterministic models can also be accommodated, as seen in a diagram used to describe the long jump by Kamnardsiri et al. (2015, fig. 1) which contained three distinct styles—the hang, stride jump, and hitch kick—within a shared ‘flight and landing’ phase.
However, while a model for the long jump has a quantifiable outcome measure regardless of style—horizontal distance travelled—a kong vault model would also need to investigate and document the outcomes that traceurs desire when using the kong vault. This may result in numerous deterministic models being constructed, each with a different outcome measure at the highest level of the model depending on the chosen goal. Finally, common coaching cues for the kong vault can also be sampled and linked to the mechanical variables they are attempting to control within the model, as seen in a systematic review of Yurchenko gymnastics vault literature by Diener and Aedo-Muñoz (2019) which linked the phases of the vault to a series of kinetic and kinematic performance indicators. Categorising the phases of a kong vault and identifying the performance indicators contributing to each phase would allow coaches and researchers to accurately describe the technique being performed in classes or studies. These indicators could then be associated with quantifiable measures and tested. For example, if maximising horizontal distance is identified as a typical desired outcome, and changing take-off velocity is believed by coaches to be a kinematic predictor of horizontal distance, there would be justification for studying the relationship between these variables.
Sampling the expertise of parkour coaches also serves to build a deeper engagement between coaches and sports science researchers. The degree to which sports science research engages with topics that sports coaches find functionally relevant and helpful has been called into question (Buchheit, 2017). Currently, academic research on parkour is sparse and not well-publicised within the parkour community, but if built collaboratively between traceurs, coaches, and researchers, could benefit the future development of the sport without imposing itself upon the diverse practice sought by traceurs.
This study will aim to take one of the most common parkour movements, the kong vault, and through interviews with parkour coaches construct one or more deterministic models that describe the technique and its permutations. This approach is proposed as a communal method to catalogue and define the varied applications of parkour techniques in the absence of any objective source of parkour definition, to allow easier qualitative and quantitative assessment of parkour movements in coaching and research environments.