SkillCorner: Analysing centre forward off-ball runs

We can use plots to differentiate players (including Harry Kane) and find the best talent to fit a particular team

We can use plots to differentiate players (including Harry Kane) and find the best talent to fit a particular team

"As a striker, you need to adapt to your team. My time in England and my time in Barcelona were completely different, and I had to change my way of playing" - Thierry Henry

Tracking data continues to make strides in surfacing key insights and trends within elite football. With the recent release of SkillCorner’s Game Intelligence product, we can now detect and classify the types of off-ball run that players make.

We can also classify the passing options created by these runs when a team-mate is in possession, alongside insight into a player’s passing execution and ability to resist pressure, across more than 75 global competitions.

This off-ball run data is key for player scouting, as it enables analysts to quickly understand a player’s run volume (how many runs they make), as well as their tactical awareness (what type of runs they make and whether they are effective).

An essential aspect of using football metrics to evaluate players is to understand their stability. In other words, can a player reproduce those runs from one season to the next, either within the same team or with a different one?

Stand by for future deep dives from the SkillCorner analysis team. For more information, please visit

SkillCorner are also a sponsor of TGG Live in Manchester on October 9/10. We will be hosting an expert panel discussion exploring, 'Advancements in Data Analysis for Supporting Squad Development.'

That’s what this article is about. We are going to analyse the off-ball runs of 1,293 centre forwards across 30 competitions, from 2018/2019 to 2022/23.


We have used SkillCorner’s Game Intelligence dataset to analyse off-ball run volumes normalised for time in-possession. The following tools were used:

  • Normalisation is done per 30 minutes of team in possession (p30 TIP)
  • Run profiles will look at distribution (% of total) of runs across 10 types
  • The Coefficient of Variation will evaluate how spread out the profiles are
  • Player-level Pearson Correlation will be used to determine changes from season to season

This methodology has its limitations, because it requires SkillCorner to be tracking the player and league from one season to the next, and having data from enough games to establish confidence in the stability of the metric.


First, it’s important to understand which runs are relevant for centre forwards. We analysed the off-ball run profiles of 202 centre forwards in the Premier League, LaLiga, Ligue 1, Bundesliga and Serie A that played at least eight matches of 60 minutes in 2022/2023.

Figure 1 looks at the distribution of runs made by players in different positions. We can see that Runs in Behind (32%) are a signature move for centre forwards, along with Cross Receiver runs (23%) and Runs Ahead of the Ball (22%).

All these run types offer threat through the central channel and penalty box and account for a total of 77% of run volume. Coming Short (asking for the ball to feet) is the next most common run type, representing 8% of all centre forward runs.

Figure 1: Distribution of run types by position.

Figure 1: Distribution of run types by position.


There are various types of centre forward, playing with different styles based on their physical attributes as well as the system or league they compete in. It is therefore important to understand how players differentiate themselves, and the coefficient of variation for each run type evaluates how large the difference is between players.

This reveals that build-up and space-creating runs (Pulling Wide, Pulling Half Space, Coming Short, Support) vary most among centre forwards, making them a less standard type of run for the position group.

Figure 2 shows that Pulling Wide runs have the highest variation . On the other hand, centre forwards all tend to deliver a more similar volume of Cross-Receiver runs.

Figure 2: Run type coefficient of variation.

Figure 2: Run type coefficient of variation.


A key question is whether these behaviours are inherent to a particular player, or reflect a stylistic or tactical choice of their team.

We analysed 30 competitions of SkillCorner data, going back to 2018/2019, looking at 962 cases where a centre forward played in the same team from one season to another, and 331 cases when a player moved club between seasons. In both scenarios, the player was required to have played at least eight matches of 60 minutes duration in each season.

We then looked at the stability of the metric over several seasons. Figure 3 evaluates the correlation of the volume of runs for a given run type across two consecutive seasons.

The chart reveals:

  1. Correlations are (predictably) higher when a player remains with the same team, as they settle and deliver consistently within a system. This means that the run type volume metrics for a player are likely to be consistent season after season, when they play in the same environment.
  2. Runs in Behind and Pulling Wide have the strongest correlation between seasons, and remain strong even among players changing teams. Using space and challenging the defensive line are often characteristic of fast players looking to use their speed to create threat. Figure 4 shows the correlation of Runs in Behind from one season to the next for players who remained in their club.
  3. Cross Receiver runs have the weakest correlation, as these types of runs are more specific to different playing styles and player profiles.

A team and its playing style will vary over time as the squad evolves, and a player transferring to a new team will have to adapt tactically.

For scouting purposes, this analysis can provide confidence that a player transferring in from a team with a similar system will have a high chance of adapting and delivering a similar run profile.

Figure 3: Run type correlations.

Figure 3: Run type correlations.

Figure 4: Runs in Behind stability.

Figure 4: Runs in Behind stability.


We might expect to see a low correlation for run types such as Coming Short, Pulling Wide and Pulling Half Space*, as these vary more greatly across our sample of centre forwards. These run types do, however, have the common attributes of receiving the ball to feet and with back to goal, linking up with teammates and contributing to the build-up phase of play.

When grouping these runs into a single category of ‘Link-Up Runs’ and re-running our analysis, Figure 5 reveals a strong correlation of 0.75 from one season to the next for players that remained with the same club (for players transferring between clubs the correlation was weaker, at 0.48).

Figure 5: Correlation between Link Up Runs from one season to the next (staying with same club).

Figure 5: Correlation between Link Up Runs from one season to the next (staying with same club).


In this study, we established:

  1. A player’s off-ball run profile remains fairly stable season after season within the same team, but varies significantly more when transferring between teams.
  2. The correlation of run types from one season to the next indicates that centre forwards do have a consistent profile. However, the weaker correlation when players move teams does show the importance of team style and tactics on a player’s off-ball movement.
  3. Runs in Behind, Pulling Wide and Link-Up Runs (a grouping of build-up and support runs) are reliable across seasons for profiling players, even when moving teams. Together they reflect a centre forward’s threat and their ability to participate with the ball into feet.

When scouting centre forwards using SkillCorner data, you can therefore use a plot like Figure 6 to differentiate players and find the best talent to fit into your team.

This study has established that not only can you differentiate players based on these run types, you can also rely on them to replicate that behaviour season after season within a similar system.

Figure 6: Runs in Behind and Link Up Runs.

Figure 6: Runs in Behind and Link Up Runs.

Read more on:

AnalyticsTGG Live

More stories

Sign up to our newsletter to get all the latest news from The Guru