How big data and quick decision making fit together?
I am Lecturer in Science, Technology & Innovation Studies at the University of Edinburgh, developing a sociology of data science. I am interested in how data is used in high-velocity environments. In the last ten years, I focused my research on business managers finding that when it comes to decision-making, there is not such a trade-off between speed and accuracy: the quicker a decision is taken, the better it is. This blog is part of a new focus on football managers. My overarching question is: how big data and quick decision making fit together?
This blog has to catch up with a lot of stuff I did since starting my fieldwork in this area!
This dates back to at least October 2017 when I attended the first hackathon ever organised by a football association (the Italian Football Association). Antonio Gagliardi, the then Head of Match Analysis of the Italian National Team was quick to single out my team as a bunch of nerds using these words: “Teams from the UK they generally have a greater mathematical approach to the problem…maybe less tactical knowledge. […] In their team they were all mathematicians. They might have done a great job from the mathematical perspective but from the tactical point of view their interpretations were not so interesting”. Keep this in mind. We will come back to our very own Gagliardi vs Chris Anderson’s debate on the end of theory.
Let’s move on: lots is happening now…
After completing the preparation of a piece for the EASST review titled “A Question of Sport: Opening a New Research Agenda in Science and Technology Studies” with Michiel Van Oudheusden at the University of Cambridge (to come out end of Oct), I was excited to come across these (two) sessions on Science, Technology, Innovation and Sport at the Virtual Forum of the Society for the History of Technology. The sessions are organised by member of the curatorial team of the Game Changer exhibition at the Lemelson Center for the Study of Invention and Innovation, National Museum of American History, Smithsonian Institution. ‘Game Changer’: what an interesting title. That’s how I wanted to call the Data Study Group on Data & Football I organised at the Turing in 2018. That’s another monumental story worth a separate blog.
The other very exciting news is that I just started the Wyscout Football Data Analyst programme. After attending a couple of quite advanced Coaching Licenses and a Match Analyst Certificate, this is my latest fieldwork adventure in footballing expertise! In just its second edition and lasting until mid-November, I hope it will be a great opportunity to take a closer look to the latest developments in football analytics. The course was introduced by a talk by Federico Smanio, CEO at Wylab Italy. He was talking about the book Football Hackers by Christoph Biermann. The back cover says “Christoph Biermann has moved in the midst of the disruptive upheavals of the football data revolution, talking to scientists, coaches, managers, scouts and psychologists in the world’s major clubs” and I go “wow…that’s exactly what I want to do!”. I bought the book straight away. I am reading it. Stay tuned for more. First impression: I like the focus on outsiders – which is what data analysts currently are. I adopted a similar approach in my talk in New Orleans: the fledgling data science community of football analysts.
The life of a football analyst
Back to the life of a football analyst. So far, the course organisers spent two weeks trying to introduce me to Python with very little success. I got a bit more excited about Tableau this week. You can immediately see what a piece of code does to a graph. After 6 hours of tutorial, however I started daydreaming about professional vision in the age of screen sharing. I switched to a Thom Lawrence presentation on the topic “Some things aren’t shots” which I found really revealing of the mixture of football knowledge and data science that is going on in football analytics. But most importantly, between minute 7 and 30 seconds and minute 8 of the clip, I found a good example of that complex chain of environmentally coupled gestures that makes football analytics so exciting: a hand that points to the screen with a video clip which points to the graph, which points to the data. I have a piece of interview with Ian Cathro (assistant coach of Nuno Espirito Santo at Wolves) where he is talking about something similar.