Marti Casals (link) kindly sent me a copy of Football Analytics: Now and Beyond. The book reports the proceedings of the second Barça Sports Analytics Summit hosted by the Barça Innovation Hub (link).
In a letter that accompanied the book, Marti wrote:
The guide reflects the state of science in the field of sports analytics, specifically football, through the contribution of international scientists and some experts in this field and it wants to encourage professionals from around the world to learn from this phenomenon.
Without Marti’s sharing I would not have found Football Analytics: Now and Beyond. It is a fascinating book. It is 184 pages long with 11 chapters. Javier Fernandez (link) said of the summit “this is a unique event within the landscape of football analytics, where data scientists, coaches, game analysts and football enthusiasts come together to discuss about real-world applications of advanced data analysis“.
Preface (13) (Raúl Peláez) introduces the book and emphasises that “it is important sports scientists work very closely with athletes and coaches in order to understand the problems they face when making decisions based on uncertainty”.
Introduction to Football Analytics (15-19) (10 references) (Angel Ric, Marti Casals, Carlos Rodriguez, Sergio Llana, Pau Madrero, Javier Fernandez) encourage an interdisciplinary approach to analytics and argue for a paradigm shift that establishes an “iterative processes … where communication between different professionals at a club encourages the formulation of questions”.
Eric Abidal (20-21) is the FC Barcelona Technical Director. His chapter is a question and answer format.
New Advances in Football Technology (23-33) (5 references) (Nicholas Evans) notes that we are entering “a new gold in the era of information”. One of the questions the chapter addresses is the question Does More and More Data Require More and More Technology in Football? The chapter also considers regulation and standardisation.
Andoni Zubizarreta (34-35) is the Technical Director at Olympique Marseille. His chapter takes the form of questions and answers.
Analysing Performance and Playing Style Using Ball Event Data (36-47) (27 references) (Jan Van Haaren, Pieter Robberechts, Tom Decroos, Lotte Bransen, Jesse Davis) notes that clubs are collecting an increasing amount of data including game data (line ups, substitutions, goals and cards) ball event data and tracking data. The chapter provides examples of these three types of data. The authors note in conclusion “this section discussed recent approaches to evaluate shots, saves passes and other types ofactions as well as approaches to analyse player behaviour and team tactics”.
Gregg Berhalter (48-49) is coach of the USA men’s team. His chapter takes the form of questions and answers.
How to Find Elementary Football Structures in Positional Data (51-65) (21 references) (Daniel Link, Steffen Lang) discusses spatio-temporal tracking data and has extensive illustrations of the points discussed in the chapter. The model proposed in the chapter looks at individual ball possession, episodes of team possession and ball actions.
Vosse de Boode (66-67) is the Head of Sport Science at AFC Ajax. Her chapter takes the form of questions and answers.
Key Performance Indicators in Professional Football (69-77) (34 references) (Daniel Memmert, Jürgen Perl and Robert Rein) note that when computers became available in the 1970s “more complex data recording and analysis became possible”. They added this included connections with “mathematical game theoretic models”. The chapter includes a discussion of key performance indicators from the Bundesliga. These indicators include: space control; passing and space control; and pressing.
Patrick Kluivert (78-79) is the FC Barcelona Director of Youth Football. His chapter takes the form of questions and answers.
Ranking players and teams (81-91) (23 references) Luca Pappalardo, Paolo Cintia, Paulo Ferragina) discusses ranking players through data-driven evaluation of performance. The chapter includes a range of graphical summaries of data including box plots and network mapping.
Clement Lenglet (92-93) is an FC Barcelona player. His chapter takes the form of questions and answers.
How much does ball possession influence match performance? (95-109) (74 references) (Julien Castellano, Filipe Clemente) provides an overview of data collected to investigate dynamic ball possession. The chapter discusses data from La Liga and includes a range of graphical representation of data.
Ernesto Valverde (110-111) is the FC Barcelona first team coach. His chapter takes the form of questions and answers.
Using tracking data from matches and training situations (113-131) (96 references) (Jaime Sampaio, Bruno Goncalves, Diego Coutinho, Sara Santos, Hugo Folgado, Bruno Travassos) provides an extensive look at tracking data and the technological advances in positional, computational and imaging tools. The chapter includes a number of graphical representations of data.
Alexia Putellas (132-133) is an FC Barcelona player. Her chapter takes the form of questions and answers.
Linking sport science and analytics in a professional football club (135-143) (45 references) discusses the field-based nature of data in a professional football club. The article proposes that “an aligned analytics and sport science has the potential to be the ‘glue’ of a football club” in the quest for a transdisciplinary sport science.
Christopher Clemens (144-145) is Germany’s team video analyst. His chapter takes the form of questions and answers.
Identifying the Match Plan (148-173) ( 47 references) (Manuel Stein, Philip Zimmermann, Markus Schopp, Jens Preussner, Michael GrossNicklaus, Tobias Schreck, Daniel Keim) note that “during matches, team movement data can be captured using computer vision techniques, while biomedical measurements about the players can be collected with smart shirts”. Professional providers provide vast amounts of data too. The authors note “this wealth of data opens up many opportunities to monitor and analyse football data from different perspectives”. The chapter includes a number of thought-provoking graphical representations of data.
Future Challenges in Football Analytics (177-184) (5 references) (Javier Fernandez, David Sumpter) aims to plot a path forward in the next decade in terms of: statistical and mathematical models of football; tactical advancements and systematic match analysis; technological developments in processing match footage. The chapter looks carefully at integrating analytics within a football club and celebrates the opportunities available to work together.
Football Analytics: Now and Beyond is a lavishly illustrated publication. There is no page identifying the publishers of the book or the ISBN. I imagine that it is available from the Barça Innovation Hub.
This is a very important book for our discussions of analytics. Its empirical focus is football but it is possible to imagine other empirical domains that could be published.
A key issue for me is how we share our practices of analytics. I applaud the authors of this book for a commitment to open sharing and mutual flourishing. I am excited too that we might explore interdisciplinary and transdisciplinary ideas. For my part I see analytics as a nomadic profession and we should lay great importance to the skills of connection and communication. These, I think, are the glue for better practice.
We do need much more sharing of practice in applied settings. I would even go so far as openly sharing data. I am hopeful that these data might become like the visualisations that lavishly illustrate this book.
I am so pleased that my postman delivered this remarkable gift.
Barca Innovation Hub (FC Barcelona, Twitter)