This study sought to understand how the application of a network analysis of rugby gameplay could inform coaches of their teams’ patterns of play in an effort to aid their teams’ performance. A qualitative case study utilizing open-ended interviews and a process of evaluation and constant comparison served as a guiding framework for this the data collection and data analysis methods incorporated during this study. Results of the study identified four key findings. First, incorporating elements of community based action research into the design of a case study provided the researcher with an opportunity to build effective working relationships with both participants. Second, providing coaches with effective feedback that informed them of their player’s performance was critical to the performance analysis (PA) process. Third, modifying the network analysis process to meet the participant’s needs was key in providing applicable analysis during the cases study. Fourth, performance analysts and coaches, like those in this case study, require video feedback, linked to the network analysis, if the network analysis process is to be considered informative. Finally, creating a PA process that is able to adapt to the coaches changing needs as well as the work cycles the organization proceeds through is a benefit of the NA process that we developed. / Graduate

Recent scholarly writing has located performance analysis firmly within the coaching process. Although the what of performance analysis regarding system design and reliability has been well documented, the how and the why or use of video-based performance analysis within the coaching process remains less understood. Therefore, this thesis sought to develop an empirically-based understanding of some of the realities of the use of video based performance analysis feedback within the coaching process. Within a broad ethnographic framework, this thesis followed three key phases of data collection and analysis. Within phase one, a grounded theory methodology, was used to explore the what and why of the delivery of video-based performance analysis in elite youth soccer. Data were collected from interviews with 14 England youth soccer coaches. Through an iterative process of constant comparison, categories regarding Contextual Factors, Delivery Approach and Targeted Outcomes were highlighted. Within phase two, coach-athletes interactions were examined in situ over the course of a 10-month English Premier League Academy season to explore the how of the delivery of video-based feedback. Data were analysed using the techniques and procedures of conversation analysis combined with a social power analysis drawing upon the work of Bertram H. Raven. Analysis of the interactions revealed that the coach attempted to exercise control over the sequential organisation of the session, via asymmetrical turn-taking allocations, an unequal opportunity to talk, control over the topic of discussion within the interactions, and the use of questioning to select speakers to take turns to talk. Within phase three, a narrative ethnographic approach was utilised to examine the how and why of the in situ narrative construction of professional knowledge and coaching identity within video-based feedback sessions. Data were collected during the same 10 months of ethnographic filed work, as presented in phase two, with a Premier League Academy Head Coach. Additionally, in-depth interviews stimulated by video-based reflection were used to explore the participant coach s early interactional practices and subsequent changes in practice in the following four years. Data analysis was conducted using theoretical concepts of identity from the work of Anselm Strauss and revealed a number of features of the development and transformation of identity of the participant coach. Here, a reflective examination of authoritarian interactional practices and the consequences of those practices were critically considered against the creation of a positive self narrative in the development of the participant coach s professional knowledge. The empirical findings of the present thesis have highlighted some the what, why and how of the use of video-based performance analysis within the coaching process. This work has furthered understanding regarding the pedagogical practices which impact upon the delivery of video-based performance analysis feedback. In addition to broadening sports coaching s theoretical and methodological repertoire, the applied value of this work is grounded in the need for coaching practitioners to become more critically reflective about the use of video-based performance analysis within the coaching process, and the impact of their interactional practices upon the coach-athlete relationship.

The aim of the current thesis was to critically examine the reflection and experience-based learning of professional football players and coaches at a football club. Specific attention was paid to the influence that the social environment had on players and coaches experiences and the extent to which they influenced each others experience-based learning and reflective practice. A case study approach using semi-structured interviews and ethnography including participant observation, informal interviews and audio/video recordings informed the current research. Schön's (1983) experience-based theory of learning and reflective practice was used to represent coaches and players reflective practice prior to the application of Foucault (1972, 1979, 1988, 1991a) as social theory. It was found that an institutionally reproduced discourse, which emphasized the importance of winning, governed both coaches and players experience-based learning at the club. Positive discourses of reflection were introduced by coaches and embodied by willing and docile players due to the added legitimacy that was associated with their knowledge. Players reflective practice represented a technology of power as it was dominated by their coaches presence and resulted in players interpretations being normalised to the extent that they became self-surveillant. Players compliance contributed to the construction and reproduction of an overarching disciplinary culture of surveillance that was initially introduced by the club s coaches and made possible through the constant assimilation of data and different forms of performance monitoring (i.e. GPS, video-based PA, physical testing).

<p>Robots have become increasingly common in various sectors, such as manufacturing, healthcare, and service industries. With the growing demand for automation and the expectation for interactive and assistive capabilities, robots must learn to adapt to unpredictable environments like humans can. This necessitates the development of learning methods that can effectively enable robots to collaborate with humans, learn from them, and provide guidance. Human experts commonly teach their collaborators to perform tasks via a few demonstrations, often followed by episodes of coaching that refine the trainee’s performance during practice. Adopting a similar approach that facilitates interactions to teaching robots is highly intuitive and enables task experts to teach the robots directly. Learning from Demonstration (LfD) is a popular method for robots to learn tasks by observing human demonstrations. However, for contact-rich tasks such as cleaning, cutting, or writing, LfD alone is insufficient to achieve a good performance. Further, LfD methods are developed to achieve observed goals while ignoring actions to maximize efficiency. By contrast, we recognize that leveraging human social learning strategies of practice and coaching in conjunction enables learning tasks with improved performance and efficacy. To address the deficiencies of learning from demonstration, we propose a Coaching by Demonstration (CbD) framework that integrates LfD-based practice with sparse coaching interactions from a human expert.</p> <p><br></p> <p>The LfD-based practice in CbD was implemented as an end-to-end off-policy reinforcement learning (RL) agent with the action space and rewards inferred from the demonstration. By modeling the reward as a similarity network trained on expert demonstrations, we eliminate the need for designing task-specific engineered rewards. Representation learning was leveraged to create a novel state feature that captures interaction markers necessary for performing contact-rich skills. This LfD-based practice was combined with coaching, where the human expert can improve or correct the objectives through a series of interactions. The dynamics of interaction in coaching are formalized using a partially observable Markov decision process. The robot aims to learn the true objectives by observing the corrective feedback from the human expert. We provide an approximate solution by reducing this to a policy parameter update using KL divergence between the RL policy and a Gaussian approximation based on coaching. The proposed framework was evaluated on a dataset of 10 contact-rich tasks from the assembly (peg-insertion), service (cleaning, writing, peeling), and medical domains (cricothyroidotomy, sonography). Compared to baselines of behavioral cloning and reinforcement learning algorithms, CbD demonstrates improved performance and efficiency.</p> <p><br></p> <p>During the learning process, the demonstrations and coaching feedback imbue the robot with expert knowledge of the task. To leverage this expertise, we develop a reverse coaching model where the robot can leverage knowledge from demonstrations and coaching corrections to provide guided feedback to human trainees to improve their performance. Providing feedback adapted to individual trainees' "style" is vital to coaching. To this end, we have proposed representing style as objectives in the task null space. Unsupervised clustering of the null-space trajectories using Gaussian mixture models allows the robot to learn different styles of executing the same skill. Given the coaching corrections and style clusters database, a style-conditioned RL agent was developed to provide feedback to human trainees by coaching their execution using virtual fixtures. The reverse coaching model was evaluated on two tasks, a simulated incision and obstacle avoidance through a haptic teleoperation interface. The model improves human trainees’ accuracy and completion time compared to a baseline without corrective feedback. Thus, by taking advantage of different human-social learning strategies, human-robot collaboration can be realized in human-centric environments. </p> <p><br></p>