How to Analyse Rivals like Bielsa in Less Than 1 Minute

Bielsa showing his analysis in press conference

Bielsa is well known for his work in studying rival teams. His attention to detail and depth in his analysis make him unique. In his Excel sheets, Powerpoint presentations and personal booklets you could find results, formations, best players and every single thing about all the teams he will face.

Thus his conclusions are excellent, his handmade processes are not allowing him to get the full potential out of his data collection. The two main issues he’s facing are:

  1. The data entry is not efficient. There is no optimized and structured tool for them to enter the data. Collaborators lose a lot of time writing simple stuff that can be taken from any data provider (goals, red cards, etc) or drawing the same formation diagram over and over again.
  2. As the information is distributed in different files, it’s really complicated to combine data from different sources. This can also lead to version problems as more people are using the same files.

In this article we will explain how to improve the process to get the same results automatically in less than 1 minute using Data Science and technology. The new methodology consists in inserting the data in a friendly web app, which saves the information in a unique database, and then get the reports automatically.

As an example of this process, let’s see how we can make the rival formations’ analysis better.

Data Entry

Inserting the information is not that different as it’s done now. We use a web app, where analysts entry the data in a friendly and easy way.

Web App Screenshot

In this web app, data is inserted as the following:

  1. Select the tournament, season, game and team.
  2. Insert the starting lineup (period = 0 and minutes = 0).
  3. Insert new formation whenever there is a change in the schema or players, always informing the period and minute when this new lineup started.

Reporting

There are many analysis that could be made out of this information, but given the scope of this post, we are going to focus on the main ones. These are:

  1. Formations & players used.
  2. Formation changes according the result.

Because of confidentiality purposes, we are going to use Wyscout’s data for this example.

1. Formations & players used

This information gives coaches an overview of how the rival team normally lines up.

To do this, we check (programmatically) every formation and players used along the season. Then we get the time that each schema was used and the players that were most used in each position.

Let’s see the result of this report applied to Estudiantes de La Plata in Superliga 2019–20. Let’s take into account that this and the following analysis were generated automatically choosing the team, league and dates.

In the following graphs, we can see that Estudiantes played in a 4–2–3–1 formation the majority of the time.

Estudiantes’ most used formations
Estudiantes’ most used formation

We could also take a deeper dive and analyse individual players and positions. Using this tool, we can get this in a matter of seconds.

For example, here we can see the most common schemes and positions in which Enzo Kalinski has played.

2. Formation changes according the result

Another useful analysis is to see how the coach reacts to results.

For this report, we divide formation changes according to the result at that point in the game (winning, drawing or losing). For reference, let’s call Schema 0 to the formation played until the moment and Schema 1 to the one that is employed after the change.

Schema changes according the result

We can see here that when winning, Estudiantes sometimes changes his formation into a five-men defense (account that 3–5–1–1 could be also read as 5–3–1–1).

This information is useful to know all the rival coach’s alternatives and be prepared to react in the case those things happen during the game.

As we’ve seen before, this report could also be extended, for example to see which players are the ones that are mostly substituted in & out.

Conclusion

It would be stupid to doubt about Bielsa’s work and capacity. Nevertheless, his handmade analysis are not efficient and consume a lot of savable time. Even though he says that he does much more than he needs just to keep himself calm, he (and his collaborators) could be dedicating that precious time to a more productive thing.

New technologies allow us to generate Bielsa’s analysis in a much faster and cheaper way. To use this tool, data entry becomes more efficient, and in seconds, we can visualize all the information and obtain graphs like the ones we’ve seen in this article.

This shows the example for formations, but it could be done and customised for different processes, depending on each coach’s work.

Data & technology is changing football paradigms and the best teams in the world are already using it to make smart decisions.

If you’re interested in incorporating data analysis in your team, please visit our website and write us to andre@bdatafutbol.com or via Twitter.

Football fan. Bringing analytics to the Beautiful Game.