The Group Games at the 2019 Rugby World Cup concluded with the Japan v Scotland fixture (link).
World Rugby provided data about each game. From these data, I recorded: penalties and free kicks conceded; kicks; passes; scrums: lineouts. My median profiles for the Group Games were:
Penalties and free kicks conceded: 16
Kicks: 58
Passes: 264
Scrums: 14
Lineouts: 25
I was also interested in ratio of passes to kicks and lineouts to scrums as a dynamic measure of each game (link). My median ratios for Group Games were: 4.55 passes to kicks and 1.79 lineouts to scrums. I used these ratios to derive a single number for each game to describe what kind of game it was. My median ratio for the Group Games was 2.55.
My ratios for the Group Games were:
Blue horizontal line is the median ratio of 2.55
Below the median:
Above the median:
Other data from the official website included officiating:
In sport, we are awash with data. A fundamental challenge for us is how we deal with these data in our everyday practice. We are trying to make sense of exabytes of data to provide a service.
I think an answer may lie in our roots as story receivers and tellers. In analytics, our relationships with coaches, players and other support staff requires us to share stories about performance. Neil Lanham (2013) points out “stories, in their natural setting, are vitally important to human understanding because they are the tools of wisdom”. He adds that the “naturally formed mindset is acutely observational, it sees metaphoric story in almost every happening, and has the language to form and relate it” (2013: 152).
Neil is an analyst and an oral historian. The combination of both domains has enabled him to think carefully about how we share and record messages. Like Robert Perks and Alistair Thomson (2015), Neil has looked carefully at individuals at work and “to see how work connects with other aspects of their lives”.
Our roots in stories enable us to connect with others and establishes trust. Through our story abilities, we are able to connect practice with the innovation and transformation that is going on in the world of sport.
With experience we build our stories with thick description. In doing so, we engage in ethnographic, qualitative activities in which stories unfold. This enables us to make more permanent our analysis and allows us to discuss how we do share and the impact we have as analysts.
Clifford Geertz sees this thick description as an interpretive act in search of meaning. His approach resonates strongly with Peter Berger and Thomas Luckmann’s view that reality is socially constructed and that the sociology of knowledge must analyse the process in which this occurs (link).
I see this qualitative dimension as vital as we seek to employ more and more Insights scientists (link). Particularly when we encourage them to produce compelling, insightful reports that can and might be triggered by data visualisations. This is where ‘science’ meets story telling and the impact we have is defined by the ways in which we share.
It is a time when we as analysts become autoethnographers and we are able to link our biographies with the insights we are sharing. As John Tetnowski and Jack Damisco (2014) (link) suggest, it is a methodology “that gets at the inner feelings and interpretations of someone involved in the phenomenon being studied”.
It enables us to transform our practice by celebrating our qualitative roots in stories.
The Football Association has advertised two positions. One is a Player Insights Scientist (link). The second is a Player Data Scientist (link).
Player Insights Scientist
The Player Insights Scientist will support the delivery of Player Profiles across the England men’s pathway and will drive key projects relating to player depth-charting, player pipeline analysis and club insights’ gathering.
The successful candidate will have these key accountabilities:
Contribute to the development of a comprehensive, robust and data driven England player profile which is linked to the success criteria for England teams.
Support the delivery of player profiles, including the analyses for future volumes of the England Player Book.
Assimilation and analysis of data from clubs on England-eligible players.
Preparation of dossiers for sharing insights on players.
Undertake quarterly analyses of the player pipeline, to understand churn, progression and quality of the England Talent pipeline.
Production of compelling insightful reports to assist in the tracking of prospective and established England players.
Development and presentation and internal research updates which effectively communicate data insights.
Support the delivery of key internal and external research projects which underpin future service development & provide insights to the talent identification process.
Contribute to the design and delivery of on-camp assessment processes to enable valid, consistent assessment of all player characteristics.
Support the Player Insights Team in the aggregation, analysis, presentation & maintenance of player profiles using information from all departments and external sources to include the creation of depth charts and other visualisations of player status.
Execute additional tasks as required in order to meet FA Group changing priorities.
Note that one of the key accountabilities is to produce compelling insightful reports (my emphasis).
To fill this post, the Football Association is looking for:
A Postgraduate Qualification in Sports and Exercise Science, or related disciplines.
Ability to assimilate and interpret data to answer performance questions.
Proven ability to generate impactful reports suitable for multiple audiences.
Experience of working in an elite sporting environment.
Advanced communication skills with the ability to communicate to various audiences.
Ability to work to tight deadlines within a dynamic environment.
Experience of working with sports performance related data.
Ability to work independently and collaboratively across departments.
Knowledge of football and the requirements of international football.
The successful candidate will be offered “an exciting and challenging role within a changing, dynamic and world-renowned sports organisation”. The post carries “attractive benefits and a competitive salary for the right candidate”.
Player Data Scientist
The second post advertised by the Football Association involves “the creation of data driven insights through the development of an advanced data model across the Player Insight department, to support the delivery of a world leading player identification and profiling service to England teams”.
The Key Accountabilities of this role are:
Review and development of appropriate data sources and models to support player insight.
Analysis and visualisation of data sources using valid robust techniques.
Modelling of data to produce predictive maps of future trends.
Use data to benchmark English talent against contemporaries in other nations and assist in valid comparisons being made.
Production of compelling insightful reports to assist in the tracking of prospective and established England players.
Contribute to the development of a comprehensive, robust and data driven England player profile which is linked to the success criteria for England teams.
Development of metrics and KPI’s which enable the ongoing monitoring of key aspects of the talent identification process.
Close and highly effective working relationships with performance analysts, national coaches and support staff.
Support the process of defining the core characteristics and benchmarks of Winning England players at all levels.
Contribute to the design and delivery of on-camp assessment processes to enable valid, consistent assessment of all player characteristics.
Contribute to the development of technology projects to support the future direction of Player Insight.
Support the training of departmental staff in the analysis and understanding of data and visualisations.
Support the Player Insights Team in the aggregation, analysis, presentation & maintenance of player profiles, using information from all departments and external sources to include the creation of depth charts and other visualisations of player status.
Execute additional tasks as required in order to meet FA Group changing priorities.
In addition to the production of compelling insightful reports there are three explicit mentions of visualisation.
The successful candidate will have these characteristics:
Postgraduate degree in Mathematics, Statistics, Computer Science
(similar related disciplines will be considered) or equivalent
experience.
Strong programming skills (Javascript, Python, R, SQL or similar languages).
Proven ability to create meaningful and impactful dashboards for multiple audiences, using software such as Tableau (or similar)
Experience working in an elite sporting environment and working with data outputs from multiple providers, e.g. SportsCode.
Knowledge of football and the requirements of international football
Ability to tailor machine learning solutions to business problems in a cross functional team.
Extensive knowledge of statistical principles and modelling techniques.
Experience in the use of multivariate statistical methods involving
information theory and clustering/network-based community finding
algorithms.
A record in data analytics with focus on interpretability and visualisation.
Ability to work to tight deadlines within a dynamic environment
Experience of working with sports performance related data
Excellent project management skills.
Knowledge of data transfer and ingestion technologies
Expertise with manipulating data to form unified data sets for analysis
Excellent written presentation skills
Good oral communications skills
Both roles advertised by the Football Association exemplify the changes in behaviour now required of analysts. Such opportunities require us to redefine our pedagogy around performance observation and analysis. The opening up of the field to a wide range of applicants is exemplified in the Player Data Scientist role in which an applicant is sought with a postgraduate Mathematics, Statistics, Computer Science qualification. There is an expectation that the applicant will be familiar with Javascript, Python, R, SQL or similar languages and visualisation through the Tableau platform.
It will be interesting to see whom the Football Association recruits for these positions. They provide us with a benchmark as to where sport might be going with Insights.
