With data visualization being adopted across various industries and businesses, Innovation Enterprise sat down with Arsenal's data visualization analyst Tolly Coburn at the recent Data Visualisation Summit in London to discuss how Arsenal leverages data visualizations to gain insights that help to optimize player performance, as the Gunners look to climb up the Premier League and ultimately win trophies.
Additionally, Coburn said that the football club has used data visualization within match reports for a number of years and it has proven thus far to be an effective strategy through various use cases.
Coburn's presentation explored three points: Data visualization for understanding, evaluation and learning. As an example, he said that data visualization, in addition to providing match insights, offered Arsenal a way to engage its fans, as well as providing a complex and holistic process when evaluating a player's performance.
We spoke to Coburn to find out more about how Arsenal uses data visualizations, its importance to the club and whether there any emerging trends developing from data visualizations in relation to the football industry.
Innovation Enterprise: What is a common use case for data visualization within the football industry?
Tolly Coburn: In a general sense, a common use case for using data visualization would be to evaluate our performance in a game. We do what is known as post-match reports – we play a match and then the next day we produce a post-match report which includes visualizations about our performance. The visualization indicates match performance across a variety of metrics and KPIs to discover whether we met expectations and benchmarks.
IE: What is the importance of tracking data within the football industry?
TC: Football is quite unique, in the sense that it is a low-scoring sport. Luck can play quite a big role as we look at quite small samples of data from a single game, football matches can be won or lost depending on a goal which is quite rare in comparison to other sports. As an example, in a sport such as basketball, players can shoot around 50 points – football matches can be decided by one goal.
It is important for us to look beyond the result, which we describe as looking at the underlying data and the underlying performance. Data visualization can provide us with an objective overall overview on how we performed; we can look at how well we performed and more accurately assess how we played.
IE: Has the use of data visualization for reporting changed?
TC: We have used reporting for a long time and visualization within reports have existed for many years, but we understand how data has grown exponentially. It's also about how our users want to understand data and want to take a more objective view of match performance, and data visualization has been the medium that we think is effective in delivering those insights.
IE: During your presentation you discussed using data visualizations for education, can you elaborate on what you meant?
TC: I think you can use visualization to educate people of what your metrics mean. As an example, to analyze players shots we visualize the shot metrics, such as how many shots were scored and from what location. Following the analysis, we thenframe them in a domain of which they are collected from, then plot the shot data to help people visually see what our metrics mean, engaging them in algorithms.
Some algorithms assume that your output is just accepted. You feed your data into your model, get an output and deliver that output to the audience. You are assuming that your user or viewer is going to accept that output without really trying to understand what methodology has been used or how you have reached that conclusion
With data visualization you can basically board a visual analytical framework in which you are allowing your user to interact and gain insights, educating them on the insights.
Visit DATAx New York, a festival of innovation featuring five stages, 100 speakers and 700 innovators under one roof, on December 12–13, 2018.
IE: With Arsenal's purchase of StatDNA, what new opportunities has it provided you in your role as a data visualization analyst?
TC: Arsenal acquired StatDNA in 2011 and my role has basically developed with them since I joined in 2016. I work closely with StatDNA as it is our internal data collection and analytics company. Lot of my work is visualizing the insights. My role alongside their work is to report that information to our users.
Because it is an internal company owned by Arsenal, we can feed the information directly to it, and work directly with it to gain insights that we think will be relevant to us. We are not relying on an external company to provide us with information, and we can perform R&D and directly communicate with our data scientists and software engineers to refine our information to make sure it is relevant to our audience.
We still accumulate data from external data providers, as do all other Premier League clubs but because we have an internal company we feel like we have an advantage.
IE: How has your role as a data analyst changed since starting at Arsenal?
TC: My role has grown a lot because I think lots of people from across the performance team have lots data of which they want to gain insight from, and I think data visualization is a great medium to explore data and visually represent the aggregated data.
Data visualization has become a fairly large part of what we do. I think as I have started to do more visualizations, our performance group has become more aware and has started to think about how their data can be visualized.
IE: Are there any emerging trends that you see developing within your industry from data visualizations?
TC: What's really interesting is the amount of work which is being done in the public domain. There are companies that provide data to people for free, so there is a lot of research being carried out by the general public who don't work in the sport industry but are publishing their work on sites like Twitter which has had a huge influence in what people and other sport clubs see. A lot of the time we don't have the ability to necessarily carry out the research, as people within the public domain can. There is some incredible work that is being done by public analysts which I think is great. It helps as a fan to better understand the game and from an industry perspective it has resulted in an increasing amount of people becoming engaged with the data within football games.
The fact that people in the public domain have general knowledge of data is certainly helping, companies like Opta are providing reports, which might be included on a segment on Sky Sports or Match of the Day – data is becoming more prevalent in the media which is certainly influencing our role. People watch Match of the Day and they say I saw this, what can you tell me about it?
Tolly Coburn spoke at the Data Visualisation Summit in London. Visit DATAx New York on December 12–13, 2018 to hear more insights from data visualization experts.
Purchase your tickets HERE