PaCMAn: A ‘principled’ framework, arising from a systematic review of the literature, to underpin design and deployment of video games for motor skill acquisition

Highlights

• Principles that underpin effective real-life motor skill interventions are identified.

• These principles are analysed to ‘makes sense’ in a gaming context.

• The PaCMAn (Principles and Conditions for Motor Acquisition) framework is developed.

• PaCMAn is potentially used to underpin design of video games for motor acquisition.

Abstract

Research indicates that commercial exergames do not support acquisition of user Fundamental Movement Skills (FMS) owing to both technical and theoretical limitations. That is, affordable 3D sensors (e.g. Kinect®) demonstrate poor accuracy detection enabling users to ‘cheat’ motor skill outputs during gameplay and, lack fundamental design principles crucial to improve motor skill outcomes over time. Accordingly, this study outlines a principled framework to support design and deployment of video games with a primary ‘play’ purpose of motor skill acquisition. A systematic review of effective real-life interventions for motor skill acquisition was undertaken. Twenty-two studies met criteria for inclusion and were then analysed for underpinning ‘principled’ ingredients. Findings were discussed alongside fundamental game design principles, framed by The Theory of Constraints, The Exertion Framework and gaming schema from gaming literature. This led to development of a generalised framework entitled ‘PaCMAn’ (Principles and Conditions for Motor Acquisition). PaCMAn is intended to support effective design of video games for motor skill acquisition and, guide non-gaming experts (teachers, clinicians, researchers) tasked with a ‘human-in-the-loop’ deployment process deemed necessary to negotiate currently existing limitations of affordable 3D sensors.

Introduction

Modern children live increasingly sedentary lifestyles spending up to six hours a day watching TV, on mobile devices or playing video games [18]. This has a negative impact on acquisition of Fundamental Movement Skills (FMS) (also referred to as fundamental motor skills) with a majority of children and adolescents in countries like the USA, Australia and Ireland now unable to hop, skip, jump or even run properly [53], [25]. Poor FMS in childhood is a predictor of non-participation in sport, an associated predictor of health problems in later life [30] and linked with difficulties in reading and mathematics [73]. Thus, well developed motor skills are ‘fundamental’ for positive health and academic outcomes. FMS include stability skills (static/dynamic balance), object control skills (kicking, throwing, catching etc.) and locomotor skills (running, skipping, jumping etc.). Developmentally, FMS are potentially mastered by the age of eight, with this 'mastery' denoted by an effortless ability to combine components of a skill to form a unique movement sequence. Components of a ‘jump’ are outlined and illustrated for the purpose of this study (Fig. 1) informed by Ulrich [67].

In recent years, a new range of ‘e-health’ technologies have emerged (mobile applications, wearable activity monitors, etc.) marketed towards tackling user sedentary behaviours [76]. From a gaming perspective, motion sensor interfaces (e.g. Kinect® and Wii® remote) were developed more than ten years ago in an effort to transform ‘finger tapping’ sedentary gameplay into full-body, interactive experiences [8]. A new genre of ‘exergames’ was born, defined by Mueller et al., [49] as games dependent on user physical effort. A number of commercial exergames are designed to promote Physical Activity (PA) and improve user physical fitness (e.g. aerobic and muscular endurance). A systematic review of the literature suggests exergames have the capacity to do both [39]. In addition, motion sensor interfaces like Kinect® present as particularly promising in terms of accessibility and, usability. First, they are affordable and typically available ‘off the shelf’. Second, they do not require objects to be attached to the user’s body. This makes them suitable not only in the home but in the school, club or clinical setting. Studies show that exergames facilitate high levels of motivation and enjoyment once there is a high level of physical challenge that matches user physical capabilities [52]. This physical challenge could include, what Mellecker, Lyons and Baronowski [45] refer to as, ‘task-control’ movement outputs. Interestingly, a number of commercial exergames call upon users to perform FMS (jump, throw, bat etc.) for task-control purposes. However, studies show that these exergames, when deployed in typical fashion, do not support improvement in motor skill outcomes owing to poor quality task-control outputs performed during gameplay [2], [12]. That is, affordable 3D sensors, like Kinect®, currently demonstrate accuracy detection limitations, particularly for fast, bi-lateral movements involving the lower limbs (e.g. locomotor skills) [20]. This means users are able to ‘cheat’ motor skill outputs, limit energy expenditure and still experience gaming success. Limitations associated with affordable 3D sensors are a potential factor in why exergames, which often involve motor activity, have yet to be explicitly designed for the purpose of motor skill acquisition.

Despite currently existing design and technical limitations, a number of studies have successfully utilised commercial exergames to bring about a lasting change in user motor skill behaviour. For example, Levac et al. [36] outline the effectiveness of popular exergames for rehabilitation of basic motor patterns (e.g. hand to mouth) in stroke patients, whilst Vernadakis et al., [69] outline the positive impact of commercial exergameplay on user object controls skills. In both studies, success is potentially attributed to a considerate deployment process in which gameplay is supported by an expert in skill acquisition who effectively ‘adapts’ the user play experience and, delivers additional feedback/instruction that the system does not. Whilst neither study provides significant detail in relation to the role of the expert in this deployment process, results highlight the possibility of negotiating currently existing sensor limitations and game design limitations through the use of an additional ‘human’ adaptive component.

Adaption and deployment are also described by Hardy et al. [26] when outlining their suite of purpose built exergames designed to target user physical activity and stability skills. Crucially, these games were developed using an authoring tool that allows gaming features to be adapted and personalised by a human ‘on the fly’. In this instance, adaption refers to modification of gameplay to support a specific goal, whereas personalisation refers to adaption informed by individual user needs [26]. Commercial exergames such as ‘Your Shape’ for the Xbox 360, typically aim to process user performance data and utilise results as inputs to inform the user play experience. However, given current inaccuracies of affordable 3D sensors, performance data is rarely indicative of a user’s true movement capabilities. Rather than simply giving up on these technologies however, we suggest negotiating currently existing limitations by considering the use of a human-in-the-loop, capable of making decisions about user movement capabilities that the system cannot. This human-in-the-loop could personalise the gameplay experience through a deployment process potentially referred to as human-in-the-loop personalisation, with the ultimate goal of supporting motor skill acquisition in a gaming environment. To that end, we know video games with a primary ‘play’ purpose of skill acquisition should be underpinned by design principles grounded in theory [19]. Whilst game designers are already equipped with a litany of theoretically underpinned design principles e.g., ‘play’ (grounded in Social Constructivist Theory [70]), ‘challenge’ (grounded in Zone of Proximate Development Theory (ZPD) [70]) and ‘choice’ (grounded in Self Determination Theory [60]), the domain lacks knowledge and understanding of important principles required to support motor skill acquisition and how to deliver these principles in a gaming environment.

The current paper is motivated by the need for a theoretical framework or ‘recipe’ that merges theories and principles from game design and motor skill acquisition to underpin design and deployment of video games with a primary play purpose of improving user motor skill outcomes. The paper is intended to build on commonalities identified by Gonzalez and Adelantado [21] between video gameplay and motor play beginning with schemas relating to ‘rules’, the formal structure of a game that underpins the user/learner experience; ‘play’, the actual user/learner experience including the physical and psychological response and ‘culture’, the overlap between the game world and its surrounding environment [66]. A core focus of the work is to identify fundamental principles or 'rules' from game design and motor skill acquisition that could be used to structure an effective game ‘play’ experience and fosters progress in motor skill outcomes. The surrounding culture/environment is also considered, particularly in terms of exploring a human-in-the-loop personalisation process, to deliver fundamental principles that the system cannot. Our starting point is to identify appropriate theories from both motor skill acquisition and game design that can be used at a macro level to frame the design and deployment process.

Literature on motor skill acquisition points to the effectiveness of The Theory of Constraints (TOC) [51] used to underpin design of numerous ‘real-life’ motor skill interventions (e.g. The SKIP program [1]) with empirical evidence to support their worth. TOC is a non-linear pedagogy based on Ecological System Theory (EST) [7] and Dynamic Systems Theory (DST) [51]. It predicts motor skill acquisition through dynamic interactions between the (i) individual (ii) task and (iii) environment. Within this model, the individual comes with structural components (e.g. level of physical fitness, motor skill competency etc.) and functional components (e.g. motivation, enjoyment etc.) [11]. The task refers to rules and goals of an activity that determine the quality, timing and product (speed, height etc.) of the motor skill [11]. When the task is repeated in order to improve performance it is described as ‘practice’. The environment refers to the world around the individual, the physical setting, availability of equipment and supports including quality teaching, instruction and feedback [11]. A constraints-led approach to motor skill acquisition proposes that constraints within the task and environment can be manipulated to suit individual learner capabilities and bring about lasting change in motor skill outcomes.

From a game design perspective, Sgro et al. [62] suggest that video games focused on improving motor skills should also consider constraints, but specifically refer to constraints in relation to ‘the four bodily lenses’, as outlined by Mueller et al. [49] in The Exertion Framework. Inspired by Jacob et al. [31] this framework views user bodily interactions in a gaming environment in four ways. First, a ‘responding body’ lens relates to user physical fitness and changes that occur in the body during gameplay, i.e. increased heart-rate, panting, sweating etc. Next, a ‘moving body’ lens relates to task-control outputs performed during gameplay (e.g. jump, kick etc.). Third, a ‘sensing body’ lens describes how the body acts/reacts to features of a gaming environment and finally, a ‘relating body’ lens is concerned with social interactions a user might experience during gameplay (e.g. multi-player options/online interactions). A gameplay experience that views the user through these four lenses and, structures the play experience to elicit effective task-control outputs, has the potential to foster enjoyable, physically active gaming experiences that also support motor skill acquisition. More recently, Mueller and Young [47] outlined a further ‘10 lenses’ or, ‘virtues’, to guide design of video games that go beyond physically active experiences and target a user’s overarching physical health. It is important to note however; motor skill capability is an associated predictor of future health status [30]. Thus, before we can tackle overarching ‘health’, gameplay must prove capable of supporting this crucial antecedent. As such, we propose beginning with a theoretical underpinning of the user’s play experience, structuring ‘moving’ and ‘responding’ bodily outputs with a primary goal of improving motor skill acquisition.

A marriage between TOC, The Exertion Framework and gaming schema from gaming literature presents as an effective way to frame design and deployment of video games for motor skill acquisition (illustrated in Fig. 2). In this marriage, ‘the individual’, as defined by TOC [51], could be considered as ‘the user’ whose ‘structural components’, including level of physical fitness and motor skill capability, are then discussed in terms of the ‘responding body’ and ‘moving body’ respectively. The ‘task’ could be considered as ‘gameplay’, with ‘rules’ set in place to prompt motor skill outputs, via the sensing body, in line with the physical and psychological needs of the user. The ‘environment’ could also be considered as the surrounding ‘culture’ providing opportunities for delivery of effective (augmented) feedback and instruction in both the real and virtual world. Lastly, designers could consider ‘multi-player’ options and other forms of peer to peer interactions in order to target a user’s ‘relating body’ and positively impact psychological determinants such as motivation and enjoyment.

Two challenges still remain, however. First, affordable 3D sensor limitations make it difficult for a system to accurately assess user motor skill capabilities and/or personalise gameplay to suit individual user needs. Whilst this is potentially negotiated through a human-in-the-loop deployment process, where a parent, teacher or clinician makes decisions about the user and gameplay that the system cannot, this human-in-the-loop requires support and guidance in order to make the ‘right’ decisions. Therefore, our second challenge is to identify a more micro set of ‘ingredients’ to inform finer details of design and deployment. To that end, there are a number of fundamental game design principles and accompanying conditions to draw from. These include aforementioned ‘rules’ and ‘play’ as well as ‘challenge’ (concerned with provision of a play experience that is challenging but achievable to the user [32]), ‘feedback and rewards’ (concerned with continuously shaping and improving user skill outcomes [23], [75]) and ‘choice’ (concerned with affording users a sense of autonomy over the learning and play experience [57]). There are also a litany of principles and conditions deemed effective for motor skill acquisition, including those specifically relating to ‘practice’ (e.g. amount, variability etc.), ‘feedback’ (e.g. frequency, focus etc.) and ‘instruction’ (e.g. type, timing etc.). An outline of principles and conditions for motor skill acquisition, informed by Davids et al. [11], Schmidt & Wrisberg [61] and Magill [40], is presented in Appendix A. The problem is, most principles are typically studied in isolation or in small clusters across the literature [77] meaning we have yet to decipher which principles/conditions and moreover, which combination of principles/conditions, prove most effective for motor skill acquisition. Ultimately, designers need to be confident that principles used to underpin game design will support improvements in user motor skill outcomes. Accordingly, this study involves a systematic review of the literature on effective ‘real-life’ interventions for motor skill acquisition with a concerted effort to identify underpinning principled ‘ingredients’. These principles are then discussed and analysed alongside fundamental gaming principles. The Theory of Constraints [51], The Exertion Framework [49], and gaming schema from gaming literature [66] are used to frame this analysis leading to the iteration of PaCMAn (Principles and Conditions for Motor Acquisition). PaCMAn is intended to contribute towards effective design and deployment of video games with a primary ‘play’ purpose of motor skill acquisition.