Shortcomings of applying data science to improve professional football performance: Takeaways from a pilot intervention study

Positional tracking data allows football practitioners to derive features that describe patterns of player behavior and quantify performance. Existing research using tracking data has mostly focused on what occurred on the pitch, such as the determinants of effective passing. There have yet to be studies attempting to use findings from data science to improve performance. Therefore, 24 professional players (mean age = 21.6 years, SD = 5.7) were divided into a control team and an intervention team which competed against each other in a pre-test match. Metrics were gathered via notational analysis (number of passes, penalty box entries, shots on goal), and positional tracking data including pass length, pass velocity, defensive disruption (D-Def), and the number of outplayed opponents (NOO). D-Def and NOO were used to extract video clips from the pre-test that were shown to the intervention team as a teaching tool for 2 weeks prior to the post-test match. The results in the post-test showed no significant improvements from the pre-test between the Intervention Team and the Control Team for D-Def (F = 1.100, p = 0.308, η² = 0.058) or NOO (F = 0.347, p = 0.563, η² = 0.019). However, the Intervention Team made greater numerical increases for number of passes, penalty box entries, and shots on goal in the post-test match. Despite a positive tendency from the intervention, results indicate the transfer of knowledge from data science to performance was lacking. Future studies should aim to include coaches' input and use the metrics to design training exercises that encourage the desired behavior.

Creating a Live and Flexible Normative Dataset for Netball

A previous research has identified large data and information sources which exist about netball performance and align with the discussion of coaches during the games. Normative data provides context to measures across many disciplines, such as fitness testing, physical conditioning, and body composition. These data are normally presented in the tables as representations of the population categorized for benchmarking. Normative data does not exist for benchmarking or contextualization in netball, yet the coaches and players use performance statistics. A systems design methodology was adopted for this study where a process for automating the organization, normalization, and contextualization of netball performance data was developed. To maintain good ecological validity, a case study utilized expert coach feedback on the understandability and usability of the visual representations of netball performance population data. This paper provides coaches with benchmarks for assessing the performances of players, across competition levels against the player positions for performance indicators. It also provides insights to a performance analyst around how to present these benchmarks in an automated “real-time” reporting tool.

Pilot Study on the Reliability of the Coach's Eye: Identifying Talent Throughout a 4-Day Cadet Judo Camp

A typical assumption found in talent identification literature is that different coaches, given the same athletes and circumstances, will identify the same subset of athletes as “talented”. However, while coaches play a major role during talent identification in practical sport settings, there is limited empirical research exploring the processes which underpin this. The purpose of this study was to explore the reliability of “the coach's eye” during the assessment of talent in a group of athletes. Specifically, this project compared inter-coach agreement between nine judo coaches (ages 35.8 ± 10.6 years) with varying levels of experience (12.9 ± 8.9 years) in the evaluation of 24 talented cadet judo athletes (13–15 years) at seven timepoints throughout a 4-day development training camp. Without discussion of their scores with other coaches, coaches provided a single score representing each athlete's “potential for future performance” on an 11-point Likert scale at each timepoint. Scores from each coach were converted into rankings from 1 to 24 to create a normalized scale to facilitate comparison of athletes. Based on their rankings at each timepoint, athletes were placed into one of three evenly distributed groups (high, medium, and low rank). Inter-coach agreement at each timepoint was determined by the number of coaches who ranked each athlete in the same group, categorized at three levels: 50, 75 or 100% agreement. Overall results showed that at completion of the camp, coaches reached 100% agreement on only two athletes, both of whom were in the high rank group. When inter-coach agreement was set at 50%, 15 athletes (62.5%) were placed into like groups. The first timepoint at which coaches were able to differentiate between the majority of athletes was Timepoint 3 (end of day 2). The findings suggest that, in isolation, coaches do not agree on the talent or potential of athletes. This indicates that the “coach's eye” is subjective and variable, and, given the same context, there is poor inter-coach agreement in the identification of talented athletes. In turn, these findings may have significant implications for both future talent identification research and athlete selection processes by sport organizations.