Using data is an essential part of modern decision making. This is true in any business and a football club is no exception to that. However, making optimal use of data can be very challenging. Without the right nuances and full context, the benefits of data use may be limited. Misuse of data may even have a negative impact. As an experienced data analysis partner, we have invested a lot of energy in designing and improving our player profiles. We look at them as the cornerstone of data analysis included in our digital scouting services.

Data analysis

Our data analysis process contains three crucial aspects.

If you want, you can use the anchor links above to scroll to this subject immediately.

Event data

To understand how we use event data, let’s explain the difference between raw data, metrics and player profiles.

Raw data

Several data companies collect data by tagging all football actions in a match. This leads to a huge set of event data from every match worldwide in professional football. It contains detailed information about the moment in the game, the position on the field, players involved and much more. For an average match this would mean around 1500 actions like passes, tackles, dribbles, shots, etcetera.

Raw data are just the first step, because in itself these do not provide relevant information to separate good players from bad players. That is obviously essential when using data for scouting purposes. Therefore, we need more advanced derived statistics to turn this information into knowledge.

Data scouting


The advanced statistics that we derive from the raw event data are called metrics. The most used and well-known example of such a metric is expected goals. This is the outcome of a model that uses event data to estimate the chance a certain shot results in a goal. Sander IJtsma, one of the founders of 11tegen11 Scouting, developed such a model almost ten years ago. It was independently tested as one of the most reliable expected goal models.

In the last couple of years, we have developed 72 different metrics that we use to analyse and categorise players. We will get back to a few more examples of such metrics later. First, let’s talk about player profiles, which we built up from the metrics we developed.

Player profiles

Obviously, a coach expects different skills from an attacking midfielder than from a fullback. But even for two players on the same position, different player roles exist. One team might need a fast striker who is most effective when making a lot of deep runs in behind, such as Jamie Vardy or Luis Suárez. Another team might be looking for a false nine who is able to link up play and create a lot of chances for his teammates, such as Karim Benzema or Roberto Firmino.

In other words, we should evaluate two players on the same position differently with data. Ultimately, they play a different role within the team. One striker adds more value with his deep runs, while the other one does that by linking up play.

Find better players

Player roles

Different player roles require different skills, so we evaluate players from multiple different perspectives for any position they play. A central midfielder should be evaluated as a ball winner, a playmaker, a box-to-box runner and various other roles to fully capture his skills. In practice, this means we use various different sets of weights for our metrics for any position. Basically, a player profiles is a weighted average of a chosen number of metrics we developed.

Another example of a position with many different roles is the fullback. For an offensive fullback we assign relatively much value to metrics such as expected goals, expected assists, key passes and touches close to the opponent’s goal. For a fullback who plays a big role in the build-up play on their own half, we put more weight on metrics such as expected goal build-up and progressive passing.

Football tactics

A crucial aspect of our philosophy is the fact that we reflect the playing style of the team we scout for in the various metrics we use in our player profiles. Football tactics is always our starting point when using data in the scouting process, not the available event data set itself. This means we use data to answer very specific football related questions, rather than just measuring things for the sake of using data.

Dribbling skills

Let’s discuss another specific example where we try to stay as close to football as we can. Again, we’ll go from raw data via metrics to player profiles.

Number and success percentage

One thing that gets tagged in raw data, is every dribble a player makes. This is also the case for the success rate of those dribbles in a binary outcome of ‘won’ or ‘lost’. Those numbers do provide some information about player quality and style, but fail to capture what’s really important. Why is that?

A player might dribble a lot, and do this with a high percentage of success. But that does not necessarily mean that his dribbling has been effective for his team. Ten dribbles per 90 minutes, with a success percentage of 75, sounds really good. At the same time, it tells us very little about his effectiveness. For all we know so far, he may just have completed all of those dribbles in his own half.

Location and output

Therefore, we developed two additional dribbling metrics that tell us more about dribbling quality. The first one is the location of the dribbles, with incremental value added for dribbles closer to the opposition’s goal. The closer to the opposition goal, the harder it gets to complete a dribble. At the same time, the return when a player succeeds improves.

The second added metric is the number of expected goals that a team creates immediately after a dribble. This captures exactly what you want by getting past a defender: opening up the game, with a good scoring chances as a direct result.

When applying the eye ball test, we learned that these two self-developed metrics significantly improved our ability to find players with good dribbling skills. In other words: the best players from a data point of view now correspond much better with our opinion from watching a lot of football. This makes our whole scouting process more efficient.

Build-up contribution

Another example where we use football tactics as a starting point for developing metrics is in measuring the contribution in the build-up play. Players such as Sergio Busquets, Marco Verratti and Thiago Alcântara don’t show up in goals or assists charts. At the same time, everyone agrees on their great contribution in the build-up play. How to look beyond goals and assists to measure their skill?

Let’s discuss two examples.

  • Their role in passing chains leading to scoring opportunities (expected goal build-up).
  • Their passing leads to possession in better zones on the pitch (progressive passing value).

Expected goal build-up

We know the expected goal value of every shot, and we know which players were a part of the passing chain before the shot. By rewarding every player that touched the ball in the build-up to the goal with the expected goal value of the shot, we gain information about their contribution to the positional play of the team. This is what we call expected goal build-up. To correct for players who are particularly involved in set pieces, we prefer to isolate open play situations when looking at expected goal build-up (xGBo).

Sure, a player can play a small role in a passing chain and still be rewarded with a high xGBo. Or a player can make a smart run to create space for his teammates, an event that is not included in the xGBo. In both situations, a player gets more or less xGBo than he actually deserves. But… when the sample size of the event data that we analyse is big enough, the influence of these rare events will be limited and the metric starts to show its value.

Progressive passing value

A second metric we use for measuring the contribution in the build-up is called progressive passing value (PPV). In short, a player receives more points for executing a lot of successful passes into good zones on the pitch. Obviously, we reward a pass towards the central zone right in front of the penalty box with more points than a pass towards the wing. In fact, we’ve developed a continuous metric to measure the value of any pass without restricting ourselves to defining debatable zones on the pitch.

Like any metric, this is not perfect. Sometimes, a short horizontal pass, or even a backwards pass, can be decisive. Think of a striker finding the third man with a bounce pass. On the long term however, we’ve experienced a very good correlation between players with a high PPV – our data point of view – and the players we rate highly when it comes to progressive passing by watching a lot of matches – our video point of view.


Since we’ve developed such a broad set of metrics, it is possible to search for very specific player types. We can even manually adapt player profiles based on the exact needs of a club. When the search criteria are really specific from a tactical point of view, we are able to develop a customised player profile by creating a weighted average based on the skills the player needs to possess. Obviously, we do this after detailed discussion with the club.

In short, we are able to create a new player profile by assigning points to the most relevant metrics. The more important a metric in a certain player profile, the more value we assign to it. This leads to a club-specific player profile. It gives a clear indication about the role a player should play on a certain position.

Search criteria

Digital scouting

Besides event data, we use extensive video scouting to narrow down the list of most interesting players even further. We use data mostly as the first filter. Then, we use Video to select the best possible players within the search criteria of the club. Besides that, we use every other available online resource to find and evaluate players.

We’d be very happy to discuss our digital scouting services, as well as the needs of your club. One way to do so, is by scheduling a 30 minute video call. This way, we can get to know each other and find out if there is a match.