HOW TO IMPROVE DATA AND DECISION-MAKING USING STATISTICAL ANALYSIS IN FOOTBALL
The tech revolution has modified our behaviour and now it is also revolutionizing sports.
One of the most popular gadgets nowadays is being used in elite sports for training. Drones have the capacity to fly over the pitch and record training sessions, to obtain an aerial view that is especially useful to evaluate performance. Specifically, in the tactical sessions, which depend on the movement of the team as a whole, and on the players’ individual behaviour.
In spite of these advantages, there is not a unanimous position among the coaches about their use. A club characterized by its effort in scientific research, like FC Barcelona, has dismissed them. While teams such as Manchester City from the Premier League, are already using them. Also, on the other side of the Atlantic, basketball and American football coaches defend its use.
What motivates these differences is the recording capacity of the device when a camera is incorporated. Regardless of whether the footage is created by a drone or a camera, it is destined to merge with the metrics generated by the trackers the players wear. This is how the technical staff can analyze, visualizing the recording and consulting specific data from it.
The two teams mentioned, FC Barcelona and Manchester City, use similar trackers, from two companies that compete between them for the leadership in this market. Barça uses Wimu, thanks to a joint development between RealTrack Systems and the Barça Innovation Hub. And Manchester City has adopted Apex from StatSports, an established athlete performing device in the Premier League. Both devices are equivalent in their outputs and allow to measure not only each player’s physical effort but also their tactical and kinetic performance. By means of specific software, they associate data with image, and for this reason, coaches choose or dismiss drones.
At Camp Nou and Ciutat Esportiva, in the highest part of the stands, under the roof, a set of cameras that rotate and move has been installed. Their capacity not to miss any detail from any point of the pitch allows to individualize on the player or a group of players of the team and compose, adding what was recorded by several of them, great panoramas. Thanks to the high resolution and zooming capacity, clear recordings are made of the training sessions, similar to that of a drone.
In the case of Manchester City, its coach Pep Guardiola has ordered the use of drones. All the technical staff have completed a course to fly a drone. A few drones fly during each training session, recording from above the work done in the training session, not only for its analysis in real-time but also to review later.
It was David Powderly, coach of the Charlton Athletic, who first used the drones in the Premier League, and it can be said that Guardiola and David influenced each other. Powderly was watching the match Bayern Munich against FC Barcelona in 2015 when the idea came up. Guardiola, at the time Bayern’s coach, deployed for the match an individual marking of each one of his players to each one of the opponent team, and the progression of this tactic happened to be spectacular in the aerial images that were broadcasted. From that day on, the coach was obsessed with having the same view of his own team. To achieve this goal he started trying a simple drone which had a camera for the training sessions.
Today, David Powderly is in charge of DPY productions, a company specialized in professional recording of training sessions and sports events using drones. He defends their use instead of fixed cameras due to their versatility, as well as assuring that they have the capacity to stay in a fixed position, where the action is taking place, or fly over a particular player. This versatility provides more information to the team coach to know if the goal he is pursuing has been fulfilled or not. As an example, it is especially useful to find a space where to penetrate through, in matches against teams that perform very close defences. It is also helpful to verify if the team moves fast enough when the ball changes side, or because of their slowness, they leave spaces. He also adds that the footage shows the mistakes players make, reinforcing the trust on the coach’s judgment.
But not all are advantages, and the limitations of these devices can influence decisively in the decision of using them. Nowadays only the most advanced and expensive models have autonomy for 30 minutes, although it is usual that they do not exceed 15 minutes. When landing the drone to replace the batteries, it interrupts the team’s task for them to wait for the drone to fly again, if that doesn’t happen, part of the training session isn’t recorded. The UK has a rainy climate so a waterproof model will be necessary, which makes it more expensive than the conventional drones. This will not impact the budget of big clubs, but if they are compared with the fixed cameras, it can favour the latter. Finally, the last limitation to consider, the drone will have to send its footage via Wi-Fi, which will end up draining its battery faster.
Powderly, and others like him, trust that these limitations will not happen with future models. Still, many professionals continue to show a great interest in their use as demonstrated by a group of Japanese researchers in the 8th International Conference on Industrial Technology and Management, organized this year. In their presentation they explained how they used these devices for a new experimental training session of a basketball team. It was not just about recording videos but processing it with an algorithm to get the same data from a tracker like Wimu or Apex. That is to say, the drone and the associated program performed all the task without the use of another device.
To do this, the drones send their footage to a computer, which processed the images using a recognition software similar to that used to identify faces. The algorithm separates the image of the players from the rest of elements identifying them by the colour of their shirt and trousers. Also, a series of colour cones were placed on the ground, to manage without the use of a GPS, the position and travel speed metrics among others. Although the results cannot yet be compared to that of a professional tracker, the researchers highlight that this is a technology that will continue to develop. With enough investment, it could advance very quickly.
All the aforementioned leads us to the conclusion that drones can not be adopted nor dismissed only by analyzing them, but together with the rest of the advantages and limitations they provide to the coach and its staff. As technical staff members incorporate new technology innovations to their daily work, they are also subject to the evolution that these will take place. A continuous re-evaluation of it is essential to know if new models can bring new advances, as well as helping the coach to achieve the goals he has in mind for the team. Like the competition itself, this is a race that is won game by game. And where everything is always to be decided.
 (1) Stephen Karungaru, Kenji Matsuura, Hiroki Tanioka, Tomohito Wada and Naka Gotoda. (2019). Ground Sports Strategy Formulation and Assistance Technology Development: Player Data Acquisition from Drone Videos. IEEE. DOI: 10.1109/ICITM.2019.8710735
Although there are several studies on this topic, many of them have analyzed these demands by looking at just a few variables or using very broad timeframes. A new study completed by physical trainers from F.C. Barcelona has analyzed several of these details more closely.
The understanding of the modifying variables of the game, based on the degrees of freedom.
Sports Analytics has grown exponentially thanks to IT sciences and it also encompasses other subareas (e.g. sports sciences, behavior sciences, medicine or data visualization) in addition to statistics with a focus that is more tactical and sports performance related.