Ready to learn Data Science for Sports? Browse courses developed by industry thought leaders and Experfy in Harvard Innovation Lab.
With the world around us being more and more data-driven, organizations are trying to catch-up. One of the most interesting challenges for organizations which have been running long before the data revolution is how to create a data-driven culture.
A quick note for our audience based in the U.S.: Throughout the post, I will be referring to "soccer" as "football" for the sake of cohesion.
Personally, I have some experience in this context working with football teams in the UK. UK football teams are an interesting case. Even though they spend millions of pounds on players, their decision making is driven by strong traditions and intuition rather than hard science. However, in the last few years things have taken a turn.
Moneyball is an excellent film for anyone interested in sports analytics and the culture of sports clubs. If you have the time it is also worth to read the book. Even though in this post I am talking about UK football teams, the culture depicted in the film is pretty similar to the one I encountered.
A great book, as well as a great movie
One such rule is that the coach has his own philosophy and has lots of control over what happens. It is valid for a coach to have an opinion against data or science, even regarding proven facts. This is not too different from the situation of many companies that have old CEOs having a very specific picture of how things should be run.
Another interesting point is the fact that football teams are under a lot of pressure. Every week is a new challenge, and the fact that there is a hierarchy in place, means that everyone is pointing the finger to someone else. If someone introduces a new way of doing things, they take the risk that in case something goes wrong, they might be blamed, solely because they tried to deviate from the tradition.
This kind of structure can be found in some companies as well, and can create an attitude of just trying to survive the week, instead of breaking boundaries. Add to that the fact that results of implementing a data-driven culture cannot be seen immediately, the problem becomes even more difficult to cope with. A long-term investment is required to implement this type of culture, which needs faith from the team in order to work out.
In professional football, the stress of the game can sometimes take its toll on the players
Another interesting point is that making a sports team data-driven requires changes across all divisions of the team. In a team you have medical and training staff of different types: physios, football coaches, weightlifting coaches, etc. Successfully making the transition to a data-driven team requires a data-driven culture across all segments of the team. The reason is that every single person that works with the athletes needs to keep detailed records of what took place in training.
Furthermore, if the team is looking to optimize its performance, the staff should not just record data, but also believe in what it’s doing, and take actions to find out what the best methods for recording the data are. This is far from trivial, given the number of errors and mistakes I’ve seen happening because the people responsible for data entry simply don't care too much.
This probably contrasts with most companies, where there are fewer people responsible for inputting data, but it poses an interesting challenge in making an organization more data driven.
Data entry can be full of mistakes
The good thing is that the situation is changing in the UK with more and more teams trying to become more data-driven. My opinion is that sports analytics will play a large role in the medical and training divisions in football clubs in the next decade, but will play a smaller role in coaching. Medicine and training are more data-driven, while coaching is strongly affected by a coach’s personal style and opinion. We will probably not see many coaches deferring a large part of their decision making to algorithms.
I’ve found two elements to be most important in changing the culture of a team, and these lessons also apply to companies. The first one is the existence of a champion, someone in an influential position who really believes in the project and can push it forward. The second one is to quickly get results which can give some initial direction to both the staff and the people high up on the hierarchy. Feeding results back to the staff helps motivation and shows how data translates to results. Feeding results up to the hierarchy will make any doubts disappear over time, eventually turning the heads of the organization into champions.
Coming back to moneyball, everyone in baseball eventually understood that the best way to play the game is through data, simply because data brings results. That’s probably the most important step towards transitioning to a data-driven culture in a company. And that’s why we will see more and more companies, organizations and sports clubs make that step in the next years.