Cédric’s introduction to R ggplot

Cédric Scherer (link) has written a delightful guide to ggplot. His post is titled A ggplot2 Tutorial for Beautiful Plotting in R (link).

I worked through his post by looking at some of the data from the FIFA Women’s World Cup in France (link) earlier this year.

My exploration of Cédric’s suggestions was definitely of the trial and improvement kind. I did find it one of the best introductory guides to ggplot I have discovered and it helped me build on my eclectic learning journey with this form of visualisation.

The csv file I used for this exploration is available on GitHub (link) and is titled RefereesWWC.csv. My brief R record is:

I looked at five examples from the official FIFA data provided in FIFA’s Match Facts (link). I was mindful that the median ball in play time during the World Cup was 55 minutes and the median time was 97 minues.

1. A geom_point of the referees who officiated at the World Cup and the FIFA record of ball in play time in minutes.

2. A geom_line and geom_point development of visualisation 1 that connects referees that officiated at more than one game at the World Cup.

3. A geom_density_ridges visualisation of ball in play time and total game time.

4. A generative additive model for less than 1000 data points. An outlier, USA v Thailand, is recorded with annotate.

5. An example of a developed geom_density-ridges plot that used the theme_economist visualisation backdrop from the ggridges package. It uses temperature data to look at goals scored in the tournament.

This visualisation provides an opportunity to record with annotation particular games and includes two 0v0 games, the 13 v 0 game and two games involving six goals.

I do recommend Cédric’s post unreservedly. It is a great way for us to develop our use of ggplot as a visualisation tool. The basic code I used for my post is available on a GitHub (link).

Technical Direction

The Daily Mirror reports that “Manchester United are stepping up their search for a new technical director” (link). The search for United’s first ever Technical Director has gone on for over a year. It involves bringing in a former player “who understands the culture of the club”.

The role was originally described as a Director of Football charged with long-term planning of recruitment. There is a suggestion that the role is now one in which the Director would form part of a recruitment team, alongside the manager, and others (link).

The Daily Mirror notes “the new Technical Director would fit in within that structure rather than take it over because there is a strong feeling at Old Trafford they had a good summer which pushed the club in the right direction” (link).

Andrew Cave (2019) notes that a number of other teams have appointed Directors.

  • Manchester City: Txiki Begiristain (link).
  • Liverpool: Michael Edwards (link).
  • Chelsea: Petr Cech (link).
  • Arsenal: Edu (link).
  • Paris Saint-Germain: Leonardo (link).
  • Barcelona: Eric Abidal (link)
  • Bayern Munich: Hasan Salihamidzic (link).
  • Juventus: Fabio Paratici (link).
  • RB Leipzig: Paul Mitchell (link).
  • Ajax: Marc Overmars (link).

Andrew notes that “Winning eras are characterised by healthy funding, shrewd signings, solid coaching, strong, characterful management and an X factor that brings to life all that dull talk about talent management and the ability to motivate and steer a team”.

It will be interesting to see how each of these Directors bring their vision to these winning eras. A key issuse for me is how successful players transition into the roles that are being created and are able to guide the transformation that all of them seek.

There is a Football Association Level 5 qualification. The course has six modules with in-club support provided by an FA tutor. There are twenty-five contact days in total as a group, with workshops based at St. George’s Park and visits to high performing organisations and visits hosted by clubs. There is a three-day study to Europe. 

The learning outcomes for the course are: 

  • Self-awareness and its impact on leadership
  • Leadership philosophy and application
  • Understanding of high-performing teams
  • Knowledge and awareness of world-leading insights into first team performance recruitment, academy, and science and medicine departments
  • Appreciation of club infrastructure and alignment of resources to deliver sustainable success
  • Understanding and application of the rules, regulations and governance
  • Long-term people development skills and succession planning

There is a prospectus available for the course (link).

Photo Credits

Red Seat at Old Trafford (Gordon Ednie, CC BY-NC-ND 2.0)

Txiki Begiristain (BBC)

Prospectus (FA)


Martin Buchheit has posted a survey about Performance Jobs (link). He writes:

The number of “performance staff” has grown exponentially over the past years in elite clubs. While there is no doubt that these positions are created to improve long-term club processes, staff communication, and in turn, players and team performances, an important confusion exists with regard to the actual roles of those professionals. There is a feeling that there may be (almost 😊) as many job titles as structures, and there are also large variations in role (job description) within the same job title! To try to shed a bit of light upon this area, I am hoping you will be willing to fill with full honesty this very short (5 min max) questionnaire. Importantly, the job profiles examined here are club-based …

He adds:

All results will obviously remain anonymous – before getting summarized in a global article on the topic that will be published in an open-access journal (likely SPSR).

Describing what we do

In conversations about people involved in data analysis, one of my colleagues in an institute of sport observed that “the biggest challenge is how we develop and mentor these people”.

I see this as a critical issue as sport expands its data science portfolios. It has encouraged me to think about the verbs we use to describe our work in data.

When I first started in performance analysis in pre-digital days, we aspired to:

  • Observe
  • Record
  • Analyse
  • Model

Guillermo Martinez Arastey (2018), amongst others, has described how this role has changed in a digital era (link). It has meant for me that performance analysts are connected and I saw this at first hand when I met Darrell and Adam in Cardiff (link).

Darrell’s Vocational Performance Analysis post and Adam’s What has changed in Performance Analysis over the last 5 years? exemplified their reflection in and on action that define connected, sensitive educators.

I was thinking about these connections and changes when I came across IBM’s AI Ladder (link). This ladder used a fourfold taxonomy of verbs:

  • Collect accessible data
  • Organise a business-ready foundation
  • Analyse with trust and transparency at scale
  • Infuse throughout the business

I would add to this feedforward (link). With artificial intelligence I think it is vital to consider where we will be and involves us in mental time travel.

In a 2012 paper, Peter Dowrick observes:

The most rapid learning by humans can be achieved by mental simulations of future events, based on reconfigured preexisting component skills. These reconsiderations of learning from the future, emphasizing learning from oneself, have coincided with developments in neurocognitive theories of mirror neurons and mental time travel.

It is these “mental simulations of future events” that strike me as very important as we consider the verbs that guide us through a dynamic domain opening up before us.

Photo Credit

Markus Spiske on Unsplash