Using data to inform progression

In his latest article for Kognia Sports Intelligence, Ryan DeFreitas provides a detailed account of his use of performance data to inform decision-making in academy football, whilst sharing a vision for how the future could - and should - look in this space.

Whether you are in an academy or first team environment, the ability to apply data to evaluate players and inform development plans or retention/release decisions has become increasingly important. If these processes are missing in your club, the chances are you’re behind your rivals who are already making more informed decisions than you.

Nine years at Leicester City as lead academy analyst 

In my previous academy role we collected data to inform decision making for players as young as seven. To ensure there was a coherent process across the academy, we had consistency of metric and definition, meaning an 8 year old and a 21 year old were measured in the same way. There was a difference in the delivery though, as different parts of that information were deemed to be important and selected to be front-facing, dependent upon the age of the respective player. 

For instance, a full picture of an U21’s player’s technical profile would be used to inform decision making and the success of their actions is important. Whereas, for your U9 player you might only be interested in the number of times they possess the ball to ensure they are getting ample opportunities to develop technically, and volume is given a greater weight than success. 

Whilst we were able to use this data for players so young at each review point across the year, it was also to support long term strategic projects and to ensure they were data informed. For example, we were trying to better understand how players develop across their entire time at the academy, not just how they have developed after training interventions from one review cycle to the next, so as to also make longer term data informed decisions. Admittedly, it was still trying to see into a crystal ball, but our crystal ball was a bit less murky.

Some questions that we were interested in answering included:

• What do future professional players look like at different points of the development pathway?

• Are there any similarities between players and profile development that we can identify?

• Do technical profiles even change as players move through the pathway?

We were able to answer some of these questions.

Almost any process of data collection can satisfy the short term touch points for your data strategy. However, consistency of data and a full data profile is key to understanding the profile development of an academy player over longer time periods. Generally an 8 year old won't sign a professional contract at a club for another 10 years, so consistency of data collection and a clear long term academy-wide strategy is needed. Not just any process of data collection can cover that.

There is such a lack of information at these ages - for a variety of reasons - that the knowledge regarding the types of questions listed above simply doesn’t exist in most environments. We were ahead in objectifying - and quantifying - young player development with a view to having an advantage over others by contributing to more informed, more objective decision making over a player’s complete time at the academy, not just the immediate touch points. 

More and more clubs are beginning to recognise the value of viewing their player monitoring and data processes as part of the same overall equation, meaning the types of questions requiring data are becoming more frequent, more complex and more commonplace since I implemented that process of data collection and insight delivery nearly 15 years ago.

Consistency vs limitations: how we collected data

Collecting academy data was done manually by our analysts and we consistently had the same problem: it was very ball centric. This was in part to make it easier for the analyst and try to remove areas of excessive subjective judgement. For example, to break a line with a pass the ball had to travel beyond the last player in that opposition line. This made it a black-or-white, yes-or-no, type of judgment. 

Did it pass the deepest midfielder? Yes. Midfield Line Break. 

This of course has its limitations but my opinion is that whatever your definition is, it's a good definition as long as: 

1.  You stick to it

2.  Everyone understands it

This becomes your working philosophy and conversations become consistent because everyone understands the definitions, is aligned and can consider the context within its limitations.

The other reason was that the process was very time intensive and we accepted that due to the level of data integration across the academy. The value outweighed the time burden, but to widen the scope to include off-the-ball concepts with an appropriate level of detail and reliability was simply not feasible. However, as players possess the ball in a match for such a small amount of time, this meant our extensive data collection had significant blind spots.

Developing the process to be more insightful

Below is a typical example of the type of profile an academy could produce in house through manual collection. You will see that it is almost entirely ball centric due to the limitations of manual collection by an analyst under significant time pressure.

While I'm proud of the system we put in place, if I were to start again at a new club my first priority would be to address those limitations that hindered us. Our process was innovative for its time and, in many cases, still far in advance of what clubs have today in their academies. Whilst we were able to iterate on the process to make it more insightful and more relative to the evolution of the academy football programme, the technology now exists to fill gaps in the information derived without additional burden to my analysts. Outsourcing the collection of data could reproduce the previous ball centric process whilst also providing the additional insights for a player’s performance when not in possession of the ball. This would make a significant difference to the quality of insights delivered and impact departmental ability to actually inform decision making and deliver insights rather than the collection of those insights. 

Redefining player evaluation

Below is a full-back profile (green) benchmarked against first-team performance standards (blue).

Each metric is deliberately selected to reflect the role I expect my full-backs to perform in our game model - not just generic positional expectations. Using tracking data, I’ve been able to include both on- and off-ball metrics.

I can still easily establish areas where the player already performs well and areas where we require development for my player to be 1st team ready. If necessary, most likely in weekly review meetings, I also have the means to link these metrics straight to video to contextualise the information.

Assessing players’ out-of-possession performance

I also want to address the issues I faced previously in our out-of-possession phases. Through use of tracking data I now assess the performance of my defenders across the entire game model. I am still reviewing the important on-the-ball moments but I am now also assessing how they defend spaces and how they allow us to defend within a team shape. For example you can see in my centre back profile template below I am assessing how they mark opponents in the box, how they clear the box after deep defending scenarios and move forward during organised pressure to see how they make the pitch smaller to make pressing scenarios more effective. 

This is information that many first team set ups don’t have access to but I am reviewing my academy players across these tactical and off-the-ball metrics to ensure their development is truly geared towards ensuring they are ready to play for my first team. I am able to make informed decisions regarding my players with more relevant and more complete information. 

Equally important for my process, I have moved the long, labour intensive job of manual coding to an external provider so that my analysts can be more present in review points and player-facing moments meaning they become more impactful by focusing on delivery.

When comparing my 1st team player (blue) and academy u18 player (green), I can see that they are performing comparably but I have some gaps in some of the metrics I consider to be most important. Now I can subsequently put a development plan in place to ensure they practice scenarios where they can defend the box and then clear it and work on their pitch position when pressing. Additionally, because I recognise these are about changing pitch position in certain scenarios, I can also take a multidisciplinary approach and work with my Sport Science and Strength & Conditioning department to ensure they have the physical capabilities to perform these types of physical movements on the pitch repeatedly.

In conclusion

The next evolution of academy analysis is not about simply collecting data, it’s about accessing the right data to drive meaningful decisions. The foundations built through consistent, long-term data collection remain critical, but the methods and technologies available today allow us to move beyond the ball-centric limitations of the past.

By embracing automated tracking data and outsourcing collection, academies can ensure both consistency and depth of insight. This shift allows analysts to spend less time coding and more time doing what truly matters: translating data into actionable insights, supporting players directly, and informing multidisciplinary development plans.

Ultimately, the goal is not just to understand what players do with the ball, but how they influence the game as a whole and ensure they develop their technical and tactical abilities to suit the needs of the club, not the general perception of what certain positions should be performing. When data is aligned with a club’s playing philosophy and long-term vision, it becomes a genuine competitive advantage. The future of player development belongs to those who can blend technology, context and human expertise to make smarter, faster and more holistic decisions.