Everyone is probably aware of the ‘Moneyball’ concept; the popularized story of American baseball club Oakland Athletics, whose General Manager Billy Beane threw out traditional scouting methods, based merely on intuition, and implemented an evidence-based statistical approach aimed at bringing in undervalued baseball players. Thriving in the 2002 season and eventually clinching the American League West title, Moneyball has become the go-to story for information goods-related approaches in the sports industry.
The worldwide sports market is a $90 billion dollar industry [1], however it has been predominantly based on intuition and gut-feeling until the early 2000s. Only then Big Data Analytics started to emerge as a tool for helping sports teams in their decision-making. The use of information within the sports industry knows three primary ways of change:
- It can help improve player and team performance; in this case it’s mostly a way of translating the data into comprehensible visualizations for the players themselves. Players will to tend to accept instructions sooner when they know there’s an underlying truth in the data.
- It has a considerable predictive value; by analysing and tracking player behaviour and by comparing that data against others, teams can develop an efficient recruitment policy. Moreover, clubs can perform predictive analyses on opponents. What will there style of play be? What will there next move be? Questions that data analytics can frequently offer an answer to. [2]
- It can help sports teams (which are essentially firms) in enabling a highly-efficient business- and information strategy. An example of this is Danish football club FC Midtjylland who went from near-bankruptcy to the Champions League in the span of one season, employing a data-driven strategy and reaping its financial benefits.[3]
A takeaway to bear in mind. More information does not necessarily mean more success. Clubs additionally need to focus on employee-recruitment, hiring skilled data analysts.
Since the global sports market is so substantial, why do you think it took relatively long for evidence-based approaches to emerge?
How can sports teams try to differentiate in handling their information strategies to gain competitive advantage?
[1] Statista (2017), https://www.statista.com/statistics/370560/worldwide-sports-market-revenue/
[2] Principa (2015), https://insights.principa.co.za/4-cool-ways-data-analytics-is-changing-the-world-of-sports
[3] The Guardian (2015), https://www.theguardian.com/football/2015/jul/27/how-fc-midtjylland-analytical-route-champions-league-brentford-matthew-benham
I’ve always been fascinated with the introduction of data analysis into sports. It leads me to think that eventually many jobs in sports will become obsolete. Often times player signings are based on videos, attending games and gut feeling. What’s to stop the decision-makers from replacing scouts with analists? I wonder if eventually the acquired data about players and tactics becomes so widely used that it normalises tactics in such a way that we all become mindless followers of data.
While it might be the most effective method eventually, it will certainly be counterproductive towards the ‘fun’ part of sports. Interesting read.
Hey Peter, nice post! I loved the movie Moneyball, and the book is also really worthwhile!
I would like to comment on your second question: how can sports teams gain a competitive advantage through data?
First of all, I think it must be noted that large differences exist between various sports. Baseball is very well-suited to be analyzed by data because it is a static sport, with clearly defined “events” that have a clear succes and fail option. If a hitter is trying to reach a base, the fail and succes of this event are completely unambigious. As a result, baseball is easy to define in a model.
As a consequence, baseball has always known a culture of keeping statistics, as mentioned in the book.
For football, it is less clear what an “event” is, and how to determine is succes and fail state. A pass from one player to another should of course reach the other player. But a succesful pass is one that not only reaches the other player, but also brings the team closer to a goal.
In fact, in football you could argue a pass with 80% of reaching the player and a 70% chance of leading to a goal is a better pass than a pass that has a 100% chance of reacing the other player, but goes back to the keeper.
Therefore, in football and many other sports, data insights should be carefully combined with insights and knoweldge from the sport itself.
In the same fashion as we try to incorporate data and business insigths at BIM 😉